Skip to content

Refactor Part 2: Integration of mlir-aie Features#88

Draft
hunhoffe wants to merge 184 commits intodevelfrom
simplifying-refactor-rebased
Draft

Refactor Part 2: Integration of mlir-aie Features#88
hunhoffe wants to merge 184 commits intodevelfrom
simplifying-refactor-rebased

Conversation

@hunhoffe
Copy link
Collaborator

@hunhoffe hunhoffe commented Mar 25, 2026

Integration of iron.tensors, mlir-aie runtime helpers, and new form of link_with, mostly.
A bunch of small code quality fixes too.

PR Merge Checklist

  1. The PR is rebased on the latest devel commit and pointing to devel.
  2. Your PR has been reviewed and approved.
  3. All checks are passing.

hunhoffe and others added 9 commits March 25, 2026 13:17
…ueError

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
… design

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
…r map lookup

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
…eck in design

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
…e count

The previous implementation counted available artifacts before/after a rule
fires. GenerateMLIRFromPythonCompilationRule works by replacing a
PythonGeneratedMLIRArtifact node with an available SourceArtifact, which
caused the available count to stay the same in some graph configurations,
triggering a false positive.

The correct check is whether the set of unavailable artifact filenames
changed. If the same set remains unavailable after a rule fires, the rule
made no genuine progress.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
The AIERepeat parameter was renamed from num_repeats to repeat during the
simplifying refactor, but llama_npu.py was not updated. Fixes the CI failure:
TypeError: AIERepeat.__init__() got an unexpected keyword argument 'num_repeats'

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
- llama_inference_harness.py: remove stray if __name__ == '__main__' guard
  (no main() function exists in this file; it is a library, not a script)
- llama_inference_harness.py: initialize state.num_preceding_tokens = 0 in
  reset_kv_cache() so the NPU decode path never hits AttributeError
- llama_npu.py: move scratch_buffer.to("cpu") outside the per-layer loop
  (it was being called n_layers times redundantly per prefill pass)

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
@github-actions
Copy link
Contributor

📊 Test Results for Small Benchmark/Test Suite

9580ba2 (2026_03_25_21_04_17)

IRONCLAD

Tested on 2026_03_25_21_04_17 at commit 9580ba2.

Test Checks Latency (mean)Bandwidth (mean)GFLOP/s (mean)
M_128-K_128-num_aie_columns_1-tile_size_input_32-tile_size_output_128 ✅ 5/5 n/a 0.10 n/a
M_1792-K_896-N_1152-num_aie_columns_8-b_col_maj_False-c_col_maj_True-m_64-k_32-n_48-trace_size_0 ✅ 5/5 2359.44 4.07 n/a
M_192-K_384-N_64-num_aie_columns_4-b_col_maj_False-c_col_maj_False-m_48-k_96-n_16-trace_size_0 ✅ 5/5 561.84 0.52 n/a
M_192-K_384-N_64-num_aie_columns_4-b_col_maj_True-c_col_maj_True-m_48-k_96-n_16-trace_size_0 ✅ 5/5 688.80 0.36 n/a
M_2048-K_2048-N_2048-num_aie_columns_1-b_col_maj_False-c_col_maj_False-m_64-k_64-n_64-trace_size_0 ✅ 5/5 50117.84 0.50 n/a
M_2048-K_2048-N_2048-num_aie_columns_2-b_col_maj_True-c_col_maj_False-m_64-k_64-n_64-trace_size_0 ✅ 5/5 29306.10 0.86 n/a
M_2048-K_2048-N_2048-num_aie_columns_8-b_col_maj_True-c_col_maj_True-m_64-k_64-n_64-trace_size_0 ✅ 5/5 7818.54 3.22 n/a
M_2048-K_8192-num_aie_columns_1-tile_size_input_1-tile_size_output_2048 ✅ 5/5 n/a 11.37 n/a
M_2048-K_8192-num_aie_columns_2-tile_size_input_1-tile_size_output_1024 ✅ 5/5 n/a 17.93 n/a
M_2048-K_8192-num_aie_columns_4-tile_size_input_1-tile_size_output_512 ✅ 5/5 n/a 28.60 n/a
M_2048-K_8192-num_aie_columns_8-tile_size_input_1-tile_size_output_256 ✅ 5/5 n/a 28.69 n/a
M_2048-N_64-aie_columns_1-channels_1-m_64-n_64-s_8 ✅ 5/5 643.66 1.09 n/a
M_2048-N_64-aie_columns_1-channels_2-m_64-n_64-s_8 ✅ 5/5 412.96 1.50 n/a
M_384-K_1536-N_1792-num_aie_columns_4-b_col_maj_True-c_col_maj_False-m_32-k_48-n_64-trace_size_0 ✅ 5/5 2298.74 3.53 n/a
M_8192-K_2048-num_aie_columns_1-tile_size_input_4-tile_size_output_1024 ✅ 5/5 n/a 11.93 n/a
M_8192-K_2048-num_aie_columns_2-tile_size_input_4-tile_size_output_1024 ✅ 5/5 n/a 17.74 n/a
M_8192-K_2048-num_aie_columns_4-tile_size_input_4-tile_size_output_1024 ✅ 5/5 n/a 28.67 n/a
M_8192-K_2048-num_aie_columns_8-tile_size_input_4-tile_size_output_1024 ✅ 5/5 n/a 32.32 n/a
M_896-K_1792-N_640-num_aie_columns_8-b_col_maj_False-c_col_maj_True-m_32-k_64-n_80-trace_size_0 ✅ 5/5 1489.42 4.90 n/a
embedding_dim_2048-hidden_dim_2048 ✅ 5/5 n/a n/a n/a
input_length_2048-num_aie_columns_1-num_channels_1-tile_size_2048 ✅ 25/25 302.94 0.03 n/a
input_length_2048-num_aie_columns_1-num_channels_1-tile_size_2048-group_size_32 ✅ 5/5 421.34 0.01 n/a
input_length_2048-num_aie_columns_1-num_channels_1-tile_size_2048-weighted_False ✅ 5/5 329.58 0.03 n/a
input_length_2048-num_aie_columns_1-num_channels_1-tile_size_2048-weighted_True ✅ 5/5 276.76 0.05 n/a
input_length_2048-num_aie_columns_1-num_channels_2-tile_size_1024 ✅ 10/10 326.47 0.03 n/a
input_length_2048-num_aie_columns_1-num_channels_2-tile_size_1024-group_size_32 ✅ 5/5 394.26 0.02 n/a
input_length_2048-num_aie_columns_1-num_channels_2-tile_size_1024-weighted_False ✅ 5/5 308.88 0.03 n/a
input_length_2048-num_aie_columns_1-num_channels_2-tile_size_2048-scalar_factor_3.0 ✅ 5/5 345.00 0.04 n/a
input_length_2048-num_aie_columns_1-tile_size_2048 ✅ 15/15 319.69 0.04 n/a
input_length_2048-num_aie_columns_2-num_channels_1-tile_size_1024 ✅ 25/25 283.88 0.03 n/a
input_length_2048-num_aie_columns_2-num_channels_1-tile_size_1024-group_size_32 ✅ 5/5 343.28 0.02 n/a
input_length_2048-num_aie_columns_2-num_channels_1-tile_size_1024-weighted_False ✅ 5/5 267.40 0.03 n/a
input_length_2048-num_aie_columns_2-num_channels_1-tile_size_1024-weighted_True ✅ 5/5 280.64 0.04 n/a
input_length_2048-num_aie_columns_2-num_channels_2-tile_size_1024-scalar_factor_3.0 ✅ 5/5 279.48 0.04 n/a
input_length_2048-num_aie_columns_2-num_channels_2-tile_size_512 ✅ 10/10 288.26 0.03 n/a
input_length_2048-num_aie_columns_2-num_channels_2-tile_size_512-group_size_32 ✅ 5/5 396.70 0.02 n/a
input_length_2048-num_aie_columns_2-num_channels_2-tile_size_512-weighted_False ✅ 5/5 397.28 0.03 n/a
input_length_2048-num_aie_columns_2-tile_size_1024 ✅ 15/15 347.67 0.04 n/a
input_length_2048-num_aie_columns_4-num_channels_1-tile_size_512 ✅ 25/25 282.28 0.03 n/a
input_length_2048-num_aie_columns_4-num_channels_1-tile_size_512-group_size_32 ✅ 5/5 345.66 0.02 n/a
input_length_2048-num_aie_columns_4-num_channels_1-tile_size_512-weighted_False ✅ 5/5 318.02 0.03 n/a
input_length_2048-num_aie_columns_4-num_channels_1-tile_size_512-weighted_True ✅ 5/5 389.42 0.03 n/a
input_length_2048-num_aie_columns_4-num_channels_2-tile_size_256 ✅ 10/10 283.75 0.03 n/a
input_length_2048-num_aie_columns_4-num_channels_2-tile_size_256-group_size_32 ✅ 5/5 382.02 0.02 n/a
input_length_2048-num_aie_columns_4-num_channels_2-tile_size_256-weighted_False ✅ 5/5 262.68 0.03 n/a
input_length_2048-num_aie_columns_4-num_channels_2-tile_size_512-scalar_factor_3.0 ✅ 5/5 272.00 0.05 n/a
input_length_2048-num_aie_columns_4-tile_size_512 ✅ 15/15 326.79 0.04 n/a
input_length_2048-num_aie_columns_8-num_channels_1-tile_size_256 ✅ 25/25 336.71 0.03 n/a
input_length_2048-num_aie_columns_8-num_channels_1-tile_size_256-group_size_32 ✅ 5/5 306.22 0.02 n/a
input_length_2048-num_aie_columns_8-num_channels_1-tile_size_256-weighted_False ✅ 5/5 333.84 0.03 n/a
input_length_2048-num_aie_columns_8-num_channels_1-tile_size_256-weighted_True ✅ 5/5 355.52 0.03 n/a
input_length_2048-num_aie_columns_8-num_channels_2-tile_size_128 ✅ 10/10 332.74 0.03 n/a
input_length_2048-num_aie_columns_8-num_channels_2-tile_size_128-group_size_32 ✅ 5/5 412.64 0.02 n/a
input_length_2048-num_aie_columns_8-num_channels_2-tile_size_128-weighted_False ✅ 5/5 427.70 0.02 n/a
input_length_2048-num_aie_columns_8-num_channels_2-tile_size_256-scalar_factor_3.0 ✅ 5/5 299.32 0.04 n/a
input_length_2048-num_aie_columns_8-tile_size_256 ✅ 15/15 359.08 0.04 n/a
input_length_2048-num_cores_1-num_channels_1-bypass_False-tile_size_2048 ✅ 5/5 266.30 0.03 n/a
input_length_2048-num_cores_16-num_channels_2-bypass_False-tile_size_128 ✅ 5/5 413.24 0.03 n/a
input_length_2048-num_cores_2-num_channels_1-bypass_False-tile_size_1024 ✅ 5/5 252.14 0.03 n/a
input_length_2048-num_cores_2-num_channels_2-bypass_False-tile_size_1024 ✅ 5/5 276.88 0.03 n/a
input_length_2048-num_cores_4-num_channels_1-bypass_False-tile_size_512 ✅ 5/5 293.76 0.03 n/a
input_length_2048-num_cores_4-num_channels_2-bypass_False-tile_size_512 ✅ 5/5 272.78 0.03 n/a
input_length_2048-num_cores_8-num_channels_1-bypass_False-tile_size_256 ✅ 5/5 268.42 0.03 n/a
input_length_2048-num_cores_8-num_channels_2-bypass_False-tile_size_256 ✅ 5/5 316.62 0.03 n/a
input_length_32768-num_aie_columns_2-num_channels_2-tile_size_1024 ✅ 5/5 321.74 0.44 n/a
input_length_32768-num_aie_columns_2-num_channels_2-tile_size_512 ✅ 5/5 404.48 0.36 n/a
input_length_32768-num_aie_columns_4-num_channels_4-tile_size_2048 ✅ 5/5 656.18 0.21 n/a
rows_32-cols_512-angle_rows_32-aie_columns_1-method_type_0 ✅ 5/5 336.92 0.33 n/a
rows_32-cols_512-angle_rows_32-aie_columns_2-method_type_0 ✅ 5/5 311.60 0.33 n/a
rows_32-cols_512-angle_rows_32-aie_columns_8-method_type_0 ✅ 5/5 385.16 0.29 n/a
rows_32-cols_512-angle_rows_8-aie_columns_1-method_type_0 ✅ 5/5 315.28 0.24 n/a
rows_32-cols_512-angle_rows_8-aie_columns_2-method_type_0 ✅ 5/5 323.20 0.23 n/a
rows_32-cols_512-angle_rows_8-aie_columns_8-method_type_0 ✅ 5/5 412.66 0.20 n/a
seq_len_16384-dim_64-num_heads_1-num_pipelines_8-num_kv_heads_0 ❌ 0/5 n/a n/a n/a
seq_len_256-embedding_dim_2048-hidden_dim_2048-prio_accuracy_False ✅ 5/5 n/a n/a n/a
📈 Trends (vs main branch) for Small Benchmark/Test Suite

9580ba2 (2026_03_25_21_04_17)

IRONCLAD Trends

M_128-K_128-num_aie_columns_1-tile_size_input_32-tile_size_output_128

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
9580ba2 — 2026-03-25 20:55:190.13 (n/a)0.10 (n/a)0.09 (n/a)0.04 (n/a)0.03 (n/a)0.13 (n/a)0.09 (n/a)0.09 (n/a)0.04 (n/a)0.03 (n/a)

M_1792-K_896-N_1152-num_aie_columns_8-b_col_maj_False-c_col_maj_True-m_64-k_32-n_48-trace_size_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
9580ba2 — 2026-03-25 20:55:194.56 (n/a)4.07 (n/a)4.51 (n/a)3.20 (n/a)0.64 (n/a)2938.40 (n/a)2359.44 (n/a)2084.20 (n/a)2061.70 (n/a)404.25 (n/a)1794.36 (n/a)1602.46 (n/a)1774.98 (n/a)1258.98 (n/a)252.55 (n/a)

M_192-K_384-N_64-num_aie_columns_4-b_col_maj_False-c_col_maj_False-m_48-k_96-n_16-trace_size_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
9580ba2 — 2026-03-25 20:55:190.79 (n/a)0.52 (n/a)0.55 (n/a)0.21 (n/a)0.27 (n/a)1065.80 (n/a)561.84 (n/a)405.40 (n/a)278.50 (n/a)345.82 (n/a)33.88 (n/a)22.22 (n/a)23.28 (n/a)8.85 (n/a)11.50 (n/a)

M_192-K_384-N_64-num_aie_columns_4-b_col_maj_True-c_col_maj_True-m_48-k_96-n_16-trace_size_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
9580ba2 — 2026-03-25 20:55:190.65 (n/a)0.36 (n/a)0.29 (n/a)0.26 (n/a)0.16 (n/a)840.50 (n/a)688.80 (n/a)751.70 (n/a)341.10 (n/a)206.65 (n/a)27.67 (n/a)15.35 (n/a)12.55 (n/a)11.23 (n/a)6.98 (n/a)

M_2048-K_2048-N_2048-num_aie_columns_1-b_col_maj_False-c_col_maj_False-m_64-k_64-n_64-trace_size_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
9580ba2 — 2026-03-25 20:55:190.50 (n/a)0.50 (n/a)0.50 (n/a)0.50 (n/a)0.00 (n/a)50614.10 (n/a)50117.84 (n/a)49992.10 (n/a)49972.00 (n/a)278.76 (n/a)343.79 (n/a)342.80 (n/a)343.65 (n/a)339.43 (n/a)1.89 (n/a)

M_2048-K_2048-N_2048-num_aie_columns_2-b_col_maj_True-c_col_maj_False-m_64-k_64-n_64-trace_size_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
9580ba2 — 2026-03-25 20:55:190.87 (n/a)0.86 (n/a)0.86 (n/a)0.85 (n/a)0.00 (n/a)29434.00 (n/a)29306.10 (n/a)29383.20 (n/a)29092.00 (n/a)143.72 (n/a)590.54 (n/a)586.23 (n/a)584.68 (n/a)583.67 (n/a)2.88 (n/a)

M_2048-K_2048-N_2048-num_aie_columns_8-b_col_maj_True-c_col_maj_True-m_64-k_64-n_64-trace_size_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
9580ba2 — 2026-03-25 20:55:193.37 (n/a)3.22 (n/a)3.25 (n/a)3.05 (n/a)0.13 (n/a)8247.80 (n/a)7818.54 (n/a)7750.40 (n/a)7462.70 (n/a)314.91 (n/a)2302.10 (n/a)2200.15 (n/a)2216.64 (n/a)2082.97 (n/a)87.78 (n/a)

M_2048-K_8192-num_aie_columns_1-tile_size_input_1-tile_size_output_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
9580ba2 — 2026-03-25 20:55:1912.45 (n/a)11.37 (n/a)11.31 (n/a)10.42 (n/a)0.74 (n/a)12.44 (n/a)11.37 (n/a)11.30 (n/a)10.42 (n/a)0.74 (n/a)

M_2048-K_8192-num_aie_columns_2-tile_size_input_1-tile_size_output_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
9580ba2 — 2026-03-25 20:55:1921.91 (n/a)17.93 (n/a)19.07 (n/a)13.49 (n/a)4.19 (n/a)21.90 (n/a)17.92 (n/a)19.05 (n/a)13.48 (n/a)4.19 (n/a)

M_2048-K_8192-num_aie_columns_4-tile_size_input_1-tile_size_output_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
9580ba2 — 2026-03-25 20:55:1933.97 (n/a)28.60 (n/a)29.29 (n/a)22.41 (n/a)4.98 (n/a)33.95 (n/a)28.58 (n/a)29.27 (n/a)22.39 (n/a)4.97 (n/a)

M_2048-K_8192-num_aie_columns_8-tile_size_input_1-tile_size_output_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
9580ba2 — 2026-03-25 20:55:1933.72 (n/a)28.69 (n/a)29.00 (n/a)21.57 (n/a)5.17 (n/a)33.70 (n/a)28.67 (n/a)28.98 (n/a)21.56 (n/a)5.17 (n/a)

M_2048-N_64-aie_columns_1-channels_1-m_64-n_64-s_8

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
9580ba2 — 2026-03-25 20:55:192.08 (n/a)1.09 (n/a)1.06 (n/a)0.40 (n/a)0.63 (n/a)1299.90 (n/a)643.66 (n/a)492.90 (n/a)251.90 (n/a)404.86 (n/a)

M_2048-N_64-aie_columns_1-channels_2-m_64-n_64-s_8

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
9580ba2 — 2026-03-25 20:55:192.00 (n/a)1.50 (n/a)1.86 (n/a)0.75 (n/a)0.61 (n/a)700.30 (n/a)412.96 (n/a)281.20 (n/a)262.10 (n/a)202.52 (n/a)

M_384-K_1536-N_1792-num_aie_columns_4-b_col_maj_True-c_col_maj_False-m_32-k_48-n_64-trace_size_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
9580ba2 — 2026-03-25 20:55:193.89 (n/a)3.53 (n/a)3.33 (n/a)3.27 (n/a)0.32 (n/a)2466.70 (n/a)2298.74 (n/a)2422.50 (n/a)2072.40 (n/a)200.50 (n/a)1020.02 (n/a)925.41 (n/a)872.62 (n/a)856.99 (n/a)83.27 (n/a)

M_8192-K_2048-num_aie_columns_1-tile_size_input_4-tile_size_output_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
9580ba2 — 2026-03-25 20:55:1912.45 (n/a)11.93 (n/a)12.39 (n/a)10.08 (n/a)1.04 (n/a)12.45 (n/a)11.92 (n/a)12.39 (n/a)10.07 (n/a)1.04 (n/a)

M_8192-K_2048-num_aie_columns_2-tile_size_input_4-tile_size_output_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
9580ba2 — 2026-03-25 20:55:1921.58 (n/a)17.74 (n/a)16.44 (n/a)13.33 (n/a)3.60 (n/a)21.56 (n/a)17.73 (n/a)16.43 (n/a)13.32 (n/a)3.60 (n/a)

M_8192-K_2048-num_aie_columns_4-tile_size_input_4-tile_size_output_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
9580ba2 — 2026-03-25 20:55:1934.54 (n/a)28.67 (n/a)27.40 (n/a)23.08 (n/a)4.95 (n/a)34.52 (n/a)28.65 (n/a)27.38 (n/a)23.07 (n/a)4.95 (n/a)

M_8192-K_2048-num_aie_columns_8-tile_size_input_4-tile_size_output_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
9580ba2 — 2026-03-25 20:55:1936.08 (n/a)32.32 (n/a)35.45 (n/a)24.79 (n/a)4.98 (n/a)36.06 (n/a)32.30 (n/a)35.42 (n/a)24.77 (n/a)4.98 (n/a)

M_896-K_1792-N_640-num_aie_columns_8-b_col_maj_False-c_col_maj_True-m_32-k_64-n_80-trace_size_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
9580ba2 — 2026-03-25 20:55:195.71 (n/a)4.90 (n/a)5.58 (n/a)2.59 (n/a)1.33 (n/a)2573.00 (n/a)1489.42 (n/a)1193.10 (n/a)1164.50 (n/a)610.52 (n/a)1764.95 (n/a)1514.06 (n/a)1722.62 (n/a)798.76 (n/a)412.01 (n/a)

axpy_1_cols_2_channels_2048_tile_2048_3.0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.08 (+17.48%)0.07 (+11.67%)0.07 (+26.71%)0.05 (+0.80%)0.01 (+108.36%)239.80 (-0.83%)190.62 (-7.91%)166.90 (-21.09%)155.40 (-14.90%)42.89 (+79.10%)
dc66da7 — 2026-03-25 19:23:450.07 (n/a)0.06 (n/a)0.06 (n/a)0.05 (n/a)0.01 (n/a)241.80 (n/a)207.00 (n/a)211.50 (n/a)182.60 (n/a)23.95 (n/a)

axpy_1_cols_2_channels_2048_tile_2048_3.0_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.16 (-3.31%)0.14 (-1.29%)0.15 (+16.18%)0.10 (-6.80%)0.03 (+3.89%)125.70 (+7.25%)93.56 (+2.12%)83.50 (-13.92%)75.00 (+3.45%)22.38 (+18.09%)
84d3478 — 2026-02-17 23:25:130.17 (n/a)0.14 (n/a)0.13 (n/a)0.10 (n/a)0.03 (n/a)117.20 (n/a)91.62 (n/a)97.00 (n/a)72.50 (n/a)18.95 (n/a)

axpy_2_cols_2_channels_2048_tile_1024_3.0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.09 (+14.45%)0.07 (-0.78%)0.07 (-0.40%)0.05 (-7.87%)0.01 (+68.50%)233.40 (+8.51%)188.20 (+2.47%)179.80 (+0.39%)143.50 (-12.66%)34.06 (+61.09%)
dc66da7 — 2026-03-25 19:23:450.07 (n/a)0.07 (n/a)0.07 (n/a)0.06 (n/a)0.01 (n/a)215.10 (n/a)183.66 (n/a)179.10 (n/a)164.30 (n/a)21.14 (n/a)

axpy_2_cols_2_channels_2048_tile_1024_3.0_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.13 (-0.73%)0.11 (-7.43%)0.11 (-8.63%)0.07 (-27.71%)0.02 (+50.39%)182.70 (+38.30%)120.86 (+11.97%)110.80 (+9.38%)96.10 (+0.73%)35.16 (+122.88%)
84d3478 — 2026-02-17 23:25:130.13 (n/a)0.12 (n/a)0.12 (n/a)0.09 (n/a)0.02 (n/a)132.10 (n/a)107.94 (n/a)101.30 (n/a)95.40 (n/a)15.78 (n/a)

axpy_4_cols_2_channels_2048_tile_512_3.0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.10 (+6.67%)0.06 (-1.56%)0.06 (-2.24%)0.05 (+3.60%)0.02 (+20.83%)251.50 (-3.49%)200.78 (+3.11%)191.70 (+2.29%)128.70 (-6.26%)50.37 (+11.30%)
dc66da7 — 2026-03-25 19:23:450.09 (n/a)0.07 (n/a)0.07 (n/a)0.05 (n/a)0.02 (n/a)260.60 (n/a)194.72 (n/a)187.40 (n/a)137.30 (n/a)45.26 (n/a)

axpy_4_cols_2_channels_2048_tile_512_3.0_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.13 (-10.91%)0.11 (-7.68%)0.12 (-0.73%)0.06 (-17.94%)0.03 (+0.99%)194.10 (+21.85%)124.52 (+10.29%)103.30 (+0.68%)97.20 (+12.24%)40.32 (+39.15%)
84d3478 — 2026-02-17 23:25:130.14 (n/a)0.11 (n/a)0.12 (n/a)0.08 (n/a)0.03 (n/a)159.30 (n/a)112.90 (n/a)102.60 (n/a)86.60 (n/a)28.97 (n/a)

axpy_8_cols_2_channels_2048_tile_256_3.0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.08 (+1.42%)0.06 (+9.86%)0.06 (+18.71%)0.03 (+13.51%)0.02 (-10.62%)376.00 (-11.90%)225.02 (-11.61%)196.20 (-15.76%)151.40 (-1.37%)87.15 (-17.99%)
dc66da7 — 2026-03-25 19:23:450.08 (n/a)0.05 (n/a)0.05 (n/a)0.03 (n/a)0.02 (n/a)426.80 (n/a)254.58 (n/a)232.90 (n/a)153.50 (n/a)106.27 (n/a)

axpy_8_cols_2_channels_2048_tile_256_3.0_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.13 (+0.01%)0.11 (+6.10%)0.11 (+0.22%)0.08 (-0.33%)0.02 (-9.83%)156.30 (+0.32%)115.26 (-6.40%)112.70 (-0.18%)95.50 (+0.00%)24.34 (-10.22%)
84d3478 — 2026-02-17 23:25:130.13 (n/a)0.10 (n/a)0.11 (n/a)0.08 (n/a)0.02 (n/a)155.80 (n/a)123.14 (n/a)112.90 (n/a)95.50 (n/a)27.11 (n/a)

dequant_1_cols_1_channels_2048_tile_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.04 (+17.85%)0.03 (+2.55%)0.03 (+1.65%)0.03 (-2.18%)0.00 (+114.37%)207.70 (+2.26%)178.64 (-0.96%)173.60 (-1.59%)141.60 (-15.16%)28.01 (+90.63%)
dc66da7 — 2026-03-25 19:23:450.03 (n/a)0.03 (n/a)0.03 (n/a)0.03 (n/a)0.00 (n/a)203.10 (n/a)180.38 (n/a)176.40 (n/a)166.90 (n/a)14.69 (n/a)

dequant_1_cols_1_channels_2048_tile_2048_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.05 (-13.00%)0.04 (-0.87%)0.04 (+1.16%)0.03 (-3.17%)0.01 (-31.33%)158.40 (+3.26%)128.12 (-0.19%)123.80 (-1.12%)110.40 (+14.88%)18.36 (-15.65%)
84d3478 — 2026-02-17 23:25:130.05 (n/a)0.04 (n/a)0.04 (n/a)0.03 (n/a)0.01 (n/a)153.40 (n/a)128.36 (n/a)125.20 (n/a)96.10 (n/a)21.77 (n/a)

dequant_1_cols_2_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.04 (+13.44%)0.03 (+8.77%)0.03 (+14.89%)0.02 (-9.15%)0.01 (+54.39%)252.70 (+10.06%)182.14 (-6.08%)174.50 (-12.97%)145.40 (-11.88%)42.38 (+56.31%)
dc66da7 — 2026-03-25 19:23:450.03 (n/a)0.03 (n/a)0.03 (n/a)0.02 (n/a)0.00 (n/a)229.60 (n/a)193.94 (n/a)200.50 (n/a)165.00 (n/a)27.11 (n/a)

dequant_1_cols_2_channels_2048_tile_1024_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.06 (-3.74%)0.05 (-12.63%)0.05 (-5.21%)0.03 (-33.47%)0.01 (+52.60%)162.70 (+50.37%)110.84 (+18.98%)98.80 (+5.44%)81.50 (+3.95%)31.02 (+149.23%)
84d3478 — 2026-02-17 23:25:130.07 (n/a)0.06 (n/a)0.06 (n/a)0.05 (n/a)0.01 (n/a)108.20 (n/a)93.16 (n/a)93.70 (n/a)78.40 (n/a)12.45 (n/a)

dequant_2_cols_1_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.05 (+21.89%)0.03 (+20.89%)0.03 (+21.41%)0.02 (+8.07%)0.01 (+32.97%)215.80 (-7.46%)168.56 (-16.36%)172.60 (-17.65%)113.20 (-17.97%)37.61 (+1.89%)
dc66da7 — 2026-03-25 19:23:450.04 (n/a)0.03 (n/a)0.03 (n/a)0.02 (n/a)0.01 (n/a)233.20 (n/a)201.52 (n/a)209.60 (n/a)138.00 (n/a)36.92 (n/a)

dequant_2_cols_1_channels_2048_tile_1024_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.06 (+8.90%)0.05 (+1.31%)0.05 (-1.85%)0.03 (+6.35%)0.01 (+7.11%)152.30 (-5.93%)115.08 (-1.20%)100.80 (+1.92%)88.10 (-8.23%)27.20 (-4.42%)
84d3478 — 2026-02-17 23:25:130.05 (n/a)0.05 (n/a)0.05 (n/a)0.03 (n/a)0.01 (n/a)161.90 (n/a)116.48 (n/a)98.90 (n/a)96.00 (n/a)28.46 (n/a)

dequant_2_cols_2_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.05 (+49.89%)0.03 (+24.89%)0.03 (+13.74%)0.02 (+15.73%)0.01 (+106.91%)219.80 (-13.60%)173.04 (-17.40%)185.70 (-12.07%)112.30 (-33.27%)41.15 (+16.74%)
dc66da7 — 2026-03-25 19:23:450.03 (n/a)0.03 (n/a)0.02 (n/a)0.02 (n/a)0.00 (n/a)254.40 (n/a)209.50 (n/a)211.20 (n/a)168.30 (n/a)35.25 (n/a)

dequant_2_cols_2_channels_2048_tile_512_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.07 (+27.96%)0.05 (+1.57%)0.04 (-18.06%)0.03 (+10.23%)0.01 (+22.79%)164.70 (-9.31%)119.12 (-1.39%)120.20 (+22.03%)75.60 (-21.82%)31.61 (-13.80%)
84d3478 — 2026-02-17 23:25:130.05 (n/a)0.05 (n/a)0.05 (n/a)0.03 (n/a)0.01 (n/a)181.60 (n/a)120.80 (n/a)98.50 (n/a)96.70 (n/a)36.67 (n/a)

dequant_4_cols_1_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.04 (+56.54%)0.03 (+15.78%)0.03 (+9.01%)0.02 (-5.58%)0.01 (+304.26%)243.40 (+5.92%)186.06 (-8.98%)182.60 (-8.29%)118.50 (-36.12%)46.36 (+164.07%)
dc66da7 — 2026-03-25 19:23:450.03 (n/a)0.03 (n/a)0.03 (n/a)0.02 (n/a)0.00 (n/a)229.80 (n/a)204.42 (n/a)199.10 (n/a)185.50 (n/a)17.56 (n/a)

dequant_4_cols_1_channels_2048_tile_512_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.05 (-2.55%)0.04 (-12.79%)0.05 (-16.25%)0.03 (-12.85%)0.01 (-1.20%)164.60 (+14.78%)122.14 (+15.14%)115.90 (+19.48%)97.50 (+2.63%)25.08 (+19.84%)
84d3478 — 2026-02-17 23:25:130.06 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.01 (n/a)143.40 (n/a)106.08 (n/a)97.00 (n/a)95.00 (n/a)20.93 (n/a)

dequant_4_cols_2_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.04 (+28.72%)0.03 (+2.89%)0.02 (-6.84%)0.02 (+6.27%)0.01 (+95.59%)243.20 (-5.88%)208.42 (-0.42%)222.90 (+7.32%)141.10 (-22.34%)41.08 (+38.16%)
dc66da7 — 2026-03-25 19:23:450.03 (n/a)0.03 (n/a)0.03 (n/a)0.02 (n/a)0.00 (n/a)258.40 (n/a)209.30 (n/a)207.70 (n/a)181.70 (n/a)29.74 (n/a)

dequant_4_cols_2_channels_2048_tile_256_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.06 (+20.83%)0.05 (+8.21%)0.04 (+1.42%)0.03 (+17.39%)0.01 (+14.84%)166.40 (-14.80%)122.34 (-7.97%)127.70 (-1.39%)84.60 (-17.22%)30.47 (-19.21%)
84d3478 — 2026-02-17 23:25:130.05 (n/a)0.04 (n/a)0.04 (n/a)0.03 (n/a)0.01 (n/a)195.30 (n/a)132.94 (n/a)129.50 (n/a)102.20 (n/a)37.71 (n/a)

dequant_8_cols_1_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.04 (-1.52%)0.03 (+2.17%)0.03 (+18.60%)0.02 (-11.90%)0.01 (+4.42%)268.50 (+13.53%)205.32 (-1.14%)191.00 (-15.67%)141.60 (+1.51%)49.39 (+23.46%)
dc66da7 — 2026-03-25 19:23:450.04 (n/a)0.03 (n/a)0.02 (n/a)0.02 (n/a)0.01 (n/a)236.50 (n/a)207.68 (n/a)226.50 (n/a)139.50 (n/a)40.00 (n/a)

dequant_8_cols_1_channels_2048_tile_256_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.05 (-25.19%)0.04 (-21.02%)0.04 (-19.70%)0.04 (-19.76%)0.00 (-37.92%)142.70 (+24.63%)126.44 (+26.24%)125.70 (+24.46%)115.50 (+33.68%)10.61 (+3.97%)
84d3478 — 2026-02-17 23:25:130.06 (n/a)0.05 (n/a)0.05 (n/a)0.05 (n/a)0.01 (n/a)114.50 (n/a)100.16 (n/a)101.00 (n/a)86.40 (n/a)10.20 (n/a)

dequant_8_cols_2_channels_2048_tile_128

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.02 (+5.64%)0.02 (+1.26%)0.02 (-1.25%)0.02 (-3.87%)0.00 (+124.04%)258.90 (+4.02%)239.08 (-0.87%)247.20 (+1.27%)215.50 (-5.36%)18.28 (+120.74%)
dc66da7 — 2026-03-25 19:23:450.02 (n/a)0.02 (n/a)0.02 (n/a)0.02 (n/a)0.00 (n/a)248.90 (n/a)241.18 (n/a)244.10 (n/a)227.70 (n/a)8.28 (n/a)

dequant_8_cols_2_channels_2048_tile_128_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.04 (-26.69%)0.04 (-11.93%)0.04 (-16.99%)0.03 (+27.28%)0.01 (-62.60%)193.10 (-21.41%)147.68 (+2.96%)136.10 (+20.44%)135.40 (+36.49%)25.41 (-58.96%)
84d3478 — 2026-02-17 23:25:130.05 (n/a)0.04 (n/a)0.05 (n/a)0.02 (n/a)0.01 (n/a)245.70 (n/a)143.44 (n/a)113.00 (n/a)99.20 (n/a)61.90 (n/a)

eltwise_add_1_cols_2_channels_2048_tile_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.07 (-32.66%)0.06 (-13.87%)0.06 (-9.19%)0.04 (-8.04%)0.01 (-49.50%)299.50 (+8.75%)223.84 (+11.90%)212.70 (+10.15%)187.60 (+48.42%)45.70 (-16.71%)
dc66da7 — 2026-03-25 19:23:450.10 (n/a)0.07 (n/a)0.06 (n/a)0.04 (n/a)0.02 (n/a)275.40 (n/a)200.04 (n/a)193.10 (n/a)126.40 (n/a)54.87 (n/a)

eltwise_add_2_cols_2_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.07 (-33.33%)0.06 (-18.97%)0.05 (-19.23%)0.05 (+18.13%)0.01 (-59.22%)270.80 (-15.37%)223.26 (+13.77%)229.40 (+23.80%)174.50 (+50.04%)37.51 (-50.17%)
dc66da7 — 2026-03-25 19:23:450.11 (n/a)0.07 (n/a)0.07 (n/a)0.04 (n/a)0.02 (n/a)320.00 (n/a)196.24 (n/a)185.30 (n/a)116.30 (n/a)75.29 (n/a)

eltwise_add_4_cols_2_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.07 (-8.97%)0.06 (-6.00%)0.06 (-2.02%)0.05 (-2.74%)0.01 (-14.89%)243.40 (+2.83%)208.60 (+6.01%)200.50 (+2.04%)178.10 (+9.87%)31.58 (-2.07%)
dc66da7 — 2026-03-25 19:23:450.08 (n/a)0.06 (n/a)0.06 (n/a)0.05 (n/a)0.01 (n/a)236.70 (n/a)196.78 (n/a)196.50 (n/a)162.10 (n/a)32.25 (n/a)

eltwise_add_8_cols_2_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.06 (-45.96%)0.05 (-24.77%)0.05 (-15.00%)0.05 (+4.78%)0.00 (-83.62%)259.60 (-4.56%)234.66 (+20.97%)224.10 (+17.64%)218.40 (+85.08%)18.77 (-71.47%)
dc66da7 — 2026-03-25 19:23:450.10 (n/a)0.07 (n/a)0.06 (n/a)0.05 (n/a)0.02 (n/a)272.00 (n/a)193.98 (n/a)190.50 (n/a)118.00 (n/a)65.77 (n/a)

eltwise_mul_1_cols_2_channels_2048_tile_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.11 (+3.13%)0.07 (+2.52%)0.06 (-1.87%)0.05 (-14.01%)0.02 (+21.62%)257.00 (+16.29%)184.10 (+0.67%)190.70 (+1.87%)114.40 (-3.05%)55.57 (+40.96%)
dc66da7 — 2026-03-25 19:23:450.10 (n/a)0.07 (n/a)0.07 (n/a)0.06 (n/a)0.02 (n/a)221.00 (n/a)182.88 (n/a)187.20 (n/a)118.00 (n/a)39.42 (n/a)

eltwise_mul_2_cols_2_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.09 (+2.43%)0.07 (+6.63%)0.08 (+30.25%)0.05 (-10.38%)0.02 (+9.28%)264.20 (+11.62%)182.10 (-4.96%)162.40 (-23.22%)137.30 (-2.42%)52.80 (+20.53%)
dc66da7 — 2026-03-25 19:23:450.09 (n/a)0.07 (n/a)0.06 (n/a)0.05 (n/a)0.02 (n/a)236.70 (n/a)191.60 (n/a)211.50 (n/a)140.70 (n/a)43.81 (n/a)

eltwise_mul_4_cols_2_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.10 (+22.79%)0.08 (+18.44%)0.09 (+34.73%)0.06 (-0.12%)0.01 (+158.99%)191.20 (+0.10%)154.44 (-13.56%)136.60 (-25.80%)126.60 (-18.53%)29.95 (+117.72%)
dc66da7 — 2026-03-25 19:23:450.08 (n/a)0.07 (n/a)0.07 (n/a)0.06 (n/a)0.01 (n/a)191.00 (n/a)178.66 (n/a)184.10 (n/a)155.40 (n/a)13.76 (n/a)

eltwise_mul_8_cols_2_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.08 (+34.81%)0.06 (+14.94%)0.06 (+6.87%)0.04 (-5.87%)0.01 (+158.26%)280.10 (+6.22%)210.00 (-9.80%)215.10 (-6.40%)150.70 (-25.80%)49.49 (+100.84%)
dc66da7 — 2026-03-25 19:23:450.06 (n/a)0.05 (n/a)0.05 (n/a)0.05 (n/a)0.01 (n/a)263.70 (n/a)232.82 (n/a)229.80 (n/a)203.10 (n/a)24.64 (n/a)

embedding_dim_2048-hidden_dim_2048

No metrics available.

gelu_1_cols_1_channels_2048_tile_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.05 (+9.18%)0.04 (-1.03%)0.04 (-2.51%)0.04 (-8.09%)0.01 (+114.58%)221.00 (+8.81%)195.42 (+1.97%)199.10 (+2.58%)163.60 (-8.40%)23.03 (+114.14%)
dc66da7 — 2026-03-25 19:23:450.05 (n/a)0.04 (n/a)0.04 (n/a)0.04 (n/a)0.00 (n/a)203.10 (n/a)191.64 (n/a)194.10 (n/a)178.60 (n/a)10.76 (n/a)

gelu_1_cols_2_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.04 (-26.60%)0.04 (-19.13%)0.04 (-18.64%)0.04 (-2.84%)0.00 (-66.68%)226.70 (+2.91%)199.16 (+20.50%)193.10 (+22.92%)186.40 (+36.16%)16.29 (-52.49%)
dc66da7 — 2026-03-25 19:23:450.06 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.01 (n/a)220.30 (n/a)165.28 (n/a)157.10 (n/a)136.90 (n/a)34.28 (n/a)

gelu_2_cols_1_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.06 (+2.08%)0.05 (+4.66%)0.04 (-1.75%)0.03 (-5.55%)0.01 (+12.71%)283.70 (+5.86%)191.18 (-3.05%)191.00 (+1.81%)137.60 (-1.99%)58.08 (+15.01%)
dc66da7 — 2026-03-25 19:23:450.06 (n/a)0.04 (n/a)0.04 (n/a)0.03 (n/a)0.01 (n/a)268.00 (n/a)197.20 (n/a)187.60 (n/a)140.40 (n/a)50.50 (n/a)

gelu_2_cols_2_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.07 (+2.20%)0.04 (-2.29%)0.04 (+1.04%)0.04 (+16.91%)0.01 (-4.76%)227.00 (-14.47%)194.56 (+0.69%)206.00 (-1.01%)119.40 (-2.21%)44.27 (-20.37%)
dc66da7 — 2026-03-25 19:23:450.07 (n/a)0.05 (n/a)0.04 (n/a)0.03 (n/a)0.01 (n/a)265.40 (n/a)193.22 (n/a)208.10 (n/a)122.10 (n/a)55.59 (n/a)

gelu_4_cols_1_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.07 (-9.75%)0.04 (-17.39%)0.04 (-20.48%)0.03 (+12.65%)0.01 (-29.41%)239.40 (-11.23%)196.08 (+15.06%)197.90 (+25.73%)124.70 (+10.84%)44.40 (-31.09%)
dc66da7 — 2026-03-25 19:23:450.07 (n/a)0.05 (n/a)0.05 (n/a)0.03 (n/a)0.02 (n/a)269.70 (n/a)170.42 (n/a)157.40 (n/a)112.50 (n/a)64.43 (n/a)

gelu_4_cols_2_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.04 (-6.29%)0.04 (+1.81%)0.04 (+0.94%)0.03 (-0.42%)0.00 (-22.18%)284.90 (+0.42%)233.36 (-2.39%)227.20 (-0.96%)204.80 (+6.72%)30.62 (-15.21%)
dc66da7 — 2026-03-25 19:23:450.04 (n/a)0.03 (n/a)0.04 (n/a)0.03 (n/a)0.01 (n/a)283.70 (n/a)239.08 (n/a)229.40 (n/a)191.90 (n/a)36.12 (n/a)

gelu_8_cols_1_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.05 (-18.24%)0.04 (-13.70%)0.04 (-4.91%)0.03 (-15.26%)0.01 (-28.46%)285.90 (+18.04%)223.38 (+14.20%)223.20 (+5.18%)156.90 (+22.29%)47.35 (+1.82%)
dc66da7 — 2026-03-25 19:23:450.06 (n/a)0.04 (n/a)0.04 (n/a)0.03 (n/a)0.01 (n/a)242.20 (n/a)195.60 (n/a)212.20 (n/a)128.30 (n/a)46.50 (n/a)

gelu_8_cols_2_channels_2048_tile_128

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.05 (+14.38%)0.04 (+10.49%)0.03 (+3.25%)0.03 (+13.03%)0.01 (+26.31%)305.70 (-11.52%)235.82 (-8.98%)241.80 (-3.12%)177.40 (-12.57%)49.18 (-6.09%)
dc66da7 — 2026-03-25 19:23:450.04 (n/a)0.03 (n/a)0.03 (n/a)0.02 (n/a)0.01 (n/a)345.50 (n/a)259.10 (n/a)249.60 (n/a)202.90 (n/a)52.37 (n/a)

gemm_1792x896x1152_64x32x48_8cols_ccolmaj

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
8b6d0f4 — 2026-03-25 20:46:394.89 (+9.71%)4.07 (+2.20%)4.10 (+2.03%)3.40 (+0.07%)0.55 (+44.15%)2767.30 (-0.07%)2344.32 (-1.52%)2291.20 (-1.99%)1922.40 (-8.85%)308.92 (+28.54%)1924.38 (+9.71%)1600.58 (+2.20%)1614.60 (+2.03%)1336.81 (+0.07%)215.84 (+44.15%)
dc66da7 — 2026-03-25 19:23:454.46 (n/a)3.98 (n/a)4.02 (n/a)3.40 (n/a)0.38 (n/a)2769.20 (n/a)2380.42 (n/a)2337.70 (n/a)2109.10 (n/a)240.33 (n/a)1754.05 (n/a)1566.13 (n/a)1582.49 (n/a)1335.88 (n/a)149.73 (n/a)

gemm_192x384x64_48x96x16_4cols

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
8b6d0f4 — 2026-03-25 20:46:391.35 (+17.71%)0.77 (-13.89%)0.65 (-34.22%)0.56 (+7.33%)0.32 (+33.14%)393.60 (-6.84%)315.66 (+18.48%)341.40 (+52.00%)164.30 (-15.05%)87.82 (-5.13%)57.45 (+17.71%)32.93 (-13.89%)27.64 (-34.22%)23.97 (+7.33%)13.81 (+33.14%)
dc66da7 — 2026-03-25 19:23:451.14 (n/a)0.90 (n/a)0.98 (n/a)0.52 (n/a)0.24 (n/a)422.50 (n/a)266.42 (n/a)224.60 (n/a)193.40 (n/a)92.57 (n/a)48.81 (n/a)38.25 (n/a)42.02 (n/a)22.34 (n/a)10.37 (n/a)

gemm_192x384x64_48x96x16_4cols_bcolmaj_ccolmaj

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
8b6d0f4 — 2026-03-25 20:46:391.18 (+31.18%)0.79 (-0.76%)0.72 (-17.59%)0.64 (+26.96%)0.22 (+35.66%)343.10 (-21.24%)294.50 (+0.95%)309.00 (+21.37%)188.10 (-23.78%)62.26 (-23.01%)50.17 (+31.18%)33.62 (-0.76%)30.54 (-17.59%)27.51 (+26.96%)9.40 (+35.66%)
dc66da7 — 2026-03-25 19:23:450.90 (n/a)0.79 (n/a)0.87 (n/a)0.51 (n/a)0.16 (n/a)435.60 (n/a)291.72 (n/a)254.60 (n/a)246.80 (n/a)80.87 (n/a)38.24 (n/a)33.88 (n/a)37.06 (n/a)21.67 (n/a)6.93 (n/a)

gemm_2048x2048x2048_64x64x32_8_cols_0_bcolmaj_0_ccolmaj_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)GFLOP/s (max)GFLOP/s (mean)GFLOP/s (median)GFLOP/s (min)GFLOP/s (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
3a053d3 — 2025-11-26 01:12:163.72 (-0.11%)3.55 (-2.75%)3.56 (-2.75%)3.36 (-5.33%)0.12 (+76.10%)2725.23 (-0.11%)2602.52 (-2.75%)2612.11 (-2.75%)2459.89 (-5.33%)86.27 (+76.10%)6984.00 (+5.63%)6608.60 (+2.91%)6577.00 (+2.83%)6304.00 (+0.11%)221.87 (+87.15%)
9e98049 — 2025-11-25 20:50:553.72 (n/a)3.65 (n/a)3.66 (n/a)3.54 (n/a)0.07 (n/a)2728.26 (n/a)2676.06 (n/a)2686.03 (n/a)2598.29 (n/a)48.99 (n/a)6612.00 (n/a)6422.00 (n/a)6396.00 (n/a)6297.00 (n/a)118.55 (n/a)

gemm_2048x2048x2048_64x64x32_8_cols_0_bcolmaj_1_ccolmaj_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)GFLOP/s (max)GFLOP/s (mean)GFLOP/s (median)GFLOP/s (min)GFLOP/s (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
3a053d3 — 2025-11-26 01:12:163.40 (-3.14%)3.38 (-0.82%)3.38 (-0.82%)3.35 (+4.93%)0.02 (-84.51%)2493.81 (-3.14%)2475.27 (-0.82%)2479.06 (-0.82%)2453.92 (+4.93%)13.33 (-84.51%)7001.00 (-4.70%)6940.80 (+0.71%)6930.00 (+0.83%)6889.00 (+3.24%)37.43 (-84.76%)
9e98049 — 2025-11-25 20:50:553.51 (n/a)3.40 (n/a)3.41 (n/a)3.19 (n/a)0.12 (n/a)2574.53 (n/a)2495.72 (n/a)2499.62 (n/a)2338.67 (n/a)86.02 (n/a)7346.00 (n/a)6892.20 (n/a)6873.00 (n/a)6673.00 (n/a)245.57 (n/a)

gemm_2048x2048x2048_64x64x32_8_cols_1_bcolmaj_0_ccolmaj_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)GFLOP/s (max)GFLOP/s (mean)GFLOP/s (median)GFLOP/s (min)GFLOP/s (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
3a053d3 — 2025-11-26 01:12:163.02 (+0.52%)2.87 (-2.05%)2.86 (-2.24%)2.70 (-6.16%)0.11 (+131.66%)2216.19 (+0.52%)2104.44 (-2.05%)2094.08 (-2.24%)1980.84 (-6.16%)80.25 (+131.67%)8673.00 (+6.56%)8175.60 (+2.21%)8204.00 (+2.29%)7752.00 (-0.51%)314.03 (+145.36%)
9e98049 — 2025-11-25 20:50:553.01 (n/a)2.93 (n/a)2.92 (n/a)2.88 (n/a)0.05 (n/a)2204.81 (n/a)2148.41 (n/a)2142.13 (n/a)2110.81 (n/a)34.64 (n/a)8139.00 (n/a)7998.60 (n/a)8020.00 (n/a)7792.00 (n/a)127.99 (n/a)

gemm_2048x2048x2048_64x64x64_1cols

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
8b6d0f4 — 2026-03-25 20:46:390.51 (+0.13%)0.51 (+0.07%)0.51 (+0.04%)0.51 (+0.11%)0.00 (+8.62%)49296.10 (-0.11%)49206.24 (-0.07%)49283.30 (-0.04%)48878.90 (-0.13%)183.15 (+8.40%)351.48 (+0.13%)349.14 (+0.07%)348.59 (+0.04%)348.50 (+0.11%)1.31 (+8.62%)
dc66da7 — 2026-03-25 19:23:450.51 (n/a)0.51 (n/a)0.51 (n/a)0.51 (n/a)0.00 (n/a)49348.10 (n/a)49241.40 (n/a)49302.10 (n/a)48941.60 (n/a)168.96 (n/a)351.03 (n/a)348.89 (n/a)348.46 (n/a)348.14 (n/a)1.20 (n/a)

gemm_2048x2048x2048_64x64x64_2_cols_0_bcolmaj_0_ccolmaj_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
130b6ea — 2025-12-05 21:15:491.04 (+0.94%)1.03 (+0.23%)1.03 (+0.09%)1.02 (+0.07%)0.00 (+197.51%)24576.00 (-0.07%)24485.82 (-0.23%)24530.40 (-0.09%)24266.40 (-0.94%)111.26 (+194.35%)
0a6c11c — 2025-12-04 00:39:101.03 (n/a)1.03 (n/a)1.03 (n/a)1.02 (n/a)0.00 (n/a)24594.30 (n/a)24542.14 (n/a)24551.90 (n/a)24495.50 (n/a)37.80 (n/a)

gemm_2048x2048x2048_64x64x64_2_cols_0_bcolmaj_0_ccolmaj_0_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
a4b6ffe — 2026-01-05 19:24:051.04 (+0.28%)1.03 (-0.15%)1.03 (+0.05%)1.02 (-0.52%)0.01 (+42.72%)24719.80 (+0.52%)24512.42 (+0.16%)24529.20 (-0.05%)24092.60 (-0.28%)252.92 (+43.15%)713.08 (+0.28%)700.92 (-0.15%)700.38 (+0.05%)694.98 (-0.52%)7.30 (+42.71%)
50ad37c — 2025-12-29 20:35:531.04 (n/a)1.03 (n/a)1.03 (n/a)1.02 (n/a)0.01 (n/a)24592.30 (n/a)24473.70 (n/a)24540.90 (n/a)24160.80 (n/a)176.68 (n/a)711.06 (n/a)702.00 (n/a)700.05 (n/a)698.59 (n/a)5.12 (n/a)

gemm_2048x2048x2048_64x64x64_2_cols_0_bcolmaj_1_ccolmaj_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
130b6ea — 2025-12-05 21:15:490.99 (+1.69%)0.97 (+0.29%)0.97 (-0.04%)0.97 (-0.15%)0.01 (+434.92%)26025.30 (+0.15%)25839.34 (-0.29%)25916.70 (+0.04%)25446.00 (-1.67%)202.05 (+425.44%)
0a6c11c — 2025-12-04 00:39:100.97 (n/a)0.97 (n/a)0.97 (n/a)0.97 (n/a)0.00 (n/a)25986.90 (n/a)25913.52 (n/a)25905.70 (n/a)25877.00 (n/a)38.45 (n/a)

gemm_2048x2048x2048_64x64x64_2_cols_0_bcolmaj_1_ccolmaj_0_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
a4b6ffe — 2026-01-05 19:24:050.97 (-0.01%)0.97 (+0.03%)0.97 (+0.04%)0.97 (+0.00%)0.00 (-6.43%)25927.90 (-0.00%)25903.20 (-0.03%)25898.60 (-0.04%)25892.80 (+0.01%)14.05 (-6.56%)663.50 (-0.01%)663.23 (+0.03%)663.35 (+0.04%)662.60 (+0.00%)0.36 (-6.43%)
50ad37c — 2025-12-29 20:35:530.97 (n/a)0.97 (n/a)0.97 (n/a)0.97 (n/a)0.00 (n/a)25928.10 (n/a)25910.14 (n/a)25908.10 (n/a)25891.20 (n/a)15.04 (n/a)663.54 (n/a)663.06 (n/a)663.11 (n/a)662.60 (n/a)0.38 (n/a)

gemm_2048x2048x2048_64x64x64_2_cols_1_bcolmaj_0_ccolmaj_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
130b6ea — 2025-12-05 21:15:490.92 (-1.02%)0.92 (-0.66%)0.92 (-0.10%)0.91 (-0.78%)0.01 (-4.54%)27751.40 (+0.79%)27480.74 (+0.66%)27354.80 (+0.10%)27320.50 (+1.03%)182.69 (-2.84%)
0a6c11c — 2025-12-04 00:39:100.93 (n/a)0.92 (n/a)0.92 (n/a)0.91 (n/a)0.01 (n/a)27534.80 (n/a)27299.32 (n/a)27326.70 (n/a)27042.30 (n/a)188.04 (n/a)

gemm_2048x2048x2048_64x64x64_2_cols_1_bcolmaj_0_ccolmaj_0_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
a4b6ffe — 2026-01-05 19:24:050.92 (+0.11%)0.91 (-0.12%)0.91 (-0.21%)0.91 (-0.24%)0.00 (+30.34%)27723.50 (+0.24%)27548.94 (+0.12%)27522.90 (+0.21%)27403.60 (-0.11%)124.26 (+30.58%)626.92 (+0.11%)623.62 (-0.12%)624.20 (-0.21%)619.69 (-0.24%)2.81 (+30.34%)
50ad37c — 2025-12-29 20:35:530.92 (n/a)0.91 (n/a)0.92 (n/a)0.91 (n/a)0.00 (n/a)27657.10 (n/a)27515.58 (n/a)27464.70 (n/a)27434.60 (n/a)95.16 (n/a)626.21 (n/a)624.37 (n/a)625.52 (n/a)621.17 (n/a)2.16 (n/a)

gemm_2048x2048x2048_64x64x64_2cols_bcolmaj

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
8b6d0f4 — 2026-03-25 20:46:390.88 (+0.21%)0.87 (-0.11%)0.87 (-0.26%)0.87 (+0.03%)0.00 (+24.60%)29019.10 (-0.03%)28862.30 (+0.12%)28862.50 (+0.26%)28665.10 (-0.21%)148.73 (+24.28%)599.33 (+0.21%)595.25 (-0.11%)595.23 (-0.26%)592.02 (+0.03%)3.07 (+24.61%)
dc66da7 — 2026-03-25 19:23:450.88 (n/a)0.87 (n/a)0.87 (n/a)0.87 (n/a)0.00 (n/a)29027.70 (n/a)28829.10 (n/a)28788.30 (n/a)28724.70 (n/a)119.67 (n/a)598.09 (n/a)595.93 (n/a)596.77 (n/a)591.84 (n/a)2.46 (n/a)

gemm_2048x2048x2048_64x64x64_8_cols_0_bcolmaj_0_ccolmaj_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)GFLOP/s (max)GFLOP/s (mean)GFLOP/s (median)GFLOP/s (min)GFLOP/s (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
31ab24e — 2025-11-22 02:01:563.65 (-1.06%)3.59 (+0.83%)3.63 (-0.26%)3.45 (+2.66%)0.07 (-44.71%)2678.08 (-1.06%)2632.13 (+0.83%)2658.60 (-0.26%)2528.31 (+2.66%)53.86 (-44.71%)6795.00 (-2.59%)6529.80 (-0.92%)6462.00 (+0.26%)6415.00 (+1.07%)137.14 (-45.27%)
d9a64c5 — 2025-11-21 20:49:353.69 (n/a)3.56 (n/a)3.64 (n/a)3.36 (n/a)0.13 (n/a)2706.77 (n/a)2610.50 (n/a)2665.61 (n/a)2462.71 (n/a)97.41 (n/a)6976.00 (n/a)6590.40 (n/a)6445.00 (n/a)6347.00 (n/a)250.58 (n/a)

gemm_2048x2048x2048_64x64x64_8_cols_0_bcolmaj_0_ccolmaj_0_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
d48746f — 2026-01-12 22:17:103.98 (+0.20%)3.84 (-2.95%)3.76 (-5.27%)3.72 (-4.97%)0.13 (+473.97%)6760.00 (+5.23%)6559.50 (+3.13%)6697.10 (+5.56%)6321.00 (-0.20%)218.68 (+499.40%)2717.91 (+0.20%)2621.44 (-2.95%)2565.27 (-5.27%)2541.40 (-4.97%)88.37 (+473.97%)
331dcca — 2026-01-08 18:34:103.97 (n/a)3.96 (n/a)3.97 (n/a)3.92 (n/a)0.02 (n/a)6423.80 (n/a)6360.32 (n/a)6344.30 (n/a)6333.90 (n/a)36.48 (n/a)2712.39 (n/a)2701.17 (n/a)2707.91 (n/a)2674.40 (n/a)15.40 (n/a)

gemm_2048x2048x2048_64x64x64_8_cols_0_bcolmaj_1_ccolmaj_0_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
d48746f — 2026-01-12 22:17:103.75 (+0.28%)3.70 (+1.25%)3.71 (-0.21%)3.65 (+4.77%)0.04 (-66.90%)6901.80 (-4.56%)6795.46 (-1.30%)6790.20 (+0.21%)6715.90 (-0.28%)67.05 (-68.40%)2558.08 (+0.28%)2528.33 (+1.25%)2530.10 (-0.21%)2489.18 (+4.77%)24.81 (-66.90%)
331dcca — 2026-01-08 18:34:103.74 (n/a)3.66 (n/a)3.71 (n/a)3.48 (n/a)0.11 (n/a)7231.20 (n/a)6885.18 (n/a)6775.70 (n/a)6734.90 (n/a)212.21 (n/a)2550.87 (n/a)2497.04 (n/a)2535.50 (n/a)2375.78 (n/a)74.96 (n/a)

gemm_2048x2048x2048_64x64x64_8_cols_1_bcolmaj_0_ccolmaj_0_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
d48746f — 2026-01-12 22:17:103.62 (-0.77%)3.54 (+0.14%)3.60 (-0.04%)3.44 (+2.17%)0.09 (-25.77%)7307.40 (-2.13%)7103.46 (-0.18%)6982.50 (+0.04%)6948.20 (+0.77%)182.95 (-26.55%)2472.55 (-0.77%)2419.80 (+0.14%)2460.41 (-0.04%)2351.04 (+2.17%)61.76 (-25.77%)
331dcca — 2026-01-08 18:34:103.65 (n/a)3.54 (n/a)3.61 (n/a)3.37 (n/a)0.12 (n/a)7466.20 (n/a)7116.40 (n/a)6979.60 (n/a)6895.10 (n/a)249.09 (n/a)2491.61 (n/a)2416.45 (n/a)2461.45 (n/a)2301.00 (n/a)83.20 (n/a)

gemm_2048x2048x2048_64x64x64_8cols_bcolmaj_ccolmaj

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
8b6d0f4 — 2026-03-25 20:46:393.47 (-0.81%)3.41 (+0.82%)3.42 (-0.23%)3.35 (+4.63%)0.05 (-63.03%)7522.10 (-4.43%)7390.72 (-0.92%)7367.60 (+0.23%)7255.80 (+0.82%)107.14 (-64.21%)2367.74 (-0.81%)2324.90 (+0.82%)2331.81 (-0.23%)2283.92 (+4.63%)33.69 (-63.03%)
dc66da7 — 2026-03-25 19:23:453.50 (n/a)3.38 (n/a)3.42 (n/a)3.20 (n/a)0.13 (n/a)7870.70 (n/a)7459.22 (n/a)7350.70 (n/a)7196.70 (n/a)299.40 (n/a)2387.20 (n/a)2306.10 (n/a)2337.18 (n/a)2182.77 (n/a)91.13 (n/a)

gemm_384x1536x1792_32x48x64_4cols_bcolmaj

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
8b6d0f4 — 2026-03-25 20:46:394.16 (+12.80%)3.50 (+5.22%)3.50 (+2.51%)2.86 (-0.71%)0.63 (+68.68%)2814.80 (+0.72%)2362.04 (-3.41%)2304.60 (-2.45%)1936.00 (-11.34%)430.86 (+52.19%)1091.93 (+12.80%)919.16 (+5.22%)917.28 (+2.51%)751.01 (-0.71%)166.31 (+68.68%)
dc66da7 — 2026-03-25 19:23:453.69 (n/a)3.33 (n/a)3.41 (n/a)2.88 (n/a)0.38 (n/a)2794.70 (n/a)2445.42 (n/a)2362.50 (n/a)2183.70 (n/a)283.10 (n/a)968.06 (n/a)873.57 (n/a)894.79 (n/a)756.39 (n/a)98.60 (n/a)

gemm_896x1792x640_32x64x80_8cols_ccolmaj

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
8b6d0f4 — 2026-03-25 20:46:396.32 (+6.48%)4.74 (+3.95%)4.60 (+2.04%)3.49 (+3.11%)1.02 (+11.51%)1906.60 (-3.02%)1453.08 (-3.45%)1446.20 (-2.01%)1052.10 (-6.09%)304.01 (+0.60%)1953.36 (+6.48%)1465.54 (+3.95%)1421.06 (+2.04%)1077.93 (+3.11%)313.81 (+11.51%)
dc66da7 — 2026-03-25 19:23:455.94 (n/a)4.56 (n/a)4.51 (n/a)3.38 (n/a)0.91 (n/a)1966.00 (n/a)1504.94 (n/a)1475.80 (n/a)1120.30 (n/a)302.18 (n/a)1834.47 (n/a)1409.81 (n/a)1392.60 (n/a)1045.38 (n/a)281.42 (n/a)

input_length_2048-num_aie_columns_1-num_channels_1-tile_size_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
9580ba2 — 2026-03-25 20:55:190.04 (n/a)0.03 (n/a)0.03 (n/a)0.01 (n/a)0.01 (n/a)777.80 (n/a)302.94 (n/a)249.00 (n/a)223.00 (n/a)145.19 (n/a)

input_length_2048-num_aie_columns_1-num_channels_1-tile_size_2048-group_size_32

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
9580ba2 — 2026-03-25 20:55:190.02 (n/a)0.01 (n/a)0.02 (n/a)0.01 (n/a)0.01 (n/a)654.70 (n/a)421.34 (n/a)305.00 (n/a)254.80 (n/a)189.71 (n/a)

input_length_2048-num_aie_columns_1-num_channels_1-tile_size_2048-weighted_False

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
9580ba2 — 2026-03-25 20:55:190.03 (n/a)0.03 (n/a)0.03 (n/a)0.01 (n/a)0.01 (n/a)622.90 (n/a)329.58 (n/a)255.90 (n/a)247.70 (n/a)164.25 (n/a)

input_length_2048-num_aie_columns_1-num_channels_1-tile_size_2048-weighted_True

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
9580ba2 — 2026-03-25 20:55:190.05 (n/a)0.05 (n/a)0.05 (n/a)0.03 (n/a)0.01 (n/a)386.00 (n/a)276.76 (n/a)250.40 (n/a)231.30 (n/a)62.49 (n/a)

input_length_2048-num_aie_columns_1-num_channels_2-tile_size_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
9580ba2 — 2026-03-25 20:55:190.04 (n/a)0.03 (n/a)0.03 (n/a)0.01 (n/a)0.01 (n/a)834.40 (n/a)326.47 (n/a)251.55 (n/a)231.20 (n/a)182.88 (n/a)

input_length_2048-num_aie_columns_1-num_channels_2-tile_size_1024-group_size_32

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
9580ba2 — 2026-03-25 20:55:190.02 (n/a)0.02 (n/a)0.02 (n/a)0.01 (n/a)0.01 (n/a)847.10 (n/a)394.26 (n/a)278.10 (n/a)224.40 (n/a)257.99 (n/a)

input_length_2048-num_aie_columns_1-num_channels_2-tile_size_1024-weighted_False

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
9580ba2 — 2026-03-25 20:55:190.03 (n/a)0.03 (n/a)0.03 (n/a)0.02 (n/a)0.01 (n/a)503.70 (n/a)308.88 (n/a)250.40 (n/a)246.20 (n/a)110.85 (n/a)

input_length_2048-num_aie_columns_1-num_channels_2-tile_size_2048-scalar_factor_3.0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
9580ba2 — 2026-03-25 20:55:190.05 (n/a)0.04 (n/a)0.05 (n/a)0.02 (n/a)0.02 (n/a)709.50 (n/a)345.00 (n/a)235.40 (n/a)228.90 (n/a)207.40 (n/a)

input_length_2048-num_aie_columns_1-tile_size_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
9580ba2 — 2026-03-25 20:55:190.05 (n/a)0.04 (n/a)0.04 (n/a)0.02 (n/a)0.01 (n/a)713.80 (n/a)319.69 (n/a)249.40 (n/a)239.10 (n/a)143.60 (n/a)

input_length_2048-num_aie_columns_2-num_channels_1-tile_size_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
9580ba2 — 2026-03-25 20:55:190.04 (n/a)0.03 (n/a)0.03 (n/a)0.01 (n/a)0.01 (n/a)688.90 (n/a)283.88 (n/a)249.50 (n/a)225.10 (n/a)96.78 (n/a)

input_length_2048-num_aie_columns_2-num_channels_1-tile_size_1024-group_size_32

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
9580ba2 — 2026-03-25 20:55:190.02 (n/a)0.02 (n/a)0.02 (n/a)0.01 (n/a)0.00 (n/a)574.20 (n/a)343.28 (n/a)297.30 (n/a)243.40 (n/a)131.44 (n/a)

input_length_2048-num_aie_columns_2-num_channels_1-tile_size_1024-weighted_False

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
9580ba2 — 2026-03-25 20:55:190.03 (n/a)0.03 (n/a)0.03 (n/a)0.02 (n/a)0.00 (n/a)345.10 (n/a)267.40 (n/a)250.80 (n/a)243.30 (n/a)43.69 (n/a)

input_length_2048-num_aie_columns_2-num_channels_1-tile_size_1024-weighted_True

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
9580ba2 — 2026-03-25 20:55:190.04 (n/a)0.04 (n/a)0.04 (n/a)0.03 (n/a)0.01 (n/a)390.80 (n/a)280.64 (n/a)250.00 (n/a)246.80 (n/a)62.16 (n/a)

input_length_2048-num_aie_columns_2-num_channels_2-tile_size_1024-scalar_factor_3.0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
9580ba2 — 2026-03-25 20:55:190.05 (n/a)0.04 (n/a)0.05 (n/a)0.04 (n/a)0.01 (n/a)323.20 (n/a)279.48 (n/a)259.30 (n/a)246.70 (n/a)34.56 (n/a)

input_length_2048-num_aie_columns_2-num_channels_2-tile_size_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
9580ba2 — 2026-03-25 20:55:190.04 (n/a)0.03 (n/a)0.03 (n/a)0.01 (n/a)0.01 (n/a)642.10 (n/a)288.26 (n/a)236.60 (n/a)226.20 (n/a)127.67 (n/a)

input_length_2048-num_aie_columns_2-num_channels_2-tile_size_512-group_size_32

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
9580ba2 — 2026-03-25 20:55:190.02 (n/a)0.02 (n/a)0.02 (n/a)0.01 (n/a)0.01 (n/a)631.70 (n/a)396.70 (n/a)273.00 (n/a)224.90 (n/a)193.30 (n/a)

input_length_2048-num_aie_columns_2-num_channels_2-tile_size_512-weighted_False

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
9580ba2 — 2026-03-25 20:55:190.04 (n/a)0.03 (n/a)0.03 (n/a)0.01 (n/a)0.01 (n/a)1005.50 (n/a)397.28 (n/a)242.80 (n/a)227.10 (n/a)340.61 (n/a)

input_length_2048-num_aie_columns_2-tile_size_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
9580ba2 — 2026-03-25 20:55:190.05 (n/a)0.04 (n/a)0.03 (n/a)0.02 (n/a)0.01 (n/a)753.90 (n/a)347.67 (n/a)250.90 (n/a)237.20 (n/a)176.26 (n/a)

input_length_2048-num_aie_columns_4-num_channels_1-tile_size_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
9580ba2 — 2026-03-25 20:55:190.04 (n/a)0.03 (n/a)0.03 (n/a)0.01 (n/a)0.01 (n/a)674.80 (n/a)282.28 (n/a)244.80 (n/a)221.50 (n/a)99.30 (n/a)

input_length_2048-num_aie_columns_4-num_channels_1-tile_size_512-group_size_32

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
9580ba2 — 2026-03-25 20:55:190.02 (n/a)0.02 (n/a)0.02 (n/a)0.01 (n/a)0.00 (n/a)589.20 (n/a)345.66 (n/a)299.80 (n/a)234.20 (n/a)139.32 (n/a)

input_length_2048-num_aie_columns_4-num_channels_1-tile_size_512-weighted_False

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
9580ba2 — 2026-03-25 20:55:190.04 (n/a)0.03 (n/a)0.03 (n/a)0.01 (n/a)0.01 (n/a)606.50 (n/a)318.02 (n/a)249.40 (n/a)230.70 (n/a)161.57 (n/a)

input_length_2048-num_aie_columns_4-num_channels_1-tile_size_512-weighted_True

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
9580ba2 — 2026-03-25 20:55:190.04 (n/a)0.03 (n/a)0.04 (n/a)0.01 (n/a)0.01 (n/a)782.80 (n/a)389.42 (n/a)231.20 (n/a)229.50 (n/a)243.85 (n/a)

input_length_2048-num_aie_columns_4-num_channels_2-tile_size_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
9580ba2 — 2026-03-25 20:55:190.03 (n/a)0.03 (n/a)0.03 (n/a)0.02 (n/a)0.01 (n/a)432.40 (n/a)283.75 (n/a)259.90 (n/a)236.70 (n/a)62.99 (n/a)

input_length_2048-num_aie_columns_4-num_channels_2-tile_size_256-group_size_32

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
9580ba2 — 2026-03-25 20:55:190.02 (n/a)0.02 (n/a)0.02 (n/a)0.01 (n/a)0.01 (n/a)775.70 (n/a)382.02 (n/a)295.50 (n/a)251.00 (n/a)220.91 (n/a)

input_length_2048-num_aie_columns_4-num_channels_2-tile_size_256-weighted_False

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
9580ba2 — 2026-03-25 20:55:190.04 (n/a)0.03 (n/a)0.04 (n/a)0.02 (n/a)0.01 (n/a)383.70 (n/a)262.68 (n/a)232.90 (n/a)231.00 (n/a)67.66 (n/a)

input_length_2048-num_aie_columns_4-num_channels_2-tile_size_512-scalar_factor_3.0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
9580ba2 — 2026-03-25 20:55:190.05 (n/a)0.05 (n/a)0.05 (n/a)0.03 (n/a)0.01 (n/a)359.20 (n/a)272.00 (n/a)254.30 (n/a)245.90 (n/a)48.92 (n/a)

input_length_2048-num_aie_columns_4-tile_size_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
9580ba2 — 2026-03-25 20:55:190.05 (n/a)0.04 (n/a)0.05 (n/a)0.01 (n/a)0.01 (n/a)740.00 (n/a)326.79 (n/a)247.30 (n/a)223.80 (n/a)167.90 (n/a)

input_length_2048-num_aie_columns_8-num_channels_1-tile_size_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
9580ba2 — 2026-03-25 20:55:190.04 (n/a)0.03 (n/a)0.03 (n/a)0.01 (n/a)0.01 (n/a)772.10 (n/a)336.71 (n/a)251.40 (n/a)231.20 (n/a)175.02 (n/a)

input_length_2048-num_aie_columns_8-num_channels_1-tile_size_256-group_size_32

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
9580ba2 — 2026-03-25 20:55:190.02 (n/a)0.02 (n/a)0.02 (n/a)0.02 (n/a)0.00 (n/a)313.40 (n/a)306.22 (n/a)305.60 (n/a)294.20 (n/a)7.76 (n/a)

input_length_2048-num_aie_columns_8-num_channels_1-tile_size_256-weighted_False

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
9580ba2 — 2026-03-25 20:55:190.04 (n/a)0.03 (n/a)0.03 (n/a)0.01 (n/a)0.01 (n/a)734.60 (n/a)333.84 (n/a)234.30 (n/a)230.10 (n/a)224.04 (n/a)

input_length_2048-num_aie_columns_8-num_channels_1-tile_size_256-weighted_True

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
9580ba2 — 2026-03-25 20:55:190.03 (n/a)0.03 (n/a)0.03 (n/a)0.01 (n/a)0.01 (n/a)727.50 (n/a)355.52 (n/a)259.60 (n/a)255.20 (n/a)208.10 (n/a)

input_length_2048-num_aie_columns_8-num_channels_2-tile_size_128

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
9580ba2 — 2026-03-25 20:55:190.03 (n/a)0.03 (n/a)0.03 (n/a)0.01 (n/a)0.01 (n/a)708.80 (n/a)332.74 (n/a)279.80 (n/a)264.00 (n/a)136.83 (n/a)

input_length_2048-num_aie_columns_8-num_channels_2-tile_size_128-group_size_32

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
9580ba2 — 2026-03-25 20:55:190.02 (n/a)0.02 (n/a)0.02 (n/a)0.01 (n/a)0.01 (n/a)872.70 (n/a)412.64 (n/a)311.40 (n/a)273.30 (n/a)258.01 (n/a)

input_length_2048-num_aie_columns_8-num_channels_2-tile_size_128-weighted_False

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
9580ba2 — 2026-03-25 20:55:190.03 (n/a)0.02 (n/a)0.03 (n/a)0.01 (n/a)0.01 (n/a)1010.80 (n/a)427.70 (n/a)280.60 (n/a)270.90 (n/a)326.21 (n/a)

input_length_2048-num_aie_columns_8-num_channels_2-tile_size_256-scalar_factor_3.0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
9580ba2 — 2026-03-25 20:55:190.05 (n/a)0.04 (n/a)0.04 (n/a)0.04 (n/a)0.01 (n/a)341.80 (n/a)299.32 (n/a)298.40 (n/a)260.50 (n/a)36.69 (n/a)

input_length_2048-num_aie_columns_8-tile_size_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
9580ba2 — 2026-03-25 20:55:190.05 (n/a)0.04 (n/a)0.04 (n/a)0.01 (n/a)0.01 (n/a)698.00 (n/a)359.08 (n/a)263.30 (n/a)232.10 (n/a)183.99 (n/a)

input_length_2048-num_cores_1-num_channels_1-bypass_False-tile_size_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
9580ba2 — 2026-03-25 20:55:190.04 (n/a)0.03 (n/a)0.04 (n/a)0.02 (n/a)0.01 (n/a)409.20 (n/a)266.30 (n/a)225.00 (n/a)218.60 (n/a)81.33 (n/a)

input_length_2048-num_cores_16-num_channels_2-bypass_False-tile_size_128

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
9580ba2 — 2026-03-25 20:55:190.03 (n/a)0.03 (n/a)0.03 (n/a)0.01 (n/a)0.01 (n/a)972.90 (n/a)413.24 (n/a)275.20 (n/a)268.40 (n/a)312.89 (n/a)

input_length_2048-num_cores_2-num_channels_1-bypass_False-tile_size_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
9580ba2 — 2026-03-25 20:55:190.04 (n/a)0.03 (n/a)0.04 (n/a)0.02 (n/a)0.01 (n/a)350.10 (n/a)252.14 (n/a)229.80 (n/a)224.40 (n/a)54.82 (n/a)

input_length_2048-num_cores_2-num_channels_2-bypass_False-tile_size_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
9580ba2 — 2026-03-25 20:55:190.03 (n/a)0.03 (n/a)0.03 (n/a)0.02 (n/a)0.01 (n/a)398.00 (n/a)276.88 (n/a)245.90 (n/a)243.80 (n/a)67.78 (n/a)

input_length_2048-num_cores_4-num_channels_1-bypass_False-tile_size_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
9580ba2 — 2026-03-25 20:55:190.03 (n/a)0.03 (n/a)0.03 (n/a)0.02 (n/a)0.01 (n/a)494.50 (n/a)293.76 (n/a)244.90 (n/a)239.40 (n/a)112.30 (n/a)

input_length_2048-num_cores_4-num_channels_2-bypass_False-tile_size_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
9580ba2 — 2026-03-25 20:55:190.03 (n/a)0.03 (n/a)0.03 (n/a)0.02 (n/a)0.01 (n/a)374.50 (n/a)272.78 (n/a)247.70 (n/a)242.20 (n/a)57.03 (n/a)

input_length_2048-num_cores_8-num_channels_1-bypass_False-tile_size_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
9580ba2 — 2026-03-25 20:55:190.03 (n/a)0.03 (n/a)0.03 (n/a)0.02 (n/a)0.00 (n/a)334.00 (n/a)268.42 (n/a)251.10 (n/a)250.60 (n/a)36.71 (n/a)

input_length_2048-num_cores_8-num_channels_2-bypass_False-tile_size_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
9580ba2 — 2026-03-25 20:55:190.03 (n/a)0.03 (n/a)0.03 (n/a)0.01 (n/a)0.01 (n/a)563.60 (n/a)316.62 (n/a)254.20 (n/a)252.40 (n/a)138.09 (n/a)

input_length_32768-num_aie_columns_2-num_channels_2-tile_size_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
9580ba2 — 2026-03-25 20:55:190.51 (n/a)0.44 (n/a)0.49 (n/a)0.25 (n/a)0.11 (n/a)531.50 (n/a)321.74 (n/a)270.20 (n/a)259.30 (n/a)117.46 (n/a)

input_length_32768-num_aie_columns_2-num_channels_2-tile_size_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
9580ba2 — 2026-03-25 20:55:190.49 (n/a)0.36 (n/a)0.37 (n/a)0.20 (n/a)0.11 (n/a)668.50 (n/a)404.48 (n/a)349.60 (n/a)265.90 (n/a)155.68 (n/a)

input_length_32768-num_aie_columns_4-num_channels_4-tile_size_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
9580ba2 — 2026-03-25 20:55:190.35 (n/a)0.21 (n/a)0.18 (n/a)0.17 (n/a)0.08 (n/a)781.80 (n/a)656.18 (n/a)723.80 (n/a)376.10 (n/a)163.09 (n/a)

layer_norm_1_cols_1_channels_2048_tile_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.07 (+35.61%)0.05 (+6.58%)0.04 (-0.65%)0.03 (-1.11%)0.01 (+98.33%)236.50 (+1.11%)182.90 (-3.12%)183.30 (+0.60%)118.50 (-26.26%)42.30 (+41.46%)
dc66da7 — 2026-03-25 19:23:450.05 (n/a)0.04 (n/a)0.04 (n/a)0.04 (n/a)0.01 (n/a)233.90 (n/a)188.80 (n/a)182.20 (n/a)160.70 (n/a)29.90 (n/a)

layer_norm_1_cols_2_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.07 (+9.47%)0.05 (+5.79%)0.05 (+5.48%)0.04 (-7.03%)0.01 (+51.12%)206.90 (+7.54%)163.06 (-2.60%)165.20 (-5.22%)113.20 (-8.71%)40.37 (+56.59%)
dc66da7 — 2026-03-25 19:23:450.07 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.01 (n/a)192.40 (n/a)167.42 (n/a)174.30 (n/a)124.00 (n/a)25.78 (n/a)

layer_norm_2_cols_1_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.06 (-0.97%)0.05 (+14.59%)0.05 (+47.43%)0.03 (+8.92%)0.01 (-0.31%)235.10 (-8.20%)183.06 (-13.03%)157.80 (-32.19%)148.10 (+1.02%)43.12 (-7.94%)
dc66da7 — 2026-03-25 19:23:450.06 (n/a)0.04 (n/a)0.04 (n/a)0.03 (n/a)0.01 (n/a)256.10 (n/a)210.48 (n/a)232.70 (n/a)146.60 (n/a)46.84 (n/a)

layer_norm_2_cols_2_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.07 (+14.65%)0.05 (+4.61%)0.05 (-7.99%)0.04 (+11.94%)0.01 (+28.58%)201.70 (-10.67%)168.58 (-3.65%)176.00 (+8.71%)122.10 (-12.79%)33.82 (+0.06%)
dc66da7 — 2026-03-25 19:23:450.06 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.01 (n/a)225.80 (n/a)174.96 (n/a)161.90 (n/a)140.00 (n/a)33.80 (n/a)

layer_norm_4_cols_1_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.06 (-31.05%)0.05 (-3.73%)0.05 (+9.03%)0.04 (+10.49%)0.01 (-63.42%)197.60 (-9.52%)167.50 (-4.83%)166.40 (-8.27%)129.00 (+45.11%)25.47 (-51.02%)
dc66da7 — 2026-03-25 19:23:450.09 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.02 (n/a)218.40 (n/a)176.00 (n/a)181.40 (n/a)88.90 (n/a)52.00 (n/a)

layer_norm_4_cols_2_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.05 (-1.52%)0.04 (-11.20%)0.04 (-25.45%)0.03 (-0.69%)0.01 (-9.35%)243.40 (+0.66%)202.20 (+12.10%)212.70 (+34.20%)156.90 (+1.55%)34.18 (-7.92%)
dc66da7 — 2026-03-25 19:23:450.05 (n/a)0.05 (n/a)0.05 (n/a)0.03 (n/a)0.01 (n/a)241.80 (n/a)180.38 (n/a)158.50 (n/a)154.50 (n/a)37.12 (n/a)

layer_norm_8_cols_1_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.05 (-6.07%)0.04 (+15.60%)0.04 (+21.55%)0.04 (+82.28%)0.01 (-55.28%)222.00 (-45.13%)190.18 (-21.85%)193.60 (-17.72%)153.10 (+6.47%)24.64 (-75.05%)
dc66da7 — 2026-03-25 19:23:450.06 (n/a)0.04 (n/a)0.03 (n/a)0.02 (n/a)0.01 (n/a)404.60 (n/a)243.34 (n/a)235.30 (n/a)143.80 (n/a)98.74 (n/a)

layer_norm_8_cols_2_channels_2048_tile_128

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.05 (+37.82%)0.04 (+27.22%)0.04 (+9.46%)0.03 (+34.60%)0.01 (+56.79%)251.50 (-25.72%)212.60 (-21.06%)226.70 (-8.66%)175.20 (-27.45%)33.04 (-18.34%)
dc66da7 — 2026-03-25 19:23:450.03 (n/a)0.03 (n/a)0.03 (n/a)0.02 (n/a)0.00 (n/a)338.60 (n/a)269.32 (n/a)248.20 (n/a)241.50 (n/a)40.46 (n/a)

matrix_vector_mul_128x128_32_1col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)
130b6ea — 2025-12-05 21:15:490.37 (+17.13%)0.30 (+24.87%)0.30 (+38.03%)0.23 (+12.47%)0.06 (+35.23%)
0a6c11c — 2025-12-04 00:39:100.32 (n/a)0.24 (n/a)0.22 (n/a)0.20 (n/a)0.04 (n/a)

matrix_vector_mul_128x128_32_1col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
331dcca — 2026-01-08 18:34:100.26 (-1.81%)0.24 (+4.06%)0.24 (+0.52%)0.21 (+16.97%)0.02 (-48.16%)0.26 (-1.81%)0.24 (+4.06%)0.24 (+0.52%)0.20 (+16.97%)0.02 (-48.16%)
a4b6ffe — 2026-01-05 19:24:050.26 (n/a)0.23 (n/a)0.24 (n/a)0.18 (n/a)0.04 (n/a)0.26 (n/a)0.23 (n/a)0.24 (n/a)0.17 (n/a)0.04 (n/a)

matrix_vector_mul_128x128_32tsi_128tso_1col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
8b6d0f4 — 2026-03-25 20:46:390.26 (+32.10%)0.19 (+11.55%)0.18 (+9.49%)0.13 (-2.22%)0.05 (+105.83%)0.25 (+32.10%)0.19 (+11.55%)0.18 (+9.49%)0.13 (-2.22%)0.05 (+105.83%)
dc66da7 — 2026-03-25 19:23:450.19 (n/a)0.17 (n/a)0.17 (n/a)0.14 (n/a)0.02 (n/a)0.19 (n/a)0.17 (n/a)0.17 (n/a)0.14 (n/a)0.02 (n/a)

matrix_vector_mul_128x128_32tsi_128tso_1col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
897d04e — 2026-03-06 22:12:250.37 (+5.44%)0.31 (-0.05%)0.34 (-1.99%)0.22 (-7.87%)0.06 (+15.13%)0.36 (+5.44%)0.31 (-0.05%)0.34 (-1.99%)0.21 (-7.87%)0.06 (+15.13%)
84d3478 — 2026-02-17 23:25:130.35 (n/a)0.31 (n/a)0.35 (n/a)0.23 (n/a)0.05 (n/a)0.35 (n/a)0.31 (n/a)0.34 (n/a)0.23 (n/a)0.05 (n/a)

matrix_vector_mul_2048x8192_1_1col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)
130b6ea — 2025-12-05 21:15:4913.57 (+3.47%)13.15 (+1.09%)13.10 (+0.16%)12.52 (-1.58%)0.37 (+153.19%)
0a6c11c — 2025-12-04 00:39:1013.11 (n/a)13.01 (n/a)13.07 (n/a)12.72 (n/a)0.15 (n/a)

matrix_vector_mul_2048x8192_1_1col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
331dcca — 2026-01-08 18:34:1013.43 (-0.07%)13.30 (+4.62%)13.27 (+1.29%)13.16 (+16.70%)0.12 (-86.69%)13.42 (-0.07%)13.29 (+4.62%)13.26 (+1.29%)13.16 (+16.70%)0.12 (-86.69%)
a4b6ffe — 2026-01-05 19:24:0513.44 (n/a)12.71 (n/a)13.10 (n/a)11.28 (n/a)0.87 (n/a)13.43 (n/a)12.70 (n/a)13.09 (n/a)11.27 (n/a)0.87 (n/a)

matrix_vector_mul_2048x8192_1_2col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)
130b6ea — 2025-12-05 21:15:4925.06 (+2.47%)23.71 (+1.81%)24.13 (+0.50%)20.90 (+5.41%)1.48 (-15.60%)
0a6c11c — 2025-12-04 00:39:1024.46 (n/a)23.29 (n/a)24.01 (n/a)19.83 (n/a)1.75 (n/a)

matrix_vector_mul_2048x8192_1_2col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
331dcca — 2026-01-08 18:34:1024.46 (-0.60%)21.07 (-8.03%)19.75 (-17.83%)18.08 (+1.18%)3.03 (+7.07%)24.44 (-0.60%)21.06 (-8.03%)19.74 (-17.83%)18.07 (+1.18%)3.03 (+7.07%)
a4b6ffe — 2026-01-05 19:24:0524.60 (n/a)22.91 (n/a)24.04 (n/a)17.87 (n/a)2.83 (n/a)24.59 (n/a)22.90 (n/a)24.02 (n/a)17.86 (n/a)2.83 (n/a)

matrix_vector_mul_2048x8192_1_4col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)
130b6ea — 2025-12-05 21:15:4942.05 (+1.54%)40.50 (+12.60%)40.49 (+2.36%)38.98 (+97.53%)0.97 (-88.09%)
0a6c11c — 2025-12-04 00:39:1041.41 (n/a)35.97 (n/a)39.55 (n/a)19.73 (n/a)8.15 (n/a)

matrix_vector_mul_2048x8192_1_4col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
331dcca — 2026-01-08 18:34:1042.58 (+1.51%)41.37 (+14.29%)41.25 (+4.89%)40.33 (+91.42%)0.92 (-89.18%)42.56 (+1.51%)41.34 (+14.29%)41.23 (+4.89%)40.30 (+91.42%)0.92 (-89.18%)
a4b6ffe — 2026-01-05 19:24:0541.95 (n/a)36.20 (n/a)39.33 (n/a)21.07 (n/a)8.54 (n/a)41.92 (n/a)36.17 (n/a)39.31 (n/a)21.05 (n/a)8.53 (n/a)

matrix_vector_mul_2048x8192_1_8col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)
130b6ea — 2025-12-05 21:15:4946.62 (+3.11%)44.82 (+0.17%)44.78 (+0.17%)42.60 (-3.80%)1.60 (+367.72%)
0a6c11c — 2025-12-04 00:39:1045.21 (n/a)44.75 (n/a)44.70 (n/a)44.29 (n/a)0.34 (n/a)

matrix_vector_mul_2048x8192_1_8col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
331dcca — 2026-01-08 18:34:1046.04 (+4.14%)45.06 (+3.47%)44.55 (+2.76%)44.30 (+2.84%)0.87 (+66.58%)46.01 (+4.14%)45.03 (+3.47%)44.52 (+2.76%)44.27 (+2.84%)0.87 (+66.58%)
a4b6ffe — 2026-01-05 19:24:0544.21 (n/a)43.55 (n/a)43.35 (n/a)43.07 (n/a)0.52 (n/a)44.18 (n/a)43.52 (n/a)43.32 (n/a)43.05 (n/a)0.52 (n/a)

matrix_vector_mul_2048x8192_1tsi_1024tso_2col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
8b6d0f4 — 2026-03-25 20:46:3925.44 (+1.95%)24.42 (-0.28%)24.79 (+0.16%)23.14 (-2.33%)1.05 (+106.47%)25.42 (+1.95%)24.41 (-0.28%)24.77 (+0.16%)23.12 (-2.33%)1.05 (+106.47%)
dc66da7 — 2026-03-25 19:23:4524.95 (n/a)24.49 (n/a)24.75 (n/a)23.69 (n/a)0.51 (n/a)24.94 (n/a)24.48 (n/a)24.73 (n/a)23.67 (n/a)0.51 (n/a)

matrix_vector_mul_2048x8192_1tsi_1024tso_2col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
897d04e — 2026-03-06 22:12:2524.47 (-3.32%)24.16 (-3.13%)24.20 (-3.09%)23.84 (-3.04%)0.25 (-14.36%)24.46 (-3.32%)24.15 (-3.13%)24.18 (-3.09%)23.83 (-3.04%)0.25 (-14.36%)
84d3478 — 2026-02-17 23:25:1325.31 (n/a)24.94 (n/a)24.97 (n/a)24.59 (n/a)0.29 (n/a)25.30 (n/a)24.93 (n/a)24.95 (n/a)24.58 (n/a)0.29 (n/a)

matrix_vector_mul_2048x8192_1tsi_2048tso_1col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
8b6d0f4 — 2026-03-25 20:46:3913.48 (-1.01%)13.15 (+1.31%)13.11 (-0.87%)12.95 (+6.77%)0.20 (-66.05%)13.47 (-1.01%)13.14 (+1.31%)13.11 (-0.87%)12.95 (+6.77%)0.20 (-66.05%)
dc66da7 — 2026-03-25 19:23:4513.62 (n/a)12.98 (n/a)13.23 (n/a)12.13 (n/a)0.58 (n/a)13.61 (n/a)12.97 (n/a)13.22 (n/a)12.13 (n/a)0.58 (n/a)

matrix_vector_mul_2048x8192_1tsi_2048tso_1col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
897d04e — 2026-03-06 22:12:2513.25 (-3.04%)13.02 (-2.92%)13.10 (-2.64%)12.56 (-4.14%)0.27 (+32.55%)13.24 (-3.04%)13.01 (-2.92%)13.09 (-2.64%)12.55 (-4.14%)0.27 (+32.55%)
84d3478 — 2026-02-17 23:25:1313.66 (n/a)13.41 (n/a)13.45 (n/a)13.10 (n/a)0.20 (n/a)13.65 (n/a)13.40 (n/a)13.44 (n/a)13.09 (n/a)0.20 (n/a)

matrix_vector_mul_2048x8192_1tsi_256tso_8col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
8b6d0f4 — 2026-03-25 20:46:3942.06 (-4.87%)40.85 (-4.53%)40.57 (-4.52%)39.78 (-4.49%)0.87 (-11.49%)42.04 (-4.87%)40.83 (-4.53%)40.55 (-4.52%)39.76 (-4.49%)0.87 (-11.49%)
dc66da7 — 2026-03-25 19:23:4544.21 (n/a)42.79 (n/a)42.49 (n/a)41.65 (n/a)0.99 (n/a)44.19 (n/a)42.76 (n/a)42.47 (n/a)41.63 (n/a)0.99 (n/a)

matrix_vector_mul_2048x8192_1tsi_256tso_8col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
897d04e — 2026-03-06 22:12:2547.71 (+3.51%)45.38 (+9.55%)45.22 (-1.27%)43.47 (+75.24%)1.52 (-83.67%)47.68 (+3.51%)45.35 (+9.55%)45.19 (-1.27%)43.45 (+75.24%)1.52 (-83.67%)
84d3478 — 2026-02-17 23:25:1346.09 (n/a)41.43 (n/a)45.80 (n/a)24.81 (n/a)9.31 (n/a)46.06 (n/a)41.40 (n/a)45.77 (n/a)24.79 (n/a)9.30 (n/a)

matrix_vector_mul_2048x8192_1tsi_512tso_4col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
8b6d0f4 — 2026-03-25 20:46:3940.94 (-0.67%)39.64 (+0.12%)39.96 (+1.02%)37.79 (+0.51%)1.25 (-21.17%)40.91 (-0.67%)39.61 (+0.12%)39.94 (+1.02%)37.77 (+0.51%)1.25 (-21.17%)
dc66da7 — 2026-03-25 19:23:4541.21 (n/a)39.59 (n/a)39.56 (n/a)37.60 (n/a)1.58 (n/a)41.19 (n/a)39.57 (n/a)39.53 (n/a)37.58 (n/a)1.58 (n/a)

matrix_vector_mul_2048x8192_1tsi_512tso_4col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
897d04e — 2026-03-06 22:12:2540.84 (-0.32%)39.53 (-1.25%)39.40 (-1.97%)38.87 (-0.69%)0.77 (-6.24%)40.81 (-0.32%)39.50 (-1.25%)39.37 (-1.97%)38.84 (-0.69%)0.77 (-6.24%)
84d3478 — 2026-02-17 23:25:1340.97 (n/a)40.03 (n/a)40.19 (n/a)39.14 (n/a)0.82 (n/a)40.95 (n/a)40.00 (n/a)40.17 (n/a)39.11 (n/a)0.82 (n/a)

matrix_vector_mul_8192x2048_4_1col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)
130b6ea — 2025-12-05 21:15:4913.29 (+1.63%)13.19 (+6.08%)13.19 (+1.17%)13.09 (+26.86%)0.08 (-92.94%)
0a6c11c — 2025-12-04 00:39:1013.08 (n/a)12.44 (n/a)13.03 (n/a)10.32 (n/a)1.07 (n/a)

matrix_vector_mul_8192x2048_4_1col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
331dcca — 2026-01-08 18:34:1013.68 (+3.42%)13.30 (+8.54%)13.38 (+7.25%)12.72 (+14.96%)0.35 (-66.09%)13.67 (+3.42%)13.29 (+8.54%)13.38 (+7.25%)12.71 (+14.96%)0.35 (-66.09%)
a4b6ffe — 2026-01-05 19:24:0513.23 (n/a)12.25 (n/a)12.48 (n/a)11.07 (n/a)1.04 (n/a)13.22 (n/a)12.24 (n/a)12.47 (n/a)11.06 (n/a)1.04 (n/a)

matrix_vector_mul_8192x2048_4_2col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)
130b6ea — 2025-12-05 21:15:4925.58 (+4.24%)24.23 (+6.59%)24.03 (+0.29%)23.58 (+37.29%)0.70 (-74.97%)
0a6c11c — 2025-12-04 00:39:1024.54 (n/a)22.73 (n/a)23.96 (n/a)17.18 (n/a)2.79 (n/a)

matrix_vector_mul_8192x2048_4_2col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
331dcca — 2026-01-08 18:34:1025.08 (+1.32%)24.68 (+13.42%)24.61 (+3.26%)24.47 (+48.32%)0.24 (-93.56%)25.07 (+1.32%)24.67 (+13.42%)24.59 (+3.26%)24.46 (+48.32%)0.24 (-93.56%)
a4b6ffe — 2026-01-05 19:24:0524.76 (n/a)21.76 (n/a)23.83 (n/a)16.50 (n/a)3.74 (n/a)24.74 (n/a)21.75 (n/a)23.82 (n/a)16.49 (n/a)3.73 (n/a)

matrix_vector_mul_8192x2048_4_4col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)
130b6ea — 2025-12-05 21:15:4940.94 (-2.19%)39.74 (-1.10%)39.76 (-0.60%)38.17 (-2.14%)1.04 (+4.39%)
0a6c11c — 2025-12-04 00:39:1041.86 (n/a)40.18 (n/a)40.00 (n/a)39.00 (n/a)1.00 (n/a)

matrix_vector_mul_8192x2048_4_4col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
331dcca — 2026-01-08 18:34:1040.74 (-1.09%)37.16 (-7.15%)40.25 (+1.47%)24.51 (-37.44%)7.07 (+736.13%)40.72 (-1.09%)37.13 (-7.15%)40.23 (+1.47%)24.50 (-37.44%)7.07 (+736.13%)
a4b6ffe — 2026-01-05 19:24:0541.19 (n/a)40.02 (n/a)39.67 (n/a)39.18 (n/a)0.85 (n/a)41.17 (n/a)39.99 (n/a)39.64 (n/a)39.16 (n/a)0.85 (n/a)

matrix_vector_mul_8192x2048_4_8col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)
130b6ea — 2025-12-05 21:15:4946.71 (-1.04%)44.91 (-2.34%)44.71 (-2.98%)43.80 (-1.44%)0.96 (+6.93%)
0a6c11c — 2025-12-04 00:39:1047.20 (n/a)45.98 (n/a)46.08 (n/a)44.44 (n/a)0.90 (n/a)

matrix_vector_mul_8192x2048_4_8col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
331dcca — 2026-01-08 18:34:1046.28 (+5.68%)44.89 (+14.59%)44.74 (+4.26%)44.20 (+62.80%)0.84 (-87.96%)46.25 (+5.68%)44.86 (+14.59%)44.71 (+4.26%)44.17 (+62.80%)0.84 (-87.96%)
a4b6ffe — 2026-01-05 19:24:0543.79 (n/a)39.17 (n/a)42.91 (n/a)27.15 (n/a)6.98 (n/a)43.77 (n/a)39.15 (n/a)42.88 (n/a)27.13 (n/a)6.98 (n/a)

matrix_vector_mul_8192x2048_4tsi_1024tso_1col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
8b6d0f4 — 2026-03-25 20:46:3913.35 (+0.83%)13.11 (+3.52%)13.10 (+3.94%)12.94 (+7.70%)0.15 (-69.46%)13.34 (+0.83%)13.10 (+3.52%)13.09 (+3.94%)12.94 (+7.70%)0.15 (-69.46%)
dc66da7 — 2026-03-25 19:23:4513.24 (n/a)12.66 (n/a)12.60 (n/a)12.02 (n/a)0.49 (n/a)13.23 (n/a)12.66 (n/a)12.59 (n/a)12.01 (n/a)0.49 (n/a)

matrix_vector_mul_8192x2048_4tsi_1024tso_1col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
897d04e — 2026-03-06 22:12:2513.34 (-0.79%)13.20 (-0.36%)13.18 (-1.55%)13.12 (+1.75%)0.08 (-65.93%)13.33 (-0.79%)13.19 (-0.36%)13.17 (-1.55%)13.11 (+1.75%)0.08 (-65.93%)
84d3478 — 2026-02-17 23:25:1313.45 (n/a)13.25 (n/a)13.39 (n/a)12.90 (n/a)0.25 (n/a)13.44 (n/a)13.24 (n/a)13.38 (n/a)12.89 (n/a)0.25 (n/a)

matrix_vector_mul_8192x2048_4tsi_1024tso_2col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
8b6d0f4 — 2026-03-25 20:46:3924.35 (-0.26%)24.07 (+0.89%)24.16 (+1.48%)23.64 (+2.17%)0.27 (-44.88%)24.34 (-0.26%)24.06 (+0.89%)24.15 (+1.48%)23.63 (+2.17%)0.27 (-44.88%)
dc66da7 — 2026-03-25 19:23:4524.41 (n/a)23.86 (n/a)23.81 (n/a)23.14 (n/a)0.49 (n/a)24.40 (n/a)23.84 (n/a)23.80 (n/a)23.12 (n/a)0.49 (n/a)

matrix_vector_mul_8192x2048_4tsi_1024tso_2col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
897d04e — 2026-03-06 22:12:2524.10 (-5.45%)23.07 (-1.85%)24.01 (-2.88%)19.26 (+7.47%)2.13 (-32.15%)24.08 (-5.45%)23.05 (-1.85%)23.99 (-2.88%)19.25 (+7.47%)2.13 (-32.15%)
84d3478 — 2026-02-17 23:25:1325.48 (n/a)23.50 (n/a)24.72 (n/a)17.92 (n/a)3.14 (n/a)25.47 (n/a)23.48 (n/a)24.71 (n/a)17.91 (n/a)3.14 (n/a)

matrix_vector_mul_8192x2048_4tsi_1024tso_4col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
8b6d0f4 — 2026-03-25 20:46:3940.84 (-1.16%)38.86 (-0.66%)38.49 (-1.61%)37.90 (+1.94%)1.14 (-23.22%)40.82 (-1.16%)38.84 (-0.66%)38.46 (-1.61%)37.87 (+1.94%)1.14 (-23.22%)
dc66da7 — 2026-03-25 19:23:4541.32 (n/a)39.12 (n/a)39.12 (n/a)37.18 (n/a)1.49 (n/a)41.30 (n/a)39.09 (n/a)39.09 (n/a)37.15 (n/a)1.49 (n/a)

matrix_vector_mul_8192x2048_4tsi_1024tso_4col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
897d04e — 2026-03-06 22:12:2540.80 (-0.16%)39.58 (-1.93%)39.75 (-1.71%)38.28 (-2.50%)1.10 (+67.33%)40.78 (-0.16%)39.55 (-1.93%)39.73 (-1.71%)38.26 (-2.50%)1.10 (+67.33%)
84d3478 — 2026-02-17 23:25:1340.87 (n/a)40.36 (n/a)40.45 (n/a)39.26 (n/a)0.66 (n/a)40.84 (n/a)40.33 (n/a)40.42 (n/a)39.24 (n/a)0.65 (n/a)

matrix_vector_mul_8192x2048_4tsi_1024tso_8col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
8b6d0f4 — 2026-03-25 20:46:3941.79 (-12.64%)40.98 (-5.83%)40.78 (-1.04%)40.36 (-0.60%)0.56 (-84.11%)41.76 (-12.64%)40.96 (-5.83%)40.76 (-1.04%)40.34 (-0.60%)0.56 (-84.11%)
dc66da7 — 2026-03-25 19:23:4547.83 (n/a)43.52 (n/a)41.21 (n/a)40.61 (n/a)3.53 (n/a)47.80 (n/a)43.50 (n/a)41.19 (n/a)40.58 (n/a)3.53 (n/a)

matrix_vector_mul_8192x2048_4tsi_1024tso_8col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
897d04e — 2026-03-06 22:12:2549.31 (+6.27%)46.12 (+1.67%)45.30 (-0.27%)44.62 (+1.87%)1.85 (+85.10%)49.28 (+6.27%)46.10 (+1.67%)45.28 (-0.27%)44.59 (+1.87%)1.85 (+85.10%)
84d3478 — 2026-02-17 23:25:1346.40 (n/a)45.37 (n/a)45.43 (n/a)43.80 (n/a)1.00 (n/a)46.37 (n/a)45.34 (n/a)45.40 (n/a)43.78 (n/a)1.00 (n/a)

mem_copy_16_cores_2_chans_2048_tile_128_False

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.05 (+7.16%)0.03 (-9.22%)0.03 (-18.40%)0.02 (-19.93%)0.01 (+42.24%)394.80 (+24.90%)264.56 (+14.35%)259.20 (+22.55%)179.80 (-6.65%)81.04 (+64.09%)
dc66da7 — 2026-03-25 19:23:450.04 (n/a)0.04 (n/a)0.04 (n/a)0.03 (n/a)0.01 (n/a)316.10 (n/a)231.36 (n/a)211.50 (n/a)192.60 (n/a)49.39 (n/a)

mem_copy_16_cores_2_chans_2048_tile_128_False0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.08 (-2.12%)0.06 (-16.44%)0.06 (-21.48%)0.03 (-23.84%)0.02 (+10.15%)247.60 (+31.28%)148.38 (+24.08%)128.30 (+27.41%)100.10 (+2.14%)58.70 (+50.73%)
84d3478 — 2026-02-17 23:25:130.08 (n/a)0.07 (n/a)0.08 (n/a)0.04 (n/a)0.02 (n/a)188.60 (n/a)119.58 (n/a)100.70 (n/a)98.00 (n/a)38.94 (n/a)

mem_copy_1_cols_1_channels_2048_tile_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
003b49d — 2025-11-14 19:30:540.05 (-16.99%)0.05 (-12.47%)0.05 (-7.32%)0.04 (-11.17%)0.00 (-29.61%)197.00 (+12.57%)171.00 (+13.70%)164.00 (+7.89%)153.00 (+20.47%)17.49 (-4.32%)
0999a9a — 2025-11-13 19:24:080.06 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.01 (n/a)175.00 (n/a)150.40 (n/a)152.00 (n/a)127.00 (n/a)18.28 (n/a)

mem_copy_1_cols_2_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
003b49d — 2025-11-14 19:30:540.06 (+5.56%)0.05 (-1.59%)0.05 (-1.39%)0.03 (-14.07%)0.01 (+25.03%)263.00 (+16.37%)166.00 (+4.53%)144.00 (+1.41%)126.00 (-5.26%)49.38 (+43.17%)
0999a9a — 2025-11-13 19:24:080.06 (n/a)0.05 (n/a)0.05 (n/a)0.03 (n/a)0.01 (n/a)226.00 (n/a)158.80 (n/a)142.00 (n/a)133.00 (n/a)34.49 (n/a)

mem_copy_1_cores_1_chans_2048_tile_2048_False

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.09 (+75.06%)0.05 (+7.82%)0.04 (-15.32%)0.03 (-13.78%)0.02 (+425.35%)249.10 (+15.97%)190.02 (+3.98%)211.70 (+18.07%)95.60 (-42.89%)62.62 (+236.69%)
dc66da7 — 2026-03-25 19:23:450.05 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.00 (n/a)214.80 (n/a)182.74 (n/a)179.30 (n/a)167.40 (n/a)18.60 (n/a)

mem_copy_1_cores_1_chans_2048_tile_2048_False0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.11 (-0.24%)0.08 (-17.53%)0.08 (-27.05%)0.06 (-15.43%)0.02 (+4.62%)148.50 (+18.23%)113.02 (+23.38%)103.60 (+37.04%)74.10 (+0.14%)31.69 (+33.33%)
84d3478 — 2026-02-17 23:25:130.11 (n/a)0.09 (n/a)0.11 (n/a)0.07 (n/a)0.02 (n/a)125.60 (n/a)91.60 (n/a)75.60 (n/a)74.00 (n/a)23.77 (n/a)

mem_copy_2_cols_1_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
003b49d — 2025-11-14 19:30:540.06 (-15.32%)0.05 (-17.09%)0.05 (-23.81%)0.04 (-15.54%)0.01 (-18.18%)193.00 (+18.40%)160.80 (+20.36%)168.00 (+31.25%)124.00 (+18.10%)22.92 (+11.39%)
0999a9a — 2025-11-13 19:24:080.07 (n/a)0.06 (n/a)0.06 (n/a)0.05 (n/a)0.01 (n/a)163.00 (n/a)133.60 (n/a)128.00 (n/a)105.00 (n/a)20.58 (n/a)

mem_copy_2_cols_2_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
003b49d — 2025-11-14 19:30:540.06 (-23.02%)0.05 (-6.98%)0.05 (-2.13%)0.04 (-15.31%)0.01 (-33.29%)196.00 (+18.07%)147.00 (+6.52%)141.00 (+2.17%)126.00 (+29.90%)25.27 (+8.13%)
0999a9a — 2025-11-13 19:24:080.08 (n/a)0.06 (n/a)0.06 (n/a)0.05 (n/a)0.01 (n/a)166.00 (n/a)138.00 (n/a)138.00 (n/a)97.00 (n/a)23.37 (n/a)

mem_copy_2_cores_1_chans_2048_tile_1024_False

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.07 (+13.70%)0.05 (-1.93%)0.05 (-7.07%)0.04 (-27.15%)0.02 (+157.23%)230.10 (+37.29%)166.58 (+9.82%)161.90 (+7.65%)109.70 (-12.03%)53.34 (+211.05%)
dc66da7 — 2026-03-25 19:23:450.07 (n/a)0.05 (n/a)0.05 (n/a)0.05 (n/a)0.01 (n/a)167.60 (n/a)151.68 (n/a)150.40 (n/a)124.70 (n/a)17.15 (n/a)

mem_copy_2_cores_1_chans_2048_tile_1024_False0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.11 (+27.03%)0.09 (+25.79%)0.11 (+27.09%)0.05 (+12.74%)0.03 (+43.93%)161.90 (-11.29%)94.34 (-18.35%)76.30 (-21.34%)74.00 (-21.36%)38.01 (+0.49%)
84d3478 — 2026-02-17 23:25:130.09 (n/a)0.08 (n/a)0.08 (n/a)0.04 (n/a)0.02 (n/a)182.50 (n/a)115.54 (n/a)97.00 (n/a)94.10 (n/a)37.82 (n/a)

mem_copy_2_cores_2_chans_2048_tile_1024_False

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.07 (-3.17%)0.05 (-7.71%)0.05 (-6.65%)0.04 (-20.60%)0.02 (+42.64%)223.40 (+25.93%)171.62 (+13.04%)172.10 (+7.09%)120.40 (+3.26%)49.35 (+82.52%)
dc66da7 — 2026-03-25 19:23:450.07 (n/a)0.06 (n/a)0.05 (n/a)0.05 (n/a)0.01 (n/a)177.40 (n/a)151.82 (n/a)160.70 (n/a)116.60 (n/a)27.04 (n/a)

mem_copy_2_cores_2_chans_2048_tile_1024_False0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.10 (+2.70%)0.08 (-5.38%)0.09 (+2.10%)0.04 (-43.30%)0.02 (+170.50%)201.60 (+76.38%)112.08 (+16.53%)90.80 (-2.05%)84.60 (-2.65%)50.14 (+375.75%)
84d3478 — 2026-02-17 23:25:130.09 (n/a)0.09 (n/a)0.09 (n/a)0.07 (n/a)0.01 (n/a)114.30 (n/a)96.18 (n/a)92.70 (n/a)86.90 (n/a)10.54 (n/a)

mem_copy_4_cols_1_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
003b49d — 2025-11-14 19:30:540.07 (+11.97%)0.05 (+3.09%)0.05 (+2.10%)0.03 (-11.35%)0.01 (+41.65%)229.00 (+12.81%)156.00 (-0.51%)143.00 (-2.05%)117.00 (-10.69%)38.21 (+48.71%)
0999a9a — 2025-11-13 19:24:080.06 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.01 (n/a)203.00 (n/a)156.80 (n/a)146.00 (n/a)131.00 (n/a)25.69 (n/a)

mem_copy_4_cols_2_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
003b49d — 2025-11-14 19:30:540.06 (+9.02%)0.05 (+5.59%)0.05 (+0.00%)0.04 (+3.63%)0.01 (+25.48%)193.00 (-3.50%)147.40 (-4.66%)147.00 (+0.00%)122.00 (-8.27%)25.24 (+7.48%)
0999a9a — 2025-11-13 19:24:080.06 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.01 (n/a)200.00 (n/a)154.60 (n/a)147.00 (n/a)133.00 (n/a)23.48 (n/a)

mem_copy_4_cores_1_chans_2048_tile_512_False

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.06 (-2.79%)0.05 (-5.43%)0.05 (+6.39%)0.03 (-12.50%)0.01 (+16.26%)242.20 (+14.25%)184.14 (+7.91%)167.80 (-6.05%)128.30 (+2.89%)46.84 (+42.55%)
dc66da7 — 2026-03-25 19:23:450.07 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.01 (n/a)212.00 (n/a)170.64 (n/a)178.60 (n/a)124.70 (n/a)32.86 (n/a)

mem_copy_4_cores_1_chans_2048_tile_512_False0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.09 (-5.87%)0.08 (-3.61%)0.08 (-0.79%)0.06 (-12.74%)0.01 (+22.04%)146.30 (+14.66%)108.02 (+4.96%)99.30 (+0.81%)95.10 (+6.26%)21.49 (+48.38%)
84d3478 — 2026-02-17 23:25:130.09 (n/a)0.08 (n/a)0.08 (n/a)0.06 (n/a)0.01 (n/a)127.60 (n/a)102.92 (n/a)98.50 (n/a)89.50 (n/a)14.49 (n/a)

mem_copy_4_cores_2_chans_2048_tile_512_False

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.05 (+7.78%)0.04 (-14.27%)0.04 (-10.58%)0.03 (-32.73%)0.01 (+109.81%)316.90 (+48.71%)226.08 (+22.64%)209.60 (+11.85%)150.20 (-7.23%)61.97 (+194.31%)
dc66da7 — 2026-03-25 19:23:450.05 (n/a)0.04 (n/a)0.04 (n/a)0.04 (n/a)0.01 (n/a)213.10 (n/a)184.34 (n/a)187.40 (n/a)161.90 (n/a)21.05 (n/a)

mem_copy_4_cores_2_chans_2048_tile_512_False0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.11 (+13.04%)0.09 (+17.19%)0.10 (+20.43%)0.06 (+17.23%)0.02 (+16.34%)142.80 (-14.70%)91.94 (-14.71%)79.00 (-17.02%)77.70 (-11.50%)28.50 (-14.82%)
84d3478 — 2026-02-17 23:25:130.09 (n/a)0.08 (n/a)0.09 (n/a)0.05 (n/a)0.02 (n/a)167.40 (n/a)107.80 (n/a)95.20 (n/a)87.80 (n/a)33.46 (n/a)

mem_copy_8_cols_1_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
003b49d — 2025-11-14 19:30:540.05 (-13.48%)0.05 (-17.79%)0.05 (-15.33%)0.03 (-25.09%)0.01 (+6.67%)267.00 (+33.50%)176.80 (+24.51%)150.00 (+18.11%)141.00 (+15.57%)47.41 (+61.41%)
0999a9a — 2025-11-13 19:24:080.06 (n/a)0.06 (n/a)0.06 (n/a)0.04 (n/a)0.01 (n/a)200.00 (n/a)142.00 (n/a)127.00 (n/a)122.00 (n/a)29.37 (n/a)

mem_copy_8_cols_2_channels_2048_tile_128

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
003b49d — 2025-11-14 19:30:540.05 (+13.38%)0.05 (+23.85%)0.05 (+37.91%)0.05 (+30.72%)0.00 (-56.85%)166.00 (-23.50%)153.40 (-20.60%)153.00 (-27.49%)142.00 (-11.80%)7.63 (-70.60%)
0999a9a — 2025-11-13 19:24:080.05 (n/a)0.04 (n/a)0.04 (n/a)0.04 (n/a)0.01 (n/a)217.00 (n/a)193.20 (n/a)211.00 (n/a)161.00 (n/a)25.96 (n/a)

mem_copy_8_cores_1_chans_2048_tile_256_False

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.06 (-5.71%)0.04 (+0.46%)0.04 (+2.21%)0.03 (-7.12%)0.01 (-7.04%)278.00 (+7.67%)209.24 (-0.41%)216.20 (-2.17%)141.90 (+6.05%)51.53 (+9.47%)
dc66da7 — 2026-03-25 19:23:450.06 (n/a)0.04 (n/a)0.04 (n/a)0.03 (n/a)0.01 (n/a)258.20 (n/a)210.10 (n/a)221.00 (n/a)133.80 (n/a)47.07 (n/a)

mem_copy_8_cores_1_chans_2048_tile_256_False0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.08 (-0.39%)0.07 (+10.07%)0.08 (+1.07%)0.06 (+38.38%)0.01 (-47.35%)136.70 (-27.75%)110.64 (-13.07%)104.70 (-1.13%)103.00 (+0.39%)14.59 (-60.78%)
84d3478 — 2026-02-17 23:25:130.08 (n/a)0.07 (n/a)0.08 (n/a)0.04 (n/a)0.02 (n/a)189.20 (n/a)127.28 (n/a)105.90 (n/a)102.60 (n/a)37.20 (n/a)

mem_copy_8_cores_2_chans_2048_tile_256_False

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.06 (+19.11%)0.04 (+1.88%)0.04 (-11.30%)0.03 (+24.20%)0.01 (+4.86%)262.00 (-19.48%)199.56 (-3.59%)200.50 (+12.70%)139.70 (-16.05%)45.93 (-31.63%)
dc66da7 — 2026-03-25 19:23:450.05 (n/a)0.04 (n/a)0.05 (n/a)0.03 (n/a)0.01 (n/a)325.40 (n/a)207.00 (n/a)177.90 (n/a)166.40 (n/a)67.19 (n/a)

mem_copy_8_cores_2_chans_2048_tile_256_False0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.08 (-17.85%)0.07 (-24.66%)0.08 (-18.40%)0.04 (-32.08%)0.02 (+21.78%)185.30 (+47.18%)127.72 (+37.87%)102.00 (+22.45%)100.40 (+21.70%)38.66 (+106.34%)
84d3478 — 2026-02-17 23:25:130.10 (n/a)0.09 (n/a)0.10 (n/a)0.07 (n/a)0.01 (n/a)125.90 (n/a)92.64 (n/a)83.30 (n/a)82.50 (n/a)18.74 (n/a)

mha

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.18 (+0.10%)0.18 (+0.09%)0.18 (+0.09%)0.18 (+0.08%)0.00 (+11.08%)47507.50 (-0.08%)47468.96 (-0.09%)47481.10 (-0.09%)47413.10 (-0.10%)37.32 (+10.96%)
dc66da7 — 2026-03-25 19:23:450.18 (n/a)0.18 (n/a)0.18 (n/a)0.18 (n/a)0.00 (n/a)47544.20 (n/a)47512.60 (n/a)47521.80 (n/a)47459.60 (n/a)33.63 (n/a)

mha0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
84d3478 — 2026-02-17 23:25:130.18 (-0.51%)0.18 (-0.36%)0.18 (-0.41%)0.18 (-0.21%)0.00 (-35.21%)47468.60 (+0.21%)47403.94 (+0.36%)47426.00 (+0.41%)47314.40 (+0.51%)68.83 (-34.73%)
1531781 — 2026-01-23 18:49:240.18 (n/a)0.18 (n/a)0.18 (n/a)0.18 (n/a)0.00 (n/a)47367.50 (n/a)47232.68 (n/a)47232.00 (n/a)47073.00 (n/a)105.45 (n/a)

mha_16384_64_1_8_0_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.18 (n/a)0.18 (n/a)0.18 (n/a)0.18 (n/a)0.00 (n/a)47533.70 (n/a)47380.56 (n/a)47337.90 (n/a)47268.00 (n/a)110.60 (n/a)

relu_1_cols_1_channels_2048_tile_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.07 (+14.06%)0.05 (+0.31%)0.05 (-3.71%)0.04 (-18.38%)0.01 (+129.69%)208.90 (+22.52%)161.96 (+2.56%)166.90 (+3.86%)120.40 (-12.31%)33.36 (+144.00%)
dc66da7 — 2026-03-25 19:23:450.06 (n/a)0.05 (n/a)0.05 (n/a)0.05 (n/a)0.00 (n/a)170.50 (n/a)157.92 (n/a)160.70 (n/a)137.30 (n/a)13.67 (n/a)

relu_2_cols_1_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.06 (-20.92%)0.05 (-6.86%)0.05 (-1.28%)0.03 (-20.49%)0.01 (-19.18%)264.20 (+25.81%)174.76 (+7.41%)149.00 (+1.29%)140.20 (+26.42%)52.29 (+24.11%)
dc66da7 — 2026-03-25 19:23:450.07 (n/a)0.05 (n/a)0.06 (n/a)0.04 (n/a)0.01 (n/a)210.00 (n/a)162.70 (n/a)147.10 (n/a)110.90 (n/a)42.13 (n/a)

relu_4_cols_1_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.05 (-8.73%)0.04 (-10.26%)0.04 (-4.04%)0.03 (-16.73%)0.01 (+5.34%)259.40 (+20.09%)206.80 (+12.91%)200.50 (+4.21%)150.20 (+9.56%)45.96 (+41.47%)
dc66da7 — 2026-03-25 19:23:450.06 (n/a)0.05 (n/a)0.04 (n/a)0.04 (n/a)0.01 (n/a)216.00 (n/a)183.16 (n/a)192.40 (n/a)137.10 (n/a)32.49 (n/a)

relu_8_cols_1_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.05 (+4.18%)0.04 (-0.33%)0.04 (-1.55%)0.02 (-20.26%)0.01 (+47.86%)369.50 (+25.38%)233.74 (+5.19%)215.30 (+1.56%)165.20 (-4.01%)80.83 (+78.85%)
dc66da7 — 2026-03-25 19:23:450.05 (n/a)0.04 (n/a)0.04 (n/a)0.03 (n/a)0.01 (n/a)294.70 (n/a)222.20 (n/a)212.00 (n/a)172.10 (n/a)45.19 (n/a)

rms_norm_1_cols_1_channels_2048_tile_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.06 (-18.06%)0.05 (-2.61%)0.05 (-1.31%)0.04 (+12.13%)0.01 (-43.39%)226.00 (-10.81%)172.66 (-1.55%)163.30 (+1.30%)140.00 (+22.06%)32.37 (-37.02%)
dc66da7 — 2026-03-25 19:23:450.07 (n/a)0.05 (n/a)0.05 (n/a)0.03 (n/a)0.01 (n/a)253.40 (n/a)175.38 (n/a)161.20 (n/a)114.70 (n/a)51.40 (n/a)

rms_norm_1_cols_2_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.08 (+16.34%)0.05 (+6.02%)0.05 (-3.17%)0.04 (+12.01%)0.02 (+27.81%)218.40 (-10.71%)162.92 (-3.80%)150.40 (+3.23%)108.00 (-14.01%)52.41 (+3.43%)
dc66da7 — 2026-03-25 19:23:450.07 (n/a)0.05 (n/a)0.06 (n/a)0.03 (n/a)0.01 (n/a)244.60 (n/a)169.36 (n/a)145.70 (n/a)125.60 (n/a)50.68 (n/a)

rms_norm_2_cols_1_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.06 (+9.52%)0.04 (-9.59%)0.04 (-17.77%)0.03 (-18.95%)0.01 (+55.38%)255.30 (+23.39%)199.88 (+13.52%)205.30 (+21.62%)137.80 (-8.68%)43.62 (+70.31%)
dc66da7 — 2026-03-25 19:23:450.05 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.01 (n/a)206.90 (n/a)176.08 (n/a)168.80 (n/a)150.90 (n/a)25.61 (n/a)

rms_norm_2_cols_2_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.06 (+6.05%)0.04 (+0.61%)0.04 (-6.97%)0.03 (-2.64%)0.01 (+44.42%)234.40 (+2.72%)195.80 (+1.42%)215.50 (+7.48%)141.90 (-5.65%)40.78 (+43.76%)
dc66da7 — 2026-03-25 19:23:450.05 (n/a)0.04 (n/a)0.04 (n/a)0.04 (n/a)0.01 (n/a)228.20 (n/a)193.06 (n/a)200.50 (n/a)150.40 (n/a)28.37 (n/a)

rms_norm_4_cols_1_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.06 (+10.73%)0.05 (-1.13%)0.05 (+5.17%)0.04 (-10.70%)0.01 (+73.44%)225.10 (+11.99%)181.84 (+3.00%)170.50 (-4.91%)140.00 (-9.68%)33.21 (+78.26%)
dc66da7 — 2026-03-25 19:23:450.05 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.00 (n/a)201.00 (n/a)176.54 (n/a)179.30 (n/a)155.00 (n/a)18.63 (n/a)

rms_norm_4_cols_2_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.06 (-0.51%)0.04 (-4.22%)0.04 (-6.40%)0.03 (+19.93%)0.01 (-12.26%)264.40 (-16.62%)208.96 (+1.64%)223.40 (+6.84%)140.70 (+0.50%)52.99 (-24.83%)
dc66da7 — 2026-03-25 19:23:450.06 (n/a)0.04 (n/a)0.04 (n/a)0.03 (n/a)0.01 (n/a)317.10 (n/a)205.58 (n/a)209.10 (n/a)140.00 (n/a)70.49 (n/a)

rms_norm_8_cols_1_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.06 (-8.19%)0.04 (-16.80%)0.03 (-25.89%)0.03 (-16.40%)0.01 (+5.02%)283.50 (+19.62%)220.72 (+22.74%)240.30 (+34.92%)128.00 (+8.94%)58.39 (+33.24%)
dc66da7 — 2026-03-25 19:23:450.07 (n/a)0.05 (n/a)0.05 (n/a)0.03 (n/a)0.01 (n/a)237.00 (n/a)179.82 (n/a)178.10 (n/a)117.50 (n/a)43.82 (n/a)

rms_norm_8_cols_2_channels_2048_tile_128

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.04 (+4.11%)0.04 (+2.57%)0.04 (-1.84%)0.03 (+8.94%)0.00 (+2.64%)250.80 (-8.20%)224.64 (-2.58%)233.20 (+1.88%)191.20 (-3.97%)25.83 (-9.64%)
dc66da7 — 2026-03-25 19:23:450.04 (n/a)0.04 (n/a)0.04 (n/a)0.03 (n/a)0.00 (n/a)273.20 (n/a)230.60 (n/a)228.90 (n/a)199.10 (n/a)28.59 (n/a)

rope_1_cols_2_channels_4096_tile_4096_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
1b2d230 — 2026-01-13 22:38:030.19 (-21.66%)0.17 (-8.53%)0.18 (+2.09%)0.14 (-6.71%)0.02 (-38.07%)174.10 (+7.20%)150.06 (+8.11%)139.60 (-2.04%)129.80 (+27.63%)20.36 (-10.65%)
d48746f — 2026-01-12 22:17:100.24 (n/a)0.18 (n/a)0.17 (n/a)0.15 (n/a)0.04 (n/a)162.40 (n/a)138.80 (n/a)142.50 (n/a)101.70 (n/a)22.79 (n/a)

rope_1c_32rows_512cols_32arows_0m

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.91 (+31.28%)0.72 (+30.97%)0.75 (+50.09%)0.52 (+23.14%)0.16 (+34.35%)189.50 (-18.81%)142.68 (-23.26%)131.80 (-33.40%)108.20 (-23.86%)33.21 (-13.83%)
dc66da7 — 2026-03-25 19:23:450.69 (n/a)0.55 (n/a)0.50 (n/a)0.42 (n/a)0.12 (n/a)233.40 (n/a)185.92 (n/a)197.90 (n/a)142.10 (n/a)38.54 (n/a)

rope_1c_32rows_512cols_8arows_0m

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.71 (+13.56%)0.53 (+16.33%)0.51 (+20.62%)0.43 (+13.26%)0.12 (+12.08%)172.60 (-11.71%)143.86 (-14.09%)145.70 (-17.07%)103.70 (-11.97%)28.08 (-12.19%)
dc66da7 — 2026-03-25 19:23:450.63 (n/a)0.46 (n/a)0.42 (n/a)0.38 (n/a)0.10 (n/a)195.50 (n/a)167.46 (n/a)175.70 (n/a)117.80 (n/a)31.98 (n/a)

rope_2_cols_2_channels_4096_tile_2048_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
1b2d230 — 2026-01-13 22:38:030.21 (-0.56%)0.17 (-10.62%)0.16 (-13.58%)0.13 (-13.33%)0.03 (+17.05%)183.70 (+15.39%)151.76 (+12.77%)155.20 (+15.73%)117.30 (+0.60%)24.24 (+35.73%)
d48746f — 2026-01-12 22:17:100.21 (n/a)0.19 (n/a)0.18 (n/a)0.15 (n/a)0.02 (n/a)159.20 (n/a)134.58 (n/a)134.10 (n/a)116.60 (n/a)17.86 (n/a)

rope_2c_32rows_512cols_32arows_0m

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.86 (+12.27%)0.68 (+4.35%)0.67 (-1.63%)0.56 (+6.41%)0.12 (+18.02%)174.80 (-6.02%)146.78 (-3.85%)146.60 (+1.66%)114.70 (-10.88%)23.97 (-1.25%)
dc66da7 — 2026-03-25 19:23:450.76 (n/a)0.66 (n/a)0.68 (n/a)0.53 (n/a)0.10 (n/a)186.00 (n/a)152.66 (n/a)144.20 (n/a)128.70 (n/a)24.27 (n/a)

rope_2c_32rows_512cols_8arows_0m

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.63 (+15.38%)0.42 (-5.03%)0.37 (-15.06%)0.26 (-23.68%)0.14 (+76.55%)289.00 (+31.07%)192.28 (+12.08%)197.90 (+17.73%)117.80 (-13.32%)64.38 (+98.12%)
dc66da7 — 2026-03-25 19:23:450.54 (n/a)0.44 (n/a)0.44 (n/a)0.33 (n/a)0.08 (n/a)220.50 (n/a)171.56 (n/a)168.10 (n/a)135.90 (n/a)32.49 (n/a)

rope_4_cols_2_channels_4096_tile_1024_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
1b2d230 — 2026-01-13 22:38:030.22 (+18.92%)0.17 (+6.24%)0.17 (+5.57%)0.14 (-0.50%)0.03 (+90.77%)171.10 (+0.47%)147.40 (-4.39%)148.10 (-5.31%)110.10 (-15.89%)23.59 (+60.67%)
d48746f — 2026-01-12 22:17:100.19 (n/a)0.16 (n/a)0.16 (n/a)0.14 (n/a)0.02 (n/a)170.30 (n/a)154.16 (n/a)156.40 (n/a)130.90 (n/a)14.69 (n/a)

rope_8_cols_2_channels_4096_tile_512_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
1b2d230 — 2026-01-13 22:38:030.17 (-1.30%)0.15 (+6.63%)0.15 (-8.78%)0.14 (+39.68%)0.01 (-67.35%)174.90 (-28.41%)163.42 (-10.36%)167.40 (+9.63%)146.50 (+1.31%)11.19 (-76.17%)
d48746f — 2026-01-12 22:17:100.17 (n/a)0.14 (n/a)0.16 (n/a)0.10 (n/a)0.03 (n/a)244.30 (n/a)182.30 (n/a)152.70 (n/a)144.60 (n/a)46.98 (n/a)

rope_8c_32rows_512cols_32arows_0m

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.60 (+0.58%)0.47 (-10.94%)0.47 (-14.67%)0.39 (-14.06%)0.09 (+57.47%)252.70 (+16.34%)213.46 (+14.31%)211.20 (+17.20%)163.30 (-0.61%)38.58 (+85.72%)
dc66da7 — 2026-03-25 19:23:450.60 (n/a)0.53 (n/a)0.55 (n/a)0.45 (n/a)0.06 (n/a)217.20 (n/a)186.74 (n/a)180.20 (n/a)164.30 (n/a)20.77 (n/a)

rope_8c_32rows_512cols_8arows_0m

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.58 (+33.33%)0.38 (-2.44%)0.32 (-20.46%)0.28 (-18.27%)0.13 (+203.34%)264.90 (+22.36%)209.28 (+9.84%)227.20 (+25.73%)126.60 (-25.00%)59.71 (+180.42%)
dc66da7 — 2026-03-25 19:23:450.44 (n/a)0.39 (n/a)0.41 (n/a)0.34 (n/a)0.04 (n/a)216.50 (n/a)190.54 (n/a)180.70 (n/a)168.80 (n/a)21.29 (n/a)

rows_32-cols_512-angle_rows_32-aie_columns_1-method_type_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
9580ba2 — 2026-03-25 20:55:190.40 (n/a)0.33 (n/a)0.38 (n/a)0.16 (n/a)0.10 (n/a)610.60 (n/a)336.92 (n/a)256.60 (n/a)245.50 (n/a)156.45 (n/a)

rows_32-cols_512-angle_rows_32-aie_columns_2-method_type_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
9580ba2 — 2026-03-25 20:55:190.40 (n/a)0.33 (n/a)0.29 (n/a)0.26 (n/a)0.07 (n/a)378.20 (n/a)311.60 (n/a)343.80 (n/a)243.00 (n/a)63.21 (n/a)

rows_32-cols_512-angle_rows_32-aie_columns_8-method_type_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
9580ba2 — 2026-03-25 20:55:190.41 (n/a)0.29 (n/a)0.32 (n/a)0.15 (n/a)0.11 (n/a)657.00 (n/a)385.16 (n/a)302.70 (n/a)242.00 (n/a)173.83 (n/a)

rows_32-cols_512-angle_rows_8-aie_columns_1-method_type_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
9580ba2 — 2026-03-25 20:55:190.25 (n/a)0.24 (n/a)0.24 (n/a)0.20 (n/a)0.02 (n/a)369.70 (n/a)315.28 (n/a)307.40 (n/a)289.80 (n/a)31.48 (n/a)

rows_32-cols_512-angle_rows_8-aie_columns_2-method_type_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
9580ba2 — 2026-03-25 20:55:190.25 (n/a)0.23 (n/a)0.24 (n/a)0.18 (n/a)0.03 (n/a)406.70 (n/a)323.20 (n/a)305.60 (n/a)293.50 (n/a)47.01 (n/a)

rows_32-cols_512-angle_rows_8-aie_columns_8-method_type_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
9580ba2 — 2026-03-25 20:55:190.27 (n/a)0.20 (n/a)0.24 (n/a)0.10 (n/a)0.07 (n/a)753.90 (n/a)412.66 (n/a)310.80 (n/a)271.30 (n/a)198.28 (n/a)

seq_len_16384-dim_64-num_heads_1-num_pipelines_8-num_kv_heads_0

No metrics available.

seq_len_256-embedding_dim_2048-hidden_dim_2048-prio_accuracy_False

No metrics available.

sigmoid_1_cols_1_channels_2048_tile_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.05 (-14.16%)0.05 (-6.29%)0.05 (-2.65%)0.04 (-7.71%)0.01 (-21.79%)225.80 (+8.35%)180.84 (+6.33%)170.90 (+2.70%)163.30 (+16.48%)25.41 (+0.55%)
dc66da7 — 2026-03-25 19:23:450.06 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.01 (n/a)208.40 (n/a)170.08 (n/a)166.40 (n/a)140.20 (n/a)25.28 (n/a)

sigmoid_2_cols_1_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.06 (+4.68%)0.05 (+2.94%)0.04 (-13.06%)0.04 (+15.35%)0.01 (+8.49%)232.90 (-13.32%)179.44 (-3.16%)197.20 (+15.05%)127.30 (-4.50%)46.58 (-13.93%)
dc66da7 — 2026-03-25 19:23:450.06 (n/a)0.05 (n/a)0.05 (n/a)0.03 (n/a)0.01 (n/a)268.70 (n/a)185.30 (n/a)171.40 (n/a)133.30 (n/a)54.12 (n/a)

sigmoid_4_cols_1_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.07 (+10.37%)0.05 (-2.49%)0.04 (-7.15%)0.04 (-7.47%)0.01 (+64.06%)210.80 (+8.10%)175.44 (+5.70%)183.30 (+7.70%)117.30 (-9.42%)39.46 (+67.78%)
dc66da7 — 2026-03-25 19:23:450.06 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.01 (n/a)195.00 (n/a)165.98 (n/a)170.20 (n/a)129.50 (n/a)23.52 (n/a)

sigmoid_8_cols_1_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.06 (+29.56%)0.05 (+22.81%)0.05 (+37.30%)0.03 (-8.45%)0.01 (+218.31%)242.70 (+9.23%)173.04 (-14.71%)150.20 (-27.16%)134.70 (-22.81%)46.84 (+169.71%)
dc66da7 — 2026-03-25 19:23:450.05 (n/a)0.04 (n/a)0.04 (n/a)0.04 (n/a)0.00 (n/a)222.20 (n/a)202.88 (n/a)206.20 (n/a)174.50 (n/a)17.37 (n/a)

silu_1_cols_1_channels_2048_tile_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.07 (+9.38%)0.05 (-4.51%)0.04 (-23.77%)0.04 (+15.02%)0.01 (+5.83%)207.90 (-13.05%)169.28 (+4.04%)188.60 (+31.15%)114.40 (-8.63%)38.46 (-17.21%)
dc66da7 — 2026-03-25 19:23:450.07 (n/a)0.05 (n/a)0.06 (n/a)0.03 (n/a)0.01 (n/a)239.10 (n/a)162.70 (n/a)143.80 (n/a)125.20 (n/a)46.46 (n/a)

silu_2_cols_1_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.06 (-4.63%)0.05 (-13.54%)0.05 (-12.20%)0.04 (-16.50%)0.01 (+19.45%)213.90 (+19.76%)180.30 (+16.99%)181.70 (+13.92%)134.00 (+4.85%)30.84 (+49.10%)
dc66da7 — 2026-03-25 19:23:450.06 (n/a)0.05 (n/a)0.05 (n/a)0.05 (n/a)0.01 (n/a)178.60 (n/a)154.12 (n/a)159.50 (n/a)127.80 (n/a)20.68 (n/a)

silu_4_cols_1_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.06 (+6.26%)0.05 (-8.81%)0.04 (-14.86%)0.04 (-17.09%)0.01 (+81.20%)214.80 (+20.61%)176.86 (+11.68%)182.90 (+17.47%)133.30 (-5.86%)30.25 (+101.37%)
dc66da7 — 2026-03-25 19:23:450.06 (n/a)0.05 (n/a)0.05 (n/a)0.05 (n/a)0.00 (n/a)178.10 (n/a)158.36 (n/a)155.70 (n/a)141.60 (n/a)15.02 (n/a)

silu_8_cols_1_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.06 (-6.11%)0.05 (-10.71%)0.04 (-10.38%)0.04 (-11.78%)0.01 (+1.86%)224.60 (+13.38%)185.30 (+12.62%)186.00 (+11.58%)140.40 (+6.53%)33.11 (+23.38%)
dc66da7 — 2026-03-25 19:23:450.06 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.01 (n/a)198.10 (n/a)164.54 (n/a)166.70 (n/a)131.80 (n/a)26.84 (n/a)

softmax_1_cols_2_channels_4096_tile_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.19 (-8.94%)0.17 (-8.66%)0.17 (-19.83%)0.14 (+4.32%)0.02 (-35.13%)115.80 (-4.14%)99.26 (+7.47%)98.60 (+24.65%)85.00 (+9.82%)13.65 (-30.86%)
84d3478 — 2026-02-17 23:25:130.21 (n/a)0.18 (n/a)0.21 (n/a)0.14 (n/a)0.04 (n/a)120.80 (n/a)92.36 (n/a)79.10 (n/a)77.40 (n/a)19.75 (n/a)

softmax_2_cols_2_channels_32768_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.64 (-42.09%)0.60 (-31.83%)0.59 (-30.19%)0.54 (-6.68%)0.05 (-76.96%)242.70 (+7.15%)221.12 (+40.50%)220.30 (+43.24%)203.40 (+72.67%)16.99 (-59.06%)
dc66da7 — 2026-03-25 19:23:451.11 (n/a)0.87 (n/a)0.85 (n/a)0.58 (n/a)0.20 (n/a)226.50 (n/a)157.38 (n/a)153.80 (n/a)117.80 (n/a)41.50 (n/a)

softmax_2_cols_2_channels_32768_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.68 (-34.93%)0.62 (-20.25%)0.60 (-17.49%)0.55 (-10.32%)0.05 (-71.29%)240.30 (+11.51%)214.32 (+20.87%)218.20 (+21.22%)193.80 (+53.69%)18.64 (-52.51%)
dc66da7 — 2026-03-25 19:23:451.04 (n/a)0.77 (n/a)0.73 (n/a)0.61 (n/a)0.18 (n/a)215.50 (n/a)177.32 (n/a)180.00 (n/a)126.10 (n/a)39.26 (n/a)

softmax_2_cols_2_channels_4096_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.20 (-0.87%)0.16 (+6.57%)0.17 (-9.55%)0.11 (+29.81%)0.03 (-44.34%)145.70 (-22.99%)104.68 (-15.00%)95.00 (+10.59%)83.80 (+0.84%)24.16 (-54.23%)
84d3478 — 2026-02-17 23:25:130.20 (n/a)0.15 (n/a)0.19 (n/a)0.09 (n/a)0.06 (n/a)189.20 (n/a)123.16 (n/a)85.90 (n/a)83.10 (n/a)52.80 (n/a)

softmax_2_cols_2_channels_4096_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.25 (+23.53%)0.18 (+9.42%)0.18 (-3.86%)0.11 (+2.99%)0.05 (+19.24%)144.10 (-2.90%)98.98 (-8.18%)92.10 (+4.07%)64.30 (-19.02%)29.58 (-7.13%)
84d3478 — 2026-02-17 23:25:130.21 (n/a)0.16 (n/a)0.19 (n/a)0.11 (n/a)0.04 (n/a)148.40 (n/a)107.80 (n/a)88.50 (n/a)79.40 (n/a)31.85 (n/a)

softmax_4_cols_4_channels_32768_tile_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.66 (+11.72%)0.50 (-0.51%)0.54 (-4.72%)0.35 (+6.44%)0.13 (+14.83%)377.70 (-6.04%)276.32 (+1.24%)242.20 (+4.94%)199.30 (-10.51%)76.50 (-0.31%)
dc66da7 — 2026-03-25 19:23:450.59 (n/a)0.51 (n/a)0.57 (n/a)0.33 (n/a)0.11 (n/a)402.00 (n/a)272.94 (n/a)230.80 (n/a)222.70 (n/a)76.73 (n/a)

swiglu

No metrics available.

swiglu_decode_1x2048x2048

No metrics available.

swiglu_decode_1x2048x2048_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.00 (-0.21%)0.00 (+7.74%)0.00 (+0.37%)0.00 (+8.84%)0.00 (-19.58%)5170.80 (-8.12%)3761.80 (-8.70%)3408.60 (-0.36%)3407.30 (+0.21%)787.66 (-22.98%)
84d3478 — 2026-02-17 23:25:130.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)5627.70 (n/a)4120.44 (n/a)3421.00 (n/a)3400.30 (n/a)1022.61 (n/a)

swiglu_prefill_256x2048x2048

No metrics available.

tanh_1_cols_1_channels_2048_tile_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.06 (-7.39%)0.04 (-16.97%)0.04 (-14.92%)0.03 (-28.80%)0.01 (+23.25%)296.40 (+40.47%)201.28 (+24.69%)188.60 (+17.51%)141.90 (+7.99%)59.45 (+90.30%)
dc66da7 — 2026-03-25 19:23:450.06 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.01 (n/a)211.00 (n/a)161.42 (n/a)160.50 (n/a)131.40 (n/a)31.24 (n/a)

tanh_2_cols_1_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.06 (-14.29%)0.05 (-6.63%)0.05 (+2.99%)0.03 (-30.73%)0.01 (+5.83%)293.30 (+44.41%)176.82 (+11.67%)159.70 (-2.92%)128.50 (+16.71%)67.26 (+86.47%)
dc66da7 — 2026-03-25 19:23:450.07 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.01 (n/a)203.10 (n/a)158.34 (n/a)164.50 (n/a)110.10 (n/a)36.07 (n/a)

tanh_4_cols_1_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.06 (-27.96%)0.05 (-4.84%)0.05 (+4.02%)0.04 (+16.65%)0.01 (-63.57%)187.60 (-14.30%)167.32 (-1.72%)171.90 (-3.86%)133.00 (+38.83%)20.69 (-55.09%)
dc66da7 — 2026-03-25 19:23:450.09 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.02 (n/a)218.90 (n/a)170.24 (n/a)178.80 (n/a)95.80 (n/a)46.06 (n/a)

tanh_8_cols_1_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.05 (-24.07%)0.04 (-15.48%)0.04 (-23.02%)0.03 (-3.09%)0.01 (-53.59%)262.30 (+3.19%)198.18 (+12.63%)183.30 (+29.91%)173.30 (+31.69%)36.55 (-34.15%)
dc66da7 — 2026-03-25 19:23:450.06 (n/a)0.05 (n/a)0.06 (n/a)0.03 (n/a)0.01 (n/a)254.20 (n/a)175.96 (n/a)141.10 (n/a)131.60 (n/a)55.51 (n/a)

transpose_2048_M_64_N_1_cols_1_channels_64_m_64_n_8_s

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:392.51 (-25.94%)2.10 (-20.30%)2.01 (-23.11%)1.76 (-13.49%)0.29 (-41.75%)298.70 (+15.60%)253.30 (+23.92%)261.10 (+30.09%)208.60 (+35.02%)33.97 (-9.04%)
dc66da7 — 2026-03-25 19:23:453.39 (n/a)2.64 (n/a)2.61 (n/a)2.03 (n/a)0.49 (n/a)258.40 (n/a)204.40 (n/a)200.70 (n/a)154.50 (n/a)37.34 (n/a)

transpose_2048_M_64_N_1_cols_1_channels_64_m_64_n_8_s0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:254.43 (+0.85%)3.73 (+3.83%)4.02 (-1.30%)2.97 (+16.77%)0.65 (-24.99%)176.70 (-14.35%)144.46 (-6.24%)130.40 (+1.32%)118.40 (-0.84%)26.60 (-35.16%)
84d3478 — 2026-02-17 23:25:134.39 (n/a)3.59 (n/a)4.07 (n/a)2.54 (n/a)0.87 (n/a)206.30 (n/a)154.08 (n/a)128.70 (n/a)119.40 (n/a)41.02 (n/a)

transpose_2048_M_64_N_1_cols_2_channels_64_m_64_n_8_s

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:392.72 (-10.04%)2.30 (-13.01%)2.26 (-13.66%)1.88 (-21.25%)0.34 (+46.90%)278.20 (+26.97%)231.60 (+16.34%)232.20 (+15.81%)192.40 (+11.15%)34.64 (+109.20%)
dc66da7 — 2026-03-25 19:23:453.03 (n/a)2.65 (n/a)2.61 (n/a)2.39 (n/a)0.23 (n/a)219.10 (n/a)199.08 (n/a)200.50 (n/a)173.10 (n/a)16.56 (n/a)

transpose_2048_M_64_N_1_cols_2_channels_64_m_64_n_8_s0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:255.27 (-14.18%)4.35 (-2.49%)4.45 (+2.15%)2.78 (-22.18%)0.98 (-3.38%)188.60 (+28.47%)126.98 (+4.20%)117.90 (-2.08%)99.40 (+16.53%)36.02 (+50.15%)
84d3478 — 2026-02-17 23:25:136.14 (n/a)4.46 (n/a)4.35 (n/a)3.57 (n/a)1.02 (n/a)146.80 (n/a)121.86 (n/a)120.40 (n/a)85.30 (n/a)23.99 (n/a)

weighted_rms_norm_1_cols_2_channels_2048_weights_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.09 (+6.38%)0.07 (-4.87%)0.07 (+1.68%)0.05 (-23.45%)0.01 (+101.20%)230.80 (+30.62%)177.68 (+7.40%)170.00 (-1.68%)138.30 (-5.98%)33.65 (+149.29%)
dc66da7 — 2026-03-25 19:23:450.08 (n/a)0.07 (n/a)0.07 (n/a)0.07 (n/a)0.01 (n/a)176.70 (n/a)165.44 (n/a)172.90 (n/a)147.10 (n/a)13.50 (n/a)

weighted_rms_norm_2_cols_2_channels_2048_weights_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.07 (-25.98%)0.06 (-18.74%)0.06 (-7.04%)0.04 (-20.93%)0.01 (-34.18%)237.00 (+26.47%)189.06 (+22.07%)179.50 (+7.55%)151.40 (+35.06%)34.27 (+14.91%)
dc66da7 — 2026-03-25 19:23:450.09 (n/a)0.07 (n/a)0.06 (n/a)0.05 (n/a)0.01 (n/a)187.40 (n/a)154.88 (n/a)166.90 (n/a)112.10 (n/a)29.83 (n/a)

weighted_rms_norm_4_cols_2_channels_2048_weights_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.06 (-13.39%)0.05 (-21.43%)0.04 (-31.53%)0.04 (-12.12%)0.01 (-0.80%)237.90 (+13.77%)205.32 (+28.15%)225.30 (+46.01%)147.80 (+15.47%)38.15 (+27.36%)
dc66da7 — 2026-03-25 19:23:450.07 (n/a)0.06 (n/a)0.06 (n/a)0.04 (n/a)0.01 (n/a)209.10 (n/a)160.22 (n/a)154.30 (n/a)128.00 (n/a)29.95 (n/a)

weighted_rms_norm_8_cols_2_channels_2048_weights_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.07 (+33.02%)0.04 (-11.48%)0.04 (-24.93%)0.02 (-32.99%)0.02 (+124.58%)371.50 (+49.26%)251.20 (+24.91%)247.70 (+33.17%)124.90 (-24.85%)88.76 (+136.07%)
dc66da7 — 2026-03-25 19:23:450.05 (n/a)0.04 (n/a)0.05 (n/a)0.03 (n/a)0.01 (n/a)248.90 (n/a)201.10 (n/a)186.00 (n/a)166.20 (n/a)37.60 (n/a)

hunhoffe and others added 3 commits March 25, 2026 15:07
…ink_with

mlir-aie 1.3.1 attaches link_with to each func.func declaration (Kernel),
and the aie-assign-core-link-files pass aggregates all referenced .o files
into a link_files ArrayAttr on each CoreOp. aiecc resolves these paths
relative to its working directory, making .a archives unnecessary.

Remove KernelArchiveArtifact, ArchiveCompilationRule, and all -L/-l: archive
flags from the compilation infrastructure. Each Kernel() now names its own
.o file directly; multiple kernels on the same core simply reference different
.o files and the pass deduplicates automatically.

- compilation/base.py: remove KernelArchiveArtifact and ArchiveCompilationRule
- compilation/context.py: remove ArchiveCompilationRule from rules list
- common/base.py: remove self.kernel_archive; pass KernelObjectArtifacts
  directly as XclbinArtifact dependencies
- common/__init__.py: remove KernelArchiveArtifact export
- common/fusion.py: remove self.kernel_archive; pass kernel objects directly
  as FullElfArtifact dependencies; remove kernel_archive from callback_kwargs
- all design.py files: remove kernel_archive param; hardcode .o name in
  each Kernel() call (parameterised names passed via callback_kwargs for
  gemv and gemm where the filename depends on op-level config)
- gemm/op.py: extract _kernel_flags_suffix property to share the flags
  encoding between get_mlir_artifact() and get_kernel_artifacts()

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
…ions

The previous commit incorrectly assigned three mha.cc-defined functions
to wrong .o files in their Kernel() declarations:

- matmul_bf16_bf16_wrapper: defined in mha.cc → mha_mha.o, not mha_mm.o
- partial_softmax: defined in mha.cc → mha_mha.o, not mha_softmax.o
- init_scale_buffer: defined in mha.cc → mha_mha.o, not mha_softmax.o

The aie-assign-core-link-files pass uses these link_with attributes to
populate link_files on each CoreOp; referencing a .o that does not contain
the symbol causes the linker to fail with an undefined symbol error when
compiling the core ELF.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
@github-actions
Copy link
Contributor

📊 Test Results for Test Example Applications

7d76781 (2026_03_25_21_47_06)

IRONCLAD

Tested on 2026_03_25_21_47_06 at commit 7d76781.

Test Checks TTFT (mean)TPS (mean)
llama_3.2_1b_prompt_1024_tokens_1 ❌ 0/5 n/a n/a
llama_3.2_1b_prompt_1024_tokens_40 ❌ 0/5 n/a n/a
llama_3.2_1b_prompt_13_tokens_1 ❌ 0/5 n/a n/a
llama_3.2_1b_prompt_13_tokens_40 ❌ 0/5 n/a n/a
📈 Trends (vs main branch) for Test Example Applications

7d76781 (2026_03_25_21_47_06)

IRONCLAD Trends

llama_3.2_1b

Commit/Date Num Tokens (max)Num Tokens (mean)Num Tokens (median)Num Tokens (min)Num Tokens (stddev)TPS (max)TPS (mean)TPS (median)TPS (min)TPS (stddev)TTFT (max)TTFT (mean)TTFT (median)TTFT (min)TTFT (stddev)Total (max)Total (mean)Total (median)Total (min)Total (stddev)
130b6ea — 2025-12-05 21:33:1240.00 (+0.00%)40.00 (+0.00%)40.00 (+0.00%)40.00 (+0.00%)0.00 (n/a)4.71 (-0.42%)4.64 (-0.09%)4.64 (+0.65%)4.55 (-0.22%)0.05 (-17.66%)4.41 (-0.34%)4.39 (-0.19%)4.38 (-0.33%)4.37 (-0.15%)0.01 (-25.90%)12.96 (-0.00%)12.80 (+0.07%)12.80 (-0.23%)12.67 (+0.44%)0.09 (-21.12%)
0a6c11c — 2025-12-03 23:35:1540.00 (n/a)40.00 (n/a)40.00 (n/a)40.00 (n/a)0.00 (n/a)4.73 (n/a)4.64 (n/a)4.61 (n/a)4.56 (n/a)0.06 (n/a)4.42 (n/a)4.40 (n/a)4.40 (n/a)4.37 (n/a)0.02 (n/a)12.96 (n/a)12.79 (n/a)12.83 (n/a)12.62 (n/a)0.12 (n/a)

llama_3.2_1b_prompt_1024_tokens_1

Commit/Date TTFT (max)TTFT (mean)TTFT (median)TTFT (min)TTFT (stddev)
7d76781 — 2026-03-25 21:45:21n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)
8b6d0f4 — 2026-03-25 21:36:352.26 (n/a)2.25 (n/a)2.25 (n/a)2.23 (n/a)0.01 (n/a)

llama_3.2_1b_prompt_1024_tokens_40

Commit/Date TPS (max)TPS (mean)TPS (median)TPS (min)TPS (stddev)TTFT (max)TTFT (mean)TTFT (median)TTFT (min)TTFT (stddev)
7d76781 — 2026-03-25 21:45:21n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)
8b6d0f4 — 2026-03-25 21:36:354.30 (n/a)4.28 (n/a)4.28 (n/a)4.26 (n/a)0.02 (n/a)2.37 (n/a)2.27 (n/a)2.25 (n/a)2.23 (n/a)0.06 (n/a)

llama_3.2_1b_prompt_13_tokens_1

Commit/Date TTFT (max)TTFT (mean)TTFT (median)TTFT (min)TTFT (stddev)
7d76781 — 2026-03-25 21:45:21n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)
8b6d0f4 — 2026-03-25 21:36:352.19 (n/a)2.17 (n/a)2.17 (n/a)2.14 (n/a)0.02 (n/a)

llama_3.2_1b_prompt_13_tokens_40

Commit/Date TPS (max)TPS (mean)TPS (median)TPS (min)TPS (stddev)TTFT (max)TTFT (mean)TTFT (median)TTFT (min)TTFT (stddev)
7d76781 — 2026-03-25 21:45:21n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)
8b6d0f4 — 2026-03-25 21:36:354.30 (n/a)4.27 (n/a)4.27 (n/a)4.25 (n/a)0.02 (n/a)2.19 (n/a)2.17 (n/a)2.16 (n/a)2.15 (n/a)0.01 (n/a)

llama_3.2_1b_prompt_2048_tokens_1

Commit/Date Num_Tokens (max)Num_Tokens (mean)Num_Tokens (median)Num_Tokens (min)Num_Tokens (stddev)TPS (max)TPS (mean)TPS (median)TPS (min)TPS (stddev)TTFT (max)TTFT (mean)TTFT (median)TTFT (min)TTFT (stddev)
897d04e — 2026-03-06 22:56:071.00 (+0.00%)1.00 (+0.00%)1.00 (+0.00%)1.00 (+0.00%)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)2.68 (-1.06%)2.68 (-1.06%)2.68 (-1.06%)2.68 (-1.06%)0.00 (n/a)
84d3478 — 2026-02-17 23:16:231.00 (n/a)1.00 (n/a)1.00 (n/a)1.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)2.70 (n/a)2.70 (n/a)2.70 (n/a)2.70 (n/a)0.00 (n/a)

llama_3.2_1b_prompt_2048_tokens_40

Commit/Date Num_Tokens (max)Num_Tokens (mean)Num_Tokens (median)Num_Tokens (min)Num_Tokens (stddev)TPS (max)TPS (mean)TPS (median)TPS (min)TPS (stddev)TTFT (max)TTFT (mean)TTFT (median)TTFT (min)TTFT (stddev)
897d04e — 2026-03-06 22:56:0740.00 (+0.00%)40.00 (+0.00%)40.00 (+0.00%)40.00 (+0.00%)0.00 (n/a)4.00 (-1.72%)4.00 (-1.72%)4.00 (-1.72%)4.00 (-1.72%)0.00 (n/a)2.70 (-0.44%)2.70 (-0.44%)2.70 (-0.44%)2.70 (-0.44%)0.00 (n/a)
84d3478 — 2026-02-17 23:16:2340.00 (n/a)40.00 (n/a)40.00 (n/a)40.00 (n/a)0.00 (n/a)4.07 (n/a)4.07 (n/a)4.07 (n/a)4.07 (n/a)0.00 (n/a)2.71 (n/a)2.71 (n/a)2.71 (n/a)2.71 (n/a)0.00 (n/a)

@github-actions
Copy link
Contributor

📊 Test Results for Small Benchmark/Test Suite

7d76781 (2026_03_25_21_59_28)

IRONCLAD

Tested on 2026_03_25_21_59_28 at commit 7d76781.

Test Checks Latency (mean)Bandwidth (mean)Throughput (mean)
M_128-K_128-num_aie_columns_1-tile_size_input_32-tile_size_output_128 ✅ 5/5 n/a 0.11 0.11
M_1792-K_896-N_1152-num_aie_columns_8-b_col_maj_False-c_col_maj_True-m_64-k_32-n_48-trace_size_0 ✅ 5/5 2359.30 4.09 1610.76
M_192-K_384-N_64-num_aie_columns_4-b_col_maj_False-c_col_maj_False-m_48-k_96-n_16-trace_size_0 ✅ 5/5 477.94 0.57 24.36
M_192-K_384-N_64-num_aie_columns_4-b_col_maj_True-c_col_maj_True-m_48-k_96-n_16-trace_size_0 ✅ 5/5 396.96 0.63 26.75
M_2048-K_2048-N_2048-num_aie_columns_1-b_col_maj_False-c_col_maj_False-m_64-k_64-n_64-trace_size_0 ✅ 5/5 49926.42 0.50 344.11
M_2048-K_2048-N_2048-num_aie_columns_2-b_col_maj_True-c_col_maj_False-m_64-k_64-n_64-trace_size_0 ✅ 5/5 29341.14 0.86 585.55
M_2048-K_2048-N_2048-num_aie_columns_8-b_col_maj_True-c_col_maj_True-m_64-k_64-n_64-trace_size_0 ✅ 5/5 8358.12 3.06 2086.76
M_2048-K_8192-num_aie_columns_1-tile_size_input_1-tile_size_output_2048 ✅ 5/5 n/a 11.22 11.22
M_2048-K_8192-num_aie_columns_2-tile_size_input_1-tile_size_output_1024 ✅ 5/5 n/a 19.78 19.77
M_2048-K_8192-num_aie_columns_4-tile_size_input_1-tile_size_output_512 ✅ 5/5 n/a 31.19 31.17
M_2048-K_8192-num_aie_columns_8-tile_size_input_1-tile_size_output_256 ✅ 5/5 n/a 33.19 33.17
M_2048-N_64-aie_columns_1-channels_1-m_64-n_64-s_8 ✅ 5/5 408.46 1.45 n/a
M_2048-N_64-aie_columns_1-channels_2-m_64-n_64-s_8 ✅ 5/5 298.86 1.79 n/a
M_384-K_1536-N_1792-num_aie_columns_4-b_col_maj_True-c_col_maj_False-m_32-k_48-n_64-trace_size_0 ✅ 5/5 2204.96 3.68 964.83
M_8192-K_2048-num_aie_columns_1-tile_size_input_4-tile_size_output_1024 ✅ 5/5 n/a 11.89 11.88
M_8192-K_2048-num_aie_columns_2-tile_size_input_4-tile_size_output_1024 ✅ 5/5 n/a 19.42 19.41
M_8192-K_2048-num_aie_columns_4-tile_size_input_4-tile_size_output_1024 ✅ 5/5 n/a 31.91 31.89
M_8192-K_2048-num_aie_columns_8-tile_size_input_4-tile_size_output_1024 ✅ 5/5 n/a 33.35 33.33
M_896-K_1792-N_640-num_aie_columns_8-b_col_maj_False-c_col_maj_True-m_32-k_64-n_80-trace_size_0 ✅ 5/5 1241.36 5.43 1677.11
embedding_dim_2048-hidden_dim_2048 ✅ 5/5 n/a n/a n/a
input_length_2048-num_aie_columns_1-num_channels_1-tile_size_2048 ✅ 25/25 287.14 0.03 n/a
input_length_2048-num_aie_columns_1-num_channels_1-tile_size_2048-group_size_32 ✅ 5/5 333.06 0.02 n/a
input_length_2048-num_aie_columns_1-num_channels_1-tile_size_2048-weighted_False ✅ 5/5 333.58 0.03 n/a
input_length_2048-num_aie_columns_1-num_channels_1-tile_size_2048-weighted_True ✅ 5/5 280.30 0.05 n/a
input_length_2048-num_aie_columns_1-num_channels_2-tile_size_1024 ✅ 10/10 326.33 0.03 n/a
input_length_2048-num_aie_columns_1-num_channels_2-tile_size_1024-group_size_32 ✅ 5/5 331.70 0.02 n/a
input_length_2048-num_aie_columns_1-num_channels_2-tile_size_1024-weighted_False ✅ 5/5 271.18 0.03 n/a
input_length_2048-num_aie_columns_1-num_channels_2-tile_size_2048-scalar_factor_3.0 ✅ 5/5 336.46 0.04 n/a
input_length_2048-num_aie_columns_1-tile_size_2048 ✅ 15/15 334.72 0.04 n/a
input_length_2048-num_aie_columns_2-num_channels_1-tile_size_1024 ✅ 25/25 271.27 0.03 n/a
input_length_2048-num_aie_columns_2-num_channels_1-tile_size_1024-group_size_32 ✅ 5/5 365.00 0.02 n/a
input_length_2048-num_aie_columns_2-num_channels_1-tile_size_1024-weighted_False ✅ 5/5 271.00 0.03 n/a
input_length_2048-num_aie_columns_2-num_channels_1-tile_size_1024-weighted_True ✅ 5/5 306.94 0.04 n/a
input_length_2048-num_aie_columns_2-num_channels_2-tile_size_1024-scalar_factor_3.0 ✅ 5/5 272.94 0.05 n/a
input_length_2048-num_aie_columns_2-num_channels_2-tile_size_512 ✅ 10/10 296.78 0.03 n/a
input_length_2048-num_aie_columns_2-num_channels_2-tile_size_512-group_size_32 ✅ 5/5 317.12 0.02 n/a
input_length_2048-num_aie_columns_2-num_channels_2-tile_size_512-weighted_False ✅ 5/5 317.38 0.03 n/a
input_length_2048-num_aie_columns_2-tile_size_1024 ✅ 15/15 319.29 0.04 n/a
input_length_2048-num_aie_columns_4-num_channels_1-tile_size_512 ✅ 25/25 284.04 0.03 n/a
input_length_2048-num_aie_columns_4-num_channels_1-tile_size_512-group_size_32 ✅ 5/5 484.46 0.01 n/a
input_length_2048-num_aie_columns_4-num_channels_1-tile_size_512-weighted_False ✅ 5/5 282.68 0.03 n/a
input_length_2048-num_aie_columns_4-num_channels_1-tile_size_512-weighted_True ✅ 5/5 259.28 0.04 n/a
input_length_2048-num_aie_columns_4-num_channels_2-tile_size_256 ✅ 10/10 362.78 0.03 n/a
input_length_2048-num_aie_columns_4-num_channels_2-tile_size_256-group_size_32 ✅ 5/5 293.52 0.02 n/a
input_length_2048-num_aie_columns_4-num_channels_2-tile_size_256-weighted_False ✅ 5/5 316.78 0.03 n/a
input_length_2048-num_aie_columns_4-num_channels_2-tile_size_512-scalar_factor_3.0 ✅ 5/5 268.00 0.05 n/a
input_length_2048-num_aie_columns_4-tile_size_512 ✅ 15/15 305.99 0.04 n/a
input_length_2048-num_aie_columns_8-num_channels_1-tile_size_256 ✅ 25/25 320.91 0.03 n/a
input_length_2048-num_aie_columns_8-num_channels_1-tile_size_256-group_size_32 ✅ 5/5 286.44 0.02 n/a
input_length_2048-num_aie_columns_8-num_channels_1-tile_size_256-weighted_False ✅ 5/5 349.34 0.03 n/a
input_length_2048-num_aie_columns_8-num_channels_1-tile_size_256-weighted_True ✅ 5/5 381.26 0.03 n/a
input_length_2048-num_aie_columns_8-num_channels_2-tile_size_128 ✅ 10/10 385.37 0.03 n/a
input_length_2048-num_aie_columns_8-num_channels_2-tile_size_128-group_size_32 ✅ 5/5 376.92 0.02 n/a
input_length_2048-num_aie_columns_8-num_channels_2-tile_size_128-weighted_False ✅ 5/5 417.18 0.03 n/a
input_length_2048-num_aie_columns_8-num_channels_2-tile_size_256-scalar_factor_3.0 ✅ 5/5 367.94 0.04 n/a
input_length_2048-num_aie_columns_8-tile_size_256 ✅ 15/15 319.69 0.04 n/a
input_length_2048-num_cores_1-num_channels_1-bypass_False-tile_size_2048 ✅ 5/5 391.58 0.03 n/a
input_length_2048-num_cores_16-num_channels_2-bypass_False-tile_size_128 ✅ 5/5 417.06 0.02 n/a
input_length_2048-num_cores_2-num_channels_1-bypass_False-tile_size_1024 ✅ 5/5 281.78 0.03 n/a
input_length_2048-num_cores_2-num_channels_2-bypass_False-tile_size_1024 ✅ 5/5 266.60 0.03 n/a
input_length_2048-num_cores_4-num_channels_1-bypass_False-tile_size_512 ✅ 5/5 262.62 0.03 n/a
input_length_2048-num_cores_4-num_channels_2-bypass_False-tile_size_512 ✅ 5/5 325.20 0.03 n/a
input_length_2048-num_cores_8-num_channels_1-bypass_False-tile_size_256 ✅ 5/5 336.50 0.03 n/a
input_length_2048-num_cores_8-num_channels_2-bypass_False-tile_size_256 ✅ 5/5 286.54 0.03 n/a
input_length_32768-num_aie_columns_2-num_channels_2-tile_size_1024 ✅ 5/5 444.76 0.35 n/a
input_length_32768-num_aie_columns_2-num_channels_2-tile_size_512 ✅ 5/5 528.36 0.29 n/a
input_length_32768-num_aie_columns_4-num_channels_4-tile_size_2048 ✅ 5/5 482.86 0.30 n/a
rows_32-cols_512-angle_rows_32-aie_columns_1-method_type_0 ✅ 5/5 288.36 0.36 n/a
rows_32-cols_512-angle_rows_32-aie_columns_2-method_type_0 ✅ 5/5 310.88 0.36 n/a
rows_32-cols_512-angle_rows_32-aie_columns_8-method_type_0 ✅ 5/5 390.38 0.34 n/a
rows_32-cols_512-angle_rows_8-aie_columns_1-method_type_0 ✅ 5/5 297.56 0.25 n/a
rows_32-cols_512-angle_rows_8-aie_columns_2-method_type_0 ✅ 5/5 319.66 0.23 n/a
rows_32-cols_512-angle_rows_8-aie_columns_8-method_type_0 ✅ 5/5 393.10 0.21 n/a
seq_len_16384-dim_64-num_heads_1-num_pipelines_8-num_kv_heads_0 ❌ 0/5 n/a n/a n/a
seq_len_256-embedding_dim_2048-hidden_dim_2048-prio_accuracy_False ✅ 5/5 n/a n/a n/a
📈 Trends (vs main branch) for Small Benchmark/Test Suite

7d76781 (2026_03_25_21_59_28)

IRONCLAD Trends

M_128-K_128-num_aie_columns_1-tile_size_input_32-tile_size_output_128

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
7d76781 — 2026-03-25 21:49:380.14 (n/a)0.11 (n/a)0.13 (n/a)0.04 (n/a)0.04 (n/a)0.14 (n/a)0.11 (n/a)0.13 (n/a)0.04 (n/a)0.04 (n/a)

M_1792-K_896-N_1152-num_aie_columns_8-b_col_maj_False-c_col_maj_True-m_64-k_32-n_48-trace_size_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
7d76781 — 2026-03-25 21:49:384.66 (n/a)4.09 (n/a)4.27 (n/a)2.94 (n/a)0.67 (n/a)3196.10 (n/a)2359.30 (n/a)2203.00 (n/a)2017.10 (n/a)477.73 (n/a)1834.02 (n/a)1610.76 (n/a)1679.27 (n/a)1157.48 (n/a)265.22 (n/a)

M_192-K_384-N_64-num_aie_columns_4-b_col_maj_False-c_col_maj_False-m_48-k_96-n_16-trace_size_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
7d76781 — 2026-03-25 21:49:380.81 (n/a)0.57 (n/a)0.53 (n/a)0.23 (n/a)0.24 (n/a)964.30 (n/a)477.94 (n/a)420.40 (n/a)272.40 (n/a)284.43 (n/a)34.64 (n/a)24.36 (n/a)22.45 (n/a)9.79 (n/a)10.42 (n/a)

M_192-K_384-N_64-num_aie_columns_4-b_col_maj_True-c_col_maj_True-m_48-k_96-n_16-trace_size_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
7d76781 — 2026-03-25 21:49:380.80 (n/a)0.63 (n/a)0.64 (n/a)0.31 (n/a)0.19 (n/a)719.60 (n/a)396.96 (n/a)346.00 (n/a)276.30 (n/a)183.31 (n/a)34.16 (n/a)26.75 (n/a)27.27 (n/a)13.12 (n/a)8.25 (n/a)

M_2048-K_2048-N_2048-num_aie_columns_1-b_col_maj_False-c_col_maj_False-m_64-k_64-n_64-trace_size_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
7d76781 — 2026-03-25 21:49:380.51 (n/a)0.50 (n/a)0.50 (n/a)0.50 (n/a)0.00 (n/a)50023.20 (n/a)49926.42 (n/a)50014.70 (n/a)49609.80 (n/a)178.48 (n/a)346.30 (n/a)344.11 (n/a)343.50 (n/a)343.44 (n/a)1.24 (n/a)

M_2048-K_2048-N_2048-num_aie_columns_2-b_col_maj_True-c_col_maj_False-m_64-k_64-n_64-trace_size_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
7d76781 — 2026-03-25 21:49:380.87 (n/a)0.86 (n/a)0.86 (n/a)0.85 (n/a)0.01 (n/a)29549.80 (n/a)29341.14 (n/a)29389.70 (n/a)28950.70 (n/a)235.27 (n/a)593.42 (n/a)585.55 (n/a)584.55 (n/a)581.39 (n/a)4.73 (n/a)

M_2048-K_2048-N_2048-num_aie_columns_8-b_col_maj_True-c_col_maj_True-m_64-k_64-n_64-trace_size_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
7d76781 — 2026-03-25 21:49:383.37 (n/a)3.06 (n/a)3.17 (n/a)2.40 (n/a)0.39 (n/a)10475.20 (n/a)8358.12 (n/a)7942.50 (n/a)7461.20 (n/a)1230.54 (n/a)2302.56 (n/a)2086.76 (n/a)2163.03 (n/a)1640.05 (n/a)266.54 (n/a)

M_2048-K_8192-num_aie_columns_1-tile_size_input_1-tile_size_output_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
7d76781 — 2026-03-25 21:49:3812.45 (n/a)11.22 (n/a)12.25 (n/a)8.62 (n/a)1.67 (n/a)12.44 (n/a)11.22 (n/a)12.25 (n/a)8.61 (n/a)1.67 (n/a)

M_2048-K_8192-num_aie_columns_2-tile_size_input_1-tile_size_output_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
7d76781 — 2026-03-25 21:49:3821.66 (n/a)19.78 (n/a)21.45 (n/a)13.66 (n/a)3.44 (n/a)21.65 (n/a)19.77 (n/a)21.43 (n/a)13.65 (n/a)3.44 (n/a)

M_2048-K_8192-num_aie_columns_4-tile_size_input_1-tile_size_output_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
7d76781 — 2026-03-25 21:49:3833.04 (n/a)31.19 (n/a)31.73 (n/a)26.63 (n/a)2.64 (n/a)33.02 (n/a)31.17 (n/a)31.71 (n/a)26.61 (n/a)2.64 (n/a)

M_2048-K_8192-num_aie_columns_8-tile_size_input_1-tile_size_output_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
7d76781 — 2026-03-25 21:49:3836.99 (n/a)33.19 (n/a)34.16 (n/a)27.19 (n/a)3.62 (n/a)36.97 (n/a)33.17 (n/a)34.14 (n/a)27.17 (n/a)3.62 (n/a)

M_2048-N_64-aie_columns_1-channels_1-m_64-n_64-s_8

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
7d76781 — 2026-03-25 21:49:381.93 (n/a)1.45 (n/a)1.60 (n/a)0.71 (n/a)0.46 (n/a)739.20 (n/a)408.46 (n/a)327.30 (n/a)272.00 (n/a)189.06 (n/a)

M_2048-N_64-aie_columns_1-channels_2-m_64-n_64-s_8

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
7d76781 — 2026-03-25 21:49:382.06 (n/a)1.79 (n/a)1.77 (n/a)1.36 (n/a)0.28 (n/a)386.30 (n/a)298.86 (n/a)297.00 (n/a)255.10 (n/a)53.11 (n/a)

M_384-K_1536-N_1792-num_aie_columns_4-b_col_maj_True-c_col_maj_False-m_32-k_48-n_64-trace_size_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
7d76781 — 2026-03-25 21:49:383.89 (n/a)3.68 (n/a)3.88 (n/a)3.18 (n/a)0.32 (n/a)2537.40 (n/a)2204.96 (n/a)2076.20 (n/a)2071.50 (n/a)203.96 (n/a)1020.48 (n/a)964.83 (n/a)1018.18 (n/a)833.10 (n/a)82.78 (n/a)

M_8192-K_2048-num_aie_columns_1-tile_size_input_4-tile_size_output_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
7d76781 — 2026-03-25 21:49:3812.50 (n/a)11.89 (n/a)12.24 (n/a)11.01 (n/a)0.63 (n/a)12.49 (n/a)11.88 (n/a)12.23 (n/a)11.00 (n/a)0.63 (n/a)

M_8192-K_2048-num_aie_columns_2-tile_size_input_4-tile_size_output_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
7d76781 — 2026-03-25 21:49:3821.70 (n/a)19.42 (n/a)20.74 (n/a)16.07 (n/a)2.64 (n/a)21.68 (n/a)19.41 (n/a)20.73 (n/a)16.06 (n/a)2.64 (n/a)

M_8192-K_2048-num_aie_columns_4-tile_size_input_4-tile_size_output_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
7d76781 — 2026-03-25 21:49:3832.97 (n/a)31.91 (n/a)31.64 (n/a)31.03 (n/a)0.81 (n/a)32.95 (n/a)31.89 (n/a)31.62 (n/a)31.01 (n/a)0.81 (n/a)

M_8192-K_2048-num_aie_columns_8-tile_size_input_4-tile_size_output_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
7d76781 — 2026-03-25 21:49:3835.50 (n/a)33.35 (n/a)34.49 (n/a)27.86 (n/a)3.18 (n/a)35.48 (n/a)33.33 (n/a)34.46 (n/a)27.84 (n/a)3.17 (n/a)

M_896-K_1792-N_640-num_aie_columns_8-b_col_maj_False-c_col_maj_True-m_32-k_64-n_80-trace_size_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
7d76781 — 2026-03-25 21:49:385.94 (n/a)5.43 (n/a)5.60 (n/a)4.32 (n/a)0.64 (n/a)1539.90 (n/a)1241.36 (n/a)1187.30 (n/a)1120.10 (n/a)169.65 (n/a)1834.88 (n/a)1677.11 (n/a)1730.95 (n/a)1334.60 (n/a)197.09 (n/a)

axpy_1_cols_2_channels_2048_tile_2048_3.0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.08 (+17.48%)0.07 (+11.67%)0.07 (+26.71%)0.05 (+0.80%)0.01 (+108.36%)239.80 (-0.83%)190.62 (-7.91%)166.90 (-21.09%)155.40 (-14.90%)42.89 (+79.10%)
dc66da7 — 2026-03-25 19:23:450.07 (n/a)0.06 (n/a)0.06 (n/a)0.05 (n/a)0.01 (n/a)241.80 (n/a)207.00 (n/a)211.50 (n/a)182.60 (n/a)23.95 (n/a)

axpy_1_cols_2_channels_2048_tile_2048_3.0_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.16 (-3.31%)0.14 (-1.29%)0.15 (+16.18%)0.10 (-6.80%)0.03 (+3.89%)125.70 (+7.25%)93.56 (+2.12%)83.50 (-13.92%)75.00 (+3.45%)22.38 (+18.09%)
84d3478 — 2026-02-17 23:25:130.17 (n/a)0.14 (n/a)0.13 (n/a)0.10 (n/a)0.03 (n/a)117.20 (n/a)91.62 (n/a)97.00 (n/a)72.50 (n/a)18.95 (n/a)

axpy_2_cols_2_channels_2048_tile_1024_3.0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.09 (+14.45%)0.07 (-0.78%)0.07 (-0.40%)0.05 (-7.87%)0.01 (+68.50%)233.40 (+8.51%)188.20 (+2.47%)179.80 (+0.39%)143.50 (-12.66%)34.06 (+61.09%)
dc66da7 — 2026-03-25 19:23:450.07 (n/a)0.07 (n/a)0.07 (n/a)0.06 (n/a)0.01 (n/a)215.10 (n/a)183.66 (n/a)179.10 (n/a)164.30 (n/a)21.14 (n/a)

axpy_2_cols_2_channels_2048_tile_1024_3.0_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.13 (-0.73%)0.11 (-7.43%)0.11 (-8.63%)0.07 (-27.71%)0.02 (+50.39%)182.70 (+38.30%)120.86 (+11.97%)110.80 (+9.38%)96.10 (+0.73%)35.16 (+122.88%)
84d3478 — 2026-02-17 23:25:130.13 (n/a)0.12 (n/a)0.12 (n/a)0.09 (n/a)0.02 (n/a)132.10 (n/a)107.94 (n/a)101.30 (n/a)95.40 (n/a)15.78 (n/a)

axpy_4_cols_2_channels_2048_tile_512_3.0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.10 (+6.67%)0.06 (-1.56%)0.06 (-2.24%)0.05 (+3.60%)0.02 (+20.83%)251.50 (-3.49%)200.78 (+3.11%)191.70 (+2.29%)128.70 (-6.26%)50.37 (+11.30%)
dc66da7 — 2026-03-25 19:23:450.09 (n/a)0.07 (n/a)0.07 (n/a)0.05 (n/a)0.02 (n/a)260.60 (n/a)194.72 (n/a)187.40 (n/a)137.30 (n/a)45.26 (n/a)

axpy_4_cols_2_channels_2048_tile_512_3.0_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.13 (-10.91%)0.11 (-7.68%)0.12 (-0.73%)0.06 (-17.94%)0.03 (+0.99%)194.10 (+21.85%)124.52 (+10.29%)103.30 (+0.68%)97.20 (+12.24%)40.32 (+39.15%)
84d3478 — 2026-02-17 23:25:130.14 (n/a)0.11 (n/a)0.12 (n/a)0.08 (n/a)0.03 (n/a)159.30 (n/a)112.90 (n/a)102.60 (n/a)86.60 (n/a)28.97 (n/a)

axpy_8_cols_2_channels_2048_tile_256_3.0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.08 (+1.42%)0.06 (+9.86%)0.06 (+18.71%)0.03 (+13.51%)0.02 (-10.62%)376.00 (-11.90%)225.02 (-11.61%)196.20 (-15.76%)151.40 (-1.37%)87.15 (-17.99%)
dc66da7 — 2026-03-25 19:23:450.08 (n/a)0.05 (n/a)0.05 (n/a)0.03 (n/a)0.02 (n/a)426.80 (n/a)254.58 (n/a)232.90 (n/a)153.50 (n/a)106.27 (n/a)

axpy_8_cols_2_channels_2048_tile_256_3.0_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.13 (+0.01%)0.11 (+6.10%)0.11 (+0.22%)0.08 (-0.33%)0.02 (-9.83%)156.30 (+0.32%)115.26 (-6.40%)112.70 (-0.18%)95.50 (+0.00%)24.34 (-10.22%)
84d3478 — 2026-02-17 23:25:130.13 (n/a)0.10 (n/a)0.11 (n/a)0.08 (n/a)0.02 (n/a)155.80 (n/a)123.14 (n/a)112.90 (n/a)95.50 (n/a)27.11 (n/a)

dequant_1_cols_1_channels_2048_tile_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.04 (+17.85%)0.03 (+2.55%)0.03 (+1.65%)0.03 (-2.18%)0.00 (+114.37%)207.70 (+2.26%)178.64 (-0.96%)173.60 (-1.59%)141.60 (-15.16%)28.01 (+90.63%)
dc66da7 — 2026-03-25 19:23:450.03 (n/a)0.03 (n/a)0.03 (n/a)0.03 (n/a)0.00 (n/a)203.10 (n/a)180.38 (n/a)176.40 (n/a)166.90 (n/a)14.69 (n/a)

dequant_1_cols_1_channels_2048_tile_2048_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.05 (-13.00%)0.04 (-0.87%)0.04 (+1.16%)0.03 (-3.17%)0.01 (-31.33%)158.40 (+3.26%)128.12 (-0.19%)123.80 (-1.12%)110.40 (+14.88%)18.36 (-15.65%)
84d3478 — 2026-02-17 23:25:130.05 (n/a)0.04 (n/a)0.04 (n/a)0.03 (n/a)0.01 (n/a)153.40 (n/a)128.36 (n/a)125.20 (n/a)96.10 (n/a)21.77 (n/a)

dequant_1_cols_2_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.04 (+13.44%)0.03 (+8.77%)0.03 (+14.89%)0.02 (-9.15%)0.01 (+54.39%)252.70 (+10.06%)182.14 (-6.08%)174.50 (-12.97%)145.40 (-11.88%)42.38 (+56.31%)
dc66da7 — 2026-03-25 19:23:450.03 (n/a)0.03 (n/a)0.03 (n/a)0.02 (n/a)0.00 (n/a)229.60 (n/a)193.94 (n/a)200.50 (n/a)165.00 (n/a)27.11 (n/a)

dequant_1_cols_2_channels_2048_tile_1024_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.06 (-3.74%)0.05 (-12.63%)0.05 (-5.21%)0.03 (-33.47%)0.01 (+52.60%)162.70 (+50.37%)110.84 (+18.98%)98.80 (+5.44%)81.50 (+3.95%)31.02 (+149.23%)
84d3478 — 2026-02-17 23:25:130.07 (n/a)0.06 (n/a)0.06 (n/a)0.05 (n/a)0.01 (n/a)108.20 (n/a)93.16 (n/a)93.70 (n/a)78.40 (n/a)12.45 (n/a)

dequant_2_cols_1_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.05 (+21.89%)0.03 (+20.89%)0.03 (+21.41%)0.02 (+8.07%)0.01 (+32.97%)215.80 (-7.46%)168.56 (-16.36%)172.60 (-17.65%)113.20 (-17.97%)37.61 (+1.89%)
dc66da7 — 2026-03-25 19:23:450.04 (n/a)0.03 (n/a)0.03 (n/a)0.02 (n/a)0.01 (n/a)233.20 (n/a)201.52 (n/a)209.60 (n/a)138.00 (n/a)36.92 (n/a)

dequant_2_cols_1_channels_2048_tile_1024_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.06 (+8.90%)0.05 (+1.31%)0.05 (-1.85%)0.03 (+6.35%)0.01 (+7.11%)152.30 (-5.93%)115.08 (-1.20%)100.80 (+1.92%)88.10 (-8.23%)27.20 (-4.42%)
84d3478 — 2026-02-17 23:25:130.05 (n/a)0.05 (n/a)0.05 (n/a)0.03 (n/a)0.01 (n/a)161.90 (n/a)116.48 (n/a)98.90 (n/a)96.00 (n/a)28.46 (n/a)

dequant_2_cols_2_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.05 (+49.89%)0.03 (+24.89%)0.03 (+13.74%)0.02 (+15.73%)0.01 (+106.91%)219.80 (-13.60%)173.04 (-17.40%)185.70 (-12.07%)112.30 (-33.27%)41.15 (+16.74%)
dc66da7 — 2026-03-25 19:23:450.03 (n/a)0.03 (n/a)0.02 (n/a)0.02 (n/a)0.00 (n/a)254.40 (n/a)209.50 (n/a)211.20 (n/a)168.30 (n/a)35.25 (n/a)

dequant_2_cols_2_channels_2048_tile_512_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.07 (+27.96%)0.05 (+1.57%)0.04 (-18.06%)0.03 (+10.23%)0.01 (+22.79%)164.70 (-9.31%)119.12 (-1.39%)120.20 (+22.03%)75.60 (-21.82%)31.61 (-13.80%)
84d3478 — 2026-02-17 23:25:130.05 (n/a)0.05 (n/a)0.05 (n/a)0.03 (n/a)0.01 (n/a)181.60 (n/a)120.80 (n/a)98.50 (n/a)96.70 (n/a)36.67 (n/a)

dequant_4_cols_1_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.04 (+56.54%)0.03 (+15.78%)0.03 (+9.01%)0.02 (-5.58%)0.01 (+304.26%)243.40 (+5.92%)186.06 (-8.98%)182.60 (-8.29%)118.50 (-36.12%)46.36 (+164.07%)
dc66da7 — 2026-03-25 19:23:450.03 (n/a)0.03 (n/a)0.03 (n/a)0.02 (n/a)0.00 (n/a)229.80 (n/a)204.42 (n/a)199.10 (n/a)185.50 (n/a)17.56 (n/a)

dequant_4_cols_1_channels_2048_tile_512_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.05 (-2.55%)0.04 (-12.79%)0.05 (-16.25%)0.03 (-12.85%)0.01 (-1.20%)164.60 (+14.78%)122.14 (+15.14%)115.90 (+19.48%)97.50 (+2.63%)25.08 (+19.84%)
84d3478 — 2026-02-17 23:25:130.06 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.01 (n/a)143.40 (n/a)106.08 (n/a)97.00 (n/a)95.00 (n/a)20.93 (n/a)

dequant_4_cols_2_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.04 (+28.72%)0.03 (+2.89%)0.02 (-6.84%)0.02 (+6.27%)0.01 (+95.59%)243.20 (-5.88%)208.42 (-0.42%)222.90 (+7.32%)141.10 (-22.34%)41.08 (+38.16%)
dc66da7 — 2026-03-25 19:23:450.03 (n/a)0.03 (n/a)0.03 (n/a)0.02 (n/a)0.00 (n/a)258.40 (n/a)209.30 (n/a)207.70 (n/a)181.70 (n/a)29.74 (n/a)

dequant_4_cols_2_channels_2048_tile_256_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.06 (+20.83%)0.05 (+8.21%)0.04 (+1.42%)0.03 (+17.39%)0.01 (+14.84%)166.40 (-14.80%)122.34 (-7.97%)127.70 (-1.39%)84.60 (-17.22%)30.47 (-19.21%)
84d3478 — 2026-02-17 23:25:130.05 (n/a)0.04 (n/a)0.04 (n/a)0.03 (n/a)0.01 (n/a)195.30 (n/a)132.94 (n/a)129.50 (n/a)102.20 (n/a)37.71 (n/a)

dequant_8_cols_1_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.04 (-1.52%)0.03 (+2.17%)0.03 (+18.60%)0.02 (-11.90%)0.01 (+4.42%)268.50 (+13.53%)205.32 (-1.14%)191.00 (-15.67%)141.60 (+1.51%)49.39 (+23.46%)
dc66da7 — 2026-03-25 19:23:450.04 (n/a)0.03 (n/a)0.02 (n/a)0.02 (n/a)0.01 (n/a)236.50 (n/a)207.68 (n/a)226.50 (n/a)139.50 (n/a)40.00 (n/a)

dequant_8_cols_1_channels_2048_tile_256_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.05 (-25.19%)0.04 (-21.02%)0.04 (-19.70%)0.04 (-19.76%)0.00 (-37.92%)142.70 (+24.63%)126.44 (+26.24%)125.70 (+24.46%)115.50 (+33.68%)10.61 (+3.97%)
84d3478 — 2026-02-17 23:25:130.06 (n/a)0.05 (n/a)0.05 (n/a)0.05 (n/a)0.01 (n/a)114.50 (n/a)100.16 (n/a)101.00 (n/a)86.40 (n/a)10.20 (n/a)

dequant_8_cols_2_channels_2048_tile_128

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.02 (+5.64%)0.02 (+1.26%)0.02 (-1.25%)0.02 (-3.87%)0.00 (+124.04%)258.90 (+4.02%)239.08 (-0.87%)247.20 (+1.27%)215.50 (-5.36%)18.28 (+120.74%)
dc66da7 — 2026-03-25 19:23:450.02 (n/a)0.02 (n/a)0.02 (n/a)0.02 (n/a)0.00 (n/a)248.90 (n/a)241.18 (n/a)244.10 (n/a)227.70 (n/a)8.28 (n/a)

dequant_8_cols_2_channels_2048_tile_128_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.04 (-26.69%)0.04 (-11.93%)0.04 (-16.99%)0.03 (+27.28%)0.01 (-62.60%)193.10 (-21.41%)147.68 (+2.96%)136.10 (+20.44%)135.40 (+36.49%)25.41 (-58.96%)
84d3478 — 2026-02-17 23:25:130.05 (n/a)0.04 (n/a)0.05 (n/a)0.02 (n/a)0.01 (n/a)245.70 (n/a)143.44 (n/a)113.00 (n/a)99.20 (n/a)61.90 (n/a)

eltwise_add_1_cols_2_channels_2048_tile_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.07 (-32.66%)0.06 (-13.87%)0.06 (-9.19%)0.04 (-8.04%)0.01 (-49.50%)299.50 (+8.75%)223.84 (+11.90%)212.70 (+10.15%)187.60 (+48.42%)45.70 (-16.71%)
dc66da7 — 2026-03-25 19:23:450.10 (n/a)0.07 (n/a)0.06 (n/a)0.04 (n/a)0.02 (n/a)275.40 (n/a)200.04 (n/a)193.10 (n/a)126.40 (n/a)54.87 (n/a)

eltwise_add_2_cols_2_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.07 (-33.33%)0.06 (-18.97%)0.05 (-19.23%)0.05 (+18.13%)0.01 (-59.22%)270.80 (-15.37%)223.26 (+13.77%)229.40 (+23.80%)174.50 (+50.04%)37.51 (-50.17%)
dc66da7 — 2026-03-25 19:23:450.11 (n/a)0.07 (n/a)0.07 (n/a)0.04 (n/a)0.02 (n/a)320.00 (n/a)196.24 (n/a)185.30 (n/a)116.30 (n/a)75.29 (n/a)

eltwise_add_4_cols_2_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.07 (-8.97%)0.06 (-6.00%)0.06 (-2.02%)0.05 (-2.74%)0.01 (-14.89%)243.40 (+2.83%)208.60 (+6.01%)200.50 (+2.04%)178.10 (+9.87%)31.58 (-2.07%)
dc66da7 — 2026-03-25 19:23:450.08 (n/a)0.06 (n/a)0.06 (n/a)0.05 (n/a)0.01 (n/a)236.70 (n/a)196.78 (n/a)196.50 (n/a)162.10 (n/a)32.25 (n/a)

eltwise_add_8_cols_2_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.06 (-45.96%)0.05 (-24.77%)0.05 (-15.00%)0.05 (+4.78%)0.00 (-83.62%)259.60 (-4.56%)234.66 (+20.97%)224.10 (+17.64%)218.40 (+85.08%)18.77 (-71.47%)
dc66da7 — 2026-03-25 19:23:450.10 (n/a)0.07 (n/a)0.06 (n/a)0.05 (n/a)0.02 (n/a)272.00 (n/a)193.98 (n/a)190.50 (n/a)118.00 (n/a)65.77 (n/a)

eltwise_mul_1_cols_2_channels_2048_tile_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.11 (+3.13%)0.07 (+2.52%)0.06 (-1.87%)0.05 (-14.01%)0.02 (+21.62%)257.00 (+16.29%)184.10 (+0.67%)190.70 (+1.87%)114.40 (-3.05%)55.57 (+40.96%)
dc66da7 — 2026-03-25 19:23:450.10 (n/a)0.07 (n/a)0.07 (n/a)0.06 (n/a)0.02 (n/a)221.00 (n/a)182.88 (n/a)187.20 (n/a)118.00 (n/a)39.42 (n/a)

eltwise_mul_2_cols_2_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.09 (+2.43%)0.07 (+6.63%)0.08 (+30.25%)0.05 (-10.38%)0.02 (+9.28%)264.20 (+11.62%)182.10 (-4.96%)162.40 (-23.22%)137.30 (-2.42%)52.80 (+20.53%)
dc66da7 — 2026-03-25 19:23:450.09 (n/a)0.07 (n/a)0.06 (n/a)0.05 (n/a)0.02 (n/a)236.70 (n/a)191.60 (n/a)211.50 (n/a)140.70 (n/a)43.81 (n/a)

eltwise_mul_4_cols_2_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.10 (+22.79%)0.08 (+18.44%)0.09 (+34.73%)0.06 (-0.12%)0.01 (+158.99%)191.20 (+0.10%)154.44 (-13.56%)136.60 (-25.80%)126.60 (-18.53%)29.95 (+117.72%)
dc66da7 — 2026-03-25 19:23:450.08 (n/a)0.07 (n/a)0.07 (n/a)0.06 (n/a)0.01 (n/a)191.00 (n/a)178.66 (n/a)184.10 (n/a)155.40 (n/a)13.76 (n/a)

eltwise_mul_8_cols_2_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.08 (+34.81%)0.06 (+14.94%)0.06 (+6.87%)0.04 (-5.87%)0.01 (+158.26%)280.10 (+6.22%)210.00 (-9.80%)215.10 (-6.40%)150.70 (-25.80%)49.49 (+100.84%)
dc66da7 — 2026-03-25 19:23:450.06 (n/a)0.05 (n/a)0.05 (n/a)0.05 (n/a)0.01 (n/a)263.70 (n/a)232.82 (n/a)229.80 (n/a)203.10 (n/a)24.64 (n/a)

embedding_dim_2048-hidden_dim_2048

No metrics available.

gelu_1_cols_1_channels_2048_tile_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.05 (+9.18%)0.04 (-1.03%)0.04 (-2.51%)0.04 (-8.09%)0.01 (+114.58%)221.00 (+8.81%)195.42 (+1.97%)199.10 (+2.58%)163.60 (-8.40%)23.03 (+114.14%)
dc66da7 — 2026-03-25 19:23:450.05 (n/a)0.04 (n/a)0.04 (n/a)0.04 (n/a)0.00 (n/a)203.10 (n/a)191.64 (n/a)194.10 (n/a)178.60 (n/a)10.76 (n/a)

gelu_1_cols_2_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.04 (-26.60%)0.04 (-19.13%)0.04 (-18.64%)0.04 (-2.84%)0.00 (-66.68%)226.70 (+2.91%)199.16 (+20.50%)193.10 (+22.92%)186.40 (+36.16%)16.29 (-52.49%)
dc66da7 — 2026-03-25 19:23:450.06 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.01 (n/a)220.30 (n/a)165.28 (n/a)157.10 (n/a)136.90 (n/a)34.28 (n/a)

gelu_2_cols_1_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.06 (+2.08%)0.05 (+4.66%)0.04 (-1.75%)0.03 (-5.55%)0.01 (+12.71%)283.70 (+5.86%)191.18 (-3.05%)191.00 (+1.81%)137.60 (-1.99%)58.08 (+15.01%)
dc66da7 — 2026-03-25 19:23:450.06 (n/a)0.04 (n/a)0.04 (n/a)0.03 (n/a)0.01 (n/a)268.00 (n/a)197.20 (n/a)187.60 (n/a)140.40 (n/a)50.50 (n/a)

gelu_2_cols_2_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.07 (+2.20%)0.04 (-2.29%)0.04 (+1.04%)0.04 (+16.91%)0.01 (-4.76%)227.00 (-14.47%)194.56 (+0.69%)206.00 (-1.01%)119.40 (-2.21%)44.27 (-20.37%)
dc66da7 — 2026-03-25 19:23:450.07 (n/a)0.05 (n/a)0.04 (n/a)0.03 (n/a)0.01 (n/a)265.40 (n/a)193.22 (n/a)208.10 (n/a)122.10 (n/a)55.59 (n/a)

gelu_4_cols_1_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.07 (-9.75%)0.04 (-17.39%)0.04 (-20.48%)0.03 (+12.65%)0.01 (-29.41%)239.40 (-11.23%)196.08 (+15.06%)197.90 (+25.73%)124.70 (+10.84%)44.40 (-31.09%)
dc66da7 — 2026-03-25 19:23:450.07 (n/a)0.05 (n/a)0.05 (n/a)0.03 (n/a)0.02 (n/a)269.70 (n/a)170.42 (n/a)157.40 (n/a)112.50 (n/a)64.43 (n/a)

gelu_4_cols_2_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.04 (-6.29%)0.04 (+1.81%)0.04 (+0.94%)0.03 (-0.42%)0.00 (-22.18%)284.90 (+0.42%)233.36 (-2.39%)227.20 (-0.96%)204.80 (+6.72%)30.62 (-15.21%)
dc66da7 — 2026-03-25 19:23:450.04 (n/a)0.03 (n/a)0.04 (n/a)0.03 (n/a)0.01 (n/a)283.70 (n/a)239.08 (n/a)229.40 (n/a)191.90 (n/a)36.12 (n/a)

gelu_8_cols_1_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.05 (-18.24%)0.04 (-13.70%)0.04 (-4.91%)0.03 (-15.26%)0.01 (-28.46%)285.90 (+18.04%)223.38 (+14.20%)223.20 (+5.18%)156.90 (+22.29%)47.35 (+1.82%)
dc66da7 — 2026-03-25 19:23:450.06 (n/a)0.04 (n/a)0.04 (n/a)0.03 (n/a)0.01 (n/a)242.20 (n/a)195.60 (n/a)212.20 (n/a)128.30 (n/a)46.50 (n/a)

gelu_8_cols_2_channels_2048_tile_128

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.05 (+14.38%)0.04 (+10.49%)0.03 (+3.25%)0.03 (+13.03%)0.01 (+26.31%)305.70 (-11.52%)235.82 (-8.98%)241.80 (-3.12%)177.40 (-12.57%)49.18 (-6.09%)
dc66da7 — 2026-03-25 19:23:450.04 (n/a)0.03 (n/a)0.03 (n/a)0.02 (n/a)0.01 (n/a)345.50 (n/a)259.10 (n/a)249.60 (n/a)202.90 (n/a)52.37 (n/a)

gemm_1792x896x1152_64x32x48_8cols_ccolmaj

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
8b6d0f4 — 2026-03-25 20:46:394.89 (+9.71%)4.07 (+2.20%)4.10 (+2.03%)3.40 (+0.07%)0.55 (+44.15%)2767.30 (-0.07%)2344.32 (-1.52%)2291.20 (-1.99%)1922.40 (-8.85%)308.92 (+28.54%)1924.38 (+9.71%)1600.58 (+2.20%)1614.60 (+2.03%)1336.81 (+0.07%)215.84 (+44.15%)
dc66da7 — 2026-03-25 19:23:454.46 (n/a)3.98 (n/a)4.02 (n/a)3.40 (n/a)0.38 (n/a)2769.20 (n/a)2380.42 (n/a)2337.70 (n/a)2109.10 (n/a)240.33 (n/a)1754.05 (n/a)1566.13 (n/a)1582.49 (n/a)1335.88 (n/a)149.73 (n/a)

gemm_192x384x64_48x96x16_4cols

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
8b6d0f4 — 2026-03-25 20:46:391.35 (+17.71%)0.77 (-13.89%)0.65 (-34.22%)0.56 (+7.33%)0.32 (+33.14%)393.60 (-6.84%)315.66 (+18.48%)341.40 (+52.00%)164.30 (-15.05%)87.82 (-5.13%)57.45 (+17.71%)32.93 (-13.89%)27.64 (-34.22%)23.97 (+7.33%)13.81 (+33.14%)
dc66da7 — 2026-03-25 19:23:451.14 (n/a)0.90 (n/a)0.98 (n/a)0.52 (n/a)0.24 (n/a)422.50 (n/a)266.42 (n/a)224.60 (n/a)193.40 (n/a)92.57 (n/a)48.81 (n/a)38.25 (n/a)42.02 (n/a)22.34 (n/a)10.37 (n/a)

gemm_192x384x64_48x96x16_4cols_bcolmaj_ccolmaj

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
8b6d0f4 — 2026-03-25 20:46:391.18 (+31.18%)0.79 (-0.76%)0.72 (-17.59%)0.64 (+26.96%)0.22 (+35.66%)343.10 (-21.24%)294.50 (+0.95%)309.00 (+21.37%)188.10 (-23.78%)62.26 (-23.01%)50.17 (+31.18%)33.62 (-0.76%)30.54 (-17.59%)27.51 (+26.96%)9.40 (+35.66%)
dc66da7 — 2026-03-25 19:23:450.90 (n/a)0.79 (n/a)0.87 (n/a)0.51 (n/a)0.16 (n/a)435.60 (n/a)291.72 (n/a)254.60 (n/a)246.80 (n/a)80.87 (n/a)38.24 (n/a)33.88 (n/a)37.06 (n/a)21.67 (n/a)6.93 (n/a)

gemm_2048x2048x2048_64x64x32_8_cols_0_bcolmaj_0_ccolmaj_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)GFLOP/s (max)GFLOP/s (mean)GFLOP/s (median)GFLOP/s (min)GFLOP/s (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
3a053d3 — 2025-11-26 01:12:163.72 (-0.11%)3.55 (-2.75%)3.56 (-2.75%)3.36 (-5.33%)0.12 (+76.10%)2725.23 (-0.11%)2602.52 (-2.75%)2612.11 (-2.75%)2459.89 (-5.33%)86.27 (+76.10%)6984.00 (+5.63%)6608.60 (+2.91%)6577.00 (+2.83%)6304.00 (+0.11%)221.87 (+87.15%)
9e98049 — 2025-11-25 20:50:553.72 (n/a)3.65 (n/a)3.66 (n/a)3.54 (n/a)0.07 (n/a)2728.26 (n/a)2676.06 (n/a)2686.03 (n/a)2598.29 (n/a)48.99 (n/a)6612.00 (n/a)6422.00 (n/a)6396.00 (n/a)6297.00 (n/a)118.55 (n/a)

gemm_2048x2048x2048_64x64x32_8_cols_0_bcolmaj_1_ccolmaj_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)GFLOP/s (max)GFLOP/s (mean)GFLOP/s (median)GFLOP/s (min)GFLOP/s (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
3a053d3 — 2025-11-26 01:12:163.40 (-3.14%)3.38 (-0.82%)3.38 (-0.82%)3.35 (+4.93%)0.02 (-84.51%)2493.81 (-3.14%)2475.27 (-0.82%)2479.06 (-0.82%)2453.92 (+4.93%)13.33 (-84.51%)7001.00 (-4.70%)6940.80 (+0.71%)6930.00 (+0.83%)6889.00 (+3.24%)37.43 (-84.76%)
9e98049 — 2025-11-25 20:50:553.51 (n/a)3.40 (n/a)3.41 (n/a)3.19 (n/a)0.12 (n/a)2574.53 (n/a)2495.72 (n/a)2499.62 (n/a)2338.67 (n/a)86.02 (n/a)7346.00 (n/a)6892.20 (n/a)6873.00 (n/a)6673.00 (n/a)245.57 (n/a)

gemm_2048x2048x2048_64x64x32_8_cols_1_bcolmaj_0_ccolmaj_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)GFLOP/s (max)GFLOP/s (mean)GFLOP/s (median)GFLOP/s (min)GFLOP/s (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
3a053d3 — 2025-11-26 01:12:163.02 (+0.52%)2.87 (-2.05%)2.86 (-2.24%)2.70 (-6.16%)0.11 (+131.66%)2216.19 (+0.52%)2104.44 (-2.05%)2094.08 (-2.24%)1980.84 (-6.16%)80.25 (+131.67%)8673.00 (+6.56%)8175.60 (+2.21%)8204.00 (+2.29%)7752.00 (-0.51%)314.03 (+145.36%)
9e98049 — 2025-11-25 20:50:553.01 (n/a)2.93 (n/a)2.92 (n/a)2.88 (n/a)0.05 (n/a)2204.81 (n/a)2148.41 (n/a)2142.13 (n/a)2110.81 (n/a)34.64 (n/a)8139.00 (n/a)7998.60 (n/a)8020.00 (n/a)7792.00 (n/a)127.99 (n/a)

gemm_2048x2048x2048_64x64x64_1cols

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
8b6d0f4 — 2026-03-25 20:46:390.51 (+0.13%)0.51 (+0.07%)0.51 (+0.04%)0.51 (+0.11%)0.00 (+8.62%)49296.10 (-0.11%)49206.24 (-0.07%)49283.30 (-0.04%)48878.90 (-0.13%)183.15 (+8.40%)351.48 (+0.13%)349.14 (+0.07%)348.59 (+0.04%)348.50 (+0.11%)1.31 (+8.62%)
dc66da7 — 2026-03-25 19:23:450.51 (n/a)0.51 (n/a)0.51 (n/a)0.51 (n/a)0.00 (n/a)49348.10 (n/a)49241.40 (n/a)49302.10 (n/a)48941.60 (n/a)168.96 (n/a)351.03 (n/a)348.89 (n/a)348.46 (n/a)348.14 (n/a)1.20 (n/a)

gemm_2048x2048x2048_64x64x64_2_cols_0_bcolmaj_0_ccolmaj_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
130b6ea — 2025-12-05 21:15:491.04 (+0.94%)1.03 (+0.23%)1.03 (+0.09%)1.02 (+0.07%)0.00 (+197.51%)24576.00 (-0.07%)24485.82 (-0.23%)24530.40 (-0.09%)24266.40 (-0.94%)111.26 (+194.35%)
0a6c11c — 2025-12-04 00:39:101.03 (n/a)1.03 (n/a)1.03 (n/a)1.02 (n/a)0.00 (n/a)24594.30 (n/a)24542.14 (n/a)24551.90 (n/a)24495.50 (n/a)37.80 (n/a)

gemm_2048x2048x2048_64x64x64_2_cols_0_bcolmaj_0_ccolmaj_0_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
a4b6ffe — 2026-01-05 19:24:051.04 (+0.28%)1.03 (-0.15%)1.03 (+0.05%)1.02 (-0.52%)0.01 (+42.72%)24719.80 (+0.52%)24512.42 (+0.16%)24529.20 (-0.05%)24092.60 (-0.28%)252.92 (+43.15%)713.08 (+0.28%)700.92 (-0.15%)700.38 (+0.05%)694.98 (-0.52%)7.30 (+42.71%)
50ad37c — 2025-12-29 20:35:531.04 (n/a)1.03 (n/a)1.03 (n/a)1.02 (n/a)0.01 (n/a)24592.30 (n/a)24473.70 (n/a)24540.90 (n/a)24160.80 (n/a)176.68 (n/a)711.06 (n/a)702.00 (n/a)700.05 (n/a)698.59 (n/a)5.12 (n/a)

gemm_2048x2048x2048_64x64x64_2_cols_0_bcolmaj_1_ccolmaj_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
130b6ea — 2025-12-05 21:15:490.99 (+1.69%)0.97 (+0.29%)0.97 (-0.04%)0.97 (-0.15%)0.01 (+434.92%)26025.30 (+0.15%)25839.34 (-0.29%)25916.70 (+0.04%)25446.00 (-1.67%)202.05 (+425.44%)
0a6c11c — 2025-12-04 00:39:100.97 (n/a)0.97 (n/a)0.97 (n/a)0.97 (n/a)0.00 (n/a)25986.90 (n/a)25913.52 (n/a)25905.70 (n/a)25877.00 (n/a)38.45 (n/a)

gemm_2048x2048x2048_64x64x64_2_cols_0_bcolmaj_1_ccolmaj_0_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
a4b6ffe — 2026-01-05 19:24:050.97 (-0.01%)0.97 (+0.03%)0.97 (+0.04%)0.97 (+0.00%)0.00 (-6.43%)25927.90 (-0.00%)25903.20 (-0.03%)25898.60 (-0.04%)25892.80 (+0.01%)14.05 (-6.56%)663.50 (-0.01%)663.23 (+0.03%)663.35 (+0.04%)662.60 (+0.00%)0.36 (-6.43%)
50ad37c — 2025-12-29 20:35:530.97 (n/a)0.97 (n/a)0.97 (n/a)0.97 (n/a)0.00 (n/a)25928.10 (n/a)25910.14 (n/a)25908.10 (n/a)25891.20 (n/a)15.04 (n/a)663.54 (n/a)663.06 (n/a)663.11 (n/a)662.60 (n/a)0.38 (n/a)

gemm_2048x2048x2048_64x64x64_2_cols_1_bcolmaj_0_ccolmaj_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
130b6ea — 2025-12-05 21:15:490.92 (-1.02%)0.92 (-0.66%)0.92 (-0.10%)0.91 (-0.78%)0.01 (-4.54%)27751.40 (+0.79%)27480.74 (+0.66%)27354.80 (+0.10%)27320.50 (+1.03%)182.69 (-2.84%)
0a6c11c — 2025-12-04 00:39:100.93 (n/a)0.92 (n/a)0.92 (n/a)0.91 (n/a)0.01 (n/a)27534.80 (n/a)27299.32 (n/a)27326.70 (n/a)27042.30 (n/a)188.04 (n/a)

gemm_2048x2048x2048_64x64x64_2_cols_1_bcolmaj_0_ccolmaj_0_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
a4b6ffe — 2026-01-05 19:24:050.92 (+0.11%)0.91 (-0.12%)0.91 (-0.21%)0.91 (-0.24%)0.00 (+30.34%)27723.50 (+0.24%)27548.94 (+0.12%)27522.90 (+0.21%)27403.60 (-0.11%)124.26 (+30.58%)626.92 (+0.11%)623.62 (-0.12%)624.20 (-0.21%)619.69 (-0.24%)2.81 (+30.34%)
50ad37c — 2025-12-29 20:35:530.92 (n/a)0.91 (n/a)0.92 (n/a)0.91 (n/a)0.00 (n/a)27657.10 (n/a)27515.58 (n/a)27464.70 (n/a)27434.60 (n/a)95.16 (n/a)626.21 (n/a)624.37 (n/a)625.52 (n/a)621.17 (n/a)2.16 (n/a)

gemm_2048x2048x2048_64x64x64_2cols_bcolmaj

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
8b6d0f4 — 2026-03-25 20:46:390.88 (+0.21%)0.87 (-0.11%)0.87 (-0.26%)0.87 (+0.03%)0.00 (+24.60%)29019.10 (-0.03%)28862.30 (+0.12%)28862.50 (+0.26%)28665.10 (-0.21%)148.73 (+24.28%)599.33 (+0.21%)595.25 (-0.11%)595.23 (-0.26%)592.02 (+0.03%)3.07 (+24.61%)
dc66da7 — 2026-03-25 19:23:450.88 (n/a)0.87 (n/a)0.87 (n/a)0.87 (n/a)0.00 (n/a)29027.70 (n/a)28829.10 (n/a)28788.30 (n/a)28724.70 (n/a)119.67 (n/a)598.09 (n/a)595.93 (n/a)596.77 (n/a)591.84 (n/a)2.46 (n/a)

gemm_2048x2048x2048_64x64x64_8_cols_0_bcolmaj_0_ccolmaj_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)GFLOP/s (max)GFLOP/s (mean)GFLOP/s (median)GFLOP/s (min)GFLOP/s (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
31ab24e — 2025-11-22 02:01:563.65 (-1.06%)3.59 (+0.83%)3.63 (-0.26%)3.45 (+2.66%)0.07 (-44.71%)2678.08 (-1.06%)2632.13 (+0.83%)2658.60 (-0.26%)2528.31 (+2.66%)53.86 (-44.71%)6795.00 (-2.59%)6529.80 (-0.92%)6462.00 (+0.26%)6415.00 (+1.07%)137.14 (-45.27%)
d9a64c5 — 2025-11-21 20:49:353.69 (n/a)3.56 (n/a)3.64 (n/a)3.36 (n/a)0.13 (n/a)2706.77 (n/a)2610.50 (n/a)2665.61 (n/a)2462.71 (n/a)97.41 (n/a)6976.00 (n/a)6590.40 (n/a)6445.00 (n/a)6347.00 (n/a)250.58 (n/a)

gemm_2048x2048x2048_64x64x64_8_cols_0_bcolmaj_0_ccolmaj_0_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
d48746f — 2026-01-12 22:17:103.98 (+0.20%)3.84 (-2.95%)3.76 (-5.27%)3.72 (-4.97%)0.13 (+473.97%)6760.00 (+5.23%)6559.50 (+3.13%)6697.10 (+5.56%)6321.00 (-0.20%)218.68 (+499.40%)2717.91 (+0.20%)2621.44 (-2.95%)2565.27 (-5.27%)2541.40 (-4.97%)88.37 (+473.97%)
331dcca — 2026-01-08 18:34:103.97 (n/a)3.96 (n/a)3.97 (n/a)3.92 (n/a)0.02 (n/a)6423.80 (n/a)6360.32 (n/a)6344.30 (n/a)6333.90 (n/a)36.48 (n/a)2712.39 (n/a)2701.17 (n/a)2707.91 (n/a)2674.40 (n/a)15.40 (n/a)

gemm_2048x2048x2048_64x64x64_8_cols_0_bcolmaj_1_ccolmaj_0_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
d48746f — 2026-01-12 22:17:103.75 (+0.28%)3.70 (+1.25%)3.71 (-0.21%)3.65 (+4.77%)0.04 (-66.90%)6901.80 (-4.56%)6795.46 (-1.30%)6790.20 (+0.21%)6715.90 (-0.28%)67.05 (-68.40%)2558.08 (+0.28%)2528.33 (+1.25%)2530.10 (-0.21%)2489.18 (+4.77%)24.81 (-66.90%)
331dcca — 2026-01-08 18:34:103.74 (n/a)3.66 (n/a)3.71 (n/a)3.48 (n/a)0.11 (n/a)7231.20 (n/a)6885.18 (n/a)6775.70 (n/a)6734.90 (n/a)212.21 (n/a)2550.87 (n/a)2497.04 (n/a)2535.50 (n/a)2375.78 (n/a)74.96 (n/a)

gemm_2048x2048x2048_64x64x64_8_cols_1_bcolmaj_0_ccolmaj_0_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
d48746f — 2026-01-12 22:17:103.62 (-0.77%)3.54 (+0.14%)3.60 (-0.04%)3.44 (+2.17%)0.09 (-25.77%)7307.40 (-2.13%)7103.46 (-0.18%)6982.50 (+0.04%)6948.20 (+0.77%)182.95 (-26.55%)2472.55 (-0.77%)2419.80 (+0.14%)2460.41 (-0.04%)2351.04 (+2.17%)61.76 (-25.77%)
331dcca — 2026-01-08 18:34:103.65 (n/a)3.54 (n/a)3.61 (n/a)3.37 (n/a)0.12 (n/a)7466.20 (n/a)7116.40 (n/a)6979.60 (n/a)6895.10 (n/a)249.09 (n/a)2491.61 (n/a)2416.45 (n/a)2461.45 (n/a)2301.00 (n/a)83.20 (n/a)

gemm_2048x2048x2048_64x64x64_8cols_bcolmaj_ccolmaj

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
8b6d0f4 — 2026-03-25 20:46:393.47 (-0.81%)3.41 (+0.82%)3.42 (-0.23%)3.35 (+4.63%)0.05 (-63.03%)7522.10 (-4.43%)7390.72 (-0.92%)7367.60 (+0.23%)7255.80 (+0.82%)107.14 (-64.21%)2367.74 (-0.81%)2324.90 (+0.82%)2331.81 (-0.23%)2283.92 (+4.63%)33.69 (-63.03%)
dc66da7 — 2026-03-25 19:23:453.50 (n/a)3.38 (n/a)3.42 (n/a)3.20 (n/a)0.13 (n/a)7870.70 (n/a)7459.22 (n/a)7350.70 (n/a)7196.70 (n/a)299.40 (n/a)2387.20 (n/a)2306.10 (n/a)2337.18 (n/a)2182.77 (n/a)91.13 (n/a)

gemm_384x1536x1792_32x48x64_4cols_bcolmaj

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
8b6d0f4 — 2026-03-25 20:46:394.16 (+12.80%)3.50 (+5.22%)3.50 (+2.51%)2.86 (-0.71%)0.63 (+68.68%)2814.80 (+0.72%)2362.04 (-3.41%)2304.60 (-2.45%)1936.00 (-11.34%)430.86 (+52.19%)1091.93 (+12.80%)919.16 (+5.22%)917.28 (+2.51%)751.01 (-0.71%)166.31 (+68.68%)
dc66da7 — 2026-03-25 19:23:453.69 (n/a)3.33 (n/a)3.41 (n/a)2.88 (n/a)0.38 (n/a)2794.70 (n/a)2445.42 (n/a)2362.50 (n/a)2183.70 (n/a)283.10 (n/a)968.06 (n/a)873.57 (n/a)894.79 (n/a)756.39 (n/a)98.60 (n/a)

gemm_896x1792x640_32x64x80_8cols_ccolmaj

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
8b6d0f4 — 2026-03-25 20:46:396.32 (+6.48%)4.74 (+3.95%)4.60 (+2.04%)3.49 (+3.11%)1.02 (+11.51%)1906.60 (-3.02%)1453.08 (-3.45%)1446.20 (-2.01%)1052.10 (-6.09%)304.01 (+0.60%)1953.36 (+6.48%)1465.54 (+3.95%)1421.06 (+2.04%)1077.93 (+3.11%)313.81 (+11.51%)
dc66da7 — 2026-03-25 19:23:455.94 (n/a)4.56 (n/a)4.51 (n/a)3.38 (n/a)0.91 (n/a)1966.00 (n/a)1504.94 (n/a)1475.80 (n/a)1120.30 (n/a)302.18 (n/a)1834.47 (n/a)1409.81 (n/a)1392.60 (n/a)1045.38 (n/a)281.42 (n/a)

input_length_2048-num_aie_columns_1-num_channels_1-tile_size_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
7d76781 — 2026-03-25 21:49:380.04 (n/a)0.03 (n/a)0.03 (n/a)0.01 (n/a)0.01 (n/a)618.40 (n/a)287.14 (n/a)240.50 (n/a)222.40 (n/a)102.53 (n/a)

input_length_2048-num_aie_columns_1-num_channels_1-tile_size_2048-group_size_32

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
7d76781 — 2026-03-25 21:49:380.02 (n/a)0.02 (n/a)0.02 (n/a)0.01 (n/a)0.00 (n/a)555.40 (n/a)333.06 (n/a)286.00 (n/a)253.10 (n/a)125.25 (n/a)

input_length_2048-num_aie_columns_1-num_channels_1-tile_size_2048-weighted_False

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
7d76781 — 2026-03-25 21:49:380.04 (n/a)0.03 (n/a)0.04 (n/a)0.01 (n/a)0.01 (n/a)698.20 (n/a)333.58 (n/a)231.70 (n/a)223.00 (n/a)205.57 (n/a)

input_length_2048-num_aie_columns_1-num_channels_1-tile_size_2048-weighted_True

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
7d76781 — 2026-03-25 21:49:380.05 (n/a)0.05 (n/a)0.05 (n/a)0.03 (n/a)0.01 (n/a)412.00 (n/a)280.30 (n/a)249.70 (n/a)238.90 (n/a)73.94 (n/a)

input_length_2048-num_aie_columns_1-num_channels_2-tile_size_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
7d76781 — 2026-03-25 21:49:380.03 (n/a)0.03 (n/a)0.03 (n/a)0.01 (n/a)0.01 (n/a)622.10 (n/a)326.33 (n/a)259.85 (n/a)240.50 (n/a)131.58 (n/a)

input_length_2048-num_aie_columns_1-num_channels_2-tile_size_1024-group_size_32

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
7d76781 — 2026-03-25 21:49:380.02 (n/a)0.02 (n/a)0.02 (n/a)0.01 (n/a)0.00 (n/a)572.30 (n/a)331.70 (n/a)277.60 (n/a)258.50 (n/a)134.73 (n/a)

input_length_2048-num_aie_columns_1-num_channels_2-tile_size_1024-weighted_False

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
7d76781 — 2026-03-25 21:49:380.03 (n/a)0.03 (n/a)0.03 (n/a)0.02 (n/a)0.00 (n/a)342.40 (n/a)271.18 (n/a)248.40 (n/a)243.70 (n/a)41.97 (n/a)

input_length_2048-num_aie_columns_1-num_channels_2-tile_size_2048-scalar_factor_3.0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
7d76781 — 2026-03-25 21:49:380.05 (n/a)0.04 (n/a)0.05 (n/a)0.02 (n/a)0.02 (n/a)700.50 (n/a)336.46 (n/a)231.80 (n/a)227.90 (n/a)205.15 (n/a)

input_length_2048-num_aie_columns_1-tile_size_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
7d76781 — 2026-03-25 21:49:380.05 (n/a)0.04 (n/a)0.04 (n/a)0.02 (n/a)0.01 (n/a)710.00 (n/a)334.72 (n/a)250.60 (n/a)221.50 (n/a)152.08 (n/a)

input_length_2048-num_aie_columns_2-num_channels_1-tile_size_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
7d76781 — 2026-03-25 21:49:380.04 (n/a)0.03 (n/a)0.03 (n/a)0.02 (n/a)0.01 (n/a)464.10 (n/a)271.27 (n/a)245.30 (n/a)213.90 (n/a)68.08 (n/a)

input_length_2048-num_aie_columns_2-num_channels_1-tile_size_1024-group_size_32

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
7d76781 — 2026-03-25 21:49:380.02 (n/a)0.02 (n/a)0.02 (n/a)0.01 (n/a)0.00 (n/a)646.10 (n/a)365.00 (n/a)300.10 (n/a)284.80 (n/a)157.27 (n/a)

input_length_2048-num_aie_columns_2-num_channels_1-tile_size_1024-weighted_False

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
7d76781 — 2026-03-25 21:49:380.04 (n/a)0.03 (n/a)0.03 (n/a)0.02 (n/a)0.01 (n/a)388.10 (n/a)271.00 (n/a)244.30 (n/a)232.20 (n/a)65.70 (n/a)

input_length_2048-num_aie_columns_2-num_channels_1-tile_size_1024-weighted_True

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
7d76781 — 2026-03-25 21:49:380.04 (n/a)0.04 (n/a)0.04 (n/a)0.02 (n/a)0.01 (n/a)475.20 (n/a)306.94 (n/a)258.50 (n/a)245.40 (n/a)96.46 (n/a)

input_length_2048-num_aie_columns_2-num_channels_2-tile_size_1024-scalar_factor_3.0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
7d76781 — 2026-03-25 21:49:380.05 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.01 (n/a)337.80 (n/a)272.94 (n/a)253.90 (n/a)246.00 (n/a)38.77 (n/a)

input_length_2048-num_aie_columns_2-num_channels_2-tile_size_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
7d76781 — 2026-03-25 21:49:380.04 (n/a)0.03 (n/a)0.03 (n/a)0.01 (n/a)0.01 (n/a)631.00 (n/a)296.78 (n/a)250.50 (n/a)228.30 (n/a)120.74 (n/a)

input_length_2048-num_aie_columns_2-num_channels_2-tile_size_512-group_size_32

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
7d76781 — 2026-03-25 21:49:380.02 (n/a)0.02 (n/a)0.02 (n/a)0.01 (n/a)0.00 (n/a)507.40 (n/a)317.12 (n/a)278.00 (n/a)248.60 (n/a)107.90 (n/a)

input_length_2048-num_aie_columns_2-num_channels_2-tile_size_512-weighted_False

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
7d76781 — 2026-03-25 21:49:380.04 (n/a)0.03 (n/a)0.04 (n/a)0.01 (n/a)0.01 (n/a)633.90 (n/a)317.38 (n/a)226.40 (n/a)223.30 (n/a)178.41 (n/a)

input_length_2048-num_aie_columns_2-tile_size_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
7d76781 — 2026-03-25 21:49:380.05 (n/a)0.04 (n/a)0.04 (n/a)0.01 (n/a)0.01 (n/a)624.40 (n/a)319.29 (n/a)258.70 (n/a)242.80 (n/a)131.99 (n/a)

input_length_2048-num_aie_columns_4-num_channels_1-tile_size_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
7d76781 — 2026-03-25 21:49:380.04 (n/a)0.03 (n/a)0.03 (n/a)0.01 (n/a)0.01 (n/a)754.60 (n/a)284.04 (n/a)251.50 (n/a)218.20 (n/a)109.06 (n/a)

input_length_2048-num_aie_columns_4-num_channels_1-tile_size_512-group_size_32

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
7d76781 — 2026-03-25 21:49:380.02 (n/a)0.01 (n/a)0.01 (n/a)0.01 (n/a)0.01 (n/a)720.30 (n/a)484.46 (n/a)571.90 (n/a)259.20 (n/a)199.84 (n/a)

input_length_2048-num_aie_columns_4-num_channels_1-tile_size_512-weighted_False

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
7d76781 — 2026-03-25 21:49:380.03 (n/a)0.03 (n/a)0.03 (n/a)0.02 (n/a)0.01 (n/a)433.70 (n/a)282.68 (n/a)247.20 (n/a)240.80 (n/a)84.48 (n/a)

input_length_2048-num_aie_columns_4-num_channels_1-tile_size_512-weighted_True

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
7d76781 — 2026-03-25 21:49:380.04 (n/a)0.04 (n/a)0.04 (n/a)0.02 (n/a)0.01 (n/a)372.70 (n/a)259.28 (n/a)231.90 (n/a)226.60 (n/a)63.49 (n/a)

input_length_2048-num_aie_columns_4-num_channels_2-tile_size_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
7d76781 — 2026-03-25 21:49:380.03 (n/a)0.03 (n/a)0.03 (n/a)0.01 (n/a)0.01 (n/a)823.70 (n/a)362.78 (n/a)253.45 (n/a)249.50 (n/a)218.01 (n/a)

input_length_2048-num_aie_columns_4-num_channels_2-tile_size_256-group_size_32

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
7d76781 — 2026-03-25 21:49:380.02 (n/a)0.02 (n/a)0.02 (n/a)0.02 (n/a)0.00 (n/a)303.20 (n/a)293.52 (n/a)290.60 (n/a)287.00 (n/a)6.87 (n/a)

input_length_2048-num_aie_columns_4-num_channels_2-tile_size_256-weighted_False

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
7d76781 — 2026-03-25 21:49:380.04 (n/a)0.03 (n/a)0.03 (n/a)0.01 (n/a)0.01 (n/a)600.50 (n/a)316.78 (n/a)235.50 (n/a)227.40 (n/a)160.48 (n/a)

input_length_2048-num_aie_columns_4-num_channels_2-tile_size_512-scalar_factor_3.0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
7d76781 — 2026-03-25 21:49:380.05 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.01 (n/a)343.80 (n/a)268.00 (n/a)249.30 (n/a)243.50 (n/a)42.59 (n/a)

input_length_2048-num_aie_columns_4-tile_size_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
7d76781 — 2026-03-25 21:49:380.05 (n/a)0.04 (n/a)0.04 (n/a)0.02 (n/a)0.01 (n/a)594.60 (n/a)305.99 (n/a)253.70 (n/a)226.10 (n/a)113.39 (n/a)

input_length_2048-num_aie_columns_8-num_channels_1-tile_size_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
7d76781 — 2026-03-25 21:49:380.03 (n/a)0.03 (n/a)0.03 (n/a)0.01 (n/a)0.01 (n/a)726.80 (n/a)320.91 (n/a)258.60 (n/a)235.10 (n/a)148.61 (n/a)

input_length_2048-num_aie_columns_8-num_channels_1-tile_size_256-group_size_32

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
7d76781 — 2026-03-25 21:49:380.02 (n/a)0.02 (n/a)0.02 (n/a)0.02 (n/a)0.00 (n/a)311.10 (n/a)286.44 (n/a)297.00 (n/a)254.90 (n/a)22.81 (n/a)

input_length_2048-num_aie_columns_8-num_channels_1-tile_size_256-weighted_False

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
7d76781 — 2026-03-25 21:49:380.03 (n/a)0.03 (n/a)0.03 (n/a)0.01 (n/a)0.01 (n/a)716.70 (n/a)349.34 (n/a)257.70 (n/a)251.70 (n/a)205.40 (n/a)

input_length_2048-num_aie_columns_8-num_channels_1-tile_size_256-weighted_True

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
7d76781 — 2026-03-25 21:49:380.03 (n/a)0.03 (n/a)0.03 (n/a)0.01 (n/a)0.01 (n/a)846.20 (n/a)381.26 (n/a)274.40 (n/a)250.80 (n/a)260.12 (n/a)

input_length_2048-num_aie_columns_8-num_channels_2-tile_size_128

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
7d76781 — 2026-03-25 21:49:380.03 (n/a)0.03 (n/a)0.03 (n/a)0.01 (n/a)0.01 (n/a)1044.70 (n/a)385.37 (n/a)279.10 (n/a)252.60 (n/a)251.82 (n/a)

input_length_2048-num_aie_columns_8-num_channels_2-tile_size_128-group_size_32

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
7d76781 — 2026-03-25 21:49:380.02 (n/a)0.02 (n/a)0.02 (n/a)0.01 (n/a)0.00 (n/a)683.30 (n/a)376.92 (n/a)295.60 (n/a)292.60 (n/a)171.57 (n/a)

input_length_2048-num_aie_columns_8-num_channels_2-tile_size_128-weighted_False

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
7d76781 — 2026-03-25 21:49:380.03 (n/a)0.03 (n/a)0.03 (n/a)0.01 (n/a)0.01 (n/a)997.90 (n/a)417.18 (n/a)270.10 (n/a)266.90 (n/a)324.71 (n/a)

input_length_2048-num_aie_columns_8-num_channels_2-tile_size_256-scalar_factor_3.0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
7d76781 — 2026-03-25 21:49:380.05 (n/a)0.04 (n/a)0.05 (n/a)0.02 (n/a)0.01 (n/a)769.20 (n/a)367.94 (n/a)261.60 (n/a)256.80 (n/a)224.79 (n/a)

input_length_2048-num_aie_columns_8-tile_size_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
7d76781 — 2026-03-25 21:49:380.05 (n/a)0.04 (n/a)0.05 (n/a)0.01 (n/a)0.01 (n/a)762.60 (n/a)319.69 (n/a)257.80 (n/a)224.70 (n/a)160.33 (n/a)

input_length_2048-num_cores_1-num_channels_1-bypass_False-tile_size_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
7d76781 — 2026-03-25 21:49:380.04 (n/a)0.03 (n/a)0.03 (n/a)0.01 (n/a)0.01 (n/a)842.70 (n/a)391.58 (n/a)259.40 (n/a)224.80 (n/a)261.86 (n/a)

input_length_2048-num_cores_16-num_channels_2-bypass_False-tile_size_128

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
7d76781 — 2026-03-25 21:49:380.03 (n/a)0.02 (n/a)0.03 (n/a)0.01 (n/a)0.01 (n/a)748.50 (n/a)417.06 (n/a)292.30 (n/a)269.90 (n/a)208.88 (n/a)

input_length_2048-num_cores_2-num_channels_1-bypass_False-tile_size_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
7d76781 — 2026-03-25 21:49:380.03 (n/a)0.03 (n/a)0.03 (n/a)0.02 (n/a)0.01 (n/a)380.20 (n/a)281.78 (n/a)251.90 (n/a)237.50 (n/a)58.90 (n/a)

input_length_2048-num_cores_2-num_channels_2-bypass_False-tile_size_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
7d76781 — 2026-03-25 21:49:380.04 (n/a)0.03 (n/a)0.03 (n/a)0.02 (n/a)0.01 (n/a)381.70 (n/a)266.60 (n/a)243.50 (n/a)226.70 (n/a)64.87 (n/a)

input_length_2048-num_cores_4-num_channels_1-bypass_False-tile_size_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
7d76781 — 2026-03-25 21:49:380.04 (n/a)0.03 (n/a)0.04 (n/a)0.02 (n/a)0.01 (n/a)382.50 (n/a)262.62 (n/a)233.60 (n/a)227.30 (n/a)67.10 (n/a)

input_length_2048-num_cores_4-num_channels_2-bypass_False-tile_size_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
7d76781 — 2026-03-25 21:49:380.04 (n/a)0.03 (n/a)0.04 (n/a)0.01 (n/a)0.01 (n/a)639.60 (n/a)325.20 (n/a)231.20 (n/a)226.70 (n/a)178.63 (n/a)

input_length_2048-num_cores_8-num_channels_1-bypass_False-tile_size_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
7d76781 — 2026-03-25 21:49:380.03 (n/a)0.03 (n/a)0.03 (n/a)0.01 (n/a)0.01 (n/a)654.40 (n/a)336.50 (n/a)257.20 (n/a)255.00 (n/a)177.72 (n/a)

input_length_2048-num_cores_8-num_channels_2-bypass_False-tile_size_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
7d76781 — 2026-03-25 21:49:380.03 (n/a)0.03 (n/a)0.03 (n/a)0.02 (n/a)0.01 (n/a)411.90 (n/a)286.54 (n/a)257.60 (n/a)247.60 (n/a)70.40 (n/a)

input_length_32768-num_aie_columns_2-num_channels_2-tile_size_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
7d76781 — 2026-03-25 21:49:380.50 (n/a)0.35 (n/a)0.41 (n/a)0.15 (n/a)0.14 (n/a)871.10 (n/a)444.76 (n/a)320.20 (n/a)260.80 (n/a)249.66 (n/a)

input_length_32768-num_aie_columns_2-num_channels_2-tile_size_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
7d76781 — 2026-03-25 21:49:380.49 (n/a)0.29 (n/a)0.22 (n/a)0.17 (n/a)0.14 (n/a)766.00 (n/a)528.36 (n/a)596.50 (n/a)267.80 (n/a)219.30 (n/a)

input_length_32768-num_aie_columns_4-num_channels_4-tile_size_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
7d76781 — 2026-03-25 21:49:380.40 (n/a)0.30 (n/a)0.28 (n/a)0.18 (n/a)0.10 (n/a)723.70 (n/a)482.86 (n/a)466.20 (n/a)327.10 (n/a)168.82 (n/a)

layer_norm_1_cols_1_channels_2048_tile_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.07 (+35.61%)0.05 (+6.58%)0.04 (-0.65%)0.03 (-1.11%)0.01 (+98.33%)236.50 (+1.11%)182.90 (-3.12%)183.30 (+0.60%)118.50 (-26.26%)42.30 (+41.46%)
dc66da7 — 2026-03-25 19:23:450.05 (n/a)0.04 (n/a)0.04 (n/a)0.04 (n/a)0.01 (n/a)233.90 (n/a)188.80 (n/a)182.20 (n/a)160.70 (n/a)29.90 (n/a)

layer_norm_1_cols_2_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.07 (+9.47%)0.05 (+5.79%)0.05 (+5.48%)0.04 (-7.03%)0.01 (+51.12%)206.90 (+7.54%)163.06 (-2.60%)165.20 (-5.22%)113.20 (-8.71%)40.37 (+56.59%)
dc66da7 — 2026-03-25 19:23:450.07 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.01 (n/a)192.40 (n/a)167.42 (n/a)174.30 (n/a)124.00 (n/a)25.78 (n/a)

layer_norm_2_cols_1_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.06 (-0.97%)0.05 (+14.59%)0.05 (+47.43%)0.03 (+8.92%)0.01 (-0.31%)235.10 (-8.20%)183.06 (-13.03%)157.80 (-32.19%)148.10 (+1.02%)43.12 (-7.94%)
dc66da7 — 2026-03-25 19:23:450.06 (n/a)0.04 (n/a)0.04 (n/a)0.03 (n/a)0.01 (n/a)256.10 (n/a)210.48 (n/a)232.70 (n/a)146.60 (n/a)46.84 (n/a)

layer_norm_2_cols_2_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.07 (+14.65%)0.05 (+4.61%)0.05 (-7.99%)0.04 (+11.94%)0.01 (+28.58%)201.70 (-10.67%)168.58 (-3.65%)176.00 (+8.71%)122.10 (-12.79%)33.82 (+0.06%)
dc66da7 — 2026-03-25 19:23:450.06 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.01 (n/a)225.80 (n/a)174.96 (n/a)161.90 (n/a)140.00 (n/a)33.80 (n/a)

layer_norm_4_cols_1_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.06 (-31.05%)0.05 (-3.73%)0.05 (+9.03%)0.04 (+10.49%)0.01 (-63.42%)197.60 (-9.52%)167.50 (-4.83%)166.40 (-8.27%)129.00 (+45.11%)25.47 (-51.02%)
dc66da7 — 2026-03-25 19:23:450.09 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.02 (n/a)218.40 (n/a)176.00 (n/a)181.40 (n/a)88.90 (n/a)52.00 (n/a)

layer_norm_4_cols_2_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.05 (-1.52%)0.04 (-11.20%)0.04 (-25.45%)0.03 (-0.69%)0.01 (-9.35%)243.40 (+0.66%)202.20 (+12.10%)212.70 (+34.20%)156.90 (+1.55%)34.18 (-7.92%)
dc66da7 — 2026-03-25 19:23:450.05 (n/a)0.05 (n/a)0.05 (n/a)0.03 (n/a)0.01 (n/a)241.80 (n/a)180.38 (n/a)158.50 (n/a)154.50 (n/a)37.12 (n/a)

layer_norm_8_cols_1_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.05 (-6.07%)0.04 (+15.60%)0.04 (+21.55%)0.04 (+82.28%)0.01 (-55.28%)222.00 (-45.13%)190.18 (-21.85%)193.60 (-17.72%)153.10 (+6.47%)24.64 (-75.05%)
dc66da7 — 2026-03-25 19:23:450.06 (n/a)0.04 (n/a)0.03 (n/a)0.02 (n/a)0.01 (n/a)404.60 (n/a)243.34 (n/a)235.30 (n/a)143.80 (n/a)98.74 (n/a)

layer_norm_8_cols_2_channels_2048_tile_128

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.05 (+37.82%)0.04 (+27.22%)0.04 (+9.46%)0.03 (+34.60%)0.01 (+56.79%)251.50 (-25.72%)212.60 (-21.06%)226.70 (-8.66%)175.20 (-27.45%)33.04 (-18.34%)
dc66da7 — 2026-03-25 19:23:450.03 (n/a)0.03 (n/a)0.03 (n/a)0.02 (n/a)0.00 (n/a)338.60 (n/a)269.32 (n/a)248.20 (n/a)241.50 (n/a)40.46 (n/a)

matrix_vector_mul_128x128_32_1col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)
130b6ea — 2025-12-05 21:15:490.37 (+17.13%)0.30 (+24.87%)0.30 (+38.03%)0.23 (+12.47%)0.06 (+35.23%)
0a6c11c — 2025-12-04 00:39:100.32 (n/a)0.24 (n/a)0.22 (n/a)0.20 (n/a)0.04 (n/a)

matrix_vector_mul_128x128_32_1col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
331dcca — 2026-01-08 18:34:100.26 (-1.81%)0.24 (+4.06%)0.24 (+0.52%)0.21 (+16.97%)0.02 (-48.16%)0.26 (-1.81%)0.24 (+4.06%)0.24 (+0.52%)0.20 (+16.97%)0.02 (-48.16%)
a4b6ffe — 2026-01-05 19:24:050.26 (n/a)0.23 (n/a)0.24 (n/a)0.18 (n/a)0.04 (n/a)0.26 (n/a)0.23 (n/a)0.24 (n/a)0.17 (n/a)0.04 (n/a)

matrix_vector_mul_128x128_32tsi_128tso_1col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
8b6d0f4 — 2026-03-25 20:46:390.26 (+32.10%)0.19 (+11.55%)0.18 (+9.49%)0.13 (-2.22%)0.05 (+105.83%)0.25 (+32.10%)0.19 (+11.55%)0.18 (+9.49%)0.13 (-2.22%)0.05 (+105.83%)
dc66da7 — 2026-03-25 19:23:450.19 (n/a)0.17 (n/a)0.17 (n/a)0.14 (n/a)0.02 (n/a)0.19 (n/a)0.17 (n/a)0.17 (n/a)0.14 (n/a)0.02 (n/a)

matrix_vector_mul_128x128_32tsi_128tso_1col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
897d04e — 2026-03-06 22:12:250.37 (+5.44%)0.31 (-0.05%)0.34 (-1.99%)0.22 (-7.87%)0.06 (+15.13%)0.36 (+5.44%)0.31 (-0.05%)0.34 (-1.99%)0.21 (-7.87%)0.06 (+15.13%)
84d3478 — 2026-02-17 23:25:130.35 (n/a)0.31 (n/a)0.35 (n/a)0.23 (n/a)0.05 (n/a)0.35 (n/a)0.31 (n/a)0.34 (n/a)0.23 (n/a)0.05 (n/a)

matrix_vector_mul_2048x8192_1_1col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)
130b6ea — 2025-12-05 21:15:4913.57 (+3.47%)13.15 (+1.09%)13.10 (+0.16%)12.52 (-1.58%)0.37 (+153.19%)
0a6c11c — 2025-12-04 00:39:1013.11 (n/a)13.01 (n/a)13.07 (n/a)12.72 (n/a)0.15 (n/a)

matrix_vector_mul_2048x8192_1_1col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
331dcca — 2026-01-08 18:34:1013.43 (-0.07%)13.30 (+4.62%)13.27 (+1.29%)13.16 (+16.70%)0.12 (-86.69%)13.42 (-0.07%)13.29 (+4.62%)13.26 (+1.29%)13.16 (+16.70%)0.12 (-86.69%)
a4b6ffe — 2026-01-05 19:24:0513.44 (n/a)12.71 (n/a)13.10 (n/a)11.28 (n/a)0.87 (n/a)13.43 (n/a)12.70 (n/a)13.09 (n/a)11.27 (n/a)0.87 (n/a)

matrix_vector_mul_2048x8192_1_2col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)
130b6ea — 2025-12-05 21:15:4925.06 (+2.47%)23.71 (+1.81%)24.13 (+0.50%)20.90 (+5.41%)1.48 (-15.60%)
0a6c11c — 2025-12-04 00:39:1024.46 (n/a)23.29 (n/a)24.01 (n/a)19.83 (n/a)1.75 (n/a)

matrix_vector_mul_2048x8192_1_2col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
331dcca — 2026-01-08 18:34:1024.46 (-0.60%)21.07 (-8.03%)19.75 (-17.83%)18.08 (+1.18%)3.03 (+7.07%)24.44 (-0.60%)21.06 (-8.03%)19.74 (-17.83%)18.07 (+1.18%)3.03 (+7.07%)
a4b6ffe — 2026-01-05 19:24:0524.60 (n/a)22.91 (n/a)24.04 (n/a)17.87 (n/a)2.83 (n/a)24.59 (n/a)22.90 (n/a)24.02 (n/a)17.86 (n/a)2.83 (n/a)

matrix_vector_mul_2048x8192_1_4col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)
130b6ea — 2025-12-05 21:15:4942.05 (+1.54%)40.50 (+12.60%)40.49 (+2.36%)38.98 (+97.53%)0.97 (-88.09%)
0a6c11c — 2025-12-04 00:39:1041.41 (n/a)35.97 (n/a)39.55 (n/a)19.73 (n/a)8.15 (n/a)

matrix_vector_mul_2048x8192_1_4col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
331dcca — 2026-01-08 18:34:1042.58 (+1.51%)41.37 (+14.29%)41.25 (+4.89%)40.33 (+91.42%)0.92 (-89.18%)42.56 (+1.51%)41.34 (+14.29%)41.23 (+4.89%)40.30 (+91.42%)0.92 (-89.18%)
a4b6ffe — 2026-01-05 19:24:0541.95 (n/a)36.20 (n/a)39.33 (n/a)21.07 (n/a)8.54 (n/a)41.92 (n/a)36.17 (n/a)39.31 (n/a)21.05 (n/a)8.53 (n/a)

matrix_vector_mul_2048x8192_1_8col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)
130b6ea — 2025-12-05 21:15:4946.62 (+3.11%)44.82 (+0.17%)44.78 (+0.17%)42.60 (-3.80%)1.60 (+367.72%)
0a6c11c — 2025-12-04 00:39:1045.21 (n/a)44.75 (n/a)44.70 (n/a)44.29 (n/a)0.34 (n/a)

matrix_vector_mul_2048x8192_1_8col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
331dcca — 2026-01-08 18:34:1046.04 (+4.14%)45.06 (+3.47%)44.55 (+2.76%)44.30 (+2.84%)0.87 (+66.58%)46.01 (+4.14%)45.03 (+3.47%)44.52 (+2.76%)44.27 (+2.84%)0.87 (+66.58%)
a4b6ffe — 2026-01-05 19:24:0544.21 (n/a)43.55 (n/a)43.35 (n/a)43.07 (n/a)0.52 (n/a)44.18 (n/a)43.52 (n/a)43.32 (n/a)43.05 (n/a)0.52 (n/a)

matrix_vector_mul_2048x8192_1tsi_1024tso_2col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
8b6d0f4 — 2026-03-25 20:46:3925.44 (+1.95%)24.42 (-0.28%)24.79 (+0.16%)23.14 (-2.33%)1.05 (+106.47%)25.42 (+1.95%)24.41 (-0.28%)24.77 (+0.16%)23.12 (-2.33%)1.05 (+106.47%)
dc66da7 — 2026-03-25 19:23:4524.95 (n/a)24.49 (n/a)24.75 (n/a)23.69 (n/a)0.51 (n/a)24.94 (n/a)24.48 (n/a)24.73 (n/a)23.67 (n/a)0.51 (n/a)

matrix_vector_mul_2048x8192_1tsi_1024tso_2col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
897d04e — 2026-03-06 22:12:2524.47 (-3.32%)24.16 (-3.13%)24.20 (-3.09%)23.84 (-3.04%)0.25 (-14.36%)24.46 (-3.32%)24.15 (-3.13%)24.18 (-3.09%)23.83 (-3.04%)0.25 (-14.36%)
84d3478 — 2026-02-17 23:25:1325.31 (n/a)24.94 (n/a)24.97 (n/a)24.59 (n/a)0.29 (n/a)25.30 (n/a)24.93 (n/a)24.95 (n/a)24.58 (n/a)0.29 (n/a)

matrix_vector_mul_2048x8192_1tsi_2048tso_1col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
8b6d0f4 — 2026-03-25 20:46:3913.48 (-1.01%)13.15 (+1.31%)13.11 (-0.87%)12.95 (+6.77%)0.20 (-66.05%)13.47 (-1.01%)13.14 (+1.31%)13.11 (-0.87%)12.95 (+6.77%)0.20 (-66.05%)
dc66da7 — 2026-03-25 19:23:4513.62 (n/a)12.98 (n/a)13.23 (n/a)12.13 (n/a)0.58 (n/a)13.61 (n/a)12.97 (n/a)13.22 (n/a)12.13 (n/a)0.58 (n/a)

matrix_vector_mul_2048x8192_1tsi_2048tso_1col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
897d04e — 2026-03-06 22:12:2513.25 (-3.04%)13.02 (-2.92%)13.10 (-2.64%)12.56 (-4.14%)0.27 (+32.55%)13.24 (-3.04%)13.01 (-2.92%)13.09 (-2.64%)12.55 (-4.14%)0.27 (+32.55%)
84d3478 — 2026-02-17 23:25:1313.66 (n/a)13.41 (n/a)13.45 (n/a)13.10 (n/a)0.20 (n/a)13.65 (n/a)13.40 (n/a)13.44 (n/a)13.09 (n/a)0.20 (n/a)

matrix_vector_mul_2048x8192_1tsi_256tso_8col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
8b6d0f4 — 2026-03-25 20:46:3942.06 (-4.87%)40.85 (-4.53%)40.57 (-4.52%)39.78 (-4.49%)0.87 (-11.49%)42.04 (-4.87%)40.83 (-4.53%)40.55 (-4.52%)39.76 (-4.49%)0.87 (-11.49%)
dc66da7 — 2026-03-25 19:23:4544.21 (n/a)42.79 (n/a)42.49 (n/a)41.65 (n/a)0.99 (n/a)44.19 (n/a)42.76 (n/a)42.47 (n/a)41.63 (n/a)0.99 (n/a)

matrix_vector_mul_2048x8192_1tsi_256tso_8col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
897d04e — 2026-03-06 22:12:2547.71 (+3.51%)45.38 (+9.55%)45.22 (-1.27%)43.47 (+75.24%)1.52 (-83.67%)47.68 (+3.51%)45.35 (+9.55%)45.19 (-1.27%)43.45 (+75.24%)1.52 (-83.67%)
84d3478 — 2026-02-17 23:25:1346.09 (n/a)41.43 (n/a)45.80 (n/a)24.81 (n/a)9.31 (n/a)46.06 (n/a)41.40 (n/a)45.77 (n/a)24.79 (n/a)9.30 (n/a)

matrix_vector_mul_2048x8192_1tsi_512tso_4col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
8b6d0f4 — 2026-03-25 20:46:3940.94 (-0.67%)39.64 (+0.12%)39.96 (+1.02%)37.79 (+0.51%)1.25 (-21.17%)40.91 (-0.67%)39.61 (+0.12%)39.94 (+1.02%)37.77 (+0.51%)1.25 (-21.17%)
dc66da7 — 2026-03-25 19:23:4541.21 (n/a)39.59 (n/a)39.56 (n/a)37.60 (n/a)1.58 (n/a)41.19 (n/a)39.57 (n/a)39.53 (n/a)37.58 (n/a)1.58 (n/a)

matrix_vector_mul_2048x8192_1tsi_512tso_4col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
897d04e — 2026-03-06 22:12:2540.84 (-0.32%)39.53 (-1.25%)39.40 (-1.97%)38.87 (-0.69%)0.77 (-6.24%)40.81 (-0.32%)39.50 (-1.25%)39.37 (-1.97%)38.84 (-0.69%)0.77 (-6.24%)
84d3478 — 2026-02-17 23:25:1340.97 (n/a)40.03 (n/a)40.19 (n/a)39.14 (n/a)0.82 (n/a)40.95 (n/a)40.00 (n/a)40.17 (n/a)39.11 (n/a)0.82 (n/a)

matrix_vector_mul_8192x2048_4_1col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)
130b6ea — 2025-12-05 21:15:4913.29 (+1.63%)13.19 (+6.08%)13.19 (+1.17%)13.09 (+26.86%)0.08 (-92.94%)
0a6c11c — 2025-12-04 00:39:1013.08 (n/a)12.44 (n/a)13.03 (n/a)10.32 (n/a)1.07 (n/a)

matrix_vector_mul_8192x2048_4_1col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
331dcca — 2026-01-08 18:34:1013.68 (+3.42%)13.30 (+8.54%)13.38 (+7.25%)12.72 (+14.96%)0.35 (-66.09%)13.67 (+3.42%)13.29 (+8.54%)13.38 (+7.25%)12.71 (+14.96%)0.35 (-66.09%)
a4b6ffe — 2026-01-05 19:24:0513.23 (n/a)12.25 (n/a)12.48 (n/a)11.07 (n/a)1.04 (n/a)13.22 (n/a)12.24 (n/a)12.47 (n/a)11.06 (n/a)1.04 (n/a)

matrix_vector_mul_8192x2048_4_2col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)
130b6ea — 2025-12-05 21:15:4925.58 (+4.24%)24.23 (+6.59%)24.03 (+0.29%)23.58 (+37.29%)0.70 (-74.97%)
0a6c11c — 2025-12-04 00:39:1024.54 (n/a)22.73 (n/a)23.96 (n/a)17.18 (n/a)2.79 (n/a)

matrix_vector_mul_8192x2048_4_2col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
331dcca — 2026-01-08 18:34:1025.08 (+1.32%)24.68 (+13.42%)24.61 (+3.26%)24.47 (+48.32%)0.24 (-93.56%)25.07 (+1.32%)24.67 (+13.42%)24.59 (+3.26%)24.46 (+48.32%)0.24 (-93.56%)
a4b6ffe — 2026-01-05 19:24:0524.76 (n/a)21.76 (n/a)23.83 (n/a)16.50 (n/a)3.74 (n/a)24.74 (n/a)21.75 (n/a)23.82 (n/a)16.49 (n/a)3.73 (n/a)

matrix_vector_mul_8192x2048_4_4col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)
130b6ea — 2025-12-05 21:15:4940.94 (-2.19%)39.74 (-1.10%)39.76 (-0.60%)38.17 (-2.14%)1.04 (+4.39%)
0a6c11c — 2025-12-04 00:39:1041.86 (n/a)40.18 (n/a)40.00 (n/a)39.00 (n/a)1.00 (n/a)

matrix_vector_mul_8192x2048_4_4col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
331dcca — 2026-01-08 18:34:1040.74 (-1.09%)37.16 (-7.15%)40.25 (+1.47%)24.51 (-37.44%)7.07 (+736.13%)40.72 (-1.09%)37.13 (-7.15%)40.23 (+1.47%)24.50 (-37.44%)7.07 (+736.13%)
a4b6ffe — 2026-01-05 19:24:0541.19 (n/a)40.02 (n/a)39.67 (n/a)39.18 (n/a)0.85 (n/a)41.17 (n/a)39.99 (n/a)39.64 (n/a)39.16 (n/a)0.85 (n/a)

matrix_vector_mul_8192x2048_4_8col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)
130b6ea — 2025-12-05 21:15:4946.71 (-1.04%)44.91 (-2.34%)44.71 (-2.98%)43.80 (-1.44%)0.96 (+6.93%)
0a6c11c — 2025-12-04 00:39:1047.20 (n/a)45.98 (n/a)46.08 (n/a)44.44 (n/a)0.90 (n/a)

matrix_vector_mul_8192x2048_4_8col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
331dcca — 2026-01-08 18:34:1046.28 (+5.68%)44.89 (+14.59%)44.74 (+4.26%)44.20 (+62.80%)0.84 (-87.96%)46.25 (+5.68%)44.86 (+14.59%)44.71 (+4.26%)44.17 (+62.80%)0.84 (-87.96%)
a4b6ffe — 2026-01-05 19:24:0543.79 (n/a)39.17 (n/a)42.91 (n/a)27.15 (n/a)6.98 (n/a)43.77 (n/a)39.15 (n/a)42.88 (n/a)27.13 (n/a)6.98 (n/a)

matrix_vector_mul_8192x2048_4tsi_1024tso_1col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
8b6d0f4 — 2026-03-25 20:46:3913.35 (+0.83%)13.11 (+3.52%)13.10 (+3.94%)12.94 (+7.70%)0.15 (-69.46%)13.34 (+0.83%)13.10 (+3.52%)13.09 (+3.94%)12.94 (+7.70%)0.15 (-69.46%)
dc66da7 — 2026-03-25 19:23:4513.24 (n/a)12.66 (n/a)12.60 (n/a)12.02 (n/a)0.49 (n/a)13.23 (n/a)12.66 (n/a)12.59 (n/a)12.01 (n/a)0.49 (n/a)

matrix_vector_mul_8192x2048_4tsi_1024tso_1col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
897d04e — 2026-03-06 22:12:2513.34 (-0.79%)13.20 (-0.36%)13.18 (-1.55%)13.12 (+1.75%)0.08 (-65.93%)13.33 (-0.79%)13.19 (-0.36%)13.17 (-1.55%)13.11 (+1.75%)0.08 (-65.93%)
84d3478 — 2026-02-17 23:25:1313.45 (n/a)13.25 (n/a)13.39 (n/a)12.90 (n/a)0.25 (n/a)13.44 (n/a)13.24 (n/a)13.38 (n/a)12.89 (n/a)0.25 (n/a)

matrix_vector_mul_8192x2048_4tsi_1024tso_2col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
8b6d0f4 — 2026-03-25 20:46:3924.35 (-0.26%)24.07 (+0.89%)24.16 (+1.48%)23.64 (+2.17%)0.27 (-44.88%)24.34 (-0.26%)24.06 (+0.89%)24.15 (+1.48%)23.63 (+2.17%)0.27 (-44.88%)
dc66da7 — 2026-03-25 19:23:4524.41 (n/a)23.86 (n/a)23.81 (n/a)23.14 (n/a)0.49 (n/a)24.40 (n/a)23.84 (n/a)23.80 (n/a)23.12 (n/a)0.49 (n/a)

matrix_vector_mul_8192x2048_4tsi_1024tso_2col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
897d04e — 2026-03-06 22:12:2524.10 (-5.45%)23.07 (-1.85%)24.01 (-2.88%)19.26 (+7.47%)2.13 (-32.15%)24.08 (-5.45%)23.05 (-1.85%)23.99 (-2.88%)19.25 (+7.47%)2.13 (-32.15%)
84d3478 — 2026-02-17 23:25:1325.48 (n/a)23.50 (n/a)24.72 (n/a)17.92 (n/a)3.14 (n/a)25.47 (n/a)23.48 (n/a)24.71 (n/a)17.91 (n/a)3.14 (n/a)

matrix_vector_mul_8192x2048_4tsi_1024tso_4col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
8b6d0f4 — 2026-03-25 20:46:3940.84 (-1.16%)38.86 (-0.66%)38.49 (-1.61%)37.90 (+1.94%)1.14 (-23.22%)40.82 (-1.16%)38.84 (-0.66%)38.46 (-1.61%)37.87 (+1.94%)1.14 (-23.22%)
dc66da7 — 2026-03-25 19:23:4541.32 (n/a)39.12 (n/a)39.12 (n/a)37.18 (n/a)1.49 (n/a)41.30 (n/a)39.09 (n/a)39.09 (n/a)37.15 (n/a)1.49 (n/a)

matrix_vector_mul_8192x2048_4tsi_1024tso_4col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
897d04e — 2026-03-06 22:12:2540.80 (-0.16%)39.58 (-1.93%)39.75 (-1.71%)38.28 (-2.50%)1.10 (+67.33%)40.78 (-0.16%)39.55 (-1.93%)39.73 (-1.71%)38.26 (-2.50%)1.10 (+67.33%)
84d3478 — 2026-02-17 23:25:1340.87 (n/a)40.36 (n/a)40.45 (n/a)39.26 (n/a)0.66 (n/a)40.84 (n/a)40.33 (n/a)40.42 (n/a)39.24 (n/a)0.65 (n/a)

matrix_vector_mul_8192x2048_4tsi_1024tso_8col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
8b6d0f4 — 2026-03-25 20:46:3941.79 (-12.64%)40.98 (-5.83%)40.78 (-1.04%)40.36 (-0.60%)0.56 (-84.11%)41.76 (-12.64%)40.96 (-5.83%)40.76 (-1.04%)40.34 (-0.60%)0.56 (-84.11%)
dc66da7 — 2026-03-25 19:23:4547.83 (n/a)43.52 (n/a)41.21 (n/a)40.61 (n/a)3.53 (n/a)47.80 (n/a)43.50 (n/a)41.19 (n/a)40.58 (n/a)3.53 (n/a)

matrix_vector_mul_8192x2048_4tsi_1024tso_8col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
897d04e — 2026-03-06 22:12:2549.31 (+6.27%)46.12 (+1.67%)45.30 (-0.27%)44.62 (+1.87%)1.85 (+85.10%)49.28 (+6.27%)46.10 (+1.67%)45.28 (-0.27%)44.59 (+1.87%)1.85 (+85.10%)
84d3478 — 2026-02-17 23:25:1346.40 (n/a)45.37 (n/a)45.43 (n/a)43.80 (n/a)1.00 (n/a)46.37 (n/a)45.34 (n/a)45.40 (n/a)43.78 (n/a)1.00 (n/a)

mem_copy_16_cores_2_chans_2048_tile_128_False

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.05 (+7.16%)0.03 (-9.22%)0.03 (-18.40%)0.02 (-19.93%)0.01 (+42.24%)394.80 (+24.90%)264.56 (+14.35%)259.20 (+22.55%)179.80 (-6.65%)81.04 (+64.09%)
dc66da7 — 2026-03-25 19:23:450.04 (n/a)0.04 (n/a)0.04 (n/a)0.03 (n/a)0.01 (n/a)316.10 (n/a)231.36 (n/a)211.50 (n/a)192.60 (n/a)49.39 (n/a)

mem_copy_16_cores_2_chans_2048_tile_128_False0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.08 (-2.12%)0.06 (-16.44%)0.06 (-21.48%)0.03 (-23.84%)0.02 (+10.15%)247.60 (+31.28%)148.38 (+24.08%)128.30 (+27.41%)100.10 (+2.14%)58.70 (+50.73%)
84d3478 — 2026-02-17 23:25:130.08 (n/a)0.07 (n/a)0.08 (n/a)0.04 (n/a)0.02 (n/a)188.60 (n/a)119.58 (n/a)100.70 (n/a)98.00 (n/a)38.94 (n/a)

mem_copy_1_cols_1_channels_2048_tile_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
003b49d — 2025-11-14 19:30:540.05 (-16.99%)0.05 (-12.47%)0.05 (-7.32%)0.04 (-11.17%)0.00 (-29.61%)197.00 (+12.57%)171.00 (+13.70%)164.00 (+7.89%)153.00 (+20.47%)17.49 (-4.32%)
0999a9a — 2025-11-13 19:24:080.06 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.01 (n/a)175.00 (n/a)150.40 (n/a)152.00 (n/a)127.00 (n/a)18.28 (n/a)

mem_copy_1_cols_2_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
003b49d — 2025-11-14 19:30:540.06 (+5.56%)0.05 (-1.59%)0.05 (-1.39%)0.03 (-14.07%)0.01 (+25.03%)263.00 (+16.37%)166.00 (+4.53%)144.00 (+1.41%)126.00 (-5.26%)49.38 (+43.17%)
0999a9a — 2025-11-13 19:24:080.06 (n/a)0.05 (n/a)0.05 (n/a)0.03 (n/a)0.01 (n/a)226.00 (n/a)158.80 (n/a)142.00 (n/a)133.00 (n/a)34.49 (n/a)

mem_copy_1_cores_1_chans_2048_tile_2048_False

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.09 (+75.06%)0.05 (+7.82%)0.04 (-15.32%)0.03 (-13.78%)0.02 (+425.35%)249.10 (+15.97%)190.02 (+3.98%)211.70 (+18.07%)95.60 (-42.89%)62.62 (+236.69%)
dc66da7 — 2026-03-25 19:23:450.05 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.00 (n/a)214.80 (n/a)182.74 (n/a)179.30 (n/a)167.40 (n/a)18.60 (n/a)

mem_copy_1_cores_1_chans_2048_tile_2048_False0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.11 (-0.24%)0.08 (-17.53%)0.08 (-27.05%)0.06 (-15.43%)0.02 (+4.62%)148.50 (+18.23%)113.02 (+23.38%)103.60 (+37.04%)74.10 (+0.14%)31.69 (+33.33%)
84d3478 — 2026-02-17 23:25:130.11 (n/a)0.09 (n/a)0.11 (n/a)0.07 (n/a)0.02 (n/a)125.60 (n/a)91.60 (n/a)75.60 (n/a)74.00 (n/a)23.77 (n/a)

mem_copy_2_cols_1_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
003b49d — 2025-11-14 19:30:540.06 (-15.32%)0.05 (-17.09%)0.05 (-23.81%)0.04 (-15.54%)0.01 (-18.18%)193.00 (+18.40%)160.80 (+20.36%)168.00 (+31.25%)124.00 (+18.10%)22.92 (+11.39%)
0999a9a — 2025-11-13 19:24:080.07 (n/a)0.06 (n/a)0.06 (n/a)0.05 (n/a)0.01 (n/a)163.00 (n/a)133.60 (n/a)128.00 (n/a)105.00 (n/a)20.58 (n/a)

mem_copy_2_cols_2_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
003b49d — 2025-11-14 19:30:540.06 (-23.02%)0.05 (-6.98%)0.05 (-2.13%)0.04 (-15.31%)0.01 (-33.29%)196.00 (+18.07%)147.00 (+6.52%)141.00 (+2.17%)126.00 (+29.90%)25.27 (+8.13%)
0999a9a — 2025-11-13 19:24:080.08 (n/a)0.06 (n/a)0.06 (n/a)0.05 (n/a)0.01 (n/a)166.00 (n/a)138.00 (n/a)138.00 (n/a)97.00 (n/a)23.37 (n/a)

mem_copy_2_cores_1_chans_2048_tile_1024_False

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.07 (+13.70%)0.05 (-1.93%)0.05 (-7.07%)0.04 (-27.15%)0.02 (+157.23%)230.10 (+37.29%)166.58 (+9.82%)161.90 (+7.65%)109.70 (-12.03%)53.34 (+211.05%)
dc66da7 — 2026-03-25 19:23:450.07 (n/a)0.05 (n/a)0.05 (n/a)0.05 (n/a)0.01 (n/a)167.60 (n/a)151.68 (n/a)150.40 (n/a)124.70 (n/a)17.15 (n/a)

mem_copy_2_cores_1_chans_2048_tile_1024_False0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.11 (+27.03%)0.09 (+25.79%)0.11 (+27.09%)0.05 (+12.74%)0.03 (+43.93%)161.90 (-11.29%)94.34 (-18.35%)76.30 (-21.34%)74.00 (-21.36%)38.01 (+0.49%)
84d3478 — 2026-02-17 23:25:130.09 (n/a)0.08 (n/a)0.08 (n/a)0.04 (n/a)0.02 (n/a)182.50 (n/a)115.54 (n/a)97.00 (n/a)94.10 (n/a)37.82 (n/a)

mem_copy_2_cores_2_chans_2048_tile_1024_False

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.07 (-3.17%)0.05 (-7.71%)0.05 (-6.65%)0.04 (-20.60%)0.02 (+42.64%)223.40 (+25.93%)171.62 (+13.04%)172.10 (+7.09%)120.40 (+3.26%)49.35 (+82.52%)
dc66da7 — 2026-03-25 19:23:450.07 (n/a)0.06 (n/a)0.05 (n/a)0.05 (n/a)0.01 (n/a)177.40 (n/a)151.82 (n/a)160.70 (n/a)116.60 (n/a)27.04 (n/a)

mem_copy_2_cores_2_chans_2048_tile_1024_False0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.10 (+2.70%)0.08 (-5.38%)0.09 (+2.10%)0.04 (-43.30%)0.02 (+170.50%)201.60 (+76.38%)112.08 (+16.53%)90.80 (-2.05%)84.60 (-2.65%)50.14 (+375.75%)
84d3478 — 2026-02-17 23:25:130.09 (n/a)0.09 (n/a)0.09 (n/a)0.07 (n/a)0.01 (n/a)114.30 (n/a)96.18 (n/a)92.70 (n/a)86.90 (n/a)10.54 (n/a)

mem_copy_4_cols_1_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
003b49d — 2025-11-14 19:30:540.07 (+11.97%)0.05 (+3.09%)0.05 (+2.10%)0.03 (-11.35%)0.01 (+41.65%)229.00 (+12.81%)156.00 (-0.51%)143.00 (-2.05%)117.00 (-10.69%)38.21 (+48.71%)
0999a9a — 2025-11-13 19:24:080.06 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.01 (n/a)203.00 (n/a)156.80 (n/a)146.00 (n/a)131.00 (n/a)25.69 (n/a)

mem_copy_4_cols_2_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
003b49d — 2025-11-14 19:30:540.06 (+9.02%)0.05 (+5.59%)0.05 (+0.00%)0.04 (+3.63%)0.01 (+25.48%)193.00 (-3.50%)147.40 (-4.66%)147.00 (+0.00%)122.00 (-8.27%)25.24 (+7.48%)
0999a9a — 2025-11-13 19:24:080.06 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.01 (n/a)200.00 (n/a)154.60 (n/a)147.00 (n/a)133.00 (n/a)23.48 (n/a)

mem_copy_4_cores_1_chans_2048_tile_512_False

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.06 (-2.79%)0.05 (-5.43%)0.05 (+6.39%)0.03 (-12.50%)0.01 (+16.26%)242.20 (+14.25%)184.14 (+7.91%)167.80 (-6.05%)128.30 (+2.89%)46.84 (+42.55%)
dc66da7 — 2026-03-25 19:23:450.07 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.01 (n/a)212.00 (n/a)170.64 (n/a)178.60 (n/a)124.70 (n/a)32.86 (n/a)

mem_copy_4_cores_1_chans_2048_tile_512_False0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.09 (-5.87%)0.08 (-3.61%)0.08 (-0.79%)0.06 (-12.74%)0.01 (+22.04%)146.30 (+14.66%)108.02 (+4.96%)99.30 (+0.81%)95.10 (+6.26%)21.49 (+48.38%)
84d3478 — 2026-02-17 23:25:130.09 (n/a)0.08 (n/a)0.08 (n/a)0.06 (n/a)0.01 (n/a)127.60 (n/a)102.92 (n/a)98.50 (n/a)89.50 (n/a)14.49 (n/a)

mem_copy_4_cores_2_chans_2048_tile_512_False

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.05 (+7.78%)0.04 (-14.27%)0.04 (-10.58%)0.03 (-32.73%)0.01 (+109.81%)316.90 (+48.71%)226.08 (+22.64%)209.60 (+11.85%)150.20 (-7.23%)61.97 (+194.31%)
dc66da7 — 2026-03-25 19:23:450.05 (n/a)0.04 (n/a)0.04 (n/a)0.04 (n/a)0.01 (n/a)213.10 (n/a)184.34 (n/a)187.40 (n/a)161.90 (n/a)21.05 (n/a)

mem_copy_4_cores_2_chans_2048_tile_512_False0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.11 (+13.04%)0.09 (+17.19%)0.10 (+20.43%)0.06 (+17.23%)0.02 (+16.34%)142.80 (-14.70%)91.94 (-14.71%)79.00 (-17.02%)77.70 (-11.50%)28.50 (-14.82%)
84d3478 — 2026-02-17 23:25:130.09 (n/a)0.08 (n/a)0.09 (n/a)0.05 (n/a)0.02 (n/a)167.40 (n/a)107.80 (n/a)95.20 (n/a)87.80 (n/a)33.46 (n/a)

mem_copy_8_cols_1_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
003b49d — 2025-11-14 19:30:540.05 (-13.48%)0.05 (-17.79%)0.05 (-15.33%)0.03 (-25.09%)0.01 (+6.67%)267.00 (+33.50%)176.80 (+24.51%)150.00 (+18.11%)141.00 (+15.57%)47.41 (+61.41%)
0999a9a — 2025-11-13 19:24:080.06 (n/a)0.06 (n/a)0.06 (n/a)0.04 (n/a)0.01 (n/a)200.00 (n/a)142.00 (n/a)127.00 (n/a)122.00 (n/a)29.37 (n/a)

mem_copy_8_cols_2_channels_2048_tile_128

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
003b49d — 2025-11-14 19:30:540.05 (+13.38%)0.05 (+23.85%)0.05 (+37.91%)0.05 (+30.72%)0.00 (-56.85%)166.00 (-23.50%)153.40 (-20.60%)153.00 (-27.49%)142.00 (-11.80%)7.63 (-70.60%)
0999a9a — 2025-11-13 19:24:080.05 (n/a)0.04 (n/a)0.04 (n/a)0.04 (n/a)0.01 (n/a)217.00 (n/a)193.20 (n/a)211.00 (n/a)161.00 (n/a)25.96 (n/a)

mem_copy_8_cores_1_chans_2048_tile_256_False

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.06 (-5.71%)0.04 (+0.46%)0.04 (+2.21%)0.03 (-7.12%)0.01 (-7.04%)278.00 (+7.67%)209.24 (-0.41%)216.20 (-2.17%)141.90 (+6.05%)51.53 (+9.47%)
dc66da7 — 2026-03-25 19:23:450.06 (n/a)0.04 (n/a)0.04 (n/a)0.03 (n/a)0.01 (n/a)258.20 (n/a)210.10 (n/a)221.00 (n/a)133.80 (n/a)47.07 (n/a)

mem_copy_8_cores_1_chans_2048_tile_256_False0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.08 (-0.39%)0.07 (+10.07%)0.08 (+1.07%)0.06 (+38.38%)0.01 (-47.35%)136.70 (-27.75%)110.64 (-13.07%)104.70 (-1.13%)103.00 (+0.39%)14.59 (-60.78%)
84d3478 — 2026-02-17 23:25:130.08 (n/a)0.07 (n/a)0.08 (n/a)0.04 (n/a)0.02 (n/a)189.20 (n/a)127.28 (n/a)105.90 (n/a)102.60 (n/a)37.20 (n/a)

mem_copy_8_cores_2_chans_2048_tile_256_False

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.06 (+19.11%)0.04 (+1.88%)0.04 (-11.30%)0.03 (+24.20%)0.01 (+4.86%)262.00 (-19.48%)199.56 (-3.59%)200.50 (+12.70%)139.70 (-16.05%)45.93 (-31.63%)
dc66da7 — 2026-03-25 19:23:450.05 (n/a)0.04 (n/a)0.05 (n/a)0.03 (n/a)0.01 (n/a)325.40 (n/a)207.00 (n/a)177.90 (n/a)166.40 (n/a)67.19 (n/a)

mem_copy_8_cores_2_chans_2048_tile_256_False0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.08 (-17.85%)0.07 (-24.66%)0.08 (-18.40%)0.04 (-32.08%)0.02 (+21.78%)185.30 (+47.18%)127.72 (+37.87%)102.00 (+22.45%)100.40 (+21.70%)38.66 (+106.34%)
84d3478 — 2026-02-17 23:25:130.10 (n/a)0.09 (n/a)0.10 (n/a)0.07 (n/a)0.01 (n/a)125.90 (n/a)92.64 (n/a)83.30 (n/a)82.50 (n/a)18.74 (n/a)

mha

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.18 (+0.10%)0.18 (+0.09%)0.18 (+0.09%)0.18 (+0.08%)0.00 (+11.08%)47507.50 (-0.08%)47468.96 (-0.09%)47481.10 (-0.09%)47413.10 (-0.10%)37.32 (+10.96%)
dc66da7 — 2026-03-25 19:23:450.18 (n/a)0.18 (n/a)0.18 (n/a)0.18 (n/a)0.00 (n/a)47544.20 (n/a)47512.60 (n/a)47521.80 (n/a)47459.60 (n/a)33.63 (n/a)

mha0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
84d3478 — 2026-02-17 23:25:130.18 (-0.51%)0.18 (-0.36%)0.18 (-0.41%)0.18 (-0.21%)0.00 (-35.21%)47468.60 (+0.21%)47403.94 (+0.36%)47426.00 (+0.41%)47314.40 (+0.51%)68.83 (-34.73%)
1531781 — 2026-01-23 18:49:240.18 (n/a)0.18 (n/a)0.18 (n/a)0.18 (n/a)0.00 (n/a)47367.50 (n/a)47232.68 (n/a)47232.00 (n/a)47073.00 (n/a)105.45 (n/a)

mha_16384_64_1_8_0_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.18 (n/a)0.18 (n/a)0.18 (n/a)0.18 (n/a)0.00 (n/a)47533.70 (n/a)47380.56 (n/a)47337.90 (n/a)47268.00 (n/a)110.60 (n/a)

relu_1_cols_1_channels_2048_tile_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.07 (+14.06%)0.05 (+0.31%)0.05 (-3.71%)0.04 (-18.38%)0.01 (+129.69%)208.90 (+22.52%)161.96 (+2.56%)166.90 (+3.86%)120.40 (-12.31%)33.36 (+144.00%)
dc66da7 — 2026-03-25 19:23:450.06 (n/a)0.05 (n/a)0.05 (n/a)0.05 (n/a)0.00 (n/a)170.50 (n/a)157.92 (n/a)160.70 (n/a)137.30 (n/a)13.67 (n/a)

relu_2_cols_1_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.06 (-20.92%)0.05 (-6.86%)0.05 (-1.28%)0.03 (-20.49%)0.01 (-19.18%)264.20 (+25.81%)174.76 (+7.41%)149.00 (+1.29%)140.20 (+26.42%)52.29 (+24.11%)
dc66da7 — 2026-03-25 19:23:450.07 (n/a)0.05 (n/a)0.06 (n/a)0.04 (n/a)0.01 (n/a)210.00 (n/a)162.70 (n/a)147.10 (n/a)110.90 (n/a)42.13 (n/a)

relu_4_cols_1_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.05 (-8.73%)0.04 (-10.26%)0.04 (-4.04%)0.03 (-16.73%)0.01 (+5.34%)259.40 (+20.09%)206.80 (+12.91%)200.50 (+4.21%)150.20 (+9.56%)45.96 (+41.47%)
dc66da7 — 2026-03-25 19:23:450.06 (n/a)0.05 (n/a)0.04 (n/a)0.04 (n/a)0.01 (n/a)216.00 (n/a)183.16 (n/a)192.40 (n/a)137.10 (n/a)32.49 (n/a)

relu_8_cols_1_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.05 (+4.18%)0.04 (-0.33%)0.04 (-1.55%)0.02 (-20.26%)0.01 (+47.86%)369.50 (+25.38%)233.74 (+5.19%)215.30 (+1.56%)165.20 (-4.01%)80.83 (+78.85%)
dc66da7 — 2026-03-25 19:23:450.05 (n/a)0.04 (n/a)0.04 (n/a)0.03 (n/a)0.01 (n/a)294.70 (n/a)222.20 (n/a)212.00 (n/a)172.10 (n/a)45.19 (n/a)

rms_norm_1_cols_1_channels_2048_tile_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.06 (-18.06%)0.05 (-2.61%)0.05 (-1.31%)0.04 (+12.13%)0.01 (-43.39%)226.00 (-10.81%)172.66 (-1.55%)163.30 (+1.30%)140.00 (+22.06%)32.37 (-37.02%)
dc66da7 — 2026-03-25 19:23:450.07 (n/a)0.05 (n/a)0.05 (n/a)0.03 (n/a)0.01 (n/a)253.40 (n/a)175.38 (n/a)161.20 (n/a)114.70 (n/a)51.40 (n/a)

rms_norm_1_cols_2_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.08 (+16.34%)0.05 (+6.02%)0.05 (-3.17%)0.04 (+12.01%)0.02 (+27.81%)218.40 (-10.71%)162.92 (-3.80%)150.40 (+3.23%)108.00 (-14.01%)52.41 (+3.43%)
dc66da7 — 2026-03-25 19:23:450.07 (n/a)0.05 (n/a)0.06 (n/a)0.03 (n/a)0.01 (n/a)244.60 (n/a)169.36 (n/a)145.70 (n/a)125.60 (n/a)50.68 (n/a)

rms_norm_2_cols_1_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.06 (+9.52%)0.04 (-9.59%)0.04 (-17.77%)0.03 (-18.95%)0.01 (+55.38%)255.30 (+23.39%)199.88 (+13.52%)205.30 (+21.62%)137.80 (-8.68%)43.62 (+70.31%)
dc66da7 — 2026-03-25 19:23:450.05 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.01 (n/a)206.90 (n/a)176.08 (n/a)168.80 (n/a)150.90 (n/a)25.61 (n/a)

rms_norm_2_cols_2_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.06 (+6.05%)0.04 (+0.61%)0.04 (-6.97%)0.03 (-2.64%)0.01 (+44.42%)234.40 (+2.72%)195.80 (+1.42%)215.50 (+7.48%)141.90 (-5.65%)40.78 (+43.76%)
dc66da7 — 2026-03-25 19:23:450.05 (n/a)0.04 (n/a)0.04 (n/a)0.04 (n/a)0.01 (n/a)228.20 (n/a)193.06 (n/a)200.50 (n/a)150.40 (n/a)28.37 (n/a)

rms_norm_4_cols_1_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.06 (+10.73%)0.05 (-1.13%)0.05 (+5.17%)0.04 (-10.70%)0.01 (+73.44%)225.10 (+11.99%)181.84 (+3.00%)170.50 (-4.91%)140.00 (-9.68%)33.21 (+78.26%)
dc66da7 — 2026-03-25 19:23:450.05 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.00 (n/a)201.00 (n/a)176.54 (n/a)179.30 (n/a)155.00 (n/a)18.63 (n/a)

rms_norm_4_cols_2_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.06 (-0.51%)0.04 (-4.22%)0.04 (-6.40%)0.03 (+19.93%)0.01 (-12.26%)264.40 (-16.62%)208.96 (+1.64%)223.40 (+6.84%)140.70 (+0.50%)52.99 (-24.83%)
dc66da7 — 2026-03-25 19:23:450.06 (n/a)0.04 (n/a)0.04 (n/a)0.03 (n/a)0.01 (n/a)317.10 (n/a)205.58 (n/a)209.10 (n/a)140.00 (n/a)70.49 (n/a)

rms_norm_8_cols_1_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.06 (-8.19%)0.04 (-16.80%)0.03 (-25.89%)0.03 (-16.40%)0.01 (+5.02%)283.50 (+19.62%)220.72 (+22.74%)240.30 (+34.92%)128.00 (+8.94%)58.39 (+33.24%)
dc66da7 — 2026-03-25 19:23:450.07 (n/a)0.05 (n/a)0.05 (n/a)0.03 (n/a)0.01 (n/a)237.00 (n/a)179.82 (n/a)178.10 (n/a)117.50 (n/a)43.82 (n/a)

rms_norm_8_cols_2_channels_2048_tile_128

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.04 (+4.11%)0.04 (+2.57%)0.04 (-1.84%)0.03 (+8.94%)0.00 (+2.64%)250.80 (-8.20%)224.64 (-2.58%)233.20 (+1.88%)191.20 (-3.97%)25.83 (-9.64%)
dc66da7 — 2026-03-25 19:23:450.04 (n/a)0.04 (n/a)0.04 (n/a)0.03 (n/a)0.00 (n/a)273.20 (n/a)230.60 (n/a)228.90 (n/a)199.10 (n/a)28.59 (n/a)

rope_1_cols_2_channels_4096_tile_4096_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
1b2d230 — 2026-01-13 22:38:030.19 (-21.66%)0.17 (-8.53%)0.18 (+2.09%)0.14 (-6.71%)0.02 (-38.07%)174.10 (+7.20%)150.06 (+8.11%)139.60 (-2.04%)129.80 (+27.63%)20.36 (-10.65%)
d48746f — 2026-01-12 22:17:100.24 (n/a)0.18 (n/a)0.17 (n/a)0.15 (n/a)0.04 (n/a)162.40 (n/a)138.80 (n/a)142.50 (n/a)101.70 (n/a)22.79 (n/a)

rope_1c_32rows_512cols_32arows_0m

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.91 (+31.28%)0.72 (+30.97%)0.75 (+50.09%)0.52 (+23.14%)0.16 (+34.35%)189.50 (-18.81%)142.68 (-23.26%)131.80 (-33.40%)108.20 (-23.86%)33.21 (-13.83%)
dc66da7 — 2026-03-25 19:23:450.69 (n/a)0.55 (n/a)0.50 (n/a)0.42 (n/a)0.12 (n/a)233.40 (n/a)185.92 (n/a)197.90 (n/a)142.10 (n/a)38.54 (n/a)

rope_1c_32rows_512cols_8arows_0m

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.71 (+13.56%)0.53 (+16.33%)0.51 (+20.62%)0.43 (+13.26%)0.12 (+12.08%)172.60 (-11.71%)143.86 (-14.09%)145.70 (-17.07%)103.70 (-11.97%)28.08 (-12.19%)
dc66da7 — 2026-03-25 19:23:450.63 (n/a)0.46 (n/a)0.42 (n/a)0.38 (n/a)0.10 (n/a)195.50 (n/a)167.46 (n/a)175.70 (n/a)117.80 (n/a)31.98 (n/a)

rope_2_cols_2_channels_4096_tile_2048_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
1b2d230 — 2026-01-13 22:38:030.21 (-0.56%)0.17 (-10.62%)0.16 (-13.58%)0.13 (-13.33%)0.03 (+17.05%)183.70 (+15.39%)151.76 (+12.77%)155.20 (+15.73%)117.30 (+0.60%)24.24 (+35.73%)
d48746f — 2026-01-12 22:17:100.21 (n/a)0.19 (n/a)0.18 (n/a)0.15 (n/a)0.02 (n/a)159.20 (n/a)134.58 (n/a)134.10 (n/a)116.60 (n/a)17.86 (n/a)

rope_2c_32rows_512cols_32arows_0m

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.86 (+12.27%)0.68 (+4.35%)0.67 (-1.63%)0.56 (+6.41%)0.12 (+18.02%)174.80 (-6.02%)146.78 (-3.85%)146.60 (+1.66%)114.70 (-10.88%)23.97 (-1.25%)
dc66da7 — 2026-03-25 19:23:450.76 (n/a)0.66 (n/a)0.68 (n/a)0.53 (n/a)0.10 (n/a)186.00 (n/a)152.66 (n/a)144.20 (n/a)128.70 (n/a)24.27 (n/a)

rope_2c_32rows_512cols_8arows_0m

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.63 (+15.38%)0.42 (-5.03%)0.37 (-15.06%)0.26 (-23.68%)0.14 (+76.55%)289.00 (+31.07%)192.28 (+12.08%)197.90 (+17.73%)117.80 (-13.32%)64.38 (+98.12%)
dc66da7 — 2026-03-25 19:23:450.54 (n/a)0.44 (n/a)0.44 (n/a)0.33 (n/a)0.08 (n/a)220.50 (n/a)171.56 (n/a)168.10 (n/a)135.90 (n/a)32.49 (n/a)

rope_4_cols_2_channels_4096_tile_1024_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
1b2d230 — 2026-01-13 22:38:030.22 (+18.92%)0.17 (+6.24%)0.17 (+5.57%)0.14 (-0.50%)0.03 (+90.77%)171.10 (+0.47%)147.40 (-4.39%)148.10 (-5.31%)110.10 (-15.89%)23.59 (+60.67%)
d48746f — 2026-01-12 22:17:100.19 (n/a)0.16 (n/a)0.16 (n/a)0.14 (n/a)0.02 (n/a)170.30 (n/a)154.16 (n/a)156.40 (n/a)130.90 (n/a)14.69 (n/a)

rope_8_cols_2_channels_4096_tile_512_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
1b2d230 — 2026-01-13 22:38:030.17 (-1.30%)0.15 (+6.63%)0.15 (-8.78%)0.14 (+39.68%)0.01 (-67.35%)174.90 (-28.41%)163.42 (-10.36%)167.40 (+9.63%)146.50 (+1.31%)11.19 (-76.17%)
d48746f — 2026-01-12 22:17:100.17 (n/a)0.14 (n/a)0.16 (n/a)0.10 (n/a)0.03 (n/a)244.30 (n/a)182.30 (n/a)152.70 (n/a)144.60 (n/a)46.98 (n/a)

rope_8c_32rows_512cols_32arows_0m

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.60 (+0.58%)0.47 (-10.94%)0.47 (-14.67%)0.39 (-14.06%)0.09 (+57.47%)252.70 (+16.34%)213.46 (+14.31%)211.20 (+17.20%)163.30 (-0.61%)38.58 (+85.72%)
dc66da7 — 2026-03-25 19:23:450.60 (n/a)0.53 (n/a)0.55 (n/a)0.45 (n/a)0.06 (n/a)217.20 (n/a)186.74 (n/a)180.20 (n/a)164.30 (n/a)20.77 (n/a)

rope_8c_32rows_512cols_8arows_0m

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.58 (+33.33%)0.38 (-2.44%)0.32 (-20.46%)0.28 (-18.27%)0.13 (+203.34%)264.90 (+22.36%)209.28 (+9.84%)227.20 (+25.73%)126.60 (-25.00%)59.71 (+180.42%)
dc66da7 — 2026-03-25 19:23:450.44 (n/a)0.39 (n/a)0.41 (n/a)0.34 (n/a)0.04 (n/a)216.50 (n/a)190.54 (n/a)180.70 (n/a)168.80 (n/a)21.29 (n/a)

rows_32-cols_512-angle_rows_32-aie_columns_1-method_type_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
7d76781 — 2026-03-25 21:49:380.41 (n/a)0.36 (n/a)0.38 (n/a)0.23 (n/a)0.07 (n/a)425.40 (n/a)288.36 (n/a)256.40 (n/a)237.40 (n/a)78.56 (n/a)

rows_32-cols_512-angle_rows_32-aie_columns_2-method_type_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
7d76781 — 2026-03-25 21:49:380.42 (n/a)0.36 (n/a)0.39 (n/a)0.17 (n/a)0.10 (n/a)572.70 (n/a)310.88 (n/a)249.00 (n/a)234.20 (n/a)146.58 (n/a)

rows_32-cols_512-angle_rows_32-aie_columns_8-method_type_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
7d76781 — 2026-03-25 21:49:380.41 (n/a)0.34 (n/a)0.40 (n/a)0.10 (n/a)0.13 (n/a)958.10 (n/a)390.38 (n/a)243.70 (n/a)238.70 (n/a)317.64 (n/a)

rows_32-cols_512-angle_rows_8-aie_columns_1-method_type_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
7d76781 — 2026-03-25 21:49:380.26 (n/a)0.25 (n/a)0.25 (n/a)0.22 (n/a)0.02 (n/a)331.10 (n/a)297.56 (n/a)290.20 (n/a)282.80 (n/a)19.76 (n/a)

rows_32-cols_512-angle_rows_8-aie_columns_2-method_type_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
7d76781 — 2026-03-25 21:49:380.26 (n/a)0.23 (n/a)0.25 (n/a)0.18 (n/a)0.03 (n/a)406.50 (n/a)319.66 (n/a)296.70 (n/a)285.40 (n/a)49.57 (n/a)

rows_32-cols_512-angle_rows_8-aie_columns_8-method_type_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
7d76781 — 2026-03-25 21:49:380.24 (n/a)0.21 (n/a)0.24 (n/a)0.10 (n/a)0.06 (n/a)731.60 (n/a)393.10 (n/a)309.60 (n/a)303.10 (n/a)189.26 (n/a)

seq_len_16384-dim_64-num_heads_1-num_pipelines_8-num_kv_heads_0

No metrics available.

seq_len_256-embedding_dim_2048-hidden_dim_2048-prio_accuracy_False

No metrics available.

sigmoid_1_cols_1_channels_2048_tile_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.05 (-14.16%)0.05 (-6.29%)0.05 (-2.65%)0.04 (-7.71%)0.01 (-21.79%)225.80 (+8.35%)180.84 (+6.33%)170.90 (+2.70%)163.30 (+16.48%)25.41 (+0.55%)
dc66da7 — 2026-03-25 19:23:450.06 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.01 (n/a)208.40 (n/a)170.08 (n/a)166.40 (n/a)140.20 (n/a)25.28 (n/a)

sigmoid_2_cols_1_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.06 (+4.68%)0.05 (+2.94%)0.04 (-13.06%)0.04 (+15.35%)0.01 (+8.49%)232.90 (-13.32%)179.44 (-3.16%)197.20 (+15.05%)127.30 (-4.50%)46.58 (-13.93%)
dc66da7 — 2026-03-25 19:23:450.06 (n/a)0.05 (n/a)0.05 (n/a)0.03 (n/a)0.01 (n/a)268.70 (n/a)185.30 (n/a)171.40 (n/a)133.30 (n/a)54.12 (n/a)

sigmoid_4_cols_1_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.07 (+10.37%)0.05 (-2.49%)0.04 (-7.15%)0.04 (-7.47%)0.01 (+64.06%)210.80 (+8.10%)175.44 (+5.70%)183.30 (+7.70%)117.30 (-9.42%)39.46 (+67.78%)
dc66da7 — 2026-03-25 19:23:450.06 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.01 (n/a)195.00 (n/a)165.98 (n/a)170.20 (n/a)129.50 (n/a)23.52 (n/a)

sigmoid_8_cols_1_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.06 (+29.56%)0.05 (+22.81%)0.05 (+37.30%)0.03 (-8.45%)0.01 (+218.31%)242.70 (+9.23%)173.04 (-14.71%)150.20 (-27.16%)134.70 (-22.81%)46.84 (+169.71%)
dc66da7 — 2026-03-25 19:23:450.05 (n/a)0.04 (n/a)0.04 (n/a)0.04 (n/a)0.00 (n/a)222.20 (n/a)202.88 (n/a)206.20 (n/a)174.50 (n/a)17.37 (n/a)

silu_1_cols_1_channels_2048_tile_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.07 (+9.38%)0.05 (-4.51%)0.04 (-23.77%)0.04 (+15.02%)0.01 (+5.83%)207.90 (-13.05%)169.28 (+4.04%)188.60 (+31.15%)114.40 (-8.63%)38.46 (-17.21%)
dc66da7 — 2026-03-25 19:23:450.07 (n/a)0.05 (n/a)0.06 (n/a)0.03 (n/a)0.01 (n/a)239.10 (n/a)162.70 (n/a)143.80 (n/a)125.20 (n/a)46.46 (n/a)

silu_2_cols_1_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.06 (-4.63%)0.05 (-13.54%)0.05 (-12.20%)0.04 (-16.50%)0.01 (+19.45%)213.90 (+19.76%)180.30 (+16.99%)181.70 (+13.92%)134.00 (+4.85%)30.84 (+49.10%)
dc66da7 — 2026-03-25 19:23:450.06 (n/a)0.05 (n/a)0.05 (n/a)0.05 (n/a)0.01 (n/a)178.60 (n/a)154.12 (n/a)159.50 (n/a)127.80 (n/a)20.68 (n/a)

silu_4_cols_1_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.06 (+6.26%)0.05 (-8.81%)0.04 (-14.86%)0.04 (-17.09%)0.01 (+81.20%)214.80 (+20.61%)176.86 (+11.68%)182.90 (+17.47%)133.30 (-5.86%)30.25 (+101.37%)
dc66da7 — 2026-03-25 19:23:450.06 (n/a)0.05 (n/a)0.05 (n/a)0.05 (n/a)0.00 (n/a)178.10 (n/a)158.36 (n/a)155.70 (n/a)141.60 (n/a)15.02 (n/a)

silu_8_cols_1_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.06 (-6.11%)0.05 (-10.71%)0.04 (-10.38%)0.04 (-11.78%)0.01 (+1.86%)224.60 (+13.38%)185.30 (+12.62%)186.00 (+11.58%)140.40 (+6.53%)33.11 (+23.38%)
dc66da7 — 2026-03-25 19:23:450.06 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.01 (n/a)198.10 (n/a)164.54 (n/a)166.70 (n/a)131.80 (n/a)26.84 (n/a)

softmax_1_cols_2_channels_4096_tile_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.19 (-8.94%)0.17 (-8.66%)0.17 (-19.83%)0.14 (+4.32%)0.02 (-35.13%)115.80 (-4.14%)99.26 (+7.47%)98.60 (+24.65%)85.00 (+9.82%)13.65 (-30.86%)
84d3478 — 2026-02-17 23:25:130.21 (n/a)0.18 (n/a)0.21 (n/a)0.14 (n/a)0.04 (n/a)120.80 (n/a)92.36 (n/a)79.10 (n/a)77.40 (n/a)19.75 (n/a)

softmax_2_cols_2_channels_32768_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.64 (-42.09%)0.60 (-31.83%)0.59 (-30.19%)0.54 (-6.68%)0.05 (-76.96%)242.70 (+7.15%)221.12 (+40.50%)220.30 (+43.24%)203.40 (+72.67%)16.99 (-59.06%)
dc66da7 — 2026-03-25 19:23:451.11 (n/a)0.87 (n/a)0.85 (n/a)0.58 (n/a)0.20 (n/a)226.50 (n/a)157.38 (n/a)153.80 (n/a)117.80 (n/a)41.50 (n/a)

softmax_2_cols_2_channels_32768_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.68 (-34.93%)0.62 (-20.25%)0.60 (-17.49%)0.55 (-10.32%)0.05 (-71.29%)240.30 (+11.51%)214.32 (+20.87%)218.20 (+21.22%)193.80 (+53.69%)18.64 (-52.51%)
dc66da7 — 2026-03-25 19:23:451.04 (n/a)0.77 (n/a)0.73 (n/a)0.61 (n/a)0.18 (n/a)215.50 (n/a)177.32 (n/a)180.00 (n/a)126.10 (n/a)39.26 (n/a)

softmax_2_cols_2_channels_4096_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.20 (-0.87%)0.16 (+6.57%)0.17 (-9.55%)0.11 (+29.81%)0.03 (-44.34%)145.70 (-22.99%)104.68 (-15.00%)95.00 (+10.59%)83.80 (+0.84%)24.16 (-54.23%)
84d3478 — 2026-02-17 23:25:130.20 (n/a)0.15 (n/a)0.19 (n/a)0.09 (n/a)0.06 (n/a)189.20 (n/a)123.16 (n/a)85.90 (n/a)83.10 (n/a)52.80 (n/a)

softmax_2_cols_2_channels_4096_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.25 (+23.53%)0.18 (+9.42%)0.18 (-3.86%)0.11 (+2.99%)0.05 (+19.24%)144.10 (-2.90%)98.98 (-8.18%)92.10 (+4.07%)64.30 (-19.02%)29.58 (-7.13%)
84d3478 — 2026-02-17 23:25:130.21 (n/a)0.16 (n/a)0.19 (n/a)0.11 (n/a)0.04 (n/a)148.40 (n/a)107.80 (n/a)88.50 (n/a)79.40 (n/a)31.85 (n/a)

softmax_4_cols_4_channels_32768_tile_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.66 (+11.72%)0.50 (-0.51%)0.54 (-4.72%)0.35 (+6.44%)0.13 (+14.83%)377.70 (-6.04%)276.32 (+1.24%)242.20 (+4.94%)199.30 (-10.51%)76.50 (-0.31%)
dc66da7 — 2026-03-25 19:23:450.59 (n/a)0.51 (n/a)0.57 (n/a)0.33 (n/a)0.11 (n/a)402.00 (n/a)272.94 (n/a)230.80 (n/a)222.70 (n/a)76.73 (n/a)

swiglu

No metrics available.

swiglu_decode_1x2048x2048

No metrics available.

swiglu_decode_1x2048x2048_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.00 (-0.21%)0.00 (+7.74%)0.00 (+0.37%)0.00 (+8.84%)0.00 (-19.58%)5170.80 (-8.12%)3761.80 (-8.70%)3408.60 (-0.36%)3407.30 (+0.21%)787.66 (-22.98%)
84d3478 — 2026-02-17 23:25:130.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)5627.70 (n/a)4120.44 (n/a)3421.00 (n/a)3400.30 (n/a)1022.61 (n/a)

swiglu_prefill_256x2048x2048

No metrics available.

tanh_1_cols_1_channels_2048_tile_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.06 (-7.39%)0.04 (-16.97%)0.04 (-14.92%)0.03 (-28.80%)0.01 (+23.25%)296.40 (+40.47%)201.28 (+24.69%)188.60 (+17.51%)141.90 (+7.99%)59.45 (+90.30%)
dc66da7 — 2026-03-25 19:23:450.06 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.01 (n/a)211.00 (n/a)161.42 (n/a)160.50 (n/a)131.40 (n/a)31.24 (n/a)

tanh_2_cols_1_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.06 (-14.29%)0.05 (-6.63%)0.05 (+2.99%)0.03 (-30.73%)0.01 (+5.83%)293.30 (+44.41%)176.82 (+11.67%)159.70 (-2.92%)128.50 (+16.71%)67.26 (+86.47%)
dc66da7 — 2026-03-25 19:23:450.07 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.01 (n/a)203.10 (n/a)158.34 (n/a)164.50 (n/a)110.10 (n/a)36.07 (n/a)

tanh_4_cols_1_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.06 (-27.96%)0.05 (-4.84%)0.05 (+4.02%)0.04 (+16.65%)0.01 (-63.57%)187.60 (-14.30%)167.32 (-1.72%)171.90 (-3.86%)133.00 (+38.83%)20.69 (-55.09%)
dc66da7 — 2026-03-25 19:23:450.09 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.02 (n/a)218.90 (n/a)170.24 (n/a)178.80 (n/a)95.80 (n/a)46.06 (n/a)

tanh_8_cols_1_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.05 (-24.07%)0.04 (-15.48%)0.04 (-23.02%)0.03 (-3.09%)0.01 (-53.59%)262.30 (+3.19%)198.18 (+12.63%)183.30 (+29.91%)173.30 (+31.69%)36.55 (-34.15%)
dc66da7 — 2026-03-25 19:23:450.06 (n/a)0.05 (n/a)0.06 (n/a)0.03 (n/a)0.01 (n/a)254.20 (n/a)175.96 (n/a)141.10 (n/a)131.60 (n/a)55.51 (n/a)

transpose_2048_M_64_N_1_cols_1_channels_64_m_64_n_8_s

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:392.51 (-25.94%)2.10 (-20.30%)2.01 (-23.11%)1.76 (-13.49%)0.29 (-41.75%)298.70 (+15.60%)253.30 (+23.92%)261.10 (+30.09%)208.60 (+35.02%)33.97 (-9.04%)
dc66da7 — 2026-03-25 19:23:453.39 (n/a)2.64 (n/a)2.61 (n/a)2.03 (n/a)0.49 (n/a)258.40 (n/a)204.40 (n/a)200.70 (n/a)154.50 (n/a)37.34 (n/a)

transpose_2048_M_64_N_1_cols_1_channels_64_m_64_n_8_s0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:254.43 (+0.85%)3.73 (+3.83%)4.02 (-1.30%)2.97 (+16.77%)0.65 (-24.99%)176.70 (-14.35%)144.46 (-6.24%)130.40 (+1.32%)118.40 (-0.84%)26.60 (-35.16%)
84d3478 — 2026-02-17 23:25:134.39 (n/a)3.59 (n/a)4.07 (n/a)2.54 (n/a)0.87 (n/a)206.30 (n/a)154.08 (n/a)128.70 (n/a)119.40 (n/a)41.02 (n/a)

transpose_2048_M_64_N_1_cols_2_channels_64_m_64_n_8_s

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:392.72 (-10.04%)2.30 (-13.01%)2.26 (-13.66%)1.88 (-21.25%)0.34 (+46.90%)278.20 (+26.97%)231.60 (+16.34%)232.20 (+15.81%)192.40 (+11.15%)34.64 (+109.20%)
dc66da7 — 2026-03-25 19:23:453.03 (n/a)2.65 (n/a)2.61 (n/a)2.39 (n/a)0.23 (n/a)219.10 (n/a)199.08 (n/a)200.50 (n/a)173.10 (n/a)16.56 (n/a)

transpose_2048_M_64_N_1_cols_2_channels_64_m_64_n_8_s0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:255.27 (-14.18%)4.35 (-2.49%)4.45 (+2.15%)2.78 (-22.18%)0.98 (-3.38%)188.60 (+28.47%)126.98 (+4.20%)117.90 (-2.08%)99.40 (+16.53%)36.02 (+50.15%)
84d3478 — 2026-02-17 23:25:136.14 (n/a)4.46 (n/a)4.35 (n/a)3.57 (n/a)1.02 (n/a)146.80 (n/a)121.86 (n/a)120.40 (n/a)85.30 (n/a)23.99 (n/a)

weighted_rms_norm_1_cols_2_channels_2048_weights_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.09 (+6.38%)0.07 (-4.87%)0.07 (+1.68%)0.05 (-23.45%)0.01 (+101.20%)230.80 (+30.62%)177.68 (+7.40%)170.00 (-1.68%)138.30 (-5.98%)33.65 (+149.29%)
dc66da7 — 2026-03-25 19:23:450.08 (n/a)0.07 (n/a)0.07 (n/a)0.07 (n/a)0.01 (n/a)176.70 (n/a)165.44 (n/a)172.90 (n/a)147.10 (n/a)13.50 (n/a)

weighted_rms_norm_2_cols_2_channels_2048_weights_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.07 (-25.98%)0.06 (-18.74%)0.06 (-7.04%)0.04 (-20.93%)0.01 (-34.18%)237.00 (+26.47%)189.06 (+22.07%)179.50 (+7.55%)151.40 (+35.06%)34.27 (+14.91%)
dc66da7 — 2026-03-25 19:23:450.09 (n/a)0.07 (n/a)0.06 (n/a)0.05 (n/a)0.01 (n/a)187.40 (n/a)154.88 (n/a)166.90 (n/a)112.10 (n/a)29.83 (n/a)

weighted_rms_norm_4_cols_2_channels_2048_weights_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.06 (-13.39%)0.05 (-21.43%)0.04 (-31.53%)0.04 (-12.12%)0.01 (-0.80%)237.90 (+13.77%)205.32 (+28.15%)225.30 (+46.01%)147.80 (+15.47%)38.15 (+27.36%)
dc66da7 — 2026-03-25 19:23:450.07 (n/a)0.06 (n/a)0.06 (n/a)0.04 (n/a)0.01 (n/a)209.10 (n/a)160.22 (n/a)154.30 (n/a)128.00 (n/a)29.95 (n/a)

weighted_rms_norm_8_cols_2_channels_2048_weights_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.07 (+33.02%)0.04 (-11.48%)0.04 (-24.93%)0.02 (-32.99%)0.02 (+124.58%)371.50 (+49.26%)251.20 (+24.91%)247.70 (+33.17%)124.90 (-24.85%)88.76 (+136.07%)
dc66da7 — 2026-03-25 19:23:450.05 (n/a)0.04 (n/a)0.05 (n/a)0.03 (n/a)0.01 (n/a)248.90 (n/a)201.10 (n/a)186.00 (n/a)166.20 (n/a)37.60 (n/a)

@hunhoffe hunhoffe changed the title Refactor Part 2 Refactor Part 2: Integration of mlir-aie Features Mar 26, 2026
Two bugs introduced by the .o-linking refactor:

1. PeanoCompilationRule._prefix_symbols/_rename_symbols hardcoded
   peano_dir/bin/llvm-objcopy, which doesn't exist in all CI
   environments (e.g. when llama_npu.py runs as a subprocess with a
   different peano_dir resolution). Add _find_tool() that searches
   peano_dir/bin, mlir_aie_dir/bin, and system PATH (both bare name and
   name-18 variant, matching the devel branch convention).

2. mha/op.py compiled softmax.cc into mha_softmax.o, but no Kernel()
   declaration in design.py references mha_softmax.o. The
   aie-assign-core-link-files pass only traces direct func.call edges,
   so mha_softmax.o was never added to any core's link_files, causing
   the linker to fail with an undefined symbol for partial_softmax_bf16
   (called by partial_softmax in mha_mha.o). Fix: remove the separate
   mha_softmax.o artifact and instead pass -include softmax.cc when
   compiling mha.cc, so partial_softmax_bf16 is compiled directly into
   mha_mha.o.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
@github-actions
Copy link
Contributor

📊 Test Results for Test Example Applications

dd8281d (2026_03_26_01_30_35)

IRONCLAD

Tested on 2026_03_26_01_30_35 at commit dd8281d.

Test Checks TTFT (mean)TPS (mean)
llama_3.2_1b_prompt_1024_tokens_1 ❌ 0/5 n/a n/a
llama_3.2_1b_prompt_1024_tokens_40 ❌ 0/5 n/a n/a
llama_3.2_1b_prompt_13_tokens_1 ❌ 0/5 n/a n/a
llama_3.2_1b_prompt_13_tokens_40 ❌ 0/5 n/a n/a
📈 Trends (vs main branch) for Test Example Applications

dd8281d (2026_03_26_01_30_35)

IRONCLAD Trends

llama_3.2_1b

Commit/Date Num Tokens (max)Num Tokens (mean)Num Tokens (median)Num Tokens (min)Num Tokens (stddev)TPS (max)TPS (mean)TPS (median)TPS (min)TPS (stddev)TTFT (max)TTFT (mean)TTFT (median)TTFT (min)TTFT (stddev)Total (max)Total (mean)Total (median)Total (min)Total (stddev)
130b6ea — 2025-12-05 21:33:1240.00 (+0.00%)40.00 (+0.00%)40.00 (+0.00%)40.00 (+0.00%)0.00 (n/a)4.71 (-0.42%)4.64 (-0.09%)4.64 (+0.65%)4.55 (-0.22%)0.05 (-17.66%)4.41 (-0.34%)4.39 (-0.19%)4.38 (-0.33%)4.37 (-0.15%)0.01 (-25.90%)12.96 (-0.00%)12.80 (+0.07%)12.80 (-0.23%)12.67 (+0.44%)0.09 (-21.12%)
0a6c11c — 2025-12-03 23:35:1540.00 (n/a)40.00 (n/a)40.00 (n/a)40.00 (n/a)0.00 (n/a)4.73 (n/a)4.64 (n/a)4.61 (n/a)4.56 (n/a)0.06 (n/a)4.42 (n/a)4.40 (n/a)4.40 (n/a)4.37 (n/a)0.02 (n/a)12.96 (n/a)12.79 (n/a)12.83 (n/a)12.62 (n/a)0.12 (n/a)

llama_3.2_1b_prompt_1024_tokens_1

Commit/Date TTFT (max)TTFT (mean)TTFT (median)TTFT (min)TTFT (stddev)
dd8281d — 2026-03-26 01:28:35n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)
8b6d0f4 — 2026-03-25 21:36:352.26 (n/a)2.25 (n/a)2.25 (n/a)2.23 (n/a)0.01 (n/a)

llama_3.2_1b_prompt_1024_tokens_40

Commit/Date TPS (max)TPS (mean)TPS (median)TPS (min)TPS (stddev)TTFT (max)TTFT (mean)TTFT (median)TTFT (min)TTFT (stddev)
dd8281d — 2026-03-26 01:28:35n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)
8b6d0f4 — 2026-03-25 21:36:354.30 (n/a)4.28 (n/a)4.28 (n/a)4.26 (n/a)0.02 (n/a)2.37 (n/a)2.27 (n/a)2.25 (n/a)2.23 (n/a)0.06 (n/a)

llama_3.2_1b_prompt_13_tokens_1

Commit/Date TTFT (max)TTFT (mean)TTFT (median)TTFT (min)TTFT (stddev)
dd8281d — 2026-03-26 01:28:35n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)
8b6d0f4 — 2026-03-25 21:36:352.19 (n/a)2.17 (n/a)2.17 (n/a)2.14 (n/a)0.02 (n/a)

llama_3.2_1b_prompt_13_tokens_40

Commit/Date TPS (max)TPS (mean)TPS (median)TPS (min)TPS (stddev)TTFT (max)TTFT (mean)TTFT (median)TTFT (min)TTFT (stddev)
dd8281d — 2026-03-26 01:28:35n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)
8b6d0f4 — 2026-03-25 21:36:354.30 (n/a)4.27 (n/a)4.27 (n/a)4.25 (n/a)0.02 (n/a)2.19 (n/a)2.17 (n/a)2.16 (n/a)2.15 (n/a)0.01 (n/a)

llama_3.2_1b_prompt_2048_tokens_1

Commit/Date Num_Tokens (max)Num_Tokens (mean)Num_Tokens (median)Num_Tokens (min)Num_Tokens (stddev)TPS (max)TPS (mean)TPS (median)TPS (min)TPS (stddev)TTFT (max)TTFT (mean)TTFT (median)TTFT (min)TTFT (stddev)
897d04e — 2026-03-06 22:56:071.00 (+0.00%)1.00 (+0.00%)1.00 (+0.00%)1.00 (+0.00%)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)2.68 (-1.06%)2.68 (-1.06%)2.68 (-1.06%)2.68 (-1.06%)0.00 (n/a)
84d3478 — 2026-02-17 23:16:231.00 (n/a)1.00 (n/a)1.00 (n/a)1.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)2.70 (n/a)2.70 (n/a)2.70 (n/a)2.70 (n/a)0.00 (n/a)

llama_3.2_1b_prompt_2048_tokens_40

Commit/Date Num_Tokens (max)Num_Tokens (mean)Num_Tokens (median)Num_Tokens (min)Num_Tokens (stddev)TPS (max)TPS (mean)TPS (median)TPS (min)TPS (stddev)TTFT (max)TTFT (mean)TTFT (median)TTFT (min)TTFT (stddev)
897d04e — 2026-03-06 22:56:0740.00 (+0.00%)40.00 (+0.00%)40.00 (+0.00%)40.00 (+0.00%)0.00 (n/a)4.00 (-1.72%)4.00 (-1.72%)4.00 (-1.72%)4.00 (-1.72%)0.00 (n/a)2.70 (-0.44%)2.70 (-0.44%)2.70 (-0.44%)2.70 (-0.44%)0.00 (n/a)
84d3478 — 2026-02-17 23:16:2340.00 (n/a)40.00 (n/a)40.00 (n/a)40.00 (n/a)0.00 (n/a)4.07 (n/a)4.07 (n/a)4.07 (n/a)4.07 (n/a)0.00 (n/a)2.71 (n/a)2.71 (n/a)2.71 (n/a)2.71 (n/a)0.00 (n/a)

@github-actions
Copy link
Contributor

📊 Test Results for Small Benchmark/Test Suite

dd8281d (2026_03_26_01_39_48)

IRONCLAD

Tested on 2026_03_26_01_39_48 at commit dd8281d.

Test Checks Latency (mean)Bandwidth (mean)Throughput (mean)
M_128-K_128-num_aie_columns_1-tile_size_input_32-tile_size_output_128 ✅ 5/5 n/a 0.11 0.11
M_1792-K_896-N_1152-num_aie_columns_8-b_col_maj_False-c_col_maj_True-m_64-k_32-n_48-trace_size_0 ✅ 5/5 2580.96 3.74 1472.00
M_192-K_384-N_64-num_aie_columns_4-b_col_maj_False-c_col_maj_False-m_48-k_96-n_16-trace_size_0 ✅ 5/5 480.12 0.50 21.50
M_192-K_384-N_64-num_aie_columns_4-b_col_maj_True-c_col_maj_True-m_48-k_96-n_16-trace_size_0 ✅ 5/5 423.68 0.65 27.90
M_2048-K_2048-N_2048-num_aie_columns_1-b_col_maj_False-c_col_maj_False-m_64-k_64-n_64-trace_size_0 ✅ 5/5 50096.64 0.50 342.95
M_2048-K_2048-N_2048-num_aie_columns_2-b_col_maj_True-c_col_maj_False-m_64-k_64-n_64-trace_size_0 ✅ 5/5 29443.40 0.85 583.51
M_2048-K_2048-N_2048-num_aie_columns_8-b_col_maj_True-c_col_maj_True-m_64-k_64-n_64-trace_size_0 ✅ 5/5 7654.26 3.30 2252.66
M_2048-K_8192-num_aie_columns_1-tile_size_input_1-tile_size_output_2048 ✅ 5/5 n/a 11.37 11.36
M_2048-K_8192-num_aie_columns_2-tile_size_input_1-tile_size_output_1024 ✅ 5/5 n/a 19.35 19.33
M_2048-K_8192-num_aie_columns_4-tile_size_input_1-tile_size_output_512 ✅ 5/5 n/a 30.51 30.49
M_2048-K_8192-num_aie_columns_8-tile_size_input_1-tile_size_output_256 ✅ 5/5 n/a 32.55 32.53
M_2048-N_64-aie_columns_1-channels_1-m_64-n_64-s_8 ✅ 5/5 410.16 1.54 n/a
M_2048-N_64-aie_columns_1-channels_2-m_64-n_64-s_8 ✅ 5/5 443.74 1.26 n/a
M_384-K_1536-N_1792-num_aie_columns_4-b_col_maj_True-c_col_maj_False-m_32-k_48-n_64-trace_size_0 ✅ 5/5 2315.44 3.51 920.56
M_8192-K_2048-num_aie_columns_1-tile_size_input_4-tile_size_output_1024 ✅ 5/5 n/a 11.94 11.94
M_8192-K_2048-num_aie_columns_2-tile_size_input_4-tile_size_output_1024 ✅ 5/5 n/a 21.01 21.00
M_8192-K_2048-num_aie_columns_4-tile_size_input_4-tile_size_output_1024 ✅ 5/5 n/a 32.18 32.16
M_8192-K_2048-num_aie_columns_8-tile_size_input_4-tile_size_output_1024 ✅ 5/5 n/a 33.44 33.42
M_896-K_1792-N_640-num_aie_columns_8-b_col_maj_False-c_col_maj_True-m_32-k_64-n_80-trace_size_0 ✅ 5/5 1364.76 4.96 1532.83
embedding_dim_2048-hidden_dim_2048 ✅ 5/5 n/a n/a n/a
input_length_2048-num_aie_columns_1-num_channels_1-tile_size_2048 ✅ 25/25 268.65 0.03 n/a
input_length_2048-num_aie_columns_1-num_channels_1-tile_size_2048-group_size_32 ✅ 5/5 322.24 0.02 n/a
input_length_2048-num_aie_columns_1-num_channels_1-tile_size_2048-weighted_False ✅ 5/5 285.78 0.03 n/a
input_length_2048-num_aie_columns_1-num_channels_1-tile_size_2048-weighted_True ✅ 5/5 398.46 0.04 n/a
input_length_2048-num_aie_columns_1-num_channels_2-tile_size_1024 ✅ 10/10 273.84 0.03 n/a
input_length_2048-num_aie_columns_1-num_channels_2-tile_size_1024-group_size_32 ✅ 5/5 319.34 0.02 n/a
input_length_2048-num_aie_columns_1-num_channels_2-tile_size_1024-weighted_False ✅ 5/5 283.56 0.03 n/a
input_length_2048-num_aie_columns_1-num_channels_2-tile_size_2048-scalar_factor_3.0 ✅ 5/5 285.34 0.05 n/a
input_length_2048-num_aie_columns_1-tile_size_2048 ✅ 15/15 314.03 0.04 n/a
input_length_2048-num_aie_columns_2-num_channels_1-tile_size_1024 ✅ 25/25 287.70 0.03 n/a
input_length_2048-num_aie_columns_2-num_channels_1-tile_size_1024-group_size_32 ✅ 5/5 344.90 0.02 n/a
input_length_2048-num_aie_columns_2-num_channels_1-tile_size_1024-weighted_False ✅ 5/5 301.96 0.03 n/a
input_length_2048-num_aie_columns_2-num_channels_1-tile_size_1024-weighted_True ✅ 5/5 352.74 0.03 n/a
input_length_2048-num_aie_columns_2-num_channels_2-tile_size_1024-scalar_factor_3.0 ✅ 5/5 370.02 0.04 n/a
input_length_2048-num_aie_columns_2-num_channels_2-tile_size_512 ✅ 10/10 314.31 0.03 n/a
input_length_2048-num_aie_columns_2-num_channels_2-tile_size_512-group_size_32 ✅ 5/5 398.26 0.02 n/a
input_length_2048-num_aie_columns_2-num_channels_2-tile_size_512-weighted_False ✅ 5/5 292.22 0.03 n/a
input_length_2048-num_aie_columns_2-tile_size_1024 ✅ 15/15 327.19 0.04 n/a
input_length_2048-num_aie_columns_4-num_channels_1-tile_size_512 ✅ 25/25 270.32 0.03 n/a
input_length_2048-num_aie_columns_4-num_channels_1-tile_size_512-group_size_32 ✅ 5/5 363.80 0.02 n/a
input_length_2048-num_aie_columns_4-num_channels_1-tile_size_512-weighted_False ✅ 5/5 268.18 0.03 n/a
input_length_2048-num_aie_columns_4-num_channels_1-tile_size_512-weighted_True ✅ 5/5 288.90 0.03 n/a
input_length_2048-num_aie_columns_4-num_channels_2-tile_size_256 ✅ 10/10 387.26 0.03 n/a
input_length_2048-num_aie_columns_4-num_channels_2-tile_size_256-group_size_32 ✅ 5/5 315.36 0.02 n/a
input_length_2048-num_aie_columns_4-num_channels_2-tile_size_256-weighted_False ✅ 5/5 337.84 0.03 n/a
input_length_2048-num_aie_columns_4-num_channels_2-tile_size_512-scalar_factor_3.0 ✅ 5/5 307.70 0.05 n/a
input_length_2048-num_aie_columns_4-tile_size_512 ✅ 15/15 315.75 0.04 n/a
input_length_2048-num_aie_columns_8-num_channels_1-tile_size_256 ✅ 25/25 359.04 0.03 n/a
input_length_2048-num_aie_columns_8-num_channels_1-tile_size_256-group_size_32 ✅ 5/5 306.76 0.02 n/a
input_length_2048-num_aie_columns_8-num_channels_1-tile_size_256-weighted_False ✅ 5/5 341.78 0.03 n/a
input_length_2048-num_aie_columns_8-num_channels_1-tile_size_256-weighted_True ✅ 5/5 320.54 0.03 n/a
input_length_2048-num_aie_columns_8-num_channels_2-tile_size_128 ✅ 10/10 329.32 0.03 n/a
input_length_2048-num_aie_columns_8-num_channels_2-tile_size_128-group_size_32 ✅ 5/5 302.94 0.02 n/a
input_length_2048-num_aie_columns_8-num_channels_2-tile_size_128-weighted_False ✅ 5/5 483.20 0.02 n/a
input_length_2048-num_aie_columns_8-num_channels_2-tile_size_256-scalar_factor_3.0 ✅ 5/5 282.86 0.04 n/a
input_length_2048-num_aie_columns_8-tile_size_256 ✅ 15/15 348.31 0.04 n/a
input_length_2048-num_cores_1-num_channels_1-bypass_False-tile_size_2048 ✅ 5/5 305.94 0.03 n/a
input_length_2048-num_cores_16-num_channels_2-bypass_False-tile_size_128 ✅ 5/5 417.32 0.02 n/a
input_length_2048-num_cores_2-num_channels_1-bypass_False-tile_size_1024 ✅ 5/5 351.38 0.03 n/a
input_length_2048-num_cores_2-num_channels_2-bypass_False-tile_size_1024 ✅ 5/5 265.62 0.03 n/a
input_length_2048-num_cores_4-num_channels_1-bypass_False-tile_size_512 ✅ 5/5 283.28 0.03 n/a
input_length_2048-num_cores_4-num_channels_2-bypass_False-tile_size_512 ✅ 5/5 256.76 0.03 n/a
input_length_2048-num_cores_8-num_channels_1-bypass_False-tile_size_256 ✅ 5/5 292.12 0.03 n/a
input_length_2048-num_cores_8-num_channels_2-bypass_False-tile_size_256 ✅ 5/5 293.32 0.03 n/a
input_length_32768-num_aie_columns_2-num_channels_2-tile_size_1024 ✅ 5/5 423.30 0.37 n/a
input_length_32768-num_aie_columns_2-num_channels_2-tile_size_512 ✅ 5/5 306.52 0.44 n/a
input_length_32768-num_aie_columns_4-num_channels_4-tile_size_2048 ✅ 5/5 473.46 0.35 n/a
rows_32-cols_512-angle_rows_32-aie_columns_1-method_type_0 ✅ 5/5 289.54 0.35 n/a
rows_32-cols_512-angle_rows_32-aie_columns_2-method_type_0 ✅ 5/5 322.96 0.33 n/a
rows_32-cols_512-angle_rows_32-aie_columns_8-method_type_0 ✅ 5/5 406.36 0.29 n/a
rows_32-cols_512-angle_rows_8-aie_columns_1-method_type_0 ✅ 5/5 325.24 0.23 n/a
rows_32-cols_512-angle_rows_8-aie_columns_2-method_type_0 ✅ 5/5 307.82 0.24 n/a
rows_32-cols_512-angle_rows_8-aie_columns_8-method_type_0 ✅ 5/5 392.00 0.21 n/a
seq_len_16384-dim_64-num_heads_1-num_pipelines_8-num_kv_heads_0 ❌ 0/5 n/a n/a n/a
seq_len_256-embedding_dim_2048-hidden_dim_2048-prio_accuracy_False ✅ 5/5 n/a n/a n/a
📈 Trends (vs main branch) for Small Benchmark/Test Suite

dd8281d (2026_03_26_01_39_48)

IRONCLAD Trends

M_128-K_128-num_aie_columns_1-tile_size_input_32-tile_size_output_128

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
dd8281d — 2026-03-26 01:33:330.13 (n/a)0.11 (n/a)0.12 (n/a)0.05 (n/a)0.03 (n/a)0.13 (n/a)0.11 (n/a)0.12 (n/a)0.05 (n/a)0.03 (n/a)

M_1792-K_896-N_1152-num_aie_columns_8-b_col_maj_False-c_col_maj_True-m_64-k_32-n_48-trace_size_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
dd8281d — 2026-03-26 01:33:334.55 (n/a)3.74 (n/a)3.76 (n/a)2.84 (n/a)0.66 (n/a)3307.00 (n/a)2580.96 (n/a)2499.70 (n/a)2066.90 (n/a)483.73 (n/a)1789.82 (n/a)1472.00 (n/a)1479.94 (n/a)1118.66 (n/a)260.51 (n/a)

M_192-K_384-N_64-num_aie_columns_4-b_col_maj_False-c_col_maj_False-m_48-k_96-n_16-trace_size_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
dd8281d — 2026-03-26 01:33:330.69 (n/a)0.50 (n/a)0.54 (n/a)0.30 (n/a)0.16 (n/a)733.50 (n/a)480.12 (n/a)413.00 (n/a)321.30 (n/a)169.11 (n/a)29.37 (n/a)21.50 (n/a)22.85 (n/a)12.87 (n/a)6.71 (n/a)

M_192-K_384-N_64-num_aie_columns_4-b_col_maj_True-c_col_maj_True-m_48-k_96-n_16-trace_size_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
dd8281d — 2026-03-26 01:33:330.84 (n/a)0.65 (n/a)0.72 (n/a)0.24 (n/a)0.25 (n/a)937.10 (n/a)423.68 (n/a)306.40 (n/a)264.10 (n/a)288.95 (n/a)35.74 (n/a)27.90 (n/a)30.80 (n/a)10.07 (n/a)10.56 (n/a)

M_2048-K_2048-N_2048-num_aie_columns_1-b_col_maj_False-c_col_maj_False-m_64-k_64-n_64-trace_size_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
dd8281d — 2026-03-26 01:33:330.51 (n/a)0.50 (n/a)0.50 (n/a)0.50 (n/a)0.00 (n/a)50643.50 (n/a)50096.64 (n/a)50020.00 (n/a)49758.70 (n/a)327.92 (n/a)345.26 (n/a)342.95 (n/a)343.46 (n/a)339.23 (n/a)2.23 (n/a)

M_2048-K_2048-N_2048-num_aie_columns_2-b_col_maj_True-c_col_maj_False-m_64-k_64-n_64-trace_size_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
dd8281d — 2026-03-26 01:33:330.86 (n/a)0.85 (n/a)0.85 (n/a)0.85 (n/a)0.01 (n/a)29604.60 (n/a)29443.40 (n/a)29512.50 (n/a)29114.10 (n/a)197.13 (n/a)590.09 (n/a)583.51 (n/a)582.12 (n/a)580.31 (n/a)3.93 (n/a)

M_2048-K_2048-N_2048-num_aie_columns_8-b_col_maj_True-c_col_maj_True-m_64-k_64-n_64-trace_size_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
dd8281d — 2026-03-26 01:33:333.54 (n/a)3.30 (n/a)3.35 (n/a)3.00 (n/a)0.22 (n/a)8384.50 (n/a)7654.26 (n/a)7517.60 (n/a)7101.70 (n/a)521.24 (n/a)2419.11 (n/a)2252.66 (n/a)2285.29 (n/a)2049.01 (n/a)150.18 (n/a)

M_2048-K_8192-num_aie_columns_1-tile_size_input_1-tile_size_output_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
dd8281d — 2026-03-26 01:33:3312.36 (n/a)11.37 (n/a)11.36 (n/a)10.13 (n/a)1.00 (n/a)12.35 (n/a)11.36 (n/a)11.35 (n/a)10.13 (n/a)1.00 (n/a)

M_2048-K_8192-num_aie_columns_2-tile_size_input_1-tile_size_output_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
dd8281d — 2026-03-26 01:33:3321.74 (n/a)19.35 (n/a)18.62 (n/a)16.30 (n/a)2.33 (n/a)21.73 (n/a)19.33 (n/a)18.61 (n/a)16.29 (n/a)2.33 (n/a)

M_2048-K_8192-num_aie_columns_4-tile_size_input_1-tile_size_output_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
dd8281d — 2026-03-26 01:33:3333.79 (n/a)30.51 (n/a)32.68 (n/a)21.22 (n/a)5.27 (n/a)33.77 (n/a)30.49 (n/a)32.66 (n/a)21.21 (n/a)5.26 (n/a)

M_2048-K_8192-num_aie_columns_8-tile_size_input_1-tile_size_output_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
dd8281d — 2026-03-26 01:33:3335.35 (n/a)32.55 (n/a)33.59 (n/a)28.18 (n/a)2.82 (n/a)35.33 (n/a)32.53 (n/a)33.57 (n/a)28.17 (n/a)2.82 (n/a)

M_2048-N_64-aie_columns_1-channels_1-m_64-n_64-s_8

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
dd8281d — 2026-03-26 01:33:332.02 (n/a)1.54 (n/a)1.55 (n/a)0.62 (n/a)0.56 (n/a)841.20 (n/a)410.16 (n/a)338.80 (n/a)259.90 (n/a)244.13 (n/a)

M_2048-N_64-aie_columns_1-channels_2-m_64-n_64-s_8

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
dd8281d — 2026-03-26 01:33:331.54 (n/a)1.26 (n/a)1.26 (n/a)0.76 (n/a)0.31 (n/a)690.10 (n/a)443.74 (n/a)416.90 (n/a)339.60 (n/a)143.24 (n/a)

M_384-K_1536-N_1792-num_aie_columns_4-b_col_maj_True-c_col_maj_False-m_32-k_48-n_64-trace_size_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
dd8281d — 2026-03-26 01:33:333.99 (n/a)3.51 (n/a)3.28 (n/a)3.20 (n/a)0.36 (n/a)2516.30 (n/a)2315.44 (n/a)2460.50 (n/a)2022.50 (n/a)230.34 (n/a)1045.19 (n/a)920.56 (n/a)859.15 (n/a)840.11 (n/a)95.40 (n/a)

M_8192-K_2048-num_aie_columns_1-tile_size_input_4-tile_size_output_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
dd8281d — 2026-03-26 01:33:3312.54 (n/a)11.94 (n/a)12.33 (n/a)10.90 (n/a)0.72 (n/a)12.53 (n/a)11.94 (n/a)12.32 (n/a)10.89 (n/a)0.72 (n/a)

M_8192-K_2048-num_aie_columns_2-tile_size_input_4-tile_size_output_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
dd8281d — 2026-03-26 01:33:3322.36 (n/a)21.01 (n/a)21.57 (n/a)18.02 (n/a)1.71 (n/a)22.35 (n/a)21.00 (n/a)21.56 (n/a)18.01 (n/a)1.71 (n/a)

M_8192-K_2048-num_aie_columns_4-tile_size_input_4-tile_size_output_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
dd8281d — 2026-03-26 01:33:3336.51 (n/a)32.18 (n/a)32.74 (n/a)26.61 (n/a)3.58 (n/a)36.49 (n/a)32.16 (n/a)32.72 (n/a)26.59 (n/a)3.57 (n/a)

M_8192-K_2048-num_aie_columns_8-tile_size_input_4-tile_size_output_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
dd8281d — 2026-03-26 01:33:3336.02 (n/a)33.44 (n/a)34.88 (n/a)26.66 (n/a)3.84 (n/a)35.99 (n/a)33.42 (n/a)34.86 (n/a)26.64 (n/a)3.84 (n/a)

M_896-K_1792-N_640-num_aie_columns_8-b_col_maj_False-c_col_maj_True-m_32-k_64-n_80-trace_size_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
dd8281d — 2026-03-26 01:33:335.61 (n/a)4.96 (n/a)5.08 (n/a)3.89 (n/a)0.70 (n/a)1708.90 (n/a)1364.76 (n/a)1309.40 (n/a)1185.10 (n/a)213.33 (n/a)1734.16 (n/a)1532.83 (n/a)1569.61 (n/a)1202.67 (n/a)216.50 (n/a)

axpy_1_cols_2_channels_2048_tile_2048_3.0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.08 (+17.48%)0.07 (+11.67%)0.07 (+26.71%)0.05 (+0.80%)0.01 (+108.36%)239.80 (-0.83%)190.62 (-7.91%)166.90 (-21.09%)155.40 (-14.90%)42.89 (+79.10%)
dc66da7 — 2026-03-25 19:23:450.07 (n/a)0.06 (n/a)0.06 (n/a)0.05 (n/a)0.01 (n/a)241.80 (n/a)207.00 (n/a)211.50 (n/a)182.60 (n/a)23.95 (n/a)

axpy_1_cols_2_channels_2048_tile_2048_3.0_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.16 (-3.31%)0.14 (-1.29%)0.15 (+16.18%)0.10 (-6.80%)0.03 (+3.89%)125.70 (+7.25%)93.56 (+2.12%)83.50 (-13.92%)75.00 (+3.45%)22.38 (+18.09%)
84d3478 — 2026-02-17 23:25:130.17 (n/a)0.14 (n/a)0.13 (n/a)0.10 (n/a)0.03 (n/a)117.20 (n/a)91.62 (n/a)97.00 (n/a)72.50 (n/a)18.95 (n/a)

axpy_2_cols_2_channels_2048_tile_1024_3.0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.09 (+14.45%)0.07 (-0.78%)0.07 (-0.40%)0.05 (-7.87%)0.01 (+68.50%)233.40 (+8.51%)188.20 (+2.47%)179.80 (+0.39%)143.50 (-12.66%)34.06 (+61.09%)
dc66da7 — 2026-03-25 19:23:450.07 (n/a)0.07 (n/a)0.07 (n/a)0.06 (n/a)0.01 (n/a)215.10 (n/a)183.66 (n/a)179.10 (n/a)164.30 (n/a)21.14 (n/a)

axpy_2_cols_2_channels_2048_tile_1024_3.0_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.13 (-0.73%)0.11 (-7.43%)0.11 (-8.63%)0.07 (-27.71%)0.02 (+50.39%)182.70 (+38.30%)120.86 (+11.97%)110.80 (+9.38%)96.10 (+0.73%)35.16 (+122.88%)
84d3478 — 2026-02-17 23:25:130.13 (n/a)0.12 (n/a)0.12 (n/a)0.09 (n/a)0.02 (n/a)132.10 (n/a)107.94 (n/a)101.30 (n/a)95.40 (n/a)15.78 (n/a)

axpy_4_cols_2_channels_2048_tile_512_3.0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.10 (+6.67%)0.06 (-1.56%)0.06 (-2.24%)0.05 (+3.60%)0.02 (+20.83%)251.50 (-3.49%)200.78 (+3.11%)191.70 (+2.29%)128.70 (-6.26%)50.37 (+11.30%)
dc66da7 — 2026-03-25 19:23:450.09 (n/a)0.07 (n/a)0.07 (n/a)0.05 (n/a)0.02 (n/a)260.60 (n/a)194.72 (n/a)187.40 (n/a)137.30 (n/a)45.26 (n/a)

axpy_4_cols_2_channels_2048_tile_512_3.0_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.13 (-10.91%)0.11 (-7.68%)0.12 (-0.73%)0.06 (-17.94%)0.03 (+0.99%)194.10 (+21.85%)124.52 (+10.29%)103.30 (+0.68%)97.20 (+12.24%)40.32 (+39.15%)
84d3478 — 2026-02-17 23:25:130.14 (n/a)0.11 (n/a)0.12 (n/a)0.08 (n/a)0.03 (n/a)159.30 (n/a)112.90 (n/a)102.60 (n/a)86.60 (n/a)28.97 (n/a)

axpy_8_cols_2_channels_2048_tile_256_3.0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.08 (+1.42%)0.06 (+9.86%)0.06 (+18.71%)0.03 (+13.51%)0.02 (-10.62%)376.00 (-11.90%)225.02 (-11.61%)196.20 (-15.76%)151.40 (-1.37%)87.15 (-17.99%)
dc66da7 — 2026-03-25 19:23:450.08 (n/a)0.05 (n/a)0.05 (n/a)0.03 (n/a)0.02 (n/a)426.80 (n/a)254.58 (n/a)232.90 (n/a)153.50 (n/a)106.27 (n/a)

axpy_8_cols_2_channels_2048_tile_256_3.0_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.13 (+0.01%)0.11 (+6.10%)0.11 (+0.22%)0.08 (-0.33%)0.02 (-9.83%)156.30 (+0.32%)115.26 (-6.40%)112.70 (-0.18%)95.50 (+0.00%)24.34 (-10.22%)
84d3478 — 2026-02-17 23:25:130.13 (n/a)0.10 (n/a)0.11 (n/a)0.08 (n/a)0.02 (n/a)155.80 (n/a)123.14 (n/a)112.90 (n/a)95.50 (n/a)27.11 (n/a)

dequant_1_cols_1_channels_2048_tile_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.04 (+17.85%)0.03 (+2.55%)0.03 (+1.65%)0.03 (-2.18%)0.00 (+114.37%)207.70 (+2.26%)178.64 (-0.96%)173.60 (-1.59%)141.60 (-15.16%)28.01 (+90.63%)
dc66da7 — 2026-03-25 19:23:450.03 (n/a)0.03 (n/a)0.03 (n/a)0.03 (n/a)0.00 (n/a)203.10 (n/a)180.38 (n/a)176.40 (n/a)166.90 (n/a)14.69 (n/a)

dequant_1_cols_1_channels_2048_tile_2048_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.05 (-13.00%)0.04 (-0.87%)0.04 (+1.16%)0.03 (-3.17%)0.01 (-31.33%)158.40 (+3.26%)128.12 (-0.19%)123.80 (-1.12%)110.40 (+14.88%)18.36 (-15.65%)
84d3478 — 2026-02-17 23:25:130.05 (n/a)0.04 (n/a)0.04 (n/a)0.03 (n/a)0.01 (n/a)153.40 (n/a)128.36 (n/a)125.20 (n/a)96.10 (n/a)21.77 (n/a)

dequant_1_cols_2_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.04 (+13.44%)0.03 (+8.77%)0.03 (+14.89%)0.02 (-9.15%)0.01 (+54.39%)252.70 (+10.06%)182.14 (-6.08%)174.50 (-12.97%)145.40 (-11.88%)42.38 (+56.31%)
dc66da7 — 2026-03-25 19:23:450.03 (n/a)0.03 (n/a)0.03 (n/a)0.02 (n/a)0.00 (n/a)229.60 (n/a)193.94 (n/a)200.50 (n/a)165.00 (n/a)27.11 (n/a)

dequant_1_cols_2_channels_2048_tile_1024_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.06 (-3.74%)0.05 (-12.63%)0.05 (-5.21%)0.03 (-33.47%)0.01 (+52.60%)162.70 (+50.37%)110.84 (+18.98%)98.80 (+5.44%)81.50 (+3.95%)31.02 (+149.23%)
84d3478 — 2026-02-17 23:25:130.07 (n/a)0.06 (n/a)0.06 (n/a)0.05 (n/a)0.01 (n/a)108.20 (n/a)93.16 (n/a)93.70 (n/a)78.40 (n/a)12.45 (n/a)

dequant_2_cols_1_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.05 (+21.89%)0.03 (+20.89%)0.03 (+21.41%)0.02 (+8.07%)0.01 (+32.97%)215.80 (-7.46%)168.56 (-16.36%)172.60 (-17.65%)113.20 (-17.97%)37.61 (+1.89%)
dc66da7 — 2026-03-25 19:23:450.04 (n/a)0.03 (n/a)0.03 (n/a)0.02 (n/a)0.01 (n/a)233.20 (n/a)201.52 (n/a)209.60 (n/a)138.00 (n/a)36.92 (n/a)

dequant_2_cols_1_channels_2048_tile_1024_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.06 (+8.90%)0.05 (+1.31%)0.05 (-1.85%)0.03 (+6.35%)0.01 (+7.11%)152.30 (-5.93%)115.08 (-1.20%)100.80 (+1.92%)88.10 (-8.23%)27.20 (-4.42%)
84d3478 — 2026-02-17 23:25:130.05 (n/a)0.05 (n/a)0.05 (n/a)0.03 (n/a)0.01 (n/a)161.90 (n/a)116.48 (n/a)98.90 (n/a)96.00 (n/a)28.46 (n/a)

dequant_2_cols_2_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.05 (+49.89%)0.03 (+24.89%)0.03 (+13.74%)0.02 (+15.73%)0.01 (+106.91%)219.80 (-13.60%)173.04 (-17.40%)185.70 (-12.07%)112.30 (-33.27%)41.15 (+16.74%)
dc66da7 — 2026-03-25 19:23:450.03 (n/a)0.03 (n/a)0.02 (n/a)0.02 (n/a)0.00 (n/a)254.40 (n/a)209.50 (n/a)211.20 (n/a)168.30 (n/a)35.25 (n/a)

dequant_2_cols_2_channels_2048_tile_512_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.07 (+27.96%)0.05 (+1.57%)0.04 (-18.06%)0.03 (+10.23%)0.01 (+22.79%)164.70 (-9.31%)119.12 (-1.39%)120.20 (+22.03%)75.60 (-21.82%)31.61 (-13.80%)
84d3478 — 2026-02-17 23:25:130.05 (n/a)0.05 (n/a)0.05 (n/a)0.03 (n/a)0.01 (n/a)181.60 (n/a)120.80 (n/a)98.50 (n/a)96.70 (n/a)36.67 (n/a)

dequant_4_cols_1_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.04 (+56.54%)0.03 (+15.78%)0.03 (+9.01%)0.02 (-5.58%)0.01 (+304.26%)243.40 (+5.92%)186.06 (-8.98%)182.60 (-8.29%)118.50 (-36.12%)46.36 (+164.07%)
dc66da7 — 2026-03-25 19:23:450.03 (n/a)0.03 (n/a)0.03 (n/a)0.02 (n/a)0.00 (n/a)229.80 (n/a)204.42 (n/a)199.10 (n/a)185.50 (n/a)17.56 (n/a)

dequant_4_cols_1_channels_2048_tile_512_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.05 (-2.55%)0.04 (-12.79%)0.05 (-16.25%)0.03 (-12.85%)0.01 (-1.20%)164.60 (+14.78%)122.14 (+15.14%)115.90 (+19.48%)97.50 (+2.63%)25.08 (+19.84%)
84d3478 — 2026-02-17 23:25:130.06 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.01 (n/a)143.40 (n/a)106.08 (n/a)97.00 (n/a)95.00 (n/a)20.93 (n/a)

dequant_4_cols_2_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.04 (+28.72%)0.03 (+2.89%)0.02 (-6.84%)0.02 (+6.27%)0.01 (+95.59%)243.20 (-5.88%)208.42 (-0.42%)222.90 (+7.32%)141.10 (-22.34%)41.08 (+38.16%)
dc66da7 — 2026-03-25 19:23:450.03 (n/a)0.03 (n/a)0.03 (n/a)0.02 (n/a)0.00 (n/a)258.40 (n/a)209.30 (n/a)207.70 (n/a)181.70 (n/a)29.74 (n/a)

dequant_4_cols_2_channels_2048_tile_256_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.06 (+20.83%)0.05 (+8.21%)0.04 (+1.42%)0.03 (+17.39%)0.01 (+14.84%)166.40 (-14.80%)122.34 (-7.97%)127.70 (-1.39%)84.60 (-17.22%)30.47 (-19.21%)
84d3478 — 2026-02-17 23:25:130.05 (n/a)0.04 (n/a)0.04 (n/a)0.03 (n/a)0.01 (n/a)195.30 (n/a)132.94 (n/a)129.50 (n/a)102.20 (n/a)37.71 (n/a)

dequant_8_cols_1_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.04 (-1.52%)0.03 (+2.17%)0.03 (+18.60%)0.02 (-11.90%)0.01 (+4.42%)268.50 (+13.53%)205.32 (-1.14%)191.00 (-15.67%)141.60 (+1.51%)49.39 (+23.46%)
dc66da7 — 2026-03-25 19:23:450.04 (n/a)0.03 (n/a)0.02 (n/a)0.02 (n/a)0.01 (n/a)236.50 (n/a)207.68 (n/a)226.50 (n/a)139.50 (n/a)40.00 (n/a)

dequant_8_cols_1_channels_2048_tile_256_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.05 (-25.19%)0.04 (-21.02%)0.04 (-19.70%)0.04 (-19.76%)0.00 (-37.92%)142.70 (+24.63%)126.44 (+26.24%)125.70 (+24.46%)115.50 (+33.68%)10.61 (+3.97%)
84d3478 — 2026-02-17 23:25:130.06 (n/a)0.05 (n/a)0.05 (n/a)0.05 (n/a)0.01 (n/a)114.50 (n/a)100.16 (n/a)101.00 (n/a)86.40 (n/a)10.20 (n/a)

dequant_8_cols_2_channels_2048_tile_128

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.02 (+5.64%)0.02 (+1.26%)0.02 (-1.25%)0.02 (-3.87%)0.00 (+124.04%)258.90 (+4.02%)239.08 (-0.87%)247.20 (+1.27%)215.50 (-5.36%)18.28 (+120.74%)
dc66da7 — 2026-03-25 19:23:450.02 (n/a)0.02 (n/a)0.02 (n/a)0.02 (n/a)0.00 (n/a)248.90 (n/a)241.18 (n/a)244.10 (n/a)227.70 (n/a)8.28 (n/a)

dequant_8_cols_2_channels_2048_tile_128_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.04 (-26.69%)0.04 (-11.93%)0.04 (-16.99%)0.03 (+27.28%)0.01 (-62.60%)193.10 (-21.41%)147.68 (+2.96%)136.10 (+20.44%)135.40 (+36.49%)25.41 (-58.96%)
84d3478 — 2026-02-17 23:25:130.05 (n/a)0.04 (n/a)0.05 (n/a)0.02 (n/a)0.01 (n/a)245.70 (n/a)143.44 (n/a)113.00 (n/a)99.20 (n/a)61.90 (n/a)

eltwise_add_1_cols_2_channels_2048_tile_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.07 (-32.66%)0.06 (-13.87%)0.06 (-9.19%)0.04 (-8.04%)0.01 (-49.50%)299.50 (+8.75%)223.84 (+11.90%)212.70 (+10.15%)187.60 (+48.42%)45.70 (-16.71%)
dc66da7 — 2026-03-25 19:23:450.10 (n/a)0.07 (n/a)0.06 (n/a)0.04 (n/a)0.02 (n/a)275.40 (n/a)200.04 (n/a)193.10 (n/a)126.40 (n/a)54.87 (n/a)

eltwise_add_2_cols_2_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.07 (-33.33%)0.06 (-18.97%)0.05 (-19.23%)0.05 (+18.13%)0.01 (-59.22%)270.80 (-15.37%)223.26 (+13.77%)229.40 (+23.80%)174.50 (+50.04%)37.51 (-50.17%)
dc66da7 — 2026-03-25 19:23:450.11 (n/a)0.07 (n/a)0.07 (n/a)0.04 (n/a)0.02 (n/a)320.00 (n/a)196.24 (n/a)185.30 (n/a)116.30 (n/a)75.29 (n/a)

eltwise_add_4_cols_2_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.07 (-8.97%)0.06 (-6.00%)0.06 (-2.02%)0.05 (-2.74%)0.01 (-14.89%)243.40 (+2.83%)208.60 (+6.01%)200.50 (+2.04%)178.10 (+9.87%)31.58 (-2.07%)
dc66da7 — 2026-03-25 19:23:450.08 (n/a)0.06 (n/a)0.06 (n/a)0.05 (n/a)0.01 (n/a)236.70 (n/a)196.78 (n/a)196.50 (n/a)162.10 (n/a)32.25 (n/a)

eltwise_add_8_cols_2_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.06 (-45.96%)0.05 (-24.77%)0.05 (-15.00%)0.05 (+4.78%)0.00 (-83.62%)259.60 (-4.56%)234.66 (+20.97%)224.10 (+17.64%)218.40 (+85.08%)18.77 (-71.47%)
dc66da7 — 2026-03-25 19:23:450.10 (n/a)0.07 (n/a)0.06 (n/a)0.05 (n/a)0.02 (n/a)272.00 (n/a)193.98 (n/a)190.50 (n/a)118.00 (n/a)65.77 (n/a)

eltwise_mul_1_cols_2_channels_2048_tile_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.11 (+3.13%)0.07 (+2.52%)0.06 (-1.87%)0.05 (-14.01%)0.02 (+21.62%)257.00 (+16.29%)184.10 (+0.67%)190.70 (+1.87%)114.40 (-3.05%)55.57 (+40.96%)
dc66da7 — 2026-03-25 19:23:450.10 (n/a)0.07 (n/a)0.07 (n/a)0.06 (n/a)0.02 (n/a)221.00 (n/a)182.88 (n/a)187.20 (n/a)118.00 (n/a)39.42 (n/a)

eltwise_mul_2_cols_2_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.09 (+2.43%)0.07 (+6.63%)0.08 (+30.25%)0.05 (-10.38%)0.02 (+9.28%)264.20 (+11.62%)182.10 (-4.96%)162.40 (-23.22%)137.30 (-2.42%)52.80 (+20.53%)
dc66da7 — 2026-03-25 19:23:450.09 (n/a)0.07 (n/a)0.06 (n/a)0.05 (n/a)0.02 (n/a)236.70 (n/a)191.60 (n/a)211.50 (n/a)140.70 (n/a)43.81 (n/a)

eltwise_mul_4_cols_2_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.10 (+22.79%)0.08 (+18.44%)0.09 (+34.73%)0.06 (-0.12%)0.01 (+158.99%)191.20 (+0.10%)154.44 (-13.56%)136.60 (-25.80%)126.60 (-18.53%)29.95 (+117.72%)
dc66da7 — 2026-03-25 19:23:450.08 (n/a)0.07 (n/a)0.07 (n/a)0.06 (n/a)0.01 (n/a)191.00 (n/a)178.66 (n/a)184.10 (n/a)155.40 (n/a)13.76 (n/a)

eltwise_mul_8_cols_2_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.08 (+34.81%)0.06 (+14.94%)0.06 (+6.87%)0.04 (-5.87%)0.01 (+158.26%)280.10 (+6.22%)210.00 (-9.80%)215.10 (-6.40%)150.70 (-25.80%)49.49 (+100.84%)
dc66da7 — 2026-03-25 19:23:450.06 (n/a)0.05 (n/a)0.05 (n/a)0.05 (n/a)0.01 (n/a)263.70 (n/a)232.82 (n/a)229.80 (n/a)203.10 (n/a)24.64 (n/a)

embedding_dim_2048-hidden_dim_2048

No metrics available.

gelu_1_cols_1_channels_2048_tile_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.05 (+9.18%)0.04 (-1.03%)0.04 (-2.51%)0.04 (-8.09%)0.01 (+114.58%)221.00 (+8.81%)195.42 (+1.97%)199.10 (+2.58%)163.60 (-8.40%)23.03 (+114.14%)
dc66da7 — 2026-03-25 19:23:450.05 (n/a)0.04 (n/a)0.04 (n/a)0.04 (n/a)0.00 (n/a)203.10 (n/a)191.64 (n/a)194.10 (n/a)178.60 (n/a)10.76 (n/a)

gelu_1_cols_2_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.04 (-26.60%)0.04 (-19.13%)0.04 (-18.64%)0.04 (-2.84%)0.00 (-66.68%)226.70 (+2.91%)199.16 (+20.50%)193.10 (+22.92%)186.40 (+36.16%)16.29 (-52.49%)
dc66da7 — 2026-03-25 19:23:450.06 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.01 (n/a)220.30 (n/a)165.28 (n/a)157.10 (n/a)136.90 (n/a)34.28 (n/a)

gelu_2_cols_1_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.06 (+2.08%)0.05 (+4.66%)0.04 (-1.75%)0.03 (-5.55%)0.01 (+12.71%)283.70 (+5.86%)191.18 (-3.05%)191.00 (+1.81%)137.60 (-1.99%)58.08 (+15.01%)
dc66da7 — 2026-03-25 19:23:450.06 (n/a)0.04 (n/a)0.04 (n/a)0.03 (n/a)0.01 (n/a)268.00 (n/a)197.20 (n/a)187.60 (n/a)140.40 (n/a)50.50 (n/a)

gelu_2_cols_2_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.07 (+2.20%)0.04 (-2.29%)0.04 (+1.04%)0.04 (+16.91%)0.01 (-4.76%)227.00 (-14.47%)194.56 (+0.69%)206.00 (-1.01%)119.40 (-2.21%)44.27 (-20.37%)
dc66da7 — 2026-03-25 19:23:450.07 (n/a)0.05 (n/a)0.04 (n/a)0.03 (n/a)0.01 (n/a)265.40 (n/a)193.22 (n/a)208.10 (n/a)122.10 (n/a)55.59 (n/a)

gelu_4_cols_1_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.07 (-9.75%)0.04 (-17.39%)0.04 (-20.48%)0.03 (+12.65%)0.01 (-29.41%)239.40 (-11.23%)196.08 (+15.06%)197.90 (+25.73%)124.70 (+10.84%)44.40 (-31.09%)
dc66da7 — 2026-03-25 19:23:450.07 (n/a)0.05 (n/a)0.05 (n/a)0.03 (n/a)0.02 (n/a)269.70 (n/a)170.42 (n/a)157.40 (n/a)112.50 (n/a)64.43 (n/a)

gelu_4_cols_2_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.04 (-6.29%)0.04 (+1.81%)0.04 (+0.94%)0.03 (-0.42%)0.00 (-22.18%)284.90 (+0.42%)233.36 (-2.39%)227.20 (-0.96%)204.80 (+6.72%)30.62 (-15.21%)
dc66da7 — 2026-03-25 19:23:450.04 (n/a)0.03 (n/a)0.04 (n/a)0.03 (n/a)0.01 (n/a)283.70 (n/a)239.08 (n/a)229.40 (n/a)191.90 (n/a)36.12 (n/a)

gelu_8_cols_1_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.05 (-18.24%)0.04 (-13.70%)0.04 (-4.91%)0.03 (-15.26%)0.01 (-28.46%)285.90 (+18.04%)223.38 (+14.20%)223.20 (+5.18%)156.90 (+22.29%)47.35 (+1.82%)
dc66da7 — 2026-03-25 19:23:450.06 (n/a)0.04 (n/a)0.04 (n/a)0.03 (n/a)0.01 (n/a)242.20 (n/a)195.60 (n/a)212.20 (n/a)128.30 (n/a)46.50 (n/a)

gelu_8_cols_2_channels_2048_tile_128

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.05 (+14.38%)0.04 (+10.49%)0.03 (+3.25%)0.03 (+13.03%)0.01 (+26.31%)305.70 (-11.52%)235.82 (-8.98%)241.80 (-3.12%)177.40 (-12.57%)49.18 (-6.09%)
dc66da7 — 2026-03-25 19:23:450.04 (n/a)0.03 (n/a)0.03 (n/a)0.02 (n/a)0.01 (n/a)345.50 (n/a)259.10 (n/a)249.60 (n/a)202.90 (n/a)52.37 (n/a)

gemm_1792x896x1152_64x32x48_8cols_ccolmaj

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
8b6d0f4 — 2026-03-25 20:46:394.89 (+9.71%)4.07 (+2.20%)4.10 (+2.03%)3.40 (+0.07%)0.55 (+44.15%)2767.30 (-0.07%)2344.32 (-1.52%)2291.20 (-1.99%)1922.40 (-8.85%)308.92 (+28.54%)1924.38 (+9.71%)1600.58 (+2.20%)1614.60 (+2.03%)1336.81 (+0.07%)215.84 (+44.15%)
dc66da7 — 2026-03-25 19:23:454.46 (n/a)3.98 (n/a)4.02 (n/a)3.40 (n/a)0.38 (n/a)2769.20 (n/a)2380.42 (n/a)2337.70 (n/a)2109.10 (n/a)240.33 (n/a)1754.05 (n/a)1566.13 (n/a)1582.49 (n/a)1335.88 (n/a)149.73 (n/a)

gemm_192x384x64_48x96x16_4cols

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
8b6d0f4 — 2026-03-25 20:46:391.35 (+17.71%)0.77 (-13.89%)0.65 (-34.22%)0.56 (+7.33%)0.32 (+33.14%)393.60 (-6.84%)315.66 (+18.48%)341.40 (+52.00%)164.30 (-15.05%)87.82 (-5.13%)57.45 (+17.71%)32.93 (-13.89%)27.64 (-34.22%)23.97 (+7.33%)13.81 (+33.14%)
dc66da7 — 2026-03-25 19:23:451.14 (n/a)0.90 (n/a)0.98 (n/a)0.52 (n/a)0.24 (n/a)422.50 (n/a)266.42 (n/a)224.60 (n/a)193.40 (n/a)92.57 (n/a)48.81 (n/a)38.25 (n/a)42.02 (n/a)22.34 (n/a)10.37 (n/a)

gemm_192x384x64_48x96x16_4cols_bcolmaj_ccolmaj

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
8b6d0f4 — 2026-03-25 20:46:391.18 (+31.18%)0.79 (-0.76%)0.72 (-17.59%)0.64 (+26.96%)0.22 (+35.66%)343.10 (-21.24%)294.50 (+0.95%)309.00 (+21.37%)188.10 (-23.78%)62.26 (-23.01%)50.17 (+31.18%)33.62 (-0.76%)30.54 (-17.59%)27.51 (+26.96%)9.40 (+35.66%)
dc66da7 — 2026-03-25 19:23:450.90 (n/a)0.79 (n/a)0.87 (n/a)0.51 (n/a)0.16 (n/a)435.60 (n/a)291.72 (n/a)254.60 (n/a)246.80 (n/a)80.87 (n/a)38.24 (n/a)33.88 (n/a)37.06 (n/a)21.67 (n/a)6.93 (n/a)

gemm_2048x2048x2048_64x64x32_8_cols_0_bcolmaj_0_ccolmaj_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)GFLOP/s (max)GFLOP/s (mean)GFLOP/s (median)GFLOP/s (min)GFLOP/s (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
3a053d3 — 2025-11-26 01:12:163.72 (-0.11%)3.55 (-2.75%)3.56 (-2.75%)3.36 (-5.33%)0.12 (+76.10%)2725.23 (-0.11%)2602.52 (-2.75%)2612.11 (-2.75%)2459.89 (-5.33%)86.27 (+76.10%)6984.00 (+5.63%)6608.60 (+2.91%)6577.00 (+2.83%)6304.00 (+0.11%)221.87 (+87.15%)
9e98049 — 2025-11-25 20:50:553.72 (n/a)3.65 (n/a)3.66 (n/a)3.54 (n/a)0.07 (n/a)2728.26 (n/a)2676.06 (n/a)2686.03 (n/a)2598.29 (n/a)48.99 (n/a)6612.00 (n/a)6422.00 (n/a)6396.00 (n/a)6297.00 (n/a)118.55 (n/a)

gemm_2048x2048x2048_64x64x32_8_cols_0_bcolmaj_1_ccolmaj_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)GFLOP/s (max)GFLOP/s (mean)GFLOP/s (median)GFLOP/s (min)GFLOP/s (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
3a053d3 — 2025-11-26 01:12:163.40 (-3.14%)3.38 (-0.82%)3.38 (-0.82%)3.35 (+4.93%)0.02 (-84.51%)2493.81 (-3.14%)2475.27 (-0.82%)2479.06 (-0.82%)2453.92 (+4.93%)13.33 (-84.51%)7001.00 (-4.70%)6940.80 (+0.71%)6930.00 (+0.83%)6889.00 (+3.24%)37.43 (-84.76%)
9e98049 — 2025-11-25 20:50:553.51 (n/a)3.40 (n/a)3.41 (n/a)3.19 (n/a)0.12 (n/a)2574.53 (n/a)2495.72 (n/a)2499.62 (n/a)2338.67 (n/a)86.02 (n/a)7346.00 (n/a)6892.20 (n/a)6873.00 (n/a)6673.00 (n/a)245.57 (n/a)

gemm_2048x2048x2048_64x64x32_8_cols_1_bcolmaj_0_ccolmaj_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)GFLOP/s (max)GFLOP/s (mean)GFLOP/s (median)GFLOP/s (min)GFLOP/s (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
3a053d3 — 2025-11-26 01:12:163.02 (+0.52%)2.87 (-2.05%)2.86 (-2.24%)2.70 (-6.16%)0.11 (+131.66%)2216.19 (+0.52%)2104.44 (-2.05%)2094.08 (-2.24%)1980.84 (-6.16%)80.25 (+131.67%)8673.00 (+6.56%)8175.60 (+2.21%)8204.00 (+2.29%)7752.00 (-0.51%)314.03 (+145.36%)
9e98049 — 2025-11-25 20:50:553.01 (n/a)2.93 (n/a)2.92 (n/a)2.88 (n/a)0.05 (n/a)2204.81 (n/a)2148.41 (n/a)2142.13 (n/a)2110.81 (n/a)34.64 (n/a)8139.00 (n/a)7998.60 (n/a)8020.00 (n/a)7792.00 (n/a)127.99 (n/a)

gemm_2048x2048x2048_64x64x64_1cols

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
8b6d0f4 — 2026-03-25 20:46:390.51 (+0.13%)0.51 (+0.07%)0.51 (+0.04%)0.51 (+0.11%)0.00 (+8.62%)49296.10 (-0.11%)49206.24 (-0.07%)49283.30 (-0.04%)48878.90 (-0.13%)183.15 (+8.40%)351.48 (+0.13%)349.14 (+0.07%)348.59 (+0.04%)348.50 (+0.11%)1.31 (+8.62%)
dc66da7 — 2026-03-25 19:23:450.51 (n/a)0.51 (n/a)0.51 (n/a)0.51 (n/a)0.00 (n/a)49348.10 (n/a)49241.40 (n/a)49302.10 (n/a)48941.60 (n/a)168.96 (n/a)351.03 (n/a)348.89 (n/a)348.46 (n/a)348.14 (n/a)1.20 (n/a)

gemm_2048x2048x2048_64x64x64_2_cols_0_bcolmaj_0_ccolmaj_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
130b6ea — 2025-12-05 21:15:491.04 (+0.94%)1.03 (+0.23%)1.03 (+0.09%)1.02 (+0.07%)0.00 (+197.51%)24576.00 (-0.07%)24485.82 (-0.23%)24530.40 (-0.09%)24266.40 (-0.94%)111.26 (+194.35%)
0a6c11c — 2025-12-04 00:39:101.03 (n/a)1.03 (n/a)1.03 (n/a)1.02 (n/a)0.00 (n/a)24594.30 (n/a)24542.14 (n/a)24551.90 (n/a)24495.50 (n/a)37.80 (n/a)

gemm_2048x2048x2048_64x64x64_2_cols_0_bcolmaj_0_ccolmaj_0_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
a4b6ffe — 2026-01-05 19:24:051.04 (+0.28%)1.03 (-0.15%)1.03 (+0.05%)1.02 (-0.52%)0.01 (+42.72%)24719.80 (+0.52%)24512.42 (+0.16%)24529.20 (-0.05%)24092.60 (-0.28%)252.92 (+43.15%)713.08 (+0.28%)700.92 (-0.15%)700.38 (+0.05%)694.98 (-0.52%)7.30 (+42.71%)
50ad37c — 2025-12-29 20:35:531.04 (n/a)1.03 (n/a)1.03 (n/a)1.02 (n/a)0.01 (n/a)24592.30 (n/a)24473.70 (n/a)24540.90 (n/a)24160.80 (n/a)176.68 (n/a)711.06 (n/a)702.00 (n/a)700.05 (n/a)698.59 (n/a)5.12 (n/a)

gemm_2048x2048x2048_64x64x64_2_cols_0_bcolmaj_1_ccolmaj_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
130b6ea — 2025-12-05 21:15:490.99 (+1.69%)0.97 (+0.29%)0.97 (-0.04%)0.97 (-0.15%)0.01 (+434.92%)26025.30 (+0.15%)25839.34 (-0.29%)25916.70 (+0.04%)25446.00 (-1.67%)202.05 (+425.44%)
0a6c11c — 2025-12-04 00:39:100.97 (n/a)0.97 (n/a)0.97 (n/a)0.97 (n/a)0.00 (n/a)25986.90 (n/a)25913.52 (n/a)25905.70 (n/a)25877.00 (n/a)38.45 (n/a)

gemm_2048x2048x2048_64x64x64_2_cols_0_bcolmaj_1_ccolmaj_0_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
a4b6ffe — 2026-01-05 19:24:050.97 (-0.01%)0.97 (+0.03%)0.97 (+0.04%)0.97 (+0.00%)0.00 (-6.43%)25927.90 (-0.00%)25903.20 (-0.03%)25898.60 (-0.04%)25892.80 (+0.01%)14.05 (-6.56%)663.50 (-0.01%)663.23 (+0.03%)663.35 (+0.04%)662.60 (+0.00%)0.36 (-6.43%)
50ad37c — 2025-12-29 20:35:530.97 (n/a)0.97 (n/a)0.97 (n/a)0.97 (n/a)0.00 (n/a)25928.10 (n/a)25910.14 (n/a)25908.10 (n/a)25891.20 (n/a)15.04 (n/a)663.54 (n/a)663.06 (n/a)663.11 (n/a)662.60 (n/a)0.38 (n/a)

gemm_2048x2048x2048_64x64x64_2_cols_1_bcolmaj_0_ccolmaj_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
130b6ea — 2025-12-05 21:15:490.92 (-1.02%)0.92 (-0.66%)0.92 (-0.10%)0.91 (-0.78%)0.01 (-4.54%)27751.40 (+0.79%)27480.74 (+0.66%)27354.80 (+0.10%)27320.50 (+1.03%)182.69 (-2.84%)
0a6c11c — 2025-12-04 00:39:100.93 (n/a)0.92 (n/a)0.92 (n/a)0.91 (n/a)0.01 (n/a)27534.80 (n/a)27299.32 (n/a)27326.70 (n/a)27042.30 (n/a)188.04 (n/a)

gemm_2048x2048x2048_64x64x64_2_cols_1_bcolmaj_0_ccolmaj_0_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
a4b6ffe — 2026-01-05 19:24:050.92 (+0.11%)0.91 (-0.12%)0.91 (-0.21%)0.91 (-0.24%)0.00 (+30.34%)27723.50 (+0.24%)27548.94 (+0.12%)27522.90 (+0.21%)27403.60 (-0.11%)124.26 (+30.58%)626.92 (+0.11%)623.62 (-0.12%)624.20 (-0.21%)619.69 (-0.24%)2.81 (+30.34%)
50ad37c — 2025-12-29 20:35:530.92 (n/a)0.91 (n/a)0.92 (n/a)0.91 (n/a)0.00 (n/a)27657.10 (n/a)27515.58 (n/a)27464.70 (n/a)27434.60 (n/a)95.16 (n/a)626.21 (n/a)624.37 (n/a)625.52 (n/a)621.17 (n/a)2.16 (n/a)

gemm_2048x2048x2048_64x64x64_2cols_bcolmaj

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
8b6d0f4 — 2026-03-25 20:46:390.88 (+0.21%)0.87 (-0.11%)0.87 (-0.26%)0.87 (+0.03%)0.00 (+24.60%)29019.10 (-0.03%)28862.30 (+0.12%)28862.50 (+0.26%)28665.10 (-0.21%)148.73 (+24.28%)599.33 (+0.21%)595.25 (-0.11%)595.23 (-0.26%)592.02 (+0.03%)3.07 (+24.61%)
dc66da7 — 2026-03-25 19:23:450.88 (n/a)0.87 (n/a)0.87 (n/a)0.87 (n/a)0.00 (n/a)29027.70 (n/a)28829.10 (n/a)28788.30 (n/a)28724.70 (n/a)119.67 (n/a)598.09 (n/a)595.93 (n/a)596.77 (n/a)591.84 (n/a)2.46 (n/a)

gemm_2048x2048x2048_64x64x64_8_cols_0_bcolmaj_0_ccolmaj_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)GFLOP/s (max)GFLOP/s (mean)GFLOP/s (median)GFLOP/s (min)GFLOP/s (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
31ab24e — 2025-11-22 02:01:563.65 (-1.06%)3.59 (+0.83%)3.63 (-0.26%)3.45 (+2.66%)0.07 (-44.71%)2678.08 (-1.06%)2632.13 (+0.83%)2658.60 (-0.26%)2528.31 (+2.66%)53.86 (-44.71%)6795.00 (-2.59%)6529.80 (-0.92%)6462.00 (+0.26%)6415.00 (+1.07%)137.14 (-45.27%)
d9a64c5 — 2025-11-21 20:49:353.69 (n/a)3.56 (n/a)3.64 (n/a)3.36 (n/a)0.13 (n/a)2706.77 (n/a)2610.50 (n/a)2665.61 (n/a)2462.71 (n/a)97.41 (n/a)6976.00 (n/a)6590.40 (n/a)6445.00 (n/a)6347.00 (n/a)250.58 (n/a)

gemm_2048x2048x2048_64x64x64_8_cols_0_bcolmaj_0_ccolmaj_0_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
d48746f — 2026-01-12 22:17:103.98 (+0.20%)3.84 (-2.95%)3.76 (-5.27%)3.72 (-4.97%)0.13 (+473.97%)6760.00 (+5.23%)6559.50 (+3.13%)6697.10 (+5.56%)6321.00 (-0.20%)218.68 (+499.40%)2717.91 (+0.20%)2621.44 (-2.95%)2565.27 (-5.27%)2541.40 (-4.97%)88.37 (+473.97%)
331dcca — 2026-01-08 18:34:103.97 (n/a)3.96 (n/a)3.97 (n/a)3.92 (n/a)0.02 (n/a)6423.80 (n/a)6360.32 (n/a)6344.30 (n/a)6333.90 (n/a)36.48 (n/a)2712.39 (n/a)2701.17 (n/a)2707.91 (n/a)2674.40 (n/a)15.40 (n/a)

gemm_2048x2048x2048_64x64x64_8_cols_0_bcolmaj_1_ccolmaj_0_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
d48746f — 2026-01-12 22:17:103.75 (+0.28%)3.70 (+1.25%)3.71 (-0.21%)3.65 (+4.77%)0.04 (-66.90%)6901.80 (-4.56%)6795.46 (-1.30%)6790.20 (+0.21%)6715.90 (-0.28%)67.05 (-68.40%)2558.08 (+0.28%)2528.33 (+1.25%)2530.10 (-0.21%)2489.18 (+4.77%)24.81 (-66.90%)
331dcca — 2026-01-08 18:34:103.74 (n/a)3.66 (n/a)3.71 (n/a)3.48 (n/a)0.11 (n/a)7231.20 (n/a)6885.18 (n/a)6775.70 (n/a)6734.90 (n/a)212.21 (n/a)2550.87 (n/a)2497.04 (n/a)2535.50 (n/a)2375.78 (n/a)74.96 (n/a)

gemm_2048x2048x2048_64x64x64_8_cols_1_bcolmaj_0_ccolmaj_0_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
d48746f — 2026-01-12 22:17:103.62 (-0.77%)3.54 (+0.14%)3.60 (-0.04%)3.44 (+2.17%)0.09 (-25.77%)7307.40 (-2.13%)7103.46 (-0.18%)6982.50 (+0.04%)6948.20 (+0.77%)182.95 (-26.55%)2472.55 (-0.77%)2419.80 (+0.14%)2460.41 (-0.04%)2351.04 (+2.17%)61.76 (-25.77%)
331dcca — 2026-01-08 18:34:103.65 (n/a)3.54 (n/a)3.61 (n/a)3.37 (n/a)0.12 (n/a)7466.20 (n/a)7116.40 (n/a)6979.60 (n/a)6895.10 (n/a)249.09 (n/a)2491.61 (n/a)2416.45 (n/a)2461.45 (n/a)2301.00 (n/a)83.20 (n/a)

gemm_2048x2048x2048_64x64x64_8cols_bcolmaj_ccolmaj

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
8b6d0f4 — 2026-03-25 20:46:393.47 (-0.81%)3.41 (+0.82%)3.42 (-0.23%)3.35 (+4.63%)0.05 (-63.03%)7522.10 (-4.43%)7390.72 (-0.92%)7367.60 (+0.23%)7255.80 (+0.82%)107.14 (-64.21%)2367.74 (-0.81%)2324.90 (+0.82%)2331.81 (-0.23%)2283.92 (+4.63%)33.69 (-63.03%)
dc66da7 — 2026-03-25 19:23:453.50 (n/a)3.38 (n/a)3.42 (n/a)3.20 (n/a)0.13 (n/a)7870.70 (n/a)7459.22 (n/a)7350.70 (n/a)7196.70 (n/a)299.40 (n/a)2387.20 (n/a)2306.10 (n/a)2337.18 (n/a)2182.77 (n/a)91.13 (n/a)

gemm_384x1536x1792_32x48x64_4cols_bcolmaj

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
8b6d0f4 — 2026-03-25 20:46:394.16 (+12.80%)3.50 (+5.22%)3.50 (+2.51%)2.86 (-0.71%)0.63 (+68.68%)2814.80 (+0.72%)2362.04 (-3.41%)2304.60 (-2.45%)1936.00 (-11.34%)430.86 (+52.19%)1091.93 (+12.80%)919.16 (+5.22%)917.28 (+2.51%)751.01 (-0.71%)166.31 (+68.68%)
dc66da7 — 2026-03-25 19:23:453.69 (n/a)3.33 (n/a)3.41 (n/a)2.88 (n/a)0.38 (n/a)2794.70 (n/a)2445.42 (n/a)2362.50 (n/a)2183.70 (n/a)283.10 (n/a)968.06 (n/a)873.57 (n/a)894.79 (n/a)756.39 (n/a)98.60 (n/a)

gemm_896x1792x640_32x64x80_8cols_ccolmaj

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
8b6d0f4 — 2026-03-25 20:46:396.32 (+6.48%)4.74 (+3.95%)4.60 (+2.04%)3.49 (+3.11%)1.02 (+11.51%)1906.60 (-3.02%)1453.08 (-3.45%)1446.20 (-2.01%)1052.10 (-6.09%)304.01 (+0.60%)1953.36 (+6.48%)1465.54 (+3.95%)1421.06 (+2.04%)1077.93 (+3.11%)313.81 (+11.51%)
dc66da7 — 2026-03-25 19:23:455.94 (n/a)4.56 (n/a)4.51 (n/a)3.38 (n/a)0.91 (n/a)1966.00 (n/a)1504.94 (n/a)1475.80 (n/a)1120.30 (n/a)302.18 (n/a)1834.47 (n/a)1409.81 (n/a)1392.60 (n/a)1045.38 (n/a)281.42 (n/a)

input_length_2048-num_aie_columns_1-num_channels_1-tile_size_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
dd8281d — 2026-03-26 01:33:330.04 (n/a)0.03 (n/a)0.03 (n/a)0.02 (n/a)0.01 (n/a)395.50 (n/a)268.65 (n/a)235.00 (n/a)213.00 (n/a)58.21 (n/a)

input_length_2048-num_aie_columns_1-num_channels_1-tile_size_2048-group_size_32

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
dd8281d — 2026-03-26 01:33:330.02 (n/a)0.02 (n/a)0.02 (n/a)0.01 (n/a)0.01 (n/a)540.10 (n/a)322.24 (n/a)287.80 (n/a)222.80 (n/a)124.94 (n/a)

input_length_2048-num_aie_columns_1-num_channels_1-tile_size_2048-weighted_False

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
dd8281d — 2026-03-26 01:33:330.03 (n/a)0.03 (n/a)0.03 (n/a)0.02 (n/a)0.00 (n/a)351.90 (n/a)285.78 (n/a)258.50 (n/a)256.70 (n/a)41.93 (n/a)

input_length_2048-num_aie_columns_1-num_channels_1-tile_size_2048-weighted_True

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
dd8281d — 2026-03-26 01:33:330.05 (n/a)0.04 (n/a)0.05 (n/a)0.01 (n/a)0.02 (n/a)944.30 (n/a)398.46 (n/a)264.20 (n/a)257.70 (n/a)305.15 (n/a)

input_length_2048-num_aie_columns_1-num_channels_2-tile_size_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
dd8281d — 2026-03-26 01:33:330.04 (n/a)0.03 (n/a)0.03 (n/a)0.02 (n/a)0.01 (n/a)409.80 (n/a)273.84 (n/a)241.70 (n/a)230.00 (n/a)67.02 (n/a)

input_length_2048-num_aie_columns_1-num_channels_2-tile_size_1024-group_size_32

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
dd8281d — 2026-03-26 01:33:330.02 (n/a)0.02 (n/a)0.02 (n/a)0.01 (n/a)0.00 (n/a)534.20 (n/a)319.34 (n/a)271.60 (n/a)254.70 (n/a)120.41 (n/a)

input_length_2048-num_aie_columns_1-num_channels_2-tile_size_1024-weighted_False

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
dd8281d — 2026-03-26 01:33:330.03 (n/a)0.03 (n/a)0.03 (n/a)0.02 (n/a)0.00 (n/a)373.60 (n/a)283.56 (n/a)251.20 (n/a)246.30 (n/a)54.61 (n/a)

input_length_2048-num_aie_columns_1-num_channels_2-tile_size_2048-scalar_factor_3.0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
dd8281d — 2026-03-26 01:33:330.06 (n/a)0.05 (n/a)0.06 (n/a)0.02 (n/a)0.01 (n/a)524.00 (n/a)285.34 (n/a)223.30 (n/a)218.40 (n/a)133.64 (n/a)

input_length_2048-num_aie_columns_1-tile_size_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
dd8281d — 2026-03-26 01:33:330.05 (n/a)0.04 (n/a)0.04 (n/a)0.02 (n/a)0.01 (n/a)654.50 (n/a)314.03 (n/a)256.10 (n/a)224.40 (n/a)132.18 (n/a)

input_length_2048-num_aie_columns_2-num_channels_1-tile_size_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
dd8281d — 2026-03-26 01:33:330.04 (n/a)0.03 (n/a)0.03 (n/a)0.01 (n/a)0.01 (n/a)653.70 (n/a)287.70 (n/a)247.40 (n/a)221.20 (n/a)91.64 (n/a)

input_length_2048-num_aie_columns_2-num_channels_1-tile_size_1024-group_size_32

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
dd8281d — 2026-03-26 01:33:330.02 (n/a)0.02 (n/a)0.02 (n/a)0.01 (n/a)0.00 (n/a)552.50 (n/a)344.90 (n/a)295.40 (n/a)280.90 (n/a)116.31 (n/a)

input_length_2048-num_aie_columns_2-num_channels_1-tile_size_1024-weighted_False

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
dd8281d — 2026-03-26 01:33:330.03 (n/a)0.03 (n/a)0.03 (n/a)0.02 (n/a)0.01 (n/a)451.70 (n/a)301.96 (n/a)255.90 (n/a)247.90 (n/a)86.20 (n/a)

input_length_2048-num_aie_columns_2-num_channels_1-tile_size_1024-weighted_True

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
dd8281d — 2026-03-26 01:33:330.04 (n/a)0.03 (n/a)0.04 (n/a)0.02 (n/a)0.01 (n/a)679.00 (n/a)352.74 (n/a)266.70 (n/a)258.00 (n/a)183.22 (n/a)

input_length_2048-num_aie_columns_2-num_channels_2-tile_size_1024-scalar_factor_3.0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
dd8281d — 2026-03-26 01:33:330.05 (n/a)0.04 (n/a)0.05 (n/a)0.02 (n/a)0.02 (n/a)803.50 (n/a)370.02 (n/a)254.50 (n/a)230.70 (n/a)244.33 (n/a)

input_length_2048-num_aie_columns_2-num_channels_2-tile_size_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
dd8281d — 2026-03-26 01:33:330.04 (n/a)0.03 (n/a)0.03 (n/a)0.01 (n/a)0.01 (n/a)700.40 (n/a)314.31 (n/a)249.15 (n/a)227.10 (n/a)149.43 (n/a)

input_length_2048-num_aie_columns_2-num_channels_2-tile_size_512-group_size_32

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
dd8281d — 2026-03-26 01:33:330.02 (n/a)0.02 (n/a)0.02 (n/a)0.01 (n/a)0.01 (n/a)852.60 (n/a)398.26 (n/a)293.90 (n/a)259.90 (n/a)254.40 (n/a)

input_length_2048-num_aie_columns_2-num_channels_2-tile_size_512-weighted_False

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
dd8281d — 2026-03-26 01:33:330.03 (n/a)0.03 (n/a)0.03 (n/a)0.02 (n/a)0.01 (n/a)463.10 (n/a)292.22 (n/a)252.20 (n/a)239.90 (n/a)95.69 (n/a)

input_length_2048-num_aie_columns_2-tile_size_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
dd8281d — 2026-03-26 01:33:330.06 (n/a)0.04 (n/a)0.04 (n/a)0.01 (n/a)0.01 (n/a)649.50 (n/a)327.19 (n/a)263.50 (n/a)219.70 (n/a)154.67 (n/a)

input_length_2048-num_aie_columns_4-num_channels_1-tile_size_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
dd8281d — 2026-03-26 01:33:330.04 (n/a)0.03 (n/a)0.03 (n/a)0.02 (n/a)0.01 (n/a)472.10 (n/a)270.32 (n/a)243.70 (n/a)221.60 (n/a)62.66 (n/a)

input_length_2048-num_aie_columns_4-num_channels_1-tile_size_512-group_size_32

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
dd8281d — 2026-03-26 01:33:330.02 (n/a)0.02 (n/a)0.02 (n/a)0.01 (n/a)0.01 (n/a)691.30 (n/a)363.80 (n/a)301.90 (n/a)252.50 (n/a)184.29 (n/a)

input_length_2048-num_aie_columns_4-num_channels_1-tile_size_512-weighted_False

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
dd8281d — 2026-03-26 01:33:330.04 (n/a)0.03 (n/a)0.03 (n/a)0.02 (n/a)0.01 (n/a)363.20 (n/a)268.18 (n/a)236.20 (n/a)224.50 (n/a)57.56 (n/a)

input_length_2048-num_aie_columns_4-num_channels_1-tile_size_512-weighted_True

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
dd8281d — 2026-03-26 01:33:330.04 (n/a)0.03 (n/a)0.04 (n/a)0.02 (n/a)0.01 (n/a)400.10 (n/a)288.90 (n/a)263.10 (n/a)257.80 (n/a)62.22 (n/a)

input_length_2048-num_aie_columns_4-num_channels_2-tile_size_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
dd8281d — 2026-03-26 01:33:330.04 (n/a)0.03 (n/a)0.03 (n/a)0.01 (n/a)0.01 (n/a)798.70 (n/a)387.26 (n/a)281.75 (n/a)230.60 (n/a)216.97 (n/a)

input_length_2048-num_aie_columns_4-num_channels_2-tile_size_256-group_size_32

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
dd8281d — 2026-03-26 01:33:330.02 (n/a)0.02 (n/a)0.02 (n/a)0.01 (n/a)0.00 (n/a)381.20 (n/a)315.36 (n/a)300.40 (n/a)291.50 (n/a)37.10 (n/a)

input_length_2048-num_aie_columns_4-num_channels_2-tile_size_256-weighted_False

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
dd8281d — 2026-03-26 01:33:330.03 (n/a)0.03 (n/a)0.03 (n/a)0.01 (n/a)0.01 (n/a)652.40 (n/a)337.84 (n/a)262.40 (n/a)253.00 (n/a)175.91 (n/a)

input_length_2048-num_aie_columns_4-num_channels_2-tile_size_512-scalar_factor_3.0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
dd8281d — 2026-03-26 01:33:330.05 (n/a)0.05 (n/a)0.05 (n/a)0.02 (n/a)0.01 (n/a)572.70 (n/a)307.70 (n/a)231.50 (n/a)223.50 (n/a)150.14 (n/a)

input_length_2048-num_aie_columns_4-tile_size_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
dd8281d — 2026-03-26 01:33:330.05 (n/a)0.04 (n/a)0.04 (n/a)0.02 (n/a)0.01 (n/a)731.90 (n/a)315.75 (n/a)234.40 (n/a)225.40 (n/a)156.13 (n/a)

input_length_2048-num_aie_columns_8-num_channels_1-tile_size_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
dd8281d — 2026-03-26 01:33:330.03 (n/a)0.03 (n/a)0.03 (n/a)0.01 (n/a)0.01 (n/a)937.70 (n/a)359.04 (n/a)253.10 (n/a)242.30 (n/a)204.55 (n/a)

input_length_2048-num_aie_columns_8-num_channels_1-tile_size_256-group_size_32

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
dd8281d — 2026-03-26 01:33:330.02 (n/a)0.02 (n/a)0.02 (n/a)0.01 (n/a)0.00 (n/a)355.80 (n/a)306.76 (n/a)296.00 (n/a)283.20 (n/a)28.33 (n/a)

input_length_2048-num_aie_columns_8-num_channels_1-tile_size_256-weighted_False

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
dd8281d — 2026-03-26 01:33:330.03 (n/a)0.03 (n/a)0.03 (n/a)0.01 (n/a)0.01 (n/a)647.20 (n/a)341.78 (n/a)279.50 (n/a)244.70 (n/a)171.86 (n/a)

input_length_2048-num_aie_columns_8-num_channels_1-tile_size_256-weighted_True

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
dd8281d — 2026-03-26 01:33:330.04 (n/a)0.03 (n/a)0.04 (n/a)0.01 (n/a)0.01 (n/a)625.90 (n/a)320.54 (n/a)247.10 (n/a)236.70 (n/a)170.76 (n/a)

input_length_2048-num_aie_columns_8-num_channels_2-tile_size_128

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
dd8281d — 2026-03-26 01:33:330.03 (n/a)0.03 (n/a)0.03 (n/a)0.01 (n/a)0.01 (n/a)658.20 (n/a)329.32 (n/a)269.60 (n/a)254.10 (n/a)130.45 (n/a)

input_length_2048-num_aie_columns_8-num_channels_2-tile_size_128-group_size_32

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
dd8281d — 2026-03-26 01:33:330.02 (n/a)0.02 (n/a)0.02 (n/a)0.01 (n/a)0.00 (n/a)382.70 (n/a)302.94 (n/a)306.10 (n/a)253.10 (n/a)51.16 (n/a)

input_length_2048-num_aie_columns_8-num_channels_2-tile_size_128-weighted_False

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
dd8281d — 2026-03-26 01:33:330.03 (n/a)0.02 (n/a)0.03 (n/a)0.01 (n/a)0.01 (n/a)1229.00 (n/a)483.20 (n/a)321.70 (n/a)261.80 (n/a)417.84 (n/a)

input_length_2048-num_aie_columns_8-num_channels_2-tile_size_256-scalar_factor_3.0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
dd8281d — 2026-03-26 01:33:330.05 (n/a)0.04 (n/a)0.05 (n/a)0.04 (n/a)0.00 (n/a)339.90 (n/a)282.86 (n/a)266.60 (n/a)262.50 (n/a)32.56 (n/a)

input_length_2048-num_aie_columns_8-tile_size_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
dd8281d — 2026-03-26 01:33:330.05 (n/a)0.04 (n/a)0.04 (n/a)0.01 (n/a)0.01 (n/a)886.10 (n/a)348.31 (n/a)259.90 (n/a)224.50 (n/a)203.38 (n/a)

input_length_2048-num_cores_1-num_channels_1-bypass_False-tile_size_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
dd8281d — 2026-03-26 01:33:330.04 (n/a)0.03 (n/a)0.03 (n/a)0.02 (n/a)0.01 (n/a)538.60 (n/a)305.94 (n/a)248.00 (n/a)222.60 (n/a)131.46 (n/a)

input_length_2048-num_cores_16-num_channels_2-bypass_False-tile_size_128

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
dd8281d — 2026-03-26 01:33:330.03 (n/a)0.02 (n/a)0.03 (n/a)0.01 (n/a)0.01 (n/a)942.00 (n/a)417.32 (n/a)276.10 (n/a)269.90 (n/a)294.09 (n/a)

input_length_2048-num_cores_2-num_channels_1-bypass_False-tile_size_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
dd8281d — 2026-03-26 01:33:330.03 (n/a)0.03 (n/a)0.03 (n/a)0.01 (n/a)0.01 (n/a)719.30 (n/a)351.38 (n/a)255.40 (n/a)235.80 (n/a)207.08 (n/a)

input_length_2048-num_cores_2-num_channels_2-bypass_False-tile_size_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
dd8281d — 2026-03-26 01:33:330.03 (n/a)0.03 (n/a)0.03 (n/a)0.02 (n/a)0.00 (n/a)333.80 (n/a)265.62 (n/a)249.60 (n/a)242.70 (n/a)38.43 (n/a)

input_length_2048-num_cores_4-num_channels_1-bypass_False-tile_size_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
dd8281d — 2026-03-26 01:33:330.03 (n/a)0.03 (n/a)0.03 (n/a)0.02 (n/a)0.01 (n/a)392.80 (n/a)283.28 (n/a)252.60 (n/a)251.90 (n/a)61.54 (n/a)

input_length_2048-num_cores_4-num_channels_2-bypass_False-tile_size_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
dd8281d — 2026-03-26 01:33:330.04 (n/a)0.03 (n/a)0.04 (n/a)0.02 (n/a)0.00 (n/a)343.90 (n/a)256.76 (n/a)233.80 (n/a)232.80 (n/a)48.79 (n/a)

input_length_2048-num_cores_8-num_channels_1-bypass_False-tile_size_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
dd8281d — 2026-03-26 01:33:330.03 (n/a)0.03 (n/a)0.03 (n/a)0.02 (n/a)0.01 (n/a)381.80 (n/a)292.12 (n/a)252.20 (n/a)237.30 (n/a)67.88 (n/a)

input_length_2048-num_cores_8-num_channels_2-bypass_False-tile_size_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
dd8281d — 2026-03-26 01:33:330.03 (n/a)0.03 (n/a)0.03 (n/a)0.02 (n/a)0.00 (n/a)387.10 (n/a)293.32 (n/a)259.80 (n/a)254.90 (n/a)57.00 (n/a)

input_length_32768-num_aie_columns_2-num_channels_2-tile_size_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
dd8281d — 2026-03-26 01:33:330.49 (n/a)0.37 (n/a)0.42 (n/a)0.15 (n/a)0.13 (n/a)876.10 (n/a)423.30 (n/a)314.20 (n/a)266.30 (n/a)255.26 (n/a)

input_length_32768-num_aie_columns_2-num_channels_2-tile_size_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
dd8281d — 2026-03-26 01:33:330.51 (n/a)0.44 (n/a)0.47 (n/a)0.32 (n/a)0.07 (n/a)405.10 (n/a)306.52 (n/a)278.20 (n/a)259.00 (n/a)59.31 (n/a)

input_length_32768-num_aie_columns_4-num_channels_4-tile_size_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
dd8281d — 2026-03-26 01:33:330.47 (n/a)0.35 (n/a)0.46 (n/a)0.15 (n/a)0.16 (n/a)851.70 (n/a)473.46 (n/a)283.50 (n/a)280.90 (n/a)269.82 (n/a)

layer_norm_1_cols_1_channels_2048_tile_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.07 (+35.61%)0.05 (+6.58%)0.04 (-0.65%)0.03 (-1.11%)0.01 (+98.33%)236.50 (+1.11%)182.90 (-3.12%)183.30 (+0.60%)118.50 (-26.26%)42.30 (+41.46%)
dc66da7 — 2026-03-25 19:23:450.05 (n/a)0.04 (n/a)0.04 (n/a)0.04 (n/a)0.01 (n/a)233.90 (n/a)188.80 (n/a)182.20 (n/a)160.70 (n/a)29.90 (n/a)

layer_norm_1_cols_2_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.07 (+9.47%)0.05 (+5.79%)0.05 (+5.48%)0.04 (-7.03%)0.01 (+51.12%)206.90 (+7.54%)163.06 (-2.60%)165.20 (-5.22%)113.20 (-8.71%)40.37 (+56.59%)
dc66da7 — 2026-03-25 19:23:450.07 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.01 (n/a)192.40 (n/a)167.42 (n/a)174.30 (n/a)124.00 (n/a)25.78 (n/a)

layer_norm_2_cols_1_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.06 (-0.97%)0.05 (+14.59%)0.05 (+47.43%)0.03 (+8.92%)0.01 (-0.31%)235.10 (-8.20%)183.06 (-13.03%)157.80 (-32.19%)148.10 (+1.02%)43.12 (-7.94%)
dc66da7 — 2026-03-25 19:23:450.06 (n/a)0.04 (n/a)0.04 (n/a)0.03 (n/a)0.01 (n/a)256.10 (n/a)210.48 (n/a)232.70 (n/a)146.60 (n/a)46.84 (n/a)

layer_norm_2_cols_2_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.07 (+14.65%)0.05 (+4.61%)0.05 (-7.99%)0.04 (+11.94%)0.01 (+28.58%)201.70 (-10.67%)168.58 (-3.65%)176.00 (+8.71%)122.10 (-12.79%)33.82 (+0.06%)
dc66da7 — 2026-03-25 19:23:450.06 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.01 (n/a)225.80 (n/a)174.96 (n/a)161.90 (n/a)140.00 (n/a)33.80 (n/a)

layer_norm_4_cols_1_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.06 (-31.05%)0.05 (-3.73%)0.05 (+9.03%)0.04 (+10.49%)0.01 (-63.42%)197.60 (-9.52%)167.50 (-4.83%)166.40 (-8.27%)129.00 (+45.11%)25.47 (-51.02%)
dc66da7 — 2026-03-25 19:23:450.09 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.02 (n/a)218.40 (n/a)176.00 (n/a)181.40 (n/a)88.90 (n/a)52.00 (n/a)

layer_norm_4_cols_2_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.05 (-1.52%)0.04 (-11.20%)0.04 (-25.45%)0.03 (-0.69%)0.01 (-9.35%)243.40 (+0.66%)202.20 (+12.10%)212.70 (+34.20%)156.90 (+1.55%)34.18 (-7.92%)
dc66da7 — 2026-03-25 19:23:450.05 (n/a)0.05 (n/a)0.05 (n/a)0.03 (n/a)0.01 (n/a)241.80 (n/a)180.38 (n/a)158.50 (n/a)154.50 (n/a)37.12 (n/a)

layer_norm_8_cols_1_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.05 (-6.07%)0.04 (+15.60%)0.04 (+21.55%)0.04 (+82.28%)0.01 (-55.28%)222.00 (-45.13%)190.18 (-21.85%)193.60 (-17.72%)153.10 (+6.47%)24.64 (-75.05%)
dc66da7 — 2026-03-25 19:23:450.06 (n/a)0.04 (n/a)0.03 (n/a)0.02 (n/a)0.01 (n/a)404.60 (n/a)243.34 (n/a)235.30 (n/a)143.80 (n/a)98.74 (n/a)

layer_norm_8_cols_2_channels_2048_tile_128

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.05 (+37.82%)0.04 (+27.22%)0.04 (+9.46%)0.03 (+34.60%)0.01 (+56.79%)251.50 (-25.72%)212.60 (-21.06%)226.70 (-8.66%)175.20 (-27.45%)33.04 (-18.34%)
dc66da7 — 2026-03-25 19:23:450.03 (n/a)0.03 (n/a)0.03 (n/a)0.02 (n/a)0.00 (n/a)338.60 (n/a)269.32 (n/a)248.20 (n/a)241.50 (n/a)40.46 (n/a)

matrix_vector_mul_128x128_32_1col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)
130b6ea — 2025-12-05 21:15:490.37 (+17.13%)0.30 (+24.87%)0.30 (+38.03%)0.23 (+12.47%)0.06 (+35.23%)
0a6c11c — 2025-12-04 00:39:100.32 (n/a)0.24 (n/a)0.22 (n/a)0.20 (n/a)0.04 (n/a)

matrix_vector_mul_128x128_32_1col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
331dcca — 2026-01-08 18:34:100.26 (-1.81%)0.24 (+4.06%)0.24 (+0.52%)0.21 (+16.97%)0.02 (-48.16%)0.26 (-1.81%)0.24 (+4.06%)0.24 (+0.52%)0.20 (+16.97%)0.02 (-48.16%)
a4b6ffe — 2026-01-05 19:24:050.26 (n/a)0.23 (n/a)0.24 (n/a)0.18 (n/a)0.04 (n/a)0.26 (n/a)0.23 (n/a)0.24 (n/a)0.17 (n/a)0.04 (n/a)

matrix_vector_mul_128x128_32tsi_128tso_1col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
8b6d0f4 — 2026-03-25 20:46:390.26 (+32.10%)0.19 (+11.55%)0.18 (+9.49%)0.13 (-2.22%)0.05 (+105.83%)0.25 (+32.10%)0.19 (+11.55%)0.18 (+9.49%)0.13 (-2.22%)0.05 (+105.83%)
dc66da7 — 2026-03-25 19:23:450.19 (n/a)0.17 (n/a)0.17 (n/a)0.14 (n/a)0.02 (n/a)0.19 (n/a)0.17 (n/a)0.17 (n/a)0.14 (n/a)0.02 (n/a)

matrix_vector_mul_128x128_32tsi_128tso_1col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
897d04e — 2026-03-06 22:12:250.37 (+5.44%)0.31 (-0.05%)0.34 (-1.99%)0.22 (-7.87%)0.06 (+15.13%)0.36 (+5.44%)0.31 (-0.05%)0.34 (-1.99%)0.21 (-7.87%)0.06 (+15.13%)
84d3478 — 2026-02-17 23:25:130.35 (n/a)0.31 (n/a)0.35 (n/a)0.23 (n/a)0.05 (n/a)0.35 (n/a)0.31 (n/a)0.34 (n/a)0.23 (n/a)0.05 (n/a)

matrix_vector_mul_2048x8192_1_1col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)
130b6ea — 2025-12-05 21:15:4913.57 (+3.47%)13.15 (+1.09%)13.10 (+0.16%)12.52 (-1.58%)0.37 (+153.19%)
0a6c11c — 2025-12-04 00:39:1013.11 (n/a)13.01 (n/a)13.07 (n/a)12.72 (n/a)0.15 (n/a)

matrix_vector_mul_2048x8192_1_1col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
331dcca — 2026-01-08 18:34:1013.43 (-0.07%)13.30 (+4.62%)13.27 (+1.29%)13.16 (+16.70%)0.12 (-86.69%)13.42 (-0.07%)13.29 (+4.62%)13.26 (+1.29%)13.16 (+16.70%)0.12 (-86.69%)
a4b6ffe — 2026-01-05 19:24:0513.44 (n/a)12.71 (n/a)13.10 (n/a)11.28 (n/a)0.87 (n/a)13.43 (n/a)12.70 (n/a)13.09 (n/a)11.27 (n/a)0.87 (n/a)

matrix_vector_mul_2048x8192_1_2col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)
130b6ea — 2025-12-05 21:15:4925.06 (+2.47%)23.71 (+1.81%)24.13 (+0.50%)20.90 (+5.41%)1.48 (-15.60%)
0a6c11c — 2025-12-04 00:39:1024.46 (n/a)23.29 (n/a)24.01 (n/a)19.83 (n/a)1.75 (n/a)

matrix_vector_mul_2048x8192_1_2col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
331dcca — 2026-01-08 18:34:1024.46 (-0.60%)21.07 (-8.03%)19.75 (-17.83%)18.08 (+1.18%)3.03 (+7.07%)24.44 (-0.60%)21.06 (-8.03%)19.74 (-17.83%)18.07 (+1.18%)3.03 (+7.07%)
a4b6ffe — 2026-01-05 19:24:0524.60 (n/a)22.91 (n/a)24.04 (n/a)17.87 (n/a)2.83 (n/a)24.59 (n/a)22.90 (n/a)24.02 (n/a)17.86 (n/a)2.83 (n/a)

matrix_vector_mul_2048x8192_1_4col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)
130b6ea — 2025-12-05 21:15:4942.05 (+1.54%)40.50 (+12.60%)40.49 (+2.36%)38.98 (+97.53%)0.97 (-88.09%)
0a6c11c — 2025-12-04 00:39:1041.41 (n/a)35.97 (n/a)39.55 (n/a)19.73 (n/a)8.15 (n/a)

matrix_vector_mul_2048x8192_1_4col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
331dcca — 2026-01-08 18:34:1042.58 (+1.51%)41.37 (+14.29%)41.25 (+4.89%)40.33 (+91.42%)0.92 (-89.18%)42.56 (+1.51%)41.34 (+14.29%)41.23 (+4.89%)40.30 (+91.42%)0.92 (-89.18%)
a4b6ffe — 2026-01-05 19:24:0541.95 (n/a)36.20 (n/a)39.33 (n/a)21.07 (n/a)8.54 (n/a)41.92 (n/a)36.17 (n/a)39.31 (n/a)21.05 (n/a)8.53 (n/a)

matrix_vector_mul_2048x8192_1_8col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)
130b6ea — 2025-12-05 21:15:4946.62 (+3.11%)44.82 (+0.17%)44.78 (+0.17%)42.60 (-3.80%)1.60 (+367.72%)
0a6c11c — 2025-12-04 00:39:1045.21 (n/a)44.75 (n/a)44.70 (n/a)44.29 (n/a)0.34 (n/a)

matrix_vector_mul_2048x8192_1_8col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
331dcca — 2026-01-08 18:34:1046.04 (+4.14%)45.06 (+3.47%)44.55 (+2.76%)44.30 (+2.84%)0.87 (+66.58%)46.01 (+4.14%)45.03 (+3.47%)44.52 (+2.76%)44.27 (+2.84%)0.87 (+66.58%)
a4b6ffe — 2026-01-05 19:24:0544.21 (n/a)43.55 (n/a)43.35 (n/a)43.07 (n/a)0.52 (n/a)44.18 (n/a)43.52 (n/a)43.32 (n/a)43.05 (n/a)0.52 (n/a)

matrix_vector_mul_2048x8192_1tsi_1024tso_2col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
8b6d0f4 — 2026-03-25 20:46:3925.44 (+1.95%)24.42 (-0.28%)24.79 (+0.16%)23.14 (-2.33%)1.05 (+106.47%)25.42 (+1.95%)24.41 (-0.28%)24.77 (+0.16%)23.12 (-2.33%)1.05 (+106.47%)
dc66da7 — 2026-03-25 19:23:4524.95 (n/a)24.49 (n/a)24.75 (n/a)23.69 (n/a)0.51 (n/a)24.94 (n/a)24.48 (n/a)24.73 (n/a)23.67 (n/a)0.51 (n/a)

matrix_vector_mul_2048x8192_1tsi_1024tso_2col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
897d04e — 2026-03-06 22:12:2524.47 (-3.32%)24.16 (-3.13%)24.20 (-3.09%)23.84 (-3.04%)0.25 (-14.36%)24.46 (-3.32%)24.15 (-3.13%)24.18 (-3.09%)23.83 (-3.04%)0.25 (-14.36%)
84d3478 — 2026-02-17 23:25:1325.31 (n/a)24.94 (n/a)24.97 (n/a)24.59 (n/a)0.29 (n/a)25.30 (n/a)24.93 (n/a)24.95 (n/a)24.58 (n/a)0.29 (n/a)

matrix_vector_mul_2048x8192_1tsi_2048tso_1col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
8b6d0f4 — 2026-03-25 20:46:3913.48 (-1.01%)13.15 (+1.31%)13.11 (-0.87%)12.95 (+6.77%)0.20 (-66.05%)13.47 (-1.01%)13.14 (+1.31%)13.11 (-0.87%)12.95 (+6.77%)0.20 (-66.05%)
dc66da7 — 2026-03-25 19:23:4513.62 (n/a)12.98 (n/a)13.23 (n/a)12.13 (n/a)0.58 (n/a)13.61 (n/a)12.97 (n/a)13.22 (n/a)12.13 (n/a)0.58 (n/a)

matrix_vector_mul_2048x8192_1tsi_2048tso_1col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
897d04e — 2026-03-06 22:12:2513.25 (-3.04%)13.02 (-2.92%)13.10 (-2.64%)12.56 (-4.14%)0.27 (+32.55%)13.24 (-3.04%)13.01 (-2.92%)13.09 (-2.64%)12.55 (-4.14%)0.27 (+32.55%)
84d3478 — 2026-02-17 23:25:1313.66 (n/a)13.41 (n/a)13.45 (n/a)13.10 (n/a)0.20 (n/a)13.65 (n/a)13.40 (n/a)13.44 (n/a)13.09 (n/a)0.20 (n/a)

matrix_vector_mul_2048x8192_1tsi_256tso_8col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
8b6d0f4 — 2026-03-25 20:46:3942.06 (-4.87%)40.85 (-4.53%)40.57 (-4.52%)39.78 (-4.49%)0.87 (-11.49%)42.04 (-4.87%)40.83 (-4.53%)40.55 (-4.52%)39.76 (-4.49%)0.87 (-11.49%)
dc66da7 — 2026-03-25 19:23:4544.21 (n/a)42.79 (n/a)42.49 (n/a)41.65 (n/a)0.99 (n/a)44.19 (n/a)42.76 (n/a)42.47 (n/a)41.63 (n/a)0.99 (n/a)

matrix_vector_mul_2048x8192_1tsi_256tso_8col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
897d04e — 2026-03-06 22:12:2547.71 (+3.51%)45.38 (+9.55%)45.22 (-1.27%)43.47 (+75.24%)1.52 (-83.67%)47.68 (+3.51%)45.35 (+9.55%)45.19 (-1.27%)43.45 (+75.24%)1.52 (-83.67%)
84d3478 — 2026-02-17 23:25:1346.09 (n/a)41.43 (n/a)45.80 (n/a)24.81 (n/a)9.31 (n/a)46.06 (n/a)41.40 (n/a)45.77 (n/a)24.79 (n/a)9.30 (n/a)

matrix_vector_mul_2048x8192_1tsi_512tso_4col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
8b6d0f4 — 2026-03-25 20:46:3940.94 (-0.67%)39.64 (+0.12%)39.96 (+1.02%)37.79 (+0.51%)1.25 (-21.17%)40.91 (-0.67%)39.61 (+0.12%)39.94 (+1.02%)37.77 (+0.51%)1.25 (-21.17%)
dc66da7 — 2026-03-25 19:23:4541.21 (n/a)39.59 (n/a)39.56 (n/a)37.60 (n/a)1.58 (n/a)41.19 (n/a)39.57 (n/a)39.53 (n/a)37.58 (n/a)1.58 (n/a)

matrix_vector_mul_2048x8192_1tsi_512tso_4col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
897d04e — 2026-03-06 22:12:2540.84 (-0.32%)39.53 (-1.25%)39.40 (-1.97%)38.87 (-0.69%)0.77 (-6.24%)40.81 (-0.32%)39.50 (-1.25%)39.37 (-1.97%)38.84 (-0.69%)0.77 (-6.24%)
84d3478 — 2026-02-17 23:25:1340.97 (n/a)40.03 (n/a)40.19 (n/a)39.14 (n/a)0.82 (n/a)40.95 (n/a)40.00 (n/a)40.17 (n/a)39.11 (n/a)0.82 (n/a)

matrix_vector_mul_8192x2048_4_1col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)
130b6ea — 2025-12-05 21:15:4913.29 (+1.63%)13.19 (+6.08%)13.19 (+1.17%)13.09 (+26.86%)0.08 (-92.94%)
0a6c11c — 2025-12-04 00:39:1013.08 (n/a)12.44 (n/a)13.03 (n/a)10.32 (n/a)1.07 (n/a)

matrix_vector_mul_8192x2048_4_1col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
331dcca — 2026-01-08 18:34:1013.68 (+3.42%)13.30 (+8.54%)13.38 (+7.25%)12.72 (+14.96%)0.35 (-66.09%)13.67 (+3.42%)13.29 (+8.54%)13.38 (+7.25%)12.71 (+14.96%)0.35 (-66.09%)
a4b6ffe — 2026-01-05 19:24:0513.23 (n/a)12.25 (n/a)12.48 (n/a)11.07 (n/a)1.04 (n/a)13.22 (n/a)12.24 (n/a)12.47 (n/a)11.06 (n/a)1.04 (n/a)

matrix_vector_mul_8192x2048_4_2col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)
130b6ea — 2025-12-05 21:15:4925.58 (+4.24%)24.23 (+6.59%)24.03 (+0.29%)23.58 (+37.29%)0.70 (-74.97%)
0a6c11c — 2025-12-04 00:39:1024.54 (n/a)22.73 (n/a)23.96 (n/a)17.18 (n/a)2.79 (n/a)

matrix_vector_mul_8192x2048_4_2col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
331dcca — 2026-01-08 18:34:1025.08 (+1.32%)24.68 (+13.42%)24.61 (+3.26%)24.47 (+48.32%)0.24 (-93.56%)25.07 (+1.32%)24.67 (+13.42%)24.59 (+3.26%)24.46 (+48.32%)0.24 (-93.56%)
a4b6ffe — 2026-01-05 19:24:0524.76 (n/a)21.76 (n/a)23.83 (n/a)16.50 (n/a)3.74 (n/a)24.74 (n/a)21.75 (n/a)23.82 (n/a)16.49 (n/a)3.73 (n/a)

matrix_vector_mul_8192x2048_4_4col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)
130b6ea — 2025-12-05 21:15:4940.94 (-2.19%)39.74 (-1.10%)39.76 (-0.60%)38.17 (-2.14%)1.04 (+4.39%)
0a6c11c — 2025-12-04 00:39:1041.86 (n/a)40.18 (n/a)40.00 (n/a)39.00 (n/a)1.00 (n/a)

matrix_vector_mul_8192x2048_4_4col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
331dcca — 2026-01-08 18:34:1040.74 (-1.09%)37.16 (-7.15%)40.25 (+1.47%)24.51 (-37.44%)7.07 (+736.13%)40.72 (-1.09%)37.13 (-7.15%)40.23 (+1.47%)24.50 (-37.44%)7.07 (+736.13%)
a4b6ffe — 2026-01-05 19:24:0541.19 (n/a)40.02 (n/a)39.67 (n/a)39.18 (n/a)0.85 (n/a)41.17 (n/a)39.99 (n/a)39.64 (n/a)39.16 (n/a)0.85 (n/a)

matrix_vector_mul_8192x2048_4_8col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)
130b6ea — 2025-12-05 21:15:4946.71 (-1.04%)44.91 (-2.34%)44.71 (-2.98%)43.80 (-1.44%)0.96 (+6.93%)
0a6c11c — 2025-12-04 00:39:1047.20 (n/a)45.98 (n/a)46.08 (n/a)44.44 (n/a)0.90 (n/a)

matrix_vector_mul_8192x2048_4_8col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
331dcca — 2026-01-08 18:34:1046.28 (+5.68%)44.89 (+14.59%)44.74 (+4.26%)44.20 (+62.80%)0.84 (-87.96%)46.25 (+5.68%)44.86 (+14.59%)44.71 (+4.26%)44.17 (+62.80%)0.84 (-87.96%)
a4b6ffe — 2026-01-05 19:24:0543.79 (n/a)39.17 (n/a)42.91 (n/a)27.15 (n/a)6.98 (n/a)43.77 (n/a)39.15 (n/a)42.88 (n/a)27.13 (n/a)6.98 (n/a)

matrix_vector_mul_8192x2048_4tsi_1024tso_1col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
8b6d0f4 — 2026-03-25 20:46:3913.35 (+0.83%)13.11 (+3.52%)13.10 (+3.94%)12.94 (+7.70%)0.15 (-69.46%)13.34 (+0.83%)13.10 (+3.52%)13.09 (+3.94%)12.94 (+7.70%)0.15 (-69.46%)
dc66da7 — 2026-03-25 19:23:4513.24 (n/a)12.66 (n/a)12.60 (n/a)12.02 (n/a)0.49 (n/a)13.23 (n/a)12.66 (n/a)12.59 (n/a)12.01 (n/a)0.49 (n/a)

matrix_vector_mul_8192x2048_4tsi_1024tso_1col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
897d04e — 2026-03-06 22:12:2513.34 (-0.79%)13.20 (-0.36%)13.18 (-1.55%)13.12 (+1.75%)0.08 (-65.93%)13.33 (-0.79%)13.19 (-0.36%)13.17 (-1.55%)13.11 (+1.75%)0.08 (-65.93%)
84d3478 — 2026-02-17 23:25:1313.45 (n/a)13.25 (n/a)13.39 (n/a)12.90 (n/a)0.25 (n/a)13.44 (n/a)13.24 (n/a)13.38 (n/a)12.89 (n/a)0.25 (n/a)

matrix_vector_mul_8192x2048_4tsi_1024tso_2col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
8b6d0f4 — 2026-03-25 20:46:3924.35 (-0.26%)24.07 (+0.89%)24.16 (+1.48%)23.64 (+2.17%)0.27 (-44.88%)24.34 (-0.26%)24.06 (+0.89%)24.15 (+1.48%)23.63 (+2.17%)0.27 (-44.88%)
dc66da7 — 2026-03-25 19:23:4524.41 (n/a)23.86 (n/a)23.81 (n/a)23.14 (n/a)0.49 (n/a)24.40 (n/a)23.84 (n/a)23.80 (n/a)23.12 (n/a)0.49 (n/a)

matrix_vector_mul_8192x2048_4tsi_1024tso_2col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
897d04e — 2026-03-06 22:12:2524.10 (-5.45%)23.07 (-1.85%)24.01 (-2.88%)19.26 (+7.47%)2.13 (-32.15%)24.08 (-5.45%)23.05 (-1.85%)23.99 (-2.88%)19.25 (+7.47%)2.13 (-32.15%)
84d3478 — 2026-02-17 23:25:1325.48 (n/a)23.50 (n/a)24.72 (n/a)17.92 (n/a)3.14 (n/a)25.47 (n/a)23.48 (n/a)24.71 (n/a)17.91 (n/a)3.14 (n/a)

matrix_vector_mul_8192x2048_4tsi_1024tso_4col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
8b6d0f4 — 2026-03-25 20:46:3940.84 (-1.16%)38.86 (-0.66%)38.49 (-1.61%)37.90 (+1.94%)1.14 (-23.22%)40.82 (-1.16%)38.84 (-0.66%)38.46 (-1.61%)37.87 (+1.94%)1.14 (-23.22%)
dc66da7 — 2026-03-25 19:23:4541.32 (n/a)39.12 (n/a)39.12 (n/a)37.18 (n/a)1.49 (n/a)41.30 (n/a)39.09 (n/a)39.09 (n/a)37.15 (n/a)1.49 (n/a)

matrix_vector_mul_8192x2048_4tsi_1024tso_4col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
897d04e — 2026-03-06 22:12:2540.80 (-0.16%)39.58 (-1.93%)39.75 (-1.71%)38.28 (-2.50%)1.10 (+67.33%)40.78 (-0.16%)39.55 (-1.93%)39.73 (-1.71%)38.26 (-2.50%)1.10 (+67.33%)
84d3478 — 2026-02-17 23:25:1340.87 (n/a)40.36 (n/a)40.45 (n/a)39.26 (n/a)0.66 (n/a)40.84 (n/a)40.33 (n/a)40.42 (n/a)39.24 (n/a)0.65 (n/a)

matrix_vector_mul_8192x2048_4tsi_1024tso_8col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
8b6d0f4 — 2026-03-25 20:46:3941.79 (-12.64%)40.98 (-5.83%)40.78 (-1.04%)40.36 (-0.60%)0.56 (-84.11%)41.76 (-12.64%)40.96 (-5.83%)40.76 (-1.04%)40.34 (-0.60%)0.56 (-84.11%)
dc66da7 — 2026-03-25 19:23:4547.83 (n/a)43.52 (n/a)41.21 (n/a)40.61 (n/a)3.53 (n/a)47.80 (n/a)43.50 (n/a)41.19 (n/a)40.58 (n/a)3.53 (n/a)

matrix_vector_mul_8192x2048_4tsi_1024tso_8col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
897d04e — 2026-03-06 22:12:2549.31 (+6.27%)46.12 (+1.67%)45.30 (-0.27%)44.62 (+1.87%)1.85 (+85.10%)49.28 (+6.27%)46.10 (+1.67%)45.28 (-0.27%)44.59 (+1.87%)1.85 (+85.10%)
84d3478 — 2026-02-17 23:25:1346.40 (n/a)45.37 (n/a)45.43 (n/a)43.80 (n/a)1.00 (n/a)46.37 (n/a)45.34 (n/a)45.40 (n/a)43.78 (n/a)1.00 (n/a)

mem_copy_16_cores_2_chans_2048_tile_128_False

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.05 (+7.16%)0.03 (-9.22%)0.03 (-18.40%)0.02 (-19.93%)0.01 (+42.24%)394.80 (+24.90%)264.56 (+14.35%)259.20 (+22.55%)179.80 (-6.65%)81.04 (+64.09%)
dc66da7 — 2026-03-25 19:23:450.04 (n/a)0.04 (n/a)0.04 (n/a)0.03 (n/a)0.01 (n/a)316.10 (n/a)231.36 (n/a)211.50 (n/a)192.60 (n/a)49.39 (n/a)

mem_copy_16_cores_2_chans_2048_tile_128_False0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.08 (-2.12%)0.06 (-16.44%)0.06 (-21.48%)0.03 (-23.84%)0.02 (+10.15%)247.60 (+31.28%)148.38 (+24.08%)128.30 (+27.41%)100.10 (+2.14%)58.70 (+50.73%)
84d3478 — 2026-02-17 23:25:130.08 (n/a)0.07 (n/a)0.08 (n/a)0.04 (n/a)0.02 (n/a)188.60 (n/a)119.58 (n/a)100.70 (n/a)98.00 (n/a)38.94 (n/a)

mem_copy_1_cols_1_channels_2048_tile_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
003b49d — 2025-11-14 19:30:540.05 (-16.99%)0.05 (-12.47%)0.05 (-7.32%)0.04 (-11.17%)0.00 (-29.61%)197.00 (+12.57%)171.00 (+13.70%)164.00 (+7.89%)153.00 (+20.47%)17.49 (-4.32%)
0999a9a — 2025-11-13 19:24:080.06 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.01 (n/a)175.00 (n/a)150.40 (n/a)152.00 (n/a)127.00 (n/a)18.28 (n/a)

mem_copy_1_cols_2_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
003b49d — 2025-11-14 19:30:540.06 (+5.56%)0.05 (-1.59%)0.05 (-1.39%)0.03 (-14.07%)0.01 (+25.03%)263.00 (+16.37%)166.00 (+4.53%)144.00 (+1.41%)126.00 (-5.26%)49.38 (+43.17%)
0999a9a — 2025-11-13 19:24:080.06 (n/a)0.05 (n/a)0.05 (n/a)0.03 (n/a)0.01 (n/a)226.00 (n/a)158.80 (n/a)142.00 (n/a)133.00 (n/a)34.49 (n/a)

mem_copy_1_cores_1_chans_2048_tile_2048_False

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.09 (+75.06%)0.05 (+7.82%)0.04 (-15.32%)0.03 (-13.78%)0.02 (+425.35%)249.10 (+15.97%)190.02 (+3.98%)211.70 (+18.07%)95.60 (-42.89%)62.62 (+236.69%)
dc66da7 — 2026-03-25 19:23:450.05 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.00 (n/a)214.80 (n/a)182.74 (n/a)179.30 (n/a)167.40 (n/a)18.60 (n/a)

mem_copy_1_cores_1_chans_2048_tile_2048_False0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.11 (-0.24%)0.08 (-17.53%)0.08 (-27.05%)0.06 (-15.43%)0.02 (+4.62%)148.50 (+18.23%)113.02 (+23.38%)103.60 (+37.04%)74.10 (+0.14%)31.69 (+33.33%)
84d3478 — 2026-02-17 23:25:130.11 (n/a)0.09 (n/a)0.11 (n/a)0.07 (n/a)0.02 (n/a)125.60 (n/a)91.60 (n/a)75.60 (n/a)74.00 (n/a)23.77 (n/a)

mem_copy_2_cols_1_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
003b49d — 2025-11-14 19:30:540.06 (-15.32%)0.05 (-17.09%)0.05 (-23.81%)0.04 (-15.54%)0.01 (-18.18%)193.00 (+18.40%)160.80 (+20.36%)168.00 (+31.25%)124.00 (+18.10%)22.92 (+11.39%)
0999a9a — 2025-11-13 19:24:080.07 (n/a)0.06 (n/a)0.06 (n/a)0.05 (n/a)0.01 (n/a)163.00 (n/a)133.60 (n/a)128.00 (n/a)105.00 (n/a)20.58 (n/a)

mem_copy_2_cols_2_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
003b49d — 2025-11-14 19:30:540.06 (-23.02%)0.05 (-6.98%)0.05 (-2.13%)0.04 (-15.31%)0.01 (-33.29%)196.00 (+18.07%)147.00 (+6.52%)141.00 (+2.17%)126.00 (+29.90%)25.27 (+8.13%)
0999a9a — 2025-11-13 19:24:080.08 (n/a)0.06 (n/a)0.06 (n/a)0.05 (n/a)0.01 (n/a)166.00 (n/a)138.00 (n/a)138.00 (n/a)97.00 (n/a)23.37 (n/a)

mem_copy_2_cores_1_chans_2048_tile_1024_False

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.07 (+13.70%)0.05 (-1.93%)0.05 (-7.07%)0.04 (-27.15%)0.02 (+157.23%)230.10 (+37.29%)166.58 (+9.82%)161.90 (+7.65%)109.70 (-12.03%)53.34 (+211.05%)
dc66da7 — 2026-03-25 19:23:450.07 (n/a)0.05 (n/a)0.05 (n/a)0.05 (n/a)0.01 (n/a)167.60 (n/a)151.68 (n/a)150.40 (n/a)124.70 (n/a)17.15 (n/a)

mem_copy_2_cores_1_chans_2048_tile_1024_False0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.11 (+27.03%)0.09 (+25.79%)0.11 (+27.09%)0.05 (+12.74%)0.03 (+43.93%)161.90 (-11.29%)94.34 (-18.35%)76.30 (-21.34%)74.00 (-21.36%)38.01 (+0.49%)
84d3478 — 2026-02-17 23:25:130.09 (n/a)0.08 (n/a)0.08 (n/a)0.04 (n/a)0.02 (n/a)182.50 (n/a)115.54 (n/a)97.00 (n/a)94.10 (n/a)37.82 (n/a)

mem_copy_2_cores_2_chans_2048_tile_1024_False

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.07 (-3.17%)0.05 (-7.71%)0.05 (-6.65%)0.04 (-20.60%)0.02 (+42.64%)223.40 (+25.93%)171.62 (+13.04%)172.10 (+7.09%)120.40 (+3.26%)49.35 (+82.52%)
dc66da7 — 2026-03-25 19:23:450.07 (n/a)0.06 (n/a)0.05 (n/a)0.05 (n/a)0.01 (n/a)177.40 (n/a)151.82 (n/a)160.70 (n/a)116.60 (n/a)27.04 (n/a)

mem_copy_2_cores_2_chans_2048_tile_1024_False0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.10 (+2.70%)0.08 (-5.38%)0.09 (+2.10%)0.04 (-43.30%)0.02 (+170.50%)201.60 (+76.38%)112.08 (+16.53%)90.80 (-2.05%)84.60 (-2.65%)50.14 (+375.75%)
84d3478 — 2026-02-17 23:25:130.09 (n/a)0.09 (n/a)0.09 (n/a)0.07 (n/a)0.01 (n/a)114.30 (n/a)96.18 (n/a)92.70 (n/a)86.90 (n/a)10.54 (n/a)

mem_copy_4_cols_1_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
003b49d — 2025-11-14 19:30:540.07 (+11.97%)0.05 (+3.09%)0.05 (+2.10%)0.03 (-11.35%)0.01 (+41.65%)229.00 (+12.81%)156.00 (-0.51%)143.00 (-2.05%)117.00 (-10.69%)38.21 (+48.71%)
0999a9a — 2025-11-13 19:24:080.06 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.01 (n/a)203.00 (n/a)156.80 (n/a)146.00 (n/a)131.00 (n/a)25.69 (n/a)

mem_copy_4_cols_2_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
003b49d — 2025-11-14 19:30:540.06 (+9.02%)0.05 (+5.59%)0.05 (+0.00%)0.04 (+3.63%)0.01 (+25.48%)193.00 (-3.50%)147.40 (-4.66%)147.00 (+0.00%)122.00 (-8.27%)25.24 (+7.48%)
0999a9a — 2025-11-13 19:24:080.06 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.01 (n/a)200.00 (n/a)154.60 (n/a)147.00 (n/a)133.00 (n/a)23.48 (n/a)

mem_copy_4_cores_1_chans_2048_tile_512_False

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.06 (-2.79%)0.05 (-5.43%)0.05 (+6.39%)0.03 (-12.50%)0.01 (+16.26%)242.20 (+14.25%)184.14 (+7.91%)167.80 (-6.05%)128.30 (+2.89%)46.84 (+42.55%)
dc66da7 — 2026-03-25 19:23:450.07 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.01 (n/a)212.00 (n/a)170.64 (n/a)178.60 (n/a)124.70 (n/a)32.86 (n/a)

mem_copy_4_cores_1_chans_2048_tile_512_False0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.09 (-5.87%)0.08 (-3.61%)0.08 (-0.79%)0.06 (-12.74%)0.01 (+22.04%)146.30 (+14.66%)108.02 (+4.96%)99.30 (+0.81%)95.10 (+6.26%)21.49 (+48.38%)
84d3478 — 2026-02-17 23:25:130.09 (n/a)0.08 (n/a)0.08 (n/a)0.06 (n/a)0.01 (n/a)127.60 (n/a)102.92 (n/a)98.50 (n/a)89.50 (n/a)14.49 (n/a)

mem_copy_4_cores_2_chans_2048_tile_512_False

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.05 (+7.78%)0.04 (-14.27%)0.04 (-10.58%)0.03 (-32.73%)0.01 (+109.81%)316.90 (+48.71%)226.08 (+22.64%)209.60 (+11.85%)150.20 (-7.23%)61.97 (+194.31%)
dc66da7 — 2026-03-25 19:23:450.05 (n/a)0.04 (n/a)0.04 (n/a)0.04 (n/a)0.01 (n/a)213.10 (n/a)184.34 (n/a)187.40 (n/a)161.90 (n/a)21.05 (n/a)

mem_copy_4_cores_2_chans_2048_tile_512_False0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.11 (+13.04%)0.09 (+17.19%)0.10 (+20.43%)0.06 (+17.23%)0.02 (+16.34%)142.80 (-14.70%)91.94 (-14.71%)79.00 (-17.02%)77.70 (-11.50%)28.50 (-14.82%)
84d3478 — 2026-02-17 23:25:130.09 (n/a)0.08 (n/a)0.09 (n/a)0.05 (n/a)0.02 (n/a)167.40 (n/a)107.80 (n/a)95.20 (n/a)87.80 (n/a)33.46 (n/a)

mem_copy_8_cols_1_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
003b49d — 2025-11-14 19:30:540.05 (-13.48%)0.05 (-17.79%)0.05 (-15.33%)0.03 (-25.09%)0.01 (+6.67%)267.00 (+33.50%)176.80 (+24.51%)150.00 (+18.11%)141.00 (+15.57%)47.41 (+61.41%)
0999a9a — 2025-11-13 19:24:080.06 (n/a)0.06 (n/a)0.06 (n/a)0.04 (n/a)0.01 (n/a)200.00 (n/a)142.00 (n/a)127.00 (n/a)122.00 (n/a)29.37 (n/a)

mem_copy_8_cols_2_channels_2048_tile_128

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
003b49d — 2025-11-14 19:30:540.05 (+13.38%)0.05 (+23.85%)0.05 (+37.91%)0.05 (+30.72%)0.00 (-56.85%)166.00 (-23.50%)153.40 (-20.60%)153.00 (-27.49%)142.00 (-11.80%)7.63 (-70.60%)
0999a9a — 2025-11-13 19:24:080.05 (n/a)0.04 (n/a)0.04 (n/a)0.04 (n/a)0.01 (n/a)217.00 (n/a)193.20 (n/a)211.00 (n/a)161.00 (n/a)25.96 (n/a)

mem_copy_8_cores_1_chans_2048_tile_256_False

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.06 (-5.71%)0.04 (+0.46%)0.04 (+2.21%)0.03 (-7.12%)0.01 (-7.04%)278.00 (+7.67%)209.24 (-0.41%)216.20 (-2.17%)141.90 (+6.05%)51.53 (+9.47%)
dc66da7 — 2026-03-25 19:23:450.06 (n/a)0.04 (n/a)0.04 (n/a)0.03 (n/a)0.01 (n/a)258.20 (n/a)210.10 (n/a)221.00 (n/a)133.80 (n/a)47.07 (n/a)

mem_copy_8_cores_1_chans_2048_tile_256_False0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.08 (-0.39%)0.07 (+10.07%)0.08 (+1.07%)0.06 (+38.38%)0.01 (-47.35%)136.70 (-27.75%)110.64 (-13.07%)104.70 (-1.13%)103.00 (+0.39%)14.59 (-60.78%)
84d3478 — 2026-02-17 23:25:130.08 (n/a)0.07 (n/a)0.08 (n/a)0.04 (n/a)0.02 (n/a)189.20 (n/a)127.28 (n/a)105.90 (n/a)102.60 (n/a)37.20 (n/a)

mem_copy_8_cores_2_chans_2048_tile_256_False

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.06 (+19.11%)0.04 (+1.88%)0.04 (-11.30%)0.03 (+24.20%)0.01 (+4.86%)262.00 (-19.48%)199.56 (-3.59%)200.50 (+12.70%)139.70 (-16.05%)45.93 (-31.63%)
dc66da7 — 2026-03-25 19:23:450.05 (n/a)0.04 (n/a)0.05 (n/a)0.03 (n/a)0.01 (n/a)325.40 (n/a)207.00 (n/a)177.90 (n/a)166.40 (n/a)67.19 (n/a)

mem_copy_8_cores_2_chans_2048_tile_256_False0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.08 (-17.85%)0.07 (-24.66%)0.08 (-18.40%)0.04 (-32.08%)0.02 (+21.78%)185.30 (+47.18%)127.72 (+37.87%)102.00 (+22.45%)100.40 (+21.70%)38.66 (+106.34%)
84d3478 — 2026-02-17 23:25:130.10 (n/a)0.09 (n/a)0.10 (n/a)0.07 (n/a)0.01 (n/a)125.90 (n/a)92.64 (n/a)83.30 (n/a)82.50 (n/a)18.74 (n/a)

mha

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.18 (+0.10%)0.18 (+0.09%)0.18 (+0.09%)0.18 (+0.08%)0.00 (+11.08%)47507.50 (-0.08%)47468.96 (-0.09%)47481.10 (-0.09%)47413.10 (-0.10%)37.32 (+10.96%)
dc66da7 — 2026-03-25 19:23:450.18 (n/a)0.18 (n/a)0.18 (n/a)0.18 (n/a)0.00 (n/a)47544.20 (n/a)47512.60 (n/a)47521.80 (n/a)47459.60 (n/a)33.63 (n/a)

mha0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
84d3478 — 2026-02-17 23:25:130.18 (-0.51%)0.18 (-0.36%)0.18 (-0.41%)0.18 (-0.21%)0.00 (-35.21%)47468.60 (+0.21%)47403.94 (+0.36%)47426.00 (+0.41%)47314.40 (+0.51%)68.83 (-34.73%)
1531781 — 2026-01-23 18:49:240.18 (n/a)0.18 (n/a)0.18 (n/a)0.18 (n/a)0.00 (n/a)47367.50 (n/a)47232.68 (n/a)47232.00 (n/a)47073.00 (n/a)105.45 (n/a)

mha_16384_64_1_8_0_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.18 (n/a)0.18 (n/a)0.18 (n/a)0.18 (n/a)0.00 (n/a)47533.70 (n/a)47380.56 (n/a)47337.90 (n/a)47268.00 (n/a)110.60 (n/a)

relu_1_cols_1_channels_2048_tile_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.07 (+14.06%)0.05 (+0.31%)0.05 (-3.71%)0.04 (-18.38%)0.01 (+129.69%)208.90 (+22.52%)161.96 (+2.56%)166.90 (+3.86%)120.40 (-12.31%)33.36 (+144.00%)
dc66da7 — 2026-03-25 19:23:450.06 (n/a)0.05 (n/a)0.05 (n/a)0.05 (n/a)0.00 (n/a)170.50 (n/a)157.92 (n/a)160.70 (n/a)137.30 (n/a)13.67 (n/a)

relu_2_cols_1_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.06 (-20.92%)0.05 (-6.86%)0.05 (-1.28%)0.03 (-20.49%)0.01 (-19.18%)264.20 (+25.81%)174.76 (+7.41%)149.00 (+1.29%)140.20 (+26.42%)52.29 (+24.11%)
dc66da7 — 2026-03-25 19:23:450.07 (n/a)0.05 (n/a)0.06 (n/a)0.04 (n/a)0.01 (n/a)210.00 (n/a)162.70 (n/a)147.10 (n/a)110.90 (n/a)42.13 (n/a)

relu_4_cols_1_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.05 (-8.73%)0.04 (-10.26%)0.04 (-4.04%)0.03 (-16.73%)0.01 (+5.34%)259.40 (+20.09%)206.80 (+12.91%)200.50 (+4.21%)150.20 (+9.56%)45.96 (+41.47%)
dc66da7 — 2026-03-25 19:23:450.06 (n/a)0.05 (n/a)0.04 (n/a)0.04 (n/a)0.01 (n/a)216.00 (n/a)183.16 (n/a)192.40 (n/a)137.10 (n/a)32.49 (n/a)

relu_8_cols_1_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.05 (+4.18%)0.04 (-0.33%)0.04 (-1.55%)0.02 (-20.26%)0.01 (+47.86%)369.50 (+25.38%)233.74 (+5.19%)215.30 (+1.56%)165.20 (-4.01%)80.83 (+78.85%)
dc66da7 — 2026-03-25 19:23:450.05 (n/a)0.04 (n/a)0.04 (n/a)0.03 (n/a)0.01 (n/a)294.70 (n/a)222.20 (n/a)212.00 (n/a)172.10 (n/a)45.19 (n/a)

rms_norm_1_cols_1_channels_2048_tile_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.06 (-18.06%)0.05 (-2.61%)0.05 (-1.31%)0.04 (+12.13%)0.01 (-43.39%)226.00 (-10.81%)172.66 (-1.55%)163.30 (+1.30%)140.00 (+22.06%)32.37 (-37.02%)
dc66da7 — 2026-03-25 19:23:450.07 (n/a)0.05 (n/a)0.05 (n/a)0.03 (n/a)0.01 (n/a)253.40 (n/a)175.38 (n/a)161.20 (n/a)114.70 (n/a)51.40 (n/a)

rms_norm_1_cols_2_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.08 (+16.34%)0.05 (+6.02%)0.05 (-3.17%)0.04 (+12.01%)0.02 (+27.81%)218.40 (-10.71%)162.92 (-3.80%)150.40 (+3.23%)108.00 (-14.01%)52.41 (+3.43%)
dc66da7 — 2026-03-25 19:23:450.07 (n/a)0.05 (n/a)0.06 (n/a)0.03 (n/a)0.01 (n/a)244.60 (n/a)169.36 (n/a)145.70 (n/a)125.60 (n/a)50.68 (n/a)

rms_norm_2_cols_1_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.06 (+9.52%)0.04 (-9.59%)0.04 (-17.77%)0.03 (-18.95%)0.01 (+55.38%)255.30 (+23.39%)199.88 (+13.52%)205.30 (+21.62%)137.80 (-8.68%)43.62 (+70.31%)
dc66da7 — 2026-03-25 19:23:450.05 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.01 (n/a)206.90 (n/a)176.08 (n/a)168.80 (n/a)150.90 (n/a)25.61 (n/a)

rms_norm_2_cols_2_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.06 (+6.05%)0.04 (+0.61%)0.04 (-6.97%)0.03 (-2.64%)0.01 (+44.42%)234.40 (+2.72%)195.80 (+1.42%)215.50 (+7.48%)141.90 (-5.65%)40.78 (+43.76%)
dc66da7 — 2026-03-25 19:23:450.05 (n/a)0.04 (n/a)0.04 (n/a)0.04 (n/a)0.01 (n/a)228.20 (n/a)193.06 (n/a)200.50 (n/a)150.40 (n/a)28.37 (n/a)

rms_norm_4_cols_1_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.06 (+10.73%)0.05 (-1.13%)0.05 (+5.17%)0.04 (-10.70%)0.01 (+73.44%)225.10 (+11.99%)181.84 (+3.00%)170.50 (-4.91%)140.00 (-9.68%)33.21 (+78.26%)
dc66da7 — 2026-03-25 19:23:450.05 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.00 (n/a)201.00 (n/a)176.54 (n/a)179.30 (n/a)155.00 (n/a)18.63 (n/a)

rms_norm_4_cols_2_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.06 (-0.51%)0.04 (-4.22%)0.04 (-6.40%)0.03 (+19.93%)0.01 (-12.26%)264.40 (-16.62%)208.96 (+1.64%)223.40 (+6.84%)140.70 (+0.50%)52.99 (-24.83%)
dc66da7 — 2026-03-25 19:23:450.06 (n/a)0.04 (n/a)0.04 (n/a)0.03 (n/a)0.01 (n/a)317.10 (n/a)205.58 (n/a)209.10 (n/a)140.00 (n/a)70.49 (n/a)

rms_norm_8_cols_1_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.06 (-8.19%)0.04 (-16.80%)0.03 (-25.89%)0.03 (-16.40%)0.01 (+5.02%)283.50 (+19.62%)220.72 (+22.74%)240.30 (+34.92%)128.00 (+8.94%)58.39 (+33.24%)
dc66da7 — 2026-03-25 19:23:450.07 (n/a)0.05 (n/a)0.05 (n/a)0.03 (n/a)0.01 (n/a)237.00 (n/a)179.82 (n/a)178.10 (n/a)117.50 (n/a)43.82 (n/a)

rms_norm_8_cols_2_channels_2048_tile_128

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.04 (+4.11%)0.04 (+2.57%)0.04 (-1.84%)0.03 (+8.94%)0.00 (+2.64%)250.80 (-8.20%)224.64 (-2.58%)233.20 (+1.88%)191.20 (-3.97%)25.83 (-9.64%)
dc66da7 — 2026-03-25 19:23:450.04 (n/a)0.04 (n/a)0.04 (n/a)0.03 (n/a)0.00 (n/a)273.20 (n/a)230.60 (n/a)228.90 (n/a)199.10 (n/a)28.59 (n/a)

rope_1_cols_2_channels_4096_tile_4096_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
1b2d230 — 2026-01-13 22:38:030.19 (-21.66%)0.17 (-8.53%)0.18 (+2.09%)0.14 (-6.71%)0.02 (-38.07%)174.10 (+7.20%)150.06 (+8.11%)139.60 (-2.04%)129.80 (+27.63%)20.36 (-10.65%)
d48746f — 2026-01-12 22:17:100.24 (n/a)0.18 (n/a)0.17 (n/a)0.15 (n/a)0.04 (n/a)162.40 (n/a)138.80 (n/a)142.50 (n/a)101.70 (n/a)22.79 (n/a)

rope_1c_32rows_512cols_32arows_0m

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.91 (+31.28%)0.72 (+30.97%)0.75 (+50.09%)0.52 (+23.14%)0.16 (+34.35%)189.50 (-18.81%)142.68 (-23.26%)131.80 (-33.40%)108.20 (-23.86%)33.21 (-13.83%)
dc66da7 — 2026-03-25 19:23:450.69 (n/a)0.55 (n/a)0.50 (n/a)0.42 (n/a)0.12 (n/a)233.40 (n/a)185.92 (n/a)197.90 (n/a)142.10 (n/a)38.54 (n/a)

rope_1c_32rows_512cols_8arows_0m

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.71 (+13.56%)0.53 (+16.33%)0.51 (+20.62%)0.43 (+13.26%)0.12 (+12.08%)172.60 (-11.71%)143.86 (-14.09%)145.70 (-17.07%)103.70 (-11.97%)28.08 (-12.19%)
dc66da7 — 2026-03-25 19:23:450.63 (n/a)0.46 (n/a)0.42 (n/a)0.38 (n/a)0.10 (n/a)195.50 (n/a)167.46 (n/a)175.70 (n/a)117.80 (n/a)31.98 (n/a)

rope_2_cols_2_channels_4096_tile_2048_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
1b2d230 — 2026-01-13 22:38:030.21 (-0.56%)0.17 (-10.62%)0.16 (-13.58%)0.13 (-13.33%)0.03 (+17.05%)183.70 (+15.39%)151.76 (+12.77%)155.20 (+15.73%)117.30 (+0.60%)24.24 (+35.73%)
d48746f — 2026-01-12 22:17:100.21 (n/a)0.19 (n/a)0.18 (n/a)0.15 (n/a)0.02 (n/a)159.20 (n/a)134.58 (n/a)134.10 (n/a)116.60 (n/a)17.86 (n/a)

rope_2c_32rows_512cols_32arows_0m

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.86 (+12.27%)0.68 (+4.35%)0.67 (-1.63%)0.56 (+6.41%)0.12 (+18.02%)174.80 (-6.02%)146.78 (-3.85%)146.60 (+1.66%)114.70 (-10.88%)23.97 (-1.25%)
dc66da7 — 2026-03-25 19:23:450.76 (n/a)0.66 (n/a)0.68 (n/a)0.53 (n/a)0.10 (n/a)186.00 (n/a)152.66 (n/a)144.20 (n/a)128.70 (n/a)24.27 (n/a)

rope_2c_32rows_512cols_8arows_0m

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.63 (+15.38%)0.42 (-5.03%)0.37 (-15.06%)0.26 (-23.68%)0.14 (+76.55%)289.00 (+31.07%)192.28 (+12.08%)197.90 (+17.73%)117.80 (-13.32%)64.38 (+98.12%)
dc66da7 — 2026-03-25 19:23:450.54 (n/a)0.44 (n/a)0.44 (n/a)0.33 (n/a)0.08 (n/a)220.50 (n/a)171.56 (n/a)168.10 (n/a)135.90 (n/a)32.49 (n/a)

rope_4_cols_2_channels_4096_tile_1024_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
1b2d230 — 2026-01-13 22:38:030.22 (+18.92%)0.17 (+6.24%)0.17 (+5.57%)0.14 (-0.50%)0.03 (+90.77%)171.10 (+0.47%)147.40 (-4.39%)148.10 (-5.31%)110.10 (-15.89%)23.59 (+60.67%)
d48746f — 2026-01-12 22:17:100.19 (n/a)0.16 (n/a)0.16 (n/a)0.14 (n/a)0.02 (n/a)170.30 (n/a)154.16 (n/a)156.40 (n/a)130.90 (n/a)14.69 (n/a)

rope_8_cols_2_channels_4096_tile_512_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
1b2d230 — 2026-01-13 22:38:030.17 (-1.30%)0.15 (+6.63%)0.15 (-8.78%)0.14 (+39.68%)0.01 (-67.35%)174.90 (-28.41%)163.42 (-10.36%)167.40 (+9.63%)146.50 (+1.31%)11.19 (-76.17%)
d48746f — 2026-01-12 22:17:100.17 (n/a)0.14 (n/a)0.16 (n/a)0.10 (n/a)0.03 (n/a)244.30 (n/a)182.30 (n/a)152.70 (n/a)144.60 (n/a)46.98 (n/a)

rope_8c_32rows_512cols_32arows_0m

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.60 (+0.58%)0.47 (-10.94%)0.47 (-14.67%)0.39 (-14.06%)0.09 (+57.47%)252.70 (+16.34%)213.46 (+14.31%)211.20 (+17.20%)163.30 (-0.61%)38.58 (+85.72%)
dc66da7 — 2026-03-25 19:23:450.60 (n/a)0.53 (n/a)0.55 (n/a)0.45 (n/a)0.06 (n/a)217.20 (n/a)186.74 (n/a)180.20 (n/a)164.30 (n/a)20.77 (n/a)

rope_8c_32rows_512cols_8arows_0m

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.58 (+33.33%)0.38 (-2.44%)0.32 (-20.46%)0.28 (-18.27%)0.13 (+203.34%)264.90 (+22.36%)209.28 (+9.84%)227.20 (+25.73%)126.60 (-25.00%)59.71 (+180.42%)
dc66da7 — 2026-03-25 19:23:450.44 (n/a)0.39 (n/a)0.41 (n/a)0.34 (n/a)0.04 (n/a)216.50 (n/a)190.54 (n/a)180.70 (n/a)168.80 (n/a)21.29 (n/a)

rows_32-cols_512-angle_rows_32-aie_columns_1-method_type_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
dd8281d — 2026-03-26 01:33:330.38 (n/a)0.35 (n/a)0.37 (n/a)0.25 (n/a)0.06 (n/a)391.80 (n/a)289.54 (n/a)264.50 (n/a)255.80 (n/a)57.56 (n/a)

rows_32-cols_512-angle_rows_32-aie_columns_2-method_type_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
dd8281d — 2026-03-26 01:33:330.42 (n/a)0.33 (n/a)0.35 (n/a)0.21 (n/a)0.09 (n/a)461.70 (n/a)322.96 (n/a)277.00 (n/a)236.80 (n/a)97.35 (n/a)

rows_32-cols_512-angle_rows_32-aie_columns_8-method_type_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
dd8281d — 2026-03-26 01:33:330.40 (n/a)0.29 (n/a)0.31 (n/a)0.12 (n/a)0.11 (n/a)788.50 (n/a)406.36 (n/a)313.10 (n/a)247.60 (n/a)223.29 (n/a)

rows_32-cols_512-angle_rows_8-aie_columns_1-method_type_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
dd8281d — 2026-03-26 01:33:330.25 (n/a)0.23 (n/a)0.25 (n/a)0.17 (n/a)0.03 (n/a)425.80 (n/a)325.24 (n/a)300.60 (n/a)292.00 (n/a)56.48 (n/a)

rows_32-cols_512-angle_rows_8-aie_columns_2-method_type_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
dd8281d — 2026-03-26 01:33:330.27 (n/a)0.24 (n/a)0.26 (n/a)0.19 (n/a)0.03 (n/a)397.80 (n/a)307.82 (n/a)288.70 (n/a)271.50 (n/a)50.98 (n/a)

rows_32-cols_512-angle_rows_8-aie_columns_8-method_type_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
dd8281d — 2026-03-26 01:33:330.25 (n/a)0.21 (n/a)0.24 (n/a)0.11 (n/a)0.06 (n/a)649.40 (n/a)392.00 (n/a)304.60 (n/a)293.40 (n/a)151.82 (n/a)

seq_len_16384-dim_64-num_heads_1-num_pipelines_8-num_kv_heads_0

No metrics available.

seq_len_256-embedding_dim_2048-hidden_dim_2048-prio_accuracy_False

No metrics available.

sigmoid_1_cols_1_channels_2048_tile_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.05 (-14.16%)0.05 (-6.29%)0.05 (-2.65%)0.04 (-7.71%)0.01 (-21.79%)225.80 (+8.35%)180.84 (+6.33%)170.90 (+2.70%)163.30 (+16.48%)25.41 (+0.55%)
dc66da7 — 2026-03-25 19:23:450.06 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.01 (n/a)208.40 (n/a)170.08 (n/a)166.40 (n/a)140.20 (n/a)25.28 (n/a)

sigmoid_2_cols_1_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.06 (+4.68%)0.05 (+2.94%)0.04 (-13.06%)0.04 (+15.35%)0.01 (+8.49%)232.90 (-13.32%)179.44 (-3.16%)197.20 (+15.05%)127.30 (-4.50%)46.58 (-13.93%)
dc66da7 — 2026-03-25 19:23:450.06 (n/a)0.05 (n/a)0.05 (n/a)0.03 (n/a)0.01 (n/a)268.70 (n/a)185.30 (n/a)171.40 (n/a)133.30 (n/a)54.12 (n/a)

sigmoid_4_cols_1_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.07 (+10.37%)0.05 (-2.49%)0.04 (-7.15%)0.04 (-7.47%)0.01 (+64.06%)210.80 (+8.10%)175.44 (+5.70%)183.30 (+7.70%)117.30 (-9.42%)39.46 (+67.78%)
dc66da7 — 2026-03-25 19:23:450.06 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.01 (n/a)195.00 (n/a)165.98 (n/a)170.20 (n/a)129.50 (n/a)23.52 (n/a)

sigmoid_8_cols_1_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.06 (+29.56%)0.05 (+22.81%)0.05 (+37.30%)0.03 (-8.45%)0.01 (+218.31%)242.70 (+9.23%)173.04 (-14.71%)150.20 (-27.16%)134.70 (-22.81%)46.84 (+169.71%)
dc66da7 — 2026-03-25 19:23:450.05 (n/a)0.04 (n/a)0.04 (n/a)0.04 (n/a)0.00 (n/a)222.20 (n/a)202.88 (n/a)206.20 (n/a)174.50 (n/a)17.37 (n/a)

silu_1_cols_1_channels_2048_tile_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.07 (+9.38%)0.05 (-4.51%)0.04 (-23.77%)0.04 (+15.02%)0.01 (+5.83%)207.90 (-13.05%)169.28 (+4.04%)188.60 (+31.15%)114.40 (-8.63%)38.46 (-17.21%)
dc66da7 — 2026-03-25 19:23:450.07 (n/a)0.05 (n/a)0.06 (n/a)0.03 (n/a)0.01 (n/a)239.10 (n/a)162.70 (n/a)143.80 (n/a)125.20 (n/a)46.46 (n/a)

silu_2_cols_1_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.06 (-4.63%)0.05 (-13.54%)0.05 (-12.20%)0.04 (-16.50%)0.01 (+19.45%)213.90 (+19.76%)180.30 (+16.99%)181.70 (+13.92%)134.00 (+4.85%)30.84 (+49.10%)
dc66da7 — 2026-03-25 19:23:450.06 (n/a)0.05 (n/a)0.05 (n/a)0.05 (n/a)0.01 (n/a)178.60 (n/a)154.12 (n/a)159.50 (n/a)127.80 (n/a)20.68 (n/a)

silu_4_cols_1_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.06 (+6.26%)0.05 (-8.81%)0.04 (-14.86%)0.04 (-17.09%)0.01 (+81.20%)214.80 (+20.61%)176.86 (+11.68%)182.90 (+17.47%)133.30 (-5.86%)30.25 (+101.37%)
dc66da7 — 2026-03-25 19:23:450.06 (n/a)0.05 (n/a)0.05 (n/a)0.05 (n/a)0.00 (n/a)178.10 (n/a)158.36 (n/a)155.70 (n/a)141.60 (n/a)15.02 (n/a)

silu_8_cols_1_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.06 (-6.11%)0.05 (-10.71%)0.04 (-10.38%)0.04 (-11.78%)0.01 (+1.86%)224.60 (+13.38%)185.30 (+12.62%)186.00 (+11.58%)140.40 (+6.53%)33.11 (+23.38%)
dc66da7 — 2026-03-25 19:23:450.06 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.01 (n/a)198.10 (n/a)164.54 (n/a)166.70 (n/a)131.80 (n/a)26.84 (n/a)

softmax_1_cols_2_channels_4096_tile_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.19 (-8.94%)0.17 (-8.66%)0.17 (-19.83%)0.14 (+4.32%)0.02 (-35.13%)115.80 (-4.14%)99.26 (+7.47%)98.60 (+24.65%)85.00 (+9.82%)13.65 (-30.86%)
84d3478 — 2026-02-17 23:25:130.21 (n/a)0.18 (n/a)0.21 (n/a)0.14 (n/a)0.04 (n/a)120.80 (n/a)92.36 (n/a)79.10 (n/a)77.40 (n/a)19.75 (n/a)

softmax_2_cols_2_channels_32768_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.64 (-42.09%)0.60 (-31.83%)0.59 (-30.19%)0.54 (-6.68%)0.05 (-76.96%)242.70 (+7.15%)221.12 (+40.50%)220.30 (+43.24%)203.40 (+72.67%)16.99 (-59.06%)
dc66da7 — 2026-03-25 19:23:451.11 (n/a)0.87 (n/a)0.85 (n/a)0.58 (n/a)0.20 (n/a)226.50 (n/a)157.38 (n/a)153.80 (n/a)117.80 (n/a)41.50 (n/a)

softmax_2_cols_2_channels_32768_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.68 (-34.93%)0.62 (-20.25%)0.60 (-17.49%)0.55 (-10.32%)0.05 (-71.29%)240.30 (+11.51%)214.32 (+20.87%)218.20 (+21.22%)193.80 (+53.69%)18.64 (-52.51%)
dc66da7 — 2026-03-25 19:23:451.04 (n/a)0.77 (n/a)0.73 (n/a)0.61 (n/a)0.18 (n/a)215.50 (n/a)177.32 (n/a)180.00 (n/a)126.10 (n/a)39.26 (n/a)

softmax_2_cols_2_channels_4096_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.20 (-0.87%)0.16 (+6.57%)0.17 (-9.55%)0.11 (+29.81%)0.03 (-44.34%)145.70 (-22.99%)104.68 (-15.00%)95.00 (+10.59%)83.80 (+0.84%)24.16 (-54.23%)
84d3478 — 2026-02-17 23:25:130.20 (n/a)0.15 (n/a)0.19 (n/a)0.09 (n/a)0.06 (n/a)189.20 (n/a)123.16 (n/a)85.90 (n/a)83.10 (n/a)52.80 (n/a)

softmax_2_cols_2_channels_4096_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.25 (+23.53%)0.18 (+9.42%)0.18 (-3.86%)0.11 (+2.99%)0.05 (+19.24%)144.10 (-2.90%)98.98 (-8.18%)92.10 (+4.07%)64.30 (-19.02%)29.58 (-7.13%)
84d3478 — 2026-02-17 23:25:130.21 (n/a)0.16 (n/a)0.19 (n/a)0.11 (n/a)0.04 (n/a)148.40 (n/a)107.80 (n/a)88.50 (n/a)79.40 (n/a)31.85 (n/a)

softmax_4_cols_4_channels_32768_tile_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.66 (+11.72%)0.50 (-0.51%)0.54 (-4.72%)0.35 (+6.44%)0.13 (+14.83%)377.70 (-6.04%)276.32 (+1.24%)242.20 (+4.94%)199.30 (-10.51%)76.50 (-0.31%)
dc66da7 — 2026-03-25 19:23:450.59 (n/a)0.51 (n/a)0.57 (n/a)0.33 (n/a)0.11 (n/a)402.00 (n/a)272.94 (n/a)230.80 (n/a)222.70 (n/a)76.73 (n/a)

swiglu

No metrics available.

swiglu_decode_1x2048x2048

No metrics available.

swiglu_decode_1x2048x2048_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.00 (-0.21%)0.00 (+7.74%)0.00 (+0.37%)0.00 (+8.84%)0.00 (-19.58%)5170.80 (-8.12%)3761.80 (-8.70%)3408.60 (-0.36%)3407.30 (+0.21%)787.66 (-22.98%)
84d3478 — 2026-02-17 23:25:130.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)5627.70 (n/a)4120.44 (n/a)3421.00 (n/a)3400.30 (n/a)1022.61 (n/a)

swiglu_prefill_256x2048x2048

No metrics available.

tanh_1_cols_1_channels_2048_tile_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.06 (-7.39%)0.04 (-16.97%)0.04 (-14.92%)0.03 (-28.80%)0.01 (+23.25%)296.40 (+40.47%)201.28 (+24.69%)188.60 (+17.51%)141.90 (+7.99%)59.45 (+90.30%)
dc66da7 — 2026-03-25 19:23:450.06 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.01 (n/a)211.00 (n/a)161.42 (n/a)160.50 (n/a)131.40 (n/a)31.24 (n/a)

tanh_2_cols_1_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.06 (-14.29%)0.05 (-6.63%)0.05 (+2.99%)0.03 (-30.73%)0.01 (+5.83%)293.30 (+44.41%)176.82 (+11.67%)159.70 (-2.92%)128.50 (+16.71%)67.26 (+86.47%)
dc66da7 — 2026-03-25 19:23:450.07 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.01 (n/a)203.10 (n/a)158.34 (n/a)164.50 (n/a)110.10 (n/a)36.07 (n/a)

tanh_4_cols_1_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.06 (-27.96%)0.05 (-4.84%)0.05 (+4.02%)0.04 (+16.65%)0.01 (-63.57%)187.60 (-14.30%)167.32 (-1.72%)171.90 (-3.86%)133.00 (+38.83%)20.69 (-55.09%)
dc66da7 — 2026-03-25 19:23:450.09 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.02 (n/a)218.90 (n/a)170.24 (n/a)178.80 (n/a)95.80 (n/a)46.06 (n/a)

tanh_8_cols_1_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.05 (-24.07%)0.04 (-15.48%)0.04 (-23.02%)0.03 (-3.09%)0.01 (-53.59%)262.30 (+3.19%)198.18 (+12.63%)183.30 (+29.91%)173.30 (+31.69%)36.55 (-34.15%)
dc66da7 — 2026-03-25 19:23:450.06 (n/a)0.05 (n/a)0.06 (n/a)0.03 (n/a)0.01 (n/a)254.20 (n/a)175.96 (n/a)141.10 (n/a)131.60 (n/a)55.51 (n/a)

transpose_2048_M_64_N_1_cols_1_channels_64_m_64_n_8_s

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:392.51 (-25.94%)2.10 (-20.30%)2.01 (-23.11%)1.76 (-13.49%)0.29 (-41.75%)298.70 (+15.60%)253.30 (+23.92%)261.10 (+30.09%)208.60 (+35.02%)33.97 (-9.04%)
dc66da7 — 2026-03-25 19:23:453.39 (n/a)2.64 (n/a)2.61 (n/a)2.03 (n/a)0.49 (n/a)258.40 (n/a)204.40 (n/a)200.70 (n/a)154.50 (n/a)37.34 (n/a)

transpose_2048_M_64_N_1_cols_1_channels_64_m_64_n_8_s0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:254.43 (+0.85%)3.73 (+3.83%)4.02 (-1.30%)2.97 (+16.77%)0.65 (-24.99%)176.70 (-14.35%)144.46 (-6.24%)130.40 (+1.32%)118.40 (-0.84%)26.60 (-35.16%)
84d3478 — 2026-02-17 23:25:134.39 (n/a)3.59 (n/a)4.07 (n/a)2.54 (n/a)0.87 (n/a)206.30 (n/a)154.08 (n/a)128.70 (n/a)119.40 (n/a)41.02 (n/a)

transpose_2048_M_64_N_1_cols_2_channels_64_m_64_n_8_s

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:392.72 (-10.04%)2.30 (-13.01%)2.26 (-13.66%)1.88 (-21.25%)0.34 (+46.90%)278.20 (+26.97%)231.60 (+16.34%)232.20 (+15.81%)192.40 (+11.15%)34.64 (+109.20%)
dc66da7 — 2026-03-25 19:23:453.03 (n/a)2.65 (n/a)2.61 (n/a)2.39 (n/a)0.23 (n/a)219.10 (n/a)199.08 (n/a)200.50 (n/a)173.10 (n/a)16.56 (n/a)

transpose_2048_M_64_N_1_cols_2_channels_64_m_64_n_8_s0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:255.27 (-14.18%)4.35 (-2.49%)4.45 (+2.15%)2.78 (-22.18%)0.98 (-3.38%)188.60 (+28.47%)126.98 (+4.20%)117.90 (-2.08%)99.40 (+16.53%)36.02 (+50.15%)
84d3478 — 2026-02-17 23:25:136.14 (n/a)4.46 (n/a)4.35 (n/a)3.57 (n/a)1.02 (n/a)146.80 (n/a)121.86 (n/a)120.40 (n/a)85.30 (n/a)23.99 (n/a)

weighted_rms_norm_1_cols_2_channels_2048_weights_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.09 (+6.38%)0.07 (-4.87%)0.07 (+1.68%)0.05 (-23.45%)0.01 (+101.20%)230.80 (+30.62%)177.68 (+7.40%)170.00 (-1.68%)138.30 (-5.98%)33.65 (+149.29%)
dc66da7 — 2026-03-25 19:23:450.08 (n/a)0.07 (n/a)0.07 (n/a)0.07 (n/a)0.01 (n/a)176.70 (n/a)165.44 (n/a)172.90 (n/a)147.10 (n/a)13.50 (n/a)

weighted_rms_norm_2_cols_2_channels_2048_weights_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.07 (-25.98%)0.06 (-18.74%)0.06 (-7.04%)0.04 (-20.93%)0.01 (-34.18%)237.00 (+26.47%)189.06 (+22.07%)179.50 (+7.55%)151.40 (+35.06%)34.27 (+14.91%)
dc66da7 — 2026-03-25 19:23:450.09 (n/a)0.07 (n/a)0.06 (n/a)0.05 (n/a)0.01 (n/a)187.40 (n/a)154.88 (n/a)166.90 (n/a)112.10 (n/a)29.83 (n/a)

weighted_rms_norm_4_cols_2_channels_2048_weights_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.06 (-13.39%)0.05 (-21.43%)0.04 (-31.53%)0.04 (-12.12%)0.01 (-0.80%)237.90 (+13.77%)205.32 (+28.15%)225.30 (+46.01%)147.80 (+15.47%)38.15 (+27.36%)
dc66da7 — 2026-03-25 19:23:450.07 (n/a)0.06 (n/a)0.06 (n/a)0.04 (n/a)0.01 (n/a)209.10 (n/a)160.22 (n/a)154.30 (n/a)128.00 (n/a)29.95 (n/a)

weighted_rms_norm_8_cols_2_channels_2048_weights_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.07 (+33.02%)0.04 (-11.48%)0.04 (-24.93%)0.02 (-32.99%)0.02 (+124.58%)371.50 (+49.26%)251.20 (+24.91%)247.70 (+33.17%)124.90 (-24.85%)88.76 (+136.07%)
dc66da7 — 2026-03-25 19:23:450.05 (n/a)0.04 (n/a)0.05 (n/a)0.03 (n/a)0.01 (n/a)248.90 (n/a)201.10 (n/a)186.00 (n/a)166.20 (n/a)37.60 (n/a)

hunhoffe and others added 2 commits March 26, 2026 08:37
Two bugs:

1. mha_mha.o: clang crashes when softmax.cc is passed via -include because
   __AIE_ARCH__ is not yet defined at preamble processing time. Fix by
   adding #include "softmax.cc" directly in mha.cc and removing the -include
   flag from op.py (softmax.cc added as a proper source dependency instead).

2. Fused operator link_with mismatch: FusedMLIROperator prefixes kernel .o
   filenames with op{idx}_ but operator design files passed the un-prefixed
   name as the link_with attribute in Kernel(), causing aiecc to look for
   files that don't exist. Fix by applying func_prefix to the .o filename in
   Kernel() calls across all affected operators (rms_norm, silu, elementwise_add,
   elementwise_mul, softmax, transpose, rope, gemv).

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
@github-actions
Copy link
Contributor

📊 Test Results for Test Example Applications

8c138a8 (2026_03_26_14_51_15)

IRONCLAD

Tested on 2026_03_26_14_51_15 at commit 8c138a8.

Test Checks TTFT (mean)TPS (mean)
llama_3.2_1b_prompt_1024_tokens_1 ❌ 0/5 n/a n/a
llama_3.2_1b_prompt_1024_tokens_40 ❌ 0/5 n/a n/a
llama_3.2_1b_prompt_13_tokens_1 ❌ 0/5 n/a n/a
llama_3.2_1b_prompt_13_tokens_40 ❌ 0/5 n/a n/a
📈 Trends (vs main branch) for Test Example Applications

8c138a8 (2026_03_26_14_51_15)

IRONCLAD Trends

llama_3.2_1b

Commit/Date Num Tokens (max)Num Tokens (mean)Num Tokens (median)Num Tokens (min)Num Tokens (stddev)TPS (max)TPS (mean)TPS (median)TPS (min)TPS (stddev)TTFT (max)TTFT (mean)TTFT (median)TTFT (min)TTFT (stddev)Total (max)Total (mean)Total (median)Total (min)Total (stddev)
130b6ea — 2025-12-05 21:33:1240.00 (+0.00%)40.00 (+0.00%)40.00 (+0.00%)40.00 (+0.00%)0.00 (n/a)4.71 (-0.42%)4.64 (-0.09%)4.64 (+0.65%)4.55 (-0.22%)0.05 (-17.66%)4.41 (-0.34%)4.39 (-0.19%)4.38 (-0.33%)4.37 (-0.15%)0.01 (-25.90%)12.96 (-0.00%)12.80 (+0.07%)12.80 (-0.23%)12.67 (+0.44%)0.09 (-21.12%)
0a6c11c — 2025-12-03 23:35:1540.00 (n/a)40.00 (n/a)40.00 (n/a)40.00 (n/a)0.00 (n/a)4.73 (n/a)4.64 (n/a)4.61 (n/a)4.56 (n/a)0.06 (n/a)4.42 (n/a)4.40 (n/a)4.40 (n/a)4.37 (n/a)0.02 (n/a)12.96 (n/a)12.79 (n/a)12.83 (n/a)12.62 (n/a)0.12 (n/a)

llama_3.2_1b_prompt_1024_tokens_1

Commit/Date TTFT (max)TTFT (mean)TTFT (median)TTFT (min)TTFT (stddev)
8c138a8 — 2026-03-26 14:49:05n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)
8b6d0f4 — 2026-03-25 21:36:352.26 (n/a)2.25 (n/a)2.25 (n/a)2.23 (n/a)0.01 (n/a)

llama_3.2_1b_prompt_1024_tokens_40

Commit/Date TPS (max)TPS (mean)TPS (median)TPS (min)TPS (stddev)TTFT (max)TTFT (mean)TTFT (median)TTFT (min)TTFT (stddev)
8c138a8 — 2026-03-26 14:49:05n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)
8b6d0f4 — 2026-03-25 21:36:354.30 (n/a)4.28 (n/a)4.28 (n/a)4.26 (n/a)0.02 (n/a)2.37 (n/a)2.27 (n/a)2.25 (n/a)2.23 (n/a)0.06 (n/a)

llama_3.2_1b_prompt_13_tokens_1

Commit/Date TTFT (max)TTFT (mean)TTFT (median)TTFT (min)TTFT (stddev)
8c138a8 — 2026-03-26 14:49:05n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)
8b6d0f4 — 2026-03-25 21:36:352.19 (n/a)2.17 (n/a)2.17 (n/a)2.14 (n/a)0.02 (n/a)

llama_3.2_1b_prompt_13_tokens_40

Commit/Date TPS (max)TPS (mean)TPS (median)TPS (min)TPS (stddev)TTFT (max)TTFT (mean)TTFT (median)TTFT (min)TTFT (stddev)
8c138a8 — 2026-03-26 14:49:05n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)
8b6d0f4 — 2026-03-25 21:36:354.30 (n/a)4.27 (n/a)4.27 (n/a)4.25 (n/a)0.02 (n/a)2.19 (n/a)2.17 (n/a)2.16 (n/a)2.15 (n/a)0.01 (n/a)

llama_3.2_1b_prompt_2048_tokens_1

Commit/Date Num_Tokens (max)Num_Tokens (mean)Num_Tokens (median)Num_Tokens (min)Num_Tokens (stddev)TPS (max)TPS (mean)TPS (median)TPS (min)TPS (stddev)TTFT (max)TTFT (mean)TTFT (median)TTFT (min)TTFT (stddev)
897d04e — 2026-03-06 22:56:071.00 (+0.00%)1.00 (+0.00%)1.00 (+0.00%)1.00 (+0.00%)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)2.68 (-1.06%)2.68 (-1.06%)2.68 (-1.06%)2.68 (-1.06%)0.00 (n/a)
84d3478 — 2026-02-17 23:16:231.00 (n/a)1.00 (n/a)1.00 (n/a)1.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)2.70 (n/a)2.70 (n/a)2.70 (n/a)2.70 (n/a)0.00 (n/a)

llama_3.2_1b_prompt_2048_tokens_40

Commit/Date Num_Tokens (max)Num_Tokens (mean)Num_Tokens (median)Num_Tokens (min)Num_Tokens (stddev)TPS (max)TPS (mean)TPS (median)TPS (min)TPS (stddev)TTFT (max)TTFT (mean)TTFT (median)TTFT (min)TTFT (stddev)
897d04e — 2026-03-06 22:56:0740.00 (+0.00%)40.00 (+0.00%)40.00 (+0.00%)40.00 (+0.00%)0.00 (n/a)4.00 (-1.72%)4.00 (-1.72%)4.00 (-1.72%)4.00 (-1.72%)0.00 (n/a)2.70 (-0.44%)2.70 (-0.44%)2.70 (-0.44%)2.70 (-0.44%)0.00 (n/a)
84d3478 — 2026-02-17 23:16:2340.00 (n/a)40.00 (n/a)40.00 (n/a)40.00 (n/a)0.00 (n/a)4.07 (n/a)4.07 (n/a)4.07 (n/a)4.07 (n/a)0.00 (n/a)2.71 (n/a)2.71 (n/a)2.71 (n/a)2.71 (n/a)0.00 (n/a)

@github-actions
Copy link
Contributor

📊 Test Results for Small Benchmark/Test Suite

8c138a8 (2026_03_26_15_00_25)

IRONCLAD

Tested on 2026_03_26_15_00_25 at commit 8c138a8.

Test Checks Latency (mean)Bandwidth (mean)Throughput (mean)
M_128-K_128-num_aie_columns_1-tile_size_input_32-tile_size_output_128 ✅ 5/5 n/a 0.11 0.11
M_1792-K_896-N_1152-num_aie_columns_8-b_col_maj_False-c_col_maj_True-m_64-k_32-n_48-trace_size_0 ✅ 5/5 2297.86 4.18 1645.15
M_192-K_384-N_64-num_aie_columns_4-b_col_maj_False-c_col_maj_False-m_48-k_96-n_16-trace_size_0 ✅ 5/5 473.66 0.59 24.98
M_192-K_384-N_64-num_aie_columns_4-b_col_maj_True-c_col_maj_True-m_48-k_96-n_16-trace_size_0 ✅ 5/5 475.78 0.60 25.81
M_2048-K_2048-N_2048-num_aie_columns_1-b_col_maj_False-c_col_maj_False-m_64-k_64-n_64-trace_size_0 ✅ 5/5 50028.00 0.50 343.41
M_2048-K_2048-N_2048-num_aie_columns_2-b_col_maj_True-c_col_maj_False-m_64-k_64-n_64-trace_size_0 ✅ 5/5 29226.50 0.86 587.84
M_2048-K_2048-N_2048-num_aie_columns_8-b_col_maj_True-c_col_maj_True-m_64-k_64-n_64-trace_size_0 ✅ 5/5 7814.92 3.23 2202.77
M_2048-K_8192-num_aie_columns_1-tile_size_input_1-tile_size_output_2048 ✅ 5/5 n/a 11.85 11.84
M_2048-K_8192-num_aie_columns_2-tile_size_input_1-tile_size_output_1024 ✅ 5/5 n/a 19.78 19.77
M_2048-K_8192-num_aie_columns_4-tile_size_input_1-tile_size_output_512 ✅ 5/5 n/a 30.11 30.09
M_2048-K_8192-num_aie_columns_8-tile_size_input_1-tile_size_output_256 ✅ 5/5 n/a 34.48 34.46
M_2048-N_64-aie_columns_1-channels_1-m_64-n_64-s_8 ✅ 5/5 392.96 1.42 n/a
M_2048-N_64-aie_columns_1-channels_2-m_64-n_64-s_8 ✅ 5/5 404.08 1.33 n/a
M_384-K_1536-N_1792-num_aie_columns_4-b_col_maj_True-c_col_maj_False-m_32-k_48-n_64-trace_size_0 ✅ 5/5 2244.16 3.62 948.44
M_8192-K_2048-num_aie_columns_1-tile_size_input_4-tile_size_output_1024 ✅ 5/5 n/a 12.23 12.22
M_8192-K_2048-num_aie_columns_2-tile_size_input_4-tile_size_output_1024 ✅ 5/5 n/a 19.45 19.43
M_8192-K_2048-num_aie_columns_4-tile_size_input_4-tile_size_output_1024 ✅ 5/5 n/a 31.28 31.26
M_8192-K_2048-num_aie_columns_8-tile_size_input_4-tile_size_output_1024 ✅ 5/5 n/a 33.74 33.72
M_896-K_1792-N_640-num_aie_columns_8-b_col_maj_False-c_col_maj_True-m_32-k_64-n_80-trace_size_0 ✅ 5/5 1527.46 4.83 1491.73
embedding_dim_2048-hidden_dim_2048 ✅ 5/5 n/a n/a n/a
input_length_2048-num_aie_columns_1-num_channels_1-tile_size_2048 ✅ 25/25 316.90 0.03 n/a
input_length_2048-num_aie_columns_1-num_channels_1-tile_size_2048-group_size_32 ✅ 5/5 333.16 0.02 n/a
input_length_2048-num_aie_columns_1-num_channels_1-tile_size_2048-weighted_False ✅ 5/5 274.30 0.03 n/a
input_length_2048-num_aie_columns_1-num_channels_1-tile_size_2048-weighted_True ✅ 5/5 283.26 0.04 n/a
input_length_2048-num_aie_columns_1-num_channels_2-tile_size_1024 ✅ 10/10 265.74 0.03 n/a
input_length_2048-num_aie_columns_1-num_channels_2-tile_size_1024-group_size_32 ✅ 5/5 342.98 0.02 n/a
input_length_2048-num_aie_columns_1-num_channels_2-tile_size_1024-weighted_False ✅ 5/5 257.72 0.03 n/a
input_length_2048-num_aie_columns_1-num_channels_2-tile_size_2048-scalar_factor_3.0 ✅ 5/5 344.86 0.05 n/a
input_length_2048-num_aie_columns_1-tile_size_2048 ✅ 15/15 282.31 0.04 n/a
input_length_2048-num_aie_columns_2-num_channels_1-tile_size_1024 ✅ 25/25 301.24 0.03 n/a
input_length_2048-num_aie_columns_2-num_channels_1-tile_size_1024-group_size_32 ✅ 5/5 299.10 0.02 n/a
input_length_2048-num_aie_columns_2-num_channels_1-tile_size_1024-weighted_False ✅ 5/5 276.54 0.03 n/a
input_length_2048-num_aie_columns_2-num_channels_1-tile_size_1024-weighted_True ✅ 5/5 312.78 0.04 n/a
input_length_2048-num_aie_columns_2-num_channels_2-tile_size_1024-scalar_factor_3.0 ✅ 5/5 308.38 0.04 n/a
input_length_2048-num_aie_columns_2-num_channels_2-tile_size_512 ✅ 10/10 271.34 0.03 n/a
input_length_2048-num_aie_columns_2-num_channels_2-tile_size_512-group_size_32 ✅ 5/5 296.08 0.02 n/a
input_length_2048-num_aie_columns_2-num_channels_2-tile_size_512-weighted_False ✅ 5/5 279.30 0.03 n/a
input_length_2048-num_aie_columns_2-tile_size_1024 ✅ 15/15 277.69 0.04 n/a
input_length_2048-num_aie_columns_4-num_channels_1-tile_size_512 ✅ 25/25 288.26 0.03 n/a
input_length_2048-num_aie_columns_4-num_channels_1-tile_size_512-group_size_32 ✅ 5/5 295.42 0.02 n/a
input_length_2048-num_aie_columns_4-num_channels_1-tile_size_512-weighted_False ✅ 5/5 264.24 0.03 n/a
input_length_2048-num_aie_columns_4-num_channels_1-tile_size_512-weighted_True ✅ 5/5 275.34 0.03 n/a
input_length_2048-num_aie_columns_4-num_channels_2-tile_size_256 ✅ 10/10 371.89 0.03 n/a
input_length_2048-num_aie_columns_4-num_channels_2-tile_size_256-group_size_32 ✅ 5/5 289.24 0.02 n/a
input_length_2048-num_aie_columns_4-num_channels_2-tile_size_256-weighted_False ✅ 5/5 306.42 0.03 n/a
input_length_2048-num_aie_columns_4-num_channels_2-tile_size_512-scalar_factor_3.0 ✅ 5/5 303.22 0.04 n/a
input_length_2048-num_aie_columns_4-tile_size_512 ✅ 15/15 341.81 0.04 n/a
input_length_2048-num_aie_columns_8-num_channels_1-tile_size_256 ✅ 25/25 315.54 0.03 n/a
input_length_2048-num_aie_columns_8-num_channels_1-tile_size_256-group_size_32 ✅ 5/5 301.72 0.02 n/a
input_length_2048-num_aie_columns_8-num_channels_1-tile_size_256-weighted_False ✅ 5/5 251.34 0.03 n/a
input_length_2048-num_aie_columns_8-num_channels_1-tile_size_256-weighted_True ✅ 5/5 364.80 0.03 n/a
input_length_2048-num_aie_columns_8-num_channels_2-tile_size_128 ✅ 10/10 312.01 0.03 n/a
input_length_2048-num_aie_columns_8-num_channels_2-tile_size_128-group_size_32 ✅ 5/5 382.44 0.01 n/a
input_length_2048-num_aie_columns_8-num_channels_2-tile_size_128-weighted_False ✅ 5/5 356.86 0.03 n/a
input_length_2048-num_aie_columns_8-num_channels_2-tile_size_256-scalar_factor_3.0 ✅ 5/5 354.26 0.04 n/a
input_length_2048-num_aie_columns_8-tile_size_256 ✅ 15/15 323.19 0.04 n/a
input_length_2048-num_cores_1-num_channels_1-bypass_False-tile_size_2048 ✅ 5/5 292.80 0.03 n/a
input_length_2048-num_cores_16-num_channels_2-bypass_False-tile_size_128 ✅ 5/5 338.14 0.03 n/a
input_length_2048-num_cores_2-num_channels_1-bypass_False-tile_size_1024 ✅ 5/5 326.46 0.03 n/a
input_length_2048-num_cores_2-num_channels_2-bypass_False-tile_size_1024 ✅ 5/5 289.62 0.03 n/a
input_length_2048-num_cores_4-num_channels_1-bypass_False-tile_size_512 ✅ 5/5 273.08 0.03 n/a
input_length_2048-num_cores_4-num_channels_2-bypass_False-tile_size_512 ✅ 5/5 302.62 0.03 n/a
input_length_2048-num_cores_8-num_channels_1-bypass_False-tile_size_256 ✅ 5/5 326.04 0.03 n/a
input_length_2048-num_cores_8-num_channels_2-bypass_False-tile_size_256 ✅ 5/5 332.06 0.03 n/a
input_length_32768-num_aie_columns_2-num_channels_2-tile_size_1024 ✅ 5/5 314.74 0.43 n/a
input_length_32768-num_aie_columns_2-num_channels_2-tile_size_512 ✅ 5/5 324.96 0.42 n/a
input_length_32768-num_aie_columns_4-num_channels_4-tile_size_2048 ✅ 5/5 391.14 0.38 n/a
rows_32-cols_512-angle_rows_32-aie_columns_1-method_type_0 ✅ 5/5 288.74 0.35 n/a
rows_32-cols_512-angle_rows_32-aie_columns_2-method_type_0 ✅ 5/5 348.02 0.30 n/a
rows_32-cols_512-angle_rows_32-aie_columns_8-method_type_0 ✅ 5/5 415.94 0.28 n/a
rows_32-cols_512-angle_rows_8-aie_columns_1-method_type_0 ✅ 5/5 306.66 0.24 n/a
rows_32-cols_512-angle_rows_8-aie_columns_2-method_type_0 ✅ 5/5 315.36 0.24 n/a
rows_32-cols_512-angle_rows_8-aie_columns_8-method_type_0 ✅ 5/5 504.40 0.18 n/a
seq_len_16384-dim_64-num_heads_1-num_pipelines_8-num_kv_heads_0 ❌ 0/5 n/a n/a n/a
seq_len_256-embedding_dim_2048-hidden_dim_2048-prio_accuracy_False ✅ 5/5 n/a n/a n/a
📈 Trends (vs main branch) for Small Benchmark/Test Suite

8c138a8 (2026_03_26_15_00_25)

IRONCLAD Trends

M_128-K_128-num_aie_columns_1-tile_size_input_32-tile_size_output_128

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
8c138a8 — 2026-03-26 14:54:200.13 (n/a)0.11 (n/a)0.13 (n/a)0.08 (n/a)0.02 (n/a)0.13 (n/a)0.11 (n/a)0.12 (n/a)0.08 (n/a)0.02 (n/a)

M_1792-K_896-N_1152-num_aie_columns_8-b_col_maj_False-c_col_maj_True-m_64-k_32-n_48-trace_size_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
8c138a8 — 2026-03-26 14:54:204.61 (n/a)4.18 (n/a)4.52 (n/a)3.13 (n/a)0.63 (n/a)3003.40 (n/a)2297.86 (n/a)2080.90 (n/a)2041.50 (n/a)410.35 (n/a)1812.12 (n/a)1645.15 (n/a)1777.82 (n/a)1231.74 (n/a)247.36 (n/a)

M_192-K_384-N_64-num_aie_columns_4-b_col_maj_False-c_col_maj_False-m_48-k_96-n_16-trace_size_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
8c138a8 — 2026-03-26 14:54:200.78 (n/a)0.59 (n/a)0.71 (n/a)0.22 (n/a)0.23 (n/a)1023.50 (n/a)473.66 (n/a)311.70 (n/a)281.90 (n/a)313.71 (n/a)33.47 (n/a)24.98 (n/a)30.28 (n/a)9.22 (n/a)9.92 (n/a)

M_192-K_384-N_64-num_aie_columns_4-b_col_maj_True-c_col_maj_True-m_48-k_96-n_16-trace_size_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
8c138a8 — 2026-03-26 14:54:200.82 (n/a)0.60 (n/a)0.79 (n/a)0.23 (n/a)0.28 (n/a)976.50 (n/a)475.78 (n/a)280.60 (n/a)269.40 (n/a)309.25 (n/a)35.03 (n/a)25.81 (n/a)33.63 (n/a)9.66 (n/a)11.95 (n/a)

M_2048-K_2048-N_2048-num_aie_columns_1-b_col_maj_False-c_col_maj_False-m_64-k_64-n_64-trace_size_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
8c138a8 — 2026-03-26 14:54:200.50 (n/a)0.50 (n/a)0.50 (n/a)0.50 (n/a)0.00 (n/a)50068.60 (n/a)50028.00 (n/a)50036.30 (n/a)49949.00 (n/a)46.88 (n/a)343.95 (n/a)343.41 (n/a)343.35 (n/a)343.13 (n/a)0.32 (n/a)

M_2048-K_2048-N_2048-num_aie_columns_2-b_col_maj_True-c_col_maj_False-m_64-k_64-n_64-trace_size_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
8c138a8 — 2026-03-26 14:54:200.87 (n/a)0.86 (n/a)0.86 (n/a)0.85 (n/a)0.01 (n/a)29482.60 (n/a)29226.50 (n/a)29246.10 (n/a)29040.10 (n/a)180.65 (n/a)591.59 (n/a)587.84 (n/a)587.42 (n/a)582.71 (n/a)3.63 (n/a)

M_2048-K_2048-N_2048-num_aie_columns_8-b_col_maj_True-c_col_maj_True-m_64-k_64-n_64-trace_size_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
8c138a8 — 2026-03-26 14:54:203.40 (n/a)3.23 (n/a)3.26 (n/a)3.05 (n/a)0.16 (n/a)8258.10 (n/a)7814.92 (n/a)7719.40 (n/a)7399.90 (n/a)393.09 (n/a)2321.65 (n/a)2202.77 (n/a)2225.54 (n/a)2080.37 (n/a)110.00 (n/a)

M_2048-K_8192-num_aie_columns_1-tile_size_input_1-tile_size_output_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
8c138a8 — 2026-03-26 14:54:2012.44 (n/a)11.85 (n/a)12.24 (n/a)11.08 (n/a)0.69 (n/a)12.44 (n/a)11.84 (n/a)12.23 (n/a)11.08 (n/a)0.69 (n/a)

M_2048-K_8192-num_aie_columns_2-tile_size_input_1-tile_size_output_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
8c138a8 — 2026-03-26 14:54:2022.23 (n/a)19.78 (n/a)21.50 (n/a)12.33 (n/a)4.18 (n/a)22.22 (n/a)19.77 (n/a)21.49 (n/a)12.32 (n/a)4.18 (n/a)

M_2048-K_8192-num_aie_columns_4-tile_size_input_1-tile_size_output_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
8c138a8 — 2026-03-26 14:54:2035.85 (n/a)30.11 (n/a)33.56 (n/a)13.40 (n/a)9.40 (n/a)35.83 (n/a)30.09 (n/a)33.54 (n/a)13.39 (n/a)9.40 (n/a)

M_2048-K_8192-num_aie_columns_8-tile_size_input_1-tile_size_output_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
8c138a8 — 2026-03-26 14:54:2035.28 (n/a)34.48 (n/a)34.56 (n/a)33.71 (n/a)0.58 (n/a)35.26 (n/a)34.46 (n/a)34.54 (n/a)33.69 (n/a)0.58 (n/a)

M_2048-N_64-aie_columns_1-channels_1-m_64-n_64-s_8

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8c138a8 — 2026-03-26 14:54:201.82 (n/a)1.42 (n/a)1.45 (n/a)0.92 (n/a)0.38 (n/a)569.90 (n/a)392.96 (n/a)360.60 (n/a)287.80 (n/a)117.84 (n/a)

M_2048-N_64-aie_columns_1-channels_2-m_64-n_64-s_8

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8c138a8 — 2026-03-26 14:54:201.80 (n/a)1.33 (n/a)1.27 (n/a)1.11 (n/a)0.27 (n/a)472.30 (n/a)404.08 (n/a)413.00 (n/a)292.00 (n/a)67.64 (n/a)

M_384-K_1536-N_1792-num_aie_columns_4-b_col_maj_True-c_col_maj_False-m_32-k_48-n_64-trace_size_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
8c138a8 — 2026-03-26 14:54:203.87 (n/a)3.62 (n/a)3.83 (n/a)3.24 (n/a)0.33 (n/a)2484.70 (n/a)2244.16 (n/a)2103.00 (n/a)2082.50 (n/a)210.76 (n/a)1015.10 (n/a)948.44 (n/a)1005.19 (n/a)850.77 (n/a)86.14 (n/a)

M_8192-K_2048-num_aie_columns_1-tile_size_input_4-tile_size_output_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
8c138a8 — 2026-03-26 14:54:2012.64 (n/a)12.23 (n/a)12.43 (n/a)11.21 (n/a)0.58 (n/a)12.63 (n/a)12.22 (n/a)12.43 (n/a)11.21 (n/a)0.58 (n/a)

M_8192-K_2048-num_aie_columns_2-tile_size_input_4-tile_size_output_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
8c138a8 — 2026-03-26 14:54:2022.17 (n/a)19.45 (n/a)21.23 (n/a)11.34 (n/a)4.55 (n/a)22.16 (n/a)19.43 (n/a)21.22 (n/a)11.33 (n/a)4.55 (n/a)

M_8192-K_2048-num_aie_columns_4-tile_size_input_4-tile_size_output_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
8c138a8 — 2026-03-26 14:54:2033.63 (n/a)31.28 (n/a)32.58 (n/a)24.62 (n/a)3.76 (n/a)33.61 (n/a)31.26 (n/a)32.56 (n/a)24.60 (n/a)3.76 (n/a)

M_8192-K_2048-num_aie_columns_8-tile_size_input_4-tile_size_output_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
8c138a8 — 2026-03-26 14:54:2036.32 (n/a)33.74 (n/a)35.36 (n/a)26.13 (n/a)4.29 (n/a)36.30 (n/a)33.72 (n/a)35.34 (n/a)26.11 (n/a)4.28 (n/a)

M_896-K_1792-N_640-num_aie_columns_8-b_col_maj_False-c_col_maj_True-m_32-k_64-n_80-trace_size_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
8c138a8 — 2026-03-26 14:54:205.62 (n/a)4.83 (n/a)5.36 (n/a)2.44 (n/a)1.35 (n/a)2725.00 (n/a)1527.46 (n/a)1240.00 (n/a)1183.80 (n/a)671.22 (n/a)1736.13 (n/a)1491.73 (n/a)1657.38 (n/a)754.19 (n/a)417.42 (n/a)

axpy_1_cols_2_channels_2048_tile_2048_3.0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.08 (+17.48%)0.07 (+11.67%)0.07 (+26.71%)0.05 (+0.80%)0.01 (+108.36%)239.80 (-0.83%)190.62 (-7.91%)166.90 (-21.09%)155.40 (-14.90%)42.89 (+79.10%)
dc66da7 — 2026-03-25 19:23:450.07 (n/a)0.06 (n/a)0.06 (n/a)0.05 (n/a)0.01 (n/a)241.80 (n/a)207.00 (n/a)211.50 (n/a)182.60 (n/a)23.95 (n/a)

axpy_1_cols_2_channels_2048_tile_2048_3.0_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.16 (-3.31%)0.14 (-1.29%)0.15 (+16.18%)0.10 (-6.80%)0.03 (+3.89%)125.70 (+7.25%)93.56 (+2.12%)83.50 (-13.92%)75.00 (+3.45%)22.38 (+18.09%)
84d3478 — 2026-02-17 23:25:130.17 (n/a)0.14 (n/a)0.13 (n/a)0.10 (n/a)0.03 (n/a)117.20 (n/a)91.62 (n/a)97.00 (n/a)72.50 (n/a)18.95 (n/a)

axpy_2_cols_2_channels_2048_tile_1024_3.0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.09 (+14.45%)0.07 (-0.78%)0.07 (-0.40%)0.05 (-7.87%)0.01 (+68.50%)233.40 (+8.51%)188.20 (+2.47%)179.80 (+0.39%)143.50 (-12.66%)34.06 (+61.09%)
dc66da7 — 2026-03-25 19:23:450.07 (n/a)0.07 (n/a)0.07 (n/a)0.06 (n/a)0.01 (n/a)215.10 (n/a)183.66 (n/a)179.10 (n/a)164.30 (n/a)21.14 (n/a)

axpy_2_cols_2_channels_2048_tile_1024_3.0_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.13 (-0.73%)0.11 (-7.43%)0.11 (-8.63%)0.07 (-27.71%)0.02 (+50.39%)182.70 (+38.30%)120.86 (+11.97%)110.80 (+9.38%)96.10 (+0.73%)35.16 (+122.88%)
84d3478 — 2026-02-17 23:25:130.13 (n/a)0.12 (n/a)0.12 (n/a)0.09 (n/a)0.02 (n/a)132.10 (n/a)107.94 (n/a)101.30 (n/a)95.40 (n/a)15.78 (n/a)

axpy_4_cols_2_channels_2048_tile_512_3.0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.10 (+6.67%)0.06 (-1.56%)0.06 (-2.24%)0.05 (+3.60%)0.02 (+20.83%)251.50 (-3.49%)200.78 (+3.11%)191.70 (+2.29%)128.70 (-6.26%)50.37 (+11.30%)
dc66da7 — 2026-03-25 19:23:450.09 (n/a)0.07 (n/a)0.07 (n/a)0.05 (n/a)0.02 (n/a)260.60 (n/a)194.72 (n/a)187.40 (n/a)137.30 (n/a)45.26 (n/a)

axpy_4_cols_2_channels_2048_tile_512_3.0_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.13 (-10.91%)0.11 (-7.68%)0.12 (-0.73%)0.06 (-17.94%)0.03 (+0.99%)194.10 (+21.85%)124.52 (+10.29%)103.30 (+0.68%)97.20 (+12.24%)40.32 (+39.15%)
84d3478 — 2026-02-17 23:25:130.14 (n/a)0.11 (n/a)0.12 (n/a)0.08 (n/a)0.03 (n/a)159.30 (n/a)112.90 (n/a)102.60 (n/a)86.60 (n/a)28.97 (n/a)

axpy_8_cols_2_channels_2048_tile_256_3.0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.08 (+1.42%)0.06 (+9.86%)0.06 (+18.71%)0.03 (+13.51%)0.02 (-10.62%)376.00 (-11.90%)225.02 (-11.61%)196.20 (-15.76%)151.40 (-1.37%)87.15 (-17.99%)
dc66da7 — 2026-03-25 19:23:450.08 (n/a)0.05 (n/a)0.05 (n/a)0.03 (n/a)0.02 (n/a)426.80 (n/a)254.58 (n/a)232.90 (n/a)153.50 (n/a)106.27 (n/a)

axpy_8_cols_2_channels_2048_tile_256_3.0_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.13 (+0.01%)0.11 (+6.10%)0.11 (+0.22%)0.08 (-0.33%)0.02 (-9.83%)156.30 (+0.32%)115.26 (-6.40%)112.70 (-0.18%)95.50 (+0.00%)24.34 (-10.22%)
84d3478 — 2026-02-17 23:25:130.13 (n/a)0.10 (n/a)0.11 (n/a)0.08 (n/a)0.02 (n/a)155.80 (n/a)123.14 (n/a)112.90 (n/a)95.50 (n/a)27.11 (n/a)

dequant_1_cols_1_channels_2048_tile_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.04 (+17.85%)0.03 (+2.55%)0.03 (+1.65%)0.03 (-2.18%)0.00 (+114.37%)207.70 (+2.26%)178.64 (-0.96%)173.60 (-1.59%)141.60 (-15.16%)28.01 (+90.63%)
dc66da7 — 2026-03-25 19:23:450.03 (n/a)0.03 (n/a)0.03 (n/a)0.03 (n/a)0.00 (n/a)203.10 (n/a)180.38 (n/a)176.40 (n/a)166.90 (n/a)14.69 (n/a)

dequant_1_cols_1_channels_2048_tile_2048_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.05 (-13.00%)0.04 (-0.87%)0.04 (+1.16%)0.03 (-3.17%)0.01 (-31.33%)158.40 (+3.26%)128.12 (-0.19%)123.80 (-1.12%)110.40 (+14.88%)18.36 (-15.65%)
84d3478 — 2026-02-17 23:25:130.05 (n/a)0.04 (n/a)0.04 (n/a)0.03 (n/a)0.01 (n/a)153.40 (n/a)128.36 (n/a)125.20 (n/a)96.10 (n/a)21.77 (n/a)

dequant_1_cols_2_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.04 (+13.44%)0.03 (+8.77%)0.03 (+14.89%)0.02 (-9.15%)0.01 (+54.39%)252.70 (+10.06%)182.14 (-6.08%)174.50 (-12.97%)145.40 (-11.88%)42.38 (+56.31%)
dc66da7 — 2026-03-25 19:23:450.03 (n/a)0.03 (n/a)0.03 (n/a)0.02 (n/a)0.00 (n/a)229.60 (n/a)193.94 (n/a)200.50 (n/a)165.00 (n/a)27.11 (n/a)

dequant_1_cols_2_channels_2048_tile_1024_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.06 (-3.74%)0.05 (-12.63%)0.05 (-5.21%)0.03 (-33.47%)0.01 (+52.60%)162.70 (+50.37%)110.84 (+18.98%)98.80 (+5.44%)81.50 (+3.95%)31.02 (+149.23%)
84d3478 — 2026-02-17 23:25:130.07 (n/a)0.06 (n/a)0.06 (n/a)0.05 (n/a)0.01 (n/a)108.20 (n/a)93.16 (n/a)93.70 (n/a)78.40 (n/a)12.45 (n/a)

dequant_2_cols_1_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.05 (+21.89%)0.03 (+20.89%)0.03 (+21.41%)0.02 (+8.07%)0.01 (+32.97%)215.80 (-7.46%)168.56 (-16.36%)172.60 (-17.65%)113.20 (-17.97%)37.61 (+1.89%)
dc66da7 — 2026-03-25 19:23:450.04 (n/a)0.03 (n/a)0.03 (n/a)0.02 (n/a)0.01 (n/a)233.20 (n/a)201.52 (n/a)209.60 (n/a)138.00 (n/a)36.92 (n/a)

dequant_2_cols_1_channels_2048_tile_1024_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.06 (+8.90%)0.05 (+1.31%)0.05 (-1.85%)0.03 (+6.35%)0.01 (+7.11%)152.30 (-5.93%)115.08 (-1.20%)100.80 (+1.92%)88.10 (-8.23%)27.20 (-4.42%)
84d3478 — 2026-02-17 23:25:130.05 (n/a)0.05 (n/a)0.05 (n/a)0.03 (n/a)0.01 (n/a)161.90 (n/a)116.48 (n/a)98.90 (n/a)96.00 (n/a)28.46 (n/a)

dequant_2_cols_2_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.05 (+49.89%)0.03 (+24.89%)0.03 (+13.74%)0.02 (+15.73%)0.01 (+106.91%)219.80 (-13.60%)173.04 (-17.40%)185.70 (-12.07%)112.30 (-33.27%)41.15 (+16.74%)
dc66da7 — 2026-03-25 19:23:450.03 (n/a)0.03 (n/a)0.02 (n/a)0.02 (n/a)0.00 (n/a)254.40 (n/a)209.50 (n/a)211.20 (n/a)168.30 (n/a)35.25 (n/a)

dequant_2_cols_2_channels_2048_tile_512_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.07 (+27.96%)0.05 (+1.57%)0.04 (-18.06%)0.03 (+10.23%)0.01 (+22.79%)164.70 (-9.31%)119.12 (-1.39%)120.20 (+22.03%)75.60 (-21.82%)31.61 (-13.80%)
84d3478 — 2026-02-17 23:25:130.05 (n/a)0.05 (n/a)0.05 (n/a)0.03 (n/a)0.01 (n/a)181.60 (n/a)120.80 (n/a)98.50 (n/a)96.70 (n/a)36.67 (n/a)

dequant_4_cols_1_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.04 (+56.54%)0.03 (+15.78%)0.03 (+9.01%)0.02 (-5.58%)0.01 (+304.26%)243.40 (+5.92%)186.06 (-8.98%)182.60 (-8.29%)118.50 (-36.12%)46.36 (+164.07%)
dc66da7 — 2026-03-25 19:23:450.03 (n/a)0.03 (n/a)0.03 (n/a)0.02 (n/a)0.00 (n/a)229.80 (n/a)204.42 (n/a)199.10 (n/a)185.50 (n/a)17.56 (n/a)

dequant_4_cols_1_channels_2048_tile_512_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.05 (-2.55%)0.04 (-12.79%)0.05 (-16.25%)0.03 (-12.85%)0.01 (-1.20%)164.60 (+14.78%)122.14 (+15.14%)115.90 (+19.48%)97.50 (+2.63%)25.08 (+19.84%)
84d3478 — 2026-02-17 23:25:130.06 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.01 (n/a)143.40 (n/a)106.08 (n/a)97.00 (n/a)95.00 (n/a)20.93 (n/a)

dequant_4_cols_2_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.04 (+28.72%)0.03 (+2.89%)0.02 (-6.84%)0.02 (+6.27%)0.01 (+95.59%)243.20 (-5.88%)208.42 (-0.42%)222.90 (+7.32%)141.10 (-22.34%)41.08 (+38.16%)
dc66da7 — 2026-03-25 19:23:450.03 (n/a)0.03 (n/a)0.03 (n/a)0.02 (n/a)0.00 (n/a)258.40 (n/a)209.30 (n/a)207.70 (n/a)181.70 (n/a)29.74 (n/a)

dequant_4_cols_2_channels_2048_tile_256_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.06 (+20.83%)0.05 (+8.21%)0.04 (+1.42%)0.03 (+17.39%)0.01 (+14.84%)166.40 (-14.80%)122.34 (-7.97%)127.70 (-1.39%)84.60 (-17.22%)30.47 (-19.21%)
84d3478 — 2026-02-17 23:25:130.05 (n/a)0.04 (n/a)0.04 (n/a)0.03 (n/a)0.01 (n/a)195.30 (n/a)132.94 (n/a)129.50 (n/a)102.20 (n/a)37.71 (n/a)

dequant_8_cols_1_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.04 (-1.52%)0.03 (+2.17%)0.03 (+18.60%)0.02 (-11.90%)0.01 (+4.42%)268.50 (+13.53%)205.32 (-1.14%)191.00 (-15.67%)141.60 (+1.51%)49.39 (+23.46%)
dc66da7 — 2026-03-25 19:23:450.04 (n/a)0.03 (n/a)0.02 (n/a)0.02 (n/a)0.01 (n/a)236.50 (n/a)207.68 (n/a)226.50 (n/a)139.50 (n/a)40.00 (n/a)

dequant_8_cols_1_channels_2048_tile_256_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.05 (-25.19%)0.04 (-21.02%)0.04 (-19.70%)0.04 (-19.76%)0.00 (-37.92%)142.70 (+24.63%)126.44 (+26.24%)125.70 (+24.46%)115.50 (+33.68%)10.61 (+3.97%)
84d3478 — 2026-02-17 23:25:130.06 (n/a)0.05 (n/a)0.05 (n/a)0.05 (n/a)0.01 (n/a)114.50 (n/a)100.16 (n/a)101.00 (n/a)86.40 (n/a)10.20 (n/a)

dequant_8_cols_2_channels_2048_tile_128

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.02 (+5.64%)0.02 (+1.26%)0.02 (-1.25%)0.02 (-3.87%)0.00 (+124.04%)258.90 (+4.02%)239.08 (-0.87%)247.20 (+1.27%)215.50 (-5.36%)18.28 (+120.74%)
dc66da7 — 2026-03-25 19:23:450.02 (n/a)0.02 (n/a)0.02 (n/a)0.02 (n/a)0.00 (n/a)248.90 (n/a)241.18 (n/a)244.10 (n/a)227.70 (n/a)8.28 (n/a)

dequant_8_cols_2_channels_2048_tile_128_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.04 (-26.69%)0.04 (-11.93%)0.04 (-16.99%)0.03 (+27.28%)0.01 (-62.60%)193.10 (-21.41%)147.68 (+2.96%)136.10 (+20.44%)135.40 (+36.49%)25.41 (-58.96%)
84d3478 — 2026-02-17 23:25:130.05 (n/a)0.04 (n/a)0.05 (n/a)0.02 (n/a)0.01 (n/a)245.70 (n/a)143.44 (n/a)113.00 (n/a)99.20 (n/a)61.90 (n/a)

eltwise_add_1_cols_2_channels_2048_tile_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.07 (-32.66%)0.06 (-13.87%)0.06 (-9.19%)0.04 (-8.04%)0.01 (-49.50%)299.50 (+8.75%)223.84 (+11.90%)212.70 (+10.15%)187.60 (+48.42%)45.70 (-16.71%)
dc66da7 — 2026-03-25 19:23:450.10 (n/a)0.07 (n/a)0.06 (n/a)0.04 (n/a)0.02 (n/a)275.40 (n/a)200.04 (n/a)193.10 (n/a)126.40 (n/a)54.87 (n/a)

eltwise_add_2_cols_2_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.07 (-33.33%)0.06 (-18.97%)0.05 (-19.23%)0.05 (+18.13%)0.01 (-59.22%)270.80 (-15.37%)223.26 (+13.77%)229.40 (+23.80%)174.50 (+50.04%)37.51 (-50.17%)
dc66da7 — 2026-03-25 19:23:450.11 (n/a)0.07 (n/a)0.07 (n/a)0.04 (n/a)0.02 (n/a)320.00 (n/a)196.24 (n/a)185.30 (n/a)116.30 (n/a)75.29 (n/a)

eltwise_add_4_cols_2_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.07 (-8.97%)0.06 (-6.00%)0.06 (-2.02%)0.05 (-2.74%)0.01 (-14.89%)243.40 (+2.83%)208.60 (+6.01%)200.50 (+2.04%)178.10 (+9.87%)31.58 (-2.07%)
dc66da7 — 2026-03-25 19:23:450.08 (n/a)0.06 (n/a)0.06 (n/a)0.05 (n/a)0.01 (n/a)236.70 (n/a)196.78 (n/a)196.50 (n/a)162.10 (n/a)32.25 (n/a)

eltwise_add_8_cols_2_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.06 (-45.96%)0.05 (-24.77%)0.05 (-15.00%)0.05 (+4.78%)0.00 (-83.62%)259.60 (-4.56%)234.66 (+20.97%)224.10 (+17.64%)218.40 (+85.08%)18.77 (-71.47%)
dc66da7 — 2026-03-25 19:23:450.10 (n/a)0.07 (n/a)0.06 (n/a)0.05 (n/a)0.02 (n/a)272.00 (n/a)193.98 (n/a)190.50 (n/a)118.00 (n/a)65.77 (n/a)

eltwise_mul_1_cols_2_channels_2048_tile_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.11 (+3.13%)0.07 (+2.52%)0.06 (-1.87%)0.05 (-14.01%)0.02 (+21.62%)257.00 (+16.29%)184.10 (+0.67%)190.70 (+1.87%)114.40 (-3.05%)55.57 (+40.96%)
dc66da7 — 2026-03-25 19:23:450.10 (n/a)0.07 (n/a)0.07 (n/a)0.06 (n/a)0.02 (n/a)221.00 (n/a)182.88 (n/a)187.20 (n/a)118.00 (n/a)39.42 (n/a)

eltwise_mul_2_cols_2_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.09 (+2.43%)0.07 (+6.63%)0.08 (+30.25%)0.05 (-10.38%)0.02 (+9.28%)264.20 (+11.62%)182.10 (-4.96%)162.40 (-23.22%)137.30 (-2.42%)52.80 (+20.53%)
dc66da7 — 2026-03-25 19:23:450.09 (n/a)0.07 (n/a)0.06 (n/a)0.05 (n/a)0.02 (n/a)236.70 (n/a)191.60 (n/a)211.50 (n/a)140.70 (n/a)43.81 (n/a)

eltwise_mul_4_cols_2_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.10 (+22.79%)0.08 (+18.44%)0.09 (+34.73%)0.06 (-0.12%)0.01 (+158.99%)191.20 (+0.10%)154.44 (-13.56%)136.60 (-25.80%)126.60 (-18.53%)29.95 (+117.72%)
dc66da7 — 2026-03-25 19:23:450.08 (n/a)0.07 (n/a)0.07 (n/a)0.06 (n/a)0.01 (n/a)191.00 (n/a)178.66 (n/a)184.10 (n/a)155.40 (n/a)13.76 (n/a)

eltwise_mul_8_cols_2_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.08 (+34.81%)0.06 (+14.94%)0.06 (+6.87%)0.04 (-5.87%)0.01 (+158.26%)280.10 (+6.22%)210.00 (-9.80%)215.10 (-6.40%)150.70 (-25.80%)49.49 (+100.84%)
dc66da7 — 2026-03-25 19:23:450.06 (n/a)0.05 (n/a)0.05 (n/a)0.05 (n/a)0.01 (n/a)263.70 (n/a)232.82 (n/a)229.80 (n/a)203.10 (n/a)24.64 (n/a)

embedding_dim_2048-hidden_dim_2048

No metrics available.

gelu_1_cols_1_channels_2048_tile_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.05 (+9.18%)0.04 (-1.03%)0.04 (-2.51%)0.04 (-8.09%)0.01 (+114.58%)221.00 (+8.81%)195.42 (+1.97%)199.10 (+2.58%)163.60 (-8.40%)23.03 (+114.14%)
dc66da7 — 2026-03-25 19:23:450.05 (n/a)0.04 (n/a)0.04 (n/a)0.04 (n/a)0.00 (n/a)203.10 (n/a)191.64 (n/a)194.10 (n/a)178.60 (n/a)10.76 (n/a)

gelu_1_cols_2_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.04 (-26.60%)0.04 (-19.13%)0.04 (-18.64%)0.04 (-2.84%)0.00 (-66.68%)226.70 (+2.91%)199.16 (+20.50%)193.10 (+22.92%)186.40 (+36.16%)16.29 (-52.49%)
dc66da7 — 2026-03-25 19:23:450.06 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.01 (n/a)220.30 (n/a)165.28 (n/a)157.10 (n/a)136.90 (n/a)34.28 (n/a)

gelu_2_cols_1_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.06 (+2.08%)0.05 (+4.66%)0.04 (-1.75%)0.03 (-5.55%)0.01 (+12.71%)283.70 (+5.86%)191.18 (-3.05%)191.00 (+1.81%)137.60 (-1.99%)58.08 (+15.01%)
dc66da7 — 2026-03-25 19:23:450.06 (n/a)0.04 (n/a)0.04 (n/a)0.03 (n/a)0.01 (n/a)268.00 (n/a)197.20 (n/a)187.60 (n/a)140.40 (n/a)50.50 (n/a)

gelu_2_cols_2_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.07 (+2.20%)0.04 (-2.29%)0.04 (+1.04%)0.04 (+16.91%)0.01 (-4.76%)227.00 (-14.47%)194.56 (+0.69%)206.00 (-1.01%)119.40 (-2.21%)44.27 (-20.37%)
dc66da7 — 2026-03-25 19:23:450.07 (n/a)0.05 (n/a)0.04 (n/a)0.03 (n/a)0.01 (n/a)265.40 (n/a)193.22 (n/a)208.10 (n/a)122.10 (n/a)55.59 (n/a)

gelu_4_cols_1_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.07 (-9.75%)0.04 (-17.39%)0.04 (-20.48%)0.03 (+12.65%)0.01 (-29.41%)239.40 (-11.23%)196.08 (+15.06%)197.90 (+25.73%)124.70 (+10.84%)44.40 (-31.09%)
dc66da7 — 2026-03-25 19:23:450.07 (n/a)0.05 (n/a)0.05 (n/a)0.03 (n/a)0.02 (n/a)269.70 (n/a)170.42 (n/a)157.40 (n/a)112.50 (n/a)64.43 (n/a)

gelu_4_cols_2_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.04 (-6.29%)0.04 (+1.81%)0.04 (+0.94%)0.03 (-0.42%)0.00 (-22.18%)284.90 (+0.42%)233.36 (-2.39%)227.20 (-0.96%)204.80 (+6.72%)30.62 (-15.21%)
dc66da7 — 2026-03-25 19:23:450.04 (n/a)0.03 (n/a)0.04 (n/a)0.03 (n/a)0.01 (n/a)283.70 (n/a)239.08 (n/a)229.40 (n/a)191.90 (n/a)36.12 (n/a)

gelu_8_cols_1_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.05 (-18.24%)0.04 (-13.70%)0.04 (-4.91%)0.03 (-15.26%)0.01 (-28.46%)285.90 (+18.04%)223.38 (+14.20%)223.20 (+5.18%)156.90 (+22.29%)47.35 (+1.82%)
dc66da7 — 2026-03-25 19:23:450.06 (n/a)0.04 (n/a)0.04 (n/a)0.03 (n/a)0.01 (n/a)242.20 (n/a)195.60 (n/a)212.20 (n/a)128.30 (n/a)46.50 (n/a)

gelu_8_cols_2_channels_2048_tile_128

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.05 (+14.38%)0.04 (+10.49%)0.03 (+3.25%)0.03 (+13.03%)0.01 (+26.31%)305.70 (-11.52%)235.82 (-8.98%)241.80 (-3.12%)177.40 (-12.57%)49.18 (-6.09%)
dc66da7 — 2026-03-25 19:23:450.04 (n/a)0.03 (n/a)0.03 (n/a)0.02 (n/a)0.01 (n/a)345.50 (n/a)259.10 (n/a)249.60 (n/a)202.90 (n/a)52.37 (n/a)

gemm_1792x896x1152_64x32x48_8cols_ccolmaj

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
8b6d0f4 — 2026-03-25 20:46:394.89 (+9.71%)4.07 (+2.20%)4.10 (+2.03%)3.40 (+0.07%)0.55 (+44.15%)2767.30 (-0.07%)2344.32 (-1.52%)2291.20 (-1.99%)1922.40 (-8.85%)308.92 (+28.54%)1924.38 (+9.71%)1600.58 (+2.20%)1614.60 (+2.03%)1336.81 (+0.07%)215.84 (+44.15%)
dc66da7 — 2026-03-25 19:23:454.46 (n/a)3.98 (n/a)4.02 (n/a)3.40 (n/a)0.38 (n/a)2769.20 (n/a)2380.42 (n/a)2337.70 (n/a)2109.10 (n/a)240.33 (n/a)1754.05 (n/a)1566.13 (n/a)1582.49 (n/a)1335.88 (n/a)149.73 (n/a)

gemm_192x384x64_48x96x16_4cols

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
8b6d0f4 — 2026-03-25 20:46:391.35 (+17.71%)0.77 (-13.89%)0.65 (-34.22%)0.56 (+7.33%)0.32 (+33.14%)393.60 (-6.84%)315.66 (+18.48%)341.40 (+52.00%)164.30 (-15.05%)87.82 (-5.13%)57.45 (+17.71%)32.93 (-13.89%)27.64 (-34.22%)23.97 (+7.33%)13.81 (+33.14%)
dc66da7 — 2026-03-25 19:23:451.14 (n/a)0.90 (n/a)0.98 (n/a)0.52 (n/a)0.24 (n/a)422.50 (n/a)266.42 (n/a)224.60 (n/a)193.40 (n/a)92.57 (n/a)48.81 (n/a)38.25 (n/a)42.02 (n/a)22.34 (n/a)10.37 (n/a)

gemm_192x384x64_48x96x16_4cols_bcolmaj_ccolmaj

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
8b6d0f4 — 2026-03-25 20:46:391.18 (+31.18%)0.79 (-0.76%)0.72 (-17.59%)0.64 (+26.96%)0.22 (+35.66%)343.10 (-21.24%)294.50 (+0.95%)309.00 (+21.37%)188.10 (-23.78%)62.26 (-23.01%)50.17 (+31.18%)33.62 (-0.76%)30.54 (-17.59%)27.51 (+26.96%)9.40 (+35.66%)
dc66da7 — 2026-03-25 19:23:450.90 (n/a)0.79 (n/a)0.87 (n/a)0.51 (n/a)0.16 (n/a)435.60 (n/a)291.72 (n/a)254.60 (n/a)246.80 (n/a)80.87 (n/a)38.24 (n/a)33.88 (n/a)37.06 (n/a)21.67 (n/a)6.93 (n/a)

gemm_2048x2048x2048_64x64x32_8_cols_0_bcolmaj_0_ccolmaj_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)GFLOP/s (max)GFLOP/s (mean)GFLOP/s (median)GFLOP/s (min)GFLOP/s (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
3a053d3 — 2025-11-26 01:12:163.72 (-0.11%)3.55 (-2.75%)3.56 (-2.75%)3.36 (-5.33%)0.12 (+76.10%)2725.23 (-0.11%)2602.52 (-2.75%)2612.11 (-2.75%)2459.89 (-5.33%)86.27 (+76.10%)6984.00 (+5.63%)6608.60 (+2.91%)6577.00 (+2.83%)6304.00 (+0.11%)221.87 (+87.15%)
9e98049 — 2025-11-25 20:50:553.72 (n/a)3.65 (n/a)3.66 (n/a)3.54 (n/a)0.07 (n/a)2728.26 (n/a)2676.06 (n/a)2686.03 (n/a)2598.29 (n/a)48.99 (n/a)6612.00 (n/a)6422.00 (n/a)6396.00 (n/a)6297.00 (n/a)118.55 (n/a)

gemm_2048x2048x2048_64x64x32_8_cols_0_bcolmaj_1_ccolmaj_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)GFLOP/s (max)GFLOP/s (mean)GFLOP/s (median)GFLOP/s (min)GFLOP/s (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
3a053d3 — 2025-11-26 01:12:163.40 (-3.14%)3.38 (-0.82%)3.38 (-0.82%)3.35 (+4.93%)0.02 (-84.51%)2493.81 (-3.14%)2475.27 (-0.82%)2479.06 (-0.82%)2453.92 (+4.93%)13.33 (-84.51%)7001.00 (-4.70%)6940.80 (+0.71%)6930.00 (+0.83%)6889.00 (+3.24%)37.43 (-84.76%)
9e98049 — 2025-11-25 20:50:553.51 (n/a)3.40 (n/a)3.41 (n/a)3.19 (n/a)0.12 (n/a)2574.53 (n/a)2495.72 (n/a)2499.62 (n/a)2338.67 (n/a)86.02 (n/a)7346.00 (n/a)6892.20 (n/a)6873.00 (n/a)6673.00 (n/a)245.57 (n/a)

gemm_2048x2048x2048_64x64x32_8_cols_1_bcolmaj_0_ccolmaj_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)GFLOP/s (max)GFLOP/s (mean)GFLOP/s (median)GFLOP/s (min)GFLOP/s (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
3a053d3 — 2025-11-26 01:12:163.02 (+0.52%)2.87 (-2.05%)2.86 (-2.24%)2.70 (-6.16%)0.11 (+131.66%)2216.19 (+0.52%)2104.44 (-2.05%)2094.08 (-2.24%)1980.84 (-6.16%)80.25 (+131.67%)8673.00 (+6.56%)8175.60 (+2.21%)8204.00 (+2.29%)7752.00 (-0.51%)314.03 (+145.36%)
9e98049 — 2025-11-25 20:50:553.01 (n/a)2.93 (n/a)2.92 (n/a)2.88 (n/a)0.05 (n/a)2204.81 (n/a)2148.41 (n/a)2142.13 (n/a)2110.81 (n/a)34.64 (n/a)8139.00 (n/a)7998.60 (n/a)8020.00 (n/a)7792.00 (n/a)127.99 (n/a)

gemm_2048x2048x2048_64x64x64_1cols

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
8b6d0f4 — 2026-03-25 20:46:390.51 (+0.13%)0.51 (+0.07%)0.51 (+0.04%)0.51 (+0.11%)0.00 (+8.62%)49296.10 (-0.11%)49206.24 (-0.07%)49283.30 (-0.04%)48878.90 (-0.13%)183.15 (+8.40%)351.48 (+0.13%)349.14 (+0.07%)348.59 (+0.04%)348.50 (+0.11%)1.31 (+8.62%)
dc66da7 — 2026-03-25 19:23:450.51 (n/a)0.51 (n/a)0.51 (n/a)0.51 (n/a)0.00 (n/a)49348.10 (n/a)49241.40 (n/a)49302.10 (n/a)48941.60 (n/a)168.96 (n/a)351.03 (n/a)348.89 (n/a)348.46 (n/a)348.14 (n/a)1.20 (n/a)

gemm_2048x2048x2048_64x64x64_2_cols_0_bcolmaj_0_ccolmaj_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
130b6ea — 2025-12-05 21:15:491.04 (+0.94%)1.03 (+0.23%)1.03 (+0.09%)1.02 (+0.07%)0.00 (+197.51%)24576.00 (-0.07%)24485.82 (-0.23%)24530.40 (-0.09%)24266.40 (-0.94%)111.26 (+194.35%)
0a6c11c — 2025-12-04 00:39:101.03 (n/a)1.03 (n/a)1.03 (n/a)1.02 (n/a)0.00 (n/a)24594.30 (n/a)24542.14 (n/a)24551.90 (n/a)24495.50 (n/a)37.80 (n/a)

gemm_2048x2048x2048_64x64x64_2_cols_0_bcolmaj_0_ccolmaj_0_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
a4b6ffe — 2026-01-05 19:24:051.04 (+0.28%)1.03 (-0.15%)1.03 (+0.05%)1.02 (-0.52%)0.01 (+42.72%)24719.80 (+0.52%)24512.42 (+0.16%)24529.20 (-0.05%)24092.60 (-0.28%)252.92 (+43.15%)713.08 (+0.28%)700.92 (-0.15%)700.38 (+0.05%)694.98 (-0.52%)7.30 (+42.71%)
50ad37c — 2025-12-29 20:35:531.04 (n/a)1.03 (n/a)1.03 (n/a)1.02 (n/a)0.01 (n/a)24592.30 (n/a)24473.70 (n/a)24540.90 (n/a)24160.80 (n/a)176.68 (n/a)711.06 (n/a)702.00 (n/a)700.05 (n/a)698.59 (n/a)5.12 (n/a)

gemm_2048x2048x2048_64x64x64_2_cols_0_bcolmaj_1_ccolmaj_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
130b6ea — 2025-12-05 21:15:490.99 (+1.69%)0.97 (+0.29%)0.97 (-0.04%)0.97 (-0.15%)0.01 (+434.92%)26025.30 (+0.15%)25839.34 (-0.29%)25916.70 (+0.04%)25446.00 (-1.67%)202.05 (+425.44%)
0a6c11c — 2025-12-04 00:39:100.97 (n/a)0.97 (n/a)0.97 (n/a)0.97 (n/a)0.00 (n/a)25986.90 (n/a)25913.52 (n/a)25905.70 (n/a)25877.00 (n/a)38.45 (n/a)

gemm_2048x2048x2048_64x64x64_2_cols_0_bcolmaj_1_ccolmaj_0_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
a4b6ffe — 2026-01-05 19:24:050.97 (-0.01%)0.97 (+0.03%)0.97 (+0.04%)0.97 (+0.00%)0.00 (-6.43%)25927.90 (-0.00%)25903.20 (-0.03%)25898.60 (-0.04%)25892.80 (+0.01%)14.05 (-6.56%)663.50 (-0.01%)663.23 (+0.03%)663.35 (+0.04%)662.60 (+0.00%)0.36 (-6.43%)
50ad37c — 2025-12-29 20:35:530.97 (n/a)0.97 (n/a)0.97 (n/a)0.97 (n/a)0.00 (n/a)25928.10 (n/a)25910.14 (n/a)25908.10 (n/a)25891.20 (n/a)15.04 (n/a)663.54 (n/a)663.06 (n/a)663.11 (n/a)662.60 (n/a)0.38 (n/a)

gemm_2048x2048x2048_64x64x64_2_cols_1_bcolmaj_0_ccolmaj_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
130b6ea — 2025-12-05 21:15:490.92 (-1.02%)0.92 (-0.66%)0.92 (-0.10%)0.91 (-0.78%)0.01 (-4.54%)27751.40 (+0.79%)27480.74 (+0.66%)27354.80 (+0.10%)27320.50 (+1.03%)182.69 (-2.84%)
0a6c11c — 2025-12-04 00:39:100.93 (n/a)0.92 (n/a)0.92 (n/a)0.91 (n/a)0.01 (n/a)27534.80 (n/a)27299.32 (n/a)27326.70 (n/a)27042.30 (n/a)188.04 (n/a)

gemm_2048x2048x2048_64x64x64_2_cols_1_bcolmaj_0_ccolmaj_0_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
a4b6ffe — 2026-01-05 19:24:050.92 (+0.11%)0.91 (-0.12%)0.91 (-0.21%)0.91 (-0.24%)0.00 (+30.34%)27723.50 (+0.24%)27548.94 (+0.12%)27522.90 (+0.21%)27403.60 (-0.11%)124.26 (+30.58%)626.92 (+0.11%)623.62 (-0.12%)624.20 (-0.21%)619.69 (-0.24%)2.81 (+30.34%)
50ad37c — 2025-12-29 20:35:530.92 (n/a)0.91 (n/a)0.92 (n/a)0.91 (n/a)0.00 (n/a)27657.10 (n/a)27515.58 (n/a)27464.70 (n/a)27434.60 (n/a)95.16 (n/a)626.21 (n/a)624.37 (n/a)625.52 (n/a)621.17 (n/a)2.16 (n/a)

gemm_2048x2048x2048_64x64x64_2cols_bcolmaj

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
8b6d0f4 — 2026-03-25 20:46:390.88 (+0.21%)0.87 (-0.11%)0.87 (-0.26%)0.87 (+0.03%)0.00 (+24.60%)29019.10 (-0.03%)28862.30 (+0.12%)28862.50 (+0.26%)28665.10 (-0.21%)148.73 (+24.28%)599.33 (+0.21%)595.25 (-0.11%)595.23 (-0.26%)592.02 (+0.03%)3.07 (+24.61%)
dc66da7 — 2026-03-25 19:23:450.88 (n/a)0.87 (n/a)0.87 (n/a)0.87 (n/a)0.00 (n/a)29027.70 (n/a)28829.10 (n/a)28788.30 (n/a)28724.70 (n/a)119.67 (n/a)598.09 (n/a)595.93 (n/a)596.77 (n/a)591.84 (n/a)2.46 (n/a)

gemm_2048x2048x2048_64x64x64_8_cols_0_bcolmaj_0_ccolmaj_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)GFLOP/s (max)GFLOP/s (mean)GFLOP/s (median)GFLOP/s (min)GFLOP/s (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
31ab24e — 2025-11-22 02:01:563.65 (-1.06%)3.59 (+0.83%)3.63 (-0.26%)3.45 (+2.66%)0.07 (-44.71%)2678.08 (-1.06%)2632.13 (+0.83%)2658.60 (-0.26%)2528.31 (+2.66%)53.86 (-44.71%)6795.00 (-2.59%)6529.80 (-0.92%)6462.00 (+0.26%)6415.00 (+1.07%)137.14 (-45.27%)
d9a64c5 — 2025-11-21 20:49:353.69 (n/a)3.56 (n/a)3.64 (n/a)3.36 (n/a)0.13 (n/a)2706.77 (n/a)2610.50 (n/a)2665.61 (n/a)2462.71 (n/a)97.41 (n/a)6976.00 (n/a)6590.40 (n/a)6445.00 (n/a)6347.00 (n/a)250.58 (n/a)

gemm_2048x2048x2048_64x64x64_8_cols_0_bcolmaj_0_ccolmaj_0_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
d48746f — 2026-01-12 22:17:103.98 (+0.20%)3.84 (-2.95%)3.76 (-5.27%)3.72 (-4.97%)0.13 (+473.97%)6760.00 (+5.23%)6559.50 (+3.13%)6697.10 (+5.56%)6321.00 (-0.20%)218.68 (+499.40%)2717.91 (+0.20%)2621.44 (-2.95%)2565.27 (-5.27%)2541.40 (-4.97%)88.37 (+473.97%)
331dcca — 2026-01-08 18:34:103.97 (n/a)3.96 (n/a)3.97 (n/a)3.92 (n/a)0.02 (n/a)6423.80 (n/a)6360.32 (n/a)6344.30 (n/a)6333.90 (n/a)36.48 (n/a)2712.39 (n/a)2701.17 (n/a)2707.91 (n/a)2674.40 (n/a)15.40 (n/a)

gemm_2048x2048x2048_64x64x64_8_cols_0_bcolmaj_1_ccolmaj_0_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
d48746f — 2026-01-12 22:17:103.75 (+0.28%)3.70 (+1.25%)3.71 (-0.21%)3.65 (+4.77%)0.04 (-66.90%)6901.80 (-4.56%)6795.46 (-1.30%)6790.20 (+0.21%)6715.90 (-0.28%)67.05 (-68.40%)2558.08 (+0.28%)2528.33 (+1.25%)2530.10 (-0.21%)2489.18 (+4.77%)24.81 (-66.90%)
331dcca — 2026-01-08 18:34:103.74 (n/a)3.66 (n/a)3.71 (n/a)3.48 (n/a)0.11 (n/a)7231.20 (n/a)6885.18 (n/a)6775.70 (n/a)6734.90 (n/a)212.21 (n/a)2550.87 (n/a)2497.04 (n/a)2535.50 (n/a)2375.78 (n/a)74.96 (n/a)

gemm_2048x2048x2048_64x64x64_8_cols_1_bcolmaj_0_ccolmaj_0_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
d48746f — 2026-01-12 22:17:103.62 (-0.77%)3.54 (+0.14%)3.60 (-0.04%)3.44 (+2.17%)0.09 (-25.77%)7307.40 (-2.13%)7103.46 (-0.18%)6982.50 (+0.04%)6948.20 (+0.77%)182.95 (-26.55%)2472.55 (-0.77%)2419.80 (+0.14%)2460.41 (-0.04%)2351.04 (+2.17%)61.76 (-25.77%)
331dcca — 2026-01-08 18:34:103.65 (n/a)3.54 (n/a)3.61 (n/a)3.37 (n/a)0.12 (n/a)7466.20 (n/a)7116.40 (n/a)6979.60 (n/a)6895.10 (n/a)249.09 (n/a)2491.61 (n/a)2416.45 (n/a)2461.45 (n/a)2301.00 (n/a)83.20 (n/a)

gemm_2048x2048x2048_64x64x64_8cols_bcolmaj_ccolmaj

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
8b6d0f4 — 2026-03-25 20:46:393.47 (-0.81%)3.41 (+0.82%)3.42 (-0.23%)3.35 (+4.63%)0.05 (-63.03%)7522.10 (-4.43%)7390.72 (-0.92%)7367.60 (+0.23%)7255.80 (+0.82%)107.14 (-64.21%)2367.74 (-0.81%)2324.90 (+0.82%)2331.81 (-0.23%)2283.92 (+4.63%)33.69 (-63.03%)
dc66da7 — 2026-03-25 19:23:453.50 (n/a)3.38 (n/a)3.42 (n/a)3.20 (n/a)0.13 (n/a)7870.70 (n/a)7459.22 (n/a)7350.70 (n/a)7196.70 (n/a)299.40 (n/a)2387.20 (n/a)2306.10 (n/a)2337.18 (n/a)2182.77 (n/a)91.13 (n/a)

gemm_384x1536x1792_32x48x64_4cols_bcolmaj

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
8b6d0f4 — 2026-03-25 20:46:394.16 (+12.80%)3.50 (+5.22%)3.50 (+2.51%)2.86 (-0.71%)0.63 (+68.68%)2814.80 (+0.72%)2362.04 (-3.41%)2304.60 (-2.45%)1936.00 (-11.34%)430.86 (+52.19%)1091.93 (+12.80%)919.16 (+5.22%)917.28 (+2.51%)751.01 (-0.71%)166.31 (+68.68%)
dc66da7 — 2026-03-25 19:23:453.69 (n/a)3.33 (n/a)3.41 (n/a)2.88 (n/a)0.38 (n/a)2794.70 (n/a)2445.42 (n/a)2362.50 (n/a)2183.70 (n/a)283.10 (n/a)968.06 (n/a)873.57 (n/a)894.79 (n/a)756.39 (n/a)98.60 (n/a)

gemm_896x1792x640_32x64x80_8cols_ccolmaj

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
8b6d0f4 — 2026-03-25 20:46:396.32 (+6.48%)4.74 (+3.95%)4.60 (+2.04%)3.49 (+3.11%)1.02 (+11.51%)1906.60 (-3.02%)1453.08 (-3.45%)1446.20 (-2.01%)1052.10 (-6.09%)304.01 (+0.60%)1953.36 (+6.48%)1465.54 (+3.95%)1421.06 (+2.04%)1077.93 (+3.11%)313.81 (+11.51%)
dc66da7 — 2026-03-25 19:23:455.94 (n/a)4.56 (n/a)4.51 (n/a)3.38 (n/a)0.91 (n/a)1966.00 (n/a)1504.94 (n/a)1475.80 (n/a)1120.30 (n/a)302.18 (n/a)1834.47 (n/a)1409.81 (n/a)1392.60 (n/a)1045.38 (n/a)281.42 (n/a)

input_length_2048-num_aie_columns_1-num_channels_1-tile_size_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8c138a8 — 2026-03-26 14:54:200.04 (n/a)0.03 (n/a)0.03 (n/a)0.01 (n/a)0.01 (n/a)939.90 (n/a)316.90 (n/a)248.30 (n/a)223.60 (n/a)172.19 (n/a)

input_length_2048-num_aie_columns_1-num_channels_1-tile_size_2048-group_size_32

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8c138a8 — 2026-03-26 14:54:200.02 (n/a)0.02 (n/a)0.02 (n/a)0.01 (n/a)0.00 (n/a)522.80 (n/a)333.16 (n/a)299.00 (n/a)245.40 (n/a)108.69 (n/a)

input_length_2048-num_aie_columns_1-num_channels_1-tile_size_2048-weighted_False

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8c138a8 — 2026-03-26 14:54:200.03 (n/a)0.03 (n/a)0.03 (n/a)0.02 (n/a)0.00 (n/a)365.20 (n/a)274.30 (n/a)250.20 (n/a)246.70 (n/a)51.22 (n/a)

input_length_2048-num_aie_columns_1-num_channels_1-tile_size_2048-weighted_True

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8c138a8 — 2026-03-26 14:54:200.05 (n/a)0.04 (n/a)0.05 (n/a)0.03 (n/a)0.01 (n/a)366.50 (n/a)283.26 (n/a)253.30 (n/a)248.00 (n/a)51.35 (n/a)

input_length_2048-num_aie_columns_1-num_channels_2-tile_size_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8c138a8 — 2026-03-26 14:54:200.04 (n/a)0.03 (n/a)0.03 (n/a)0.02 (n/a)0.00 (n/a)373.50 (n/a)265.74 (n/a)253.55 (n/a)230.00 (n/a)44.37 (n/a)

input_length_2048-num_aie_columns_1-num_channels_2-tile_size_1024-group_size_32

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8c138a8 — 2026-03-26 14:54:200.02 (n/a)0.02 (n/a)0.02 (n/a)0.01 (n/a)0.00 (n/a)570.00 (n/a)342.98 (n/a)292.70 (n/a)265.60 (n/a)127.49 (n/a)

input_length_2048-num_aie_columns_1-num_channels_2-tile_size_1024-weighted_False

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8c138a8 — 2026-03-26 14:54:200.04 (n/a)0.03 (n/a)0.04 (n/a)0.02 (n/a)0.01 (n/a)352.30 (n/a)257.72 (n/a)233.40 (n/a)225.80 (n/a)53.58 (n/a)

input_length_2048-num_aie_columns_1-num_channels_2-tile_size_2048-scalar_factor_3.0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8c138a8 — 2026-03-26 14:54:200.05 (n/a)0.05 (n/a)0.05 (n/a)0.02 (n/a)0.02 (n/a)799.40 (n/a)344.86 (n/a)235.70 (n/a)223.50 (n/a)254.16 (n/a)

input_length_2048-num_aie_columns_1-tile_size_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8c138a8 — 2026-03-26 14:54:200.05 (n/a)0.04 (n/a)0.04 (n/a)0.02 (n/a)0.01 (n/a)546.20 (n/a)282.31 (n/a)251.80 (n/a)211.40 (n/a)83.27 (n/a)

input_length_2048-num_aie_columns_2-num_channels_1-tile_size_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8c138a8 — 2026-03-26 14:54:200.04 (n/a)0.03 (n/a)0.03 (n/a)0.01 (n/a)0.01 (n/a)808.60 (n/a)301.24 (n/a)245.60 (n/a)222.50 (n/a)155.00 (n/a)

input_length_2048-num_aie_columns_2-num_channels_1-tile_size_1024-group_size_32

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8c138a8 — 2026-03-26 14:54:200.02 (n/a)0.02 (n/a)0.02 (n/a)0.02 (n/a)0.00 (n/a)347.30 (n/a)299.10 (n/a)297.50 (n/a)261.60 (n/a)31.04 (n/a)

input_length_2048-num_aie_columns_2-num_channels_1-tile_size_1024-weighted_False

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8c138a8 — 2026-03-26 14:54:200.03 (n/a)0.03 (n/a)0.03 (n/a)0.02 (n/a)0.00 (n/a)365.50 (n/a)276.54 (n/a)255.30 (n/a)250.80 (n/a)49.78 (n/a)

input_length_2048-num_aie_columns_2-num_channels_1-tile_size_1024-weighted_True

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8c138a8 — 2026-03-26 14:54:200.04 (n/a)0.04 (n/a)0.04 (n/a)0.02 (n/a)0.01 (n/a)514.60 (n/a)312.78 (n/a)251.60 (n/a)243.70 (n/a)115.28 (n/a)

input_length_2048-num_aie_columns_2-num_channels_2-tile_size_1024-scalar_factor_3.0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8c138a8 — 2026-03-26 14:54:200.05 (n/a)0.04 (n/a)0.05 (n/a)0.02 (n/a)0.01 (n/a)532.90 (n/a)308.38 (n/a)256.50 (n/a)242.20 (n/a)125.66 (n/a)

input_length_2048-num_aie_columns_2-num_channels_2-tile_size_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8c138a8 — 2026-03-26 14:54:200.03 (n/a)0.03 (n/a)0.03 (n/a)0.02 (n/a)0.00 (n/a)401.90 (n/a)271.34 (n/a)249.80 (n/a)237.80 (n/a)52.21 (n/a)

input_length_2048-num_aie_columns_2-num_channels_2-tile_size_512-group_size_32

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8c138a8 — 2026-03-26 14:54:200.02 (n/a)0.02 (n/a)0.02 (n/a)0.02 (n/a)0.00 (n/a)313.10 (n/a)296.08 (n/a)292.40 (n/a)275.10 (n/a)16.43 (n/a)

input_length_2048-num_aie_columns_2-num_channels_2-tile_size_512-weighted_False

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8c138a8 — 2026-03-26 14:54:200.04 (n/a)0.03 (n/a)0.03 (n/a)0.02 (n/a)0.01 (n/a)460.90 (n/a)279.30 (n/a)234.80 (n/a)231.20 (n/a)101.53 (n/a)

input_length_2048-num_aie_columns_2-tile_size_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8c138a8 — 2026-03-26 14:54:200.06 (n/a)0.04 (n/a)0.04 (n/a)0.02 (n/a)0.01 (n/a)539.80 (n/a)277.69 (n/a)251.10 (n/a)221.00 (n/a)83.18 (n/a)

input_length_2048-num_aie_columns_4-num_channels_1-tile_size_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8c138a8 — 2026-03-26 14:54:200.04 (n/a)0.03 (n/a)0.03 (n/a)0.01 (n/a)0.01 (n/a)630.50 (n/a)288.26 (n/a)248.30 (n/a)221.20 (n/a)96.49 (n/a)

input_length_2048-num_aie_columns_4-num_channels_1-tile_size_512-group_size_32

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8c138a8 — 2026-03-26 14:54:200.02 (n/a)0.02 (n/a)0.02 (n/a)0.02 (n/a)0.00 (n/a)323.10 (n/a)295.42 (n/a)297.40 (n/a)264.20 (n/a)20.92 (n/a)

input_length_2048-num_aie_columns_4-num_channels_1-tile_size_512-weighted_False

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8c138a8 — 2026-03-26 14:54:200.04 (n/a)0.03 (n/a)0.03 (n/a)0.02 (n/a)0.00 (n/a)340.80 (n/a)264.24 (n/a)248.80 (n/a)229.20 (n/a)43.82 (n/a)

input_length_2048-num_aie_columns_4-num_channels_1-tile_size_512-weighted_True

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8c138a8 — 2026-03-26 14:54:200.04 (n/a)0.03 (n/a)0.03 (n/a)0.03 (n/a)0.01 (n/a)356.70 (n/a)275.34 (n/a)263.40 (n/a)234.20 (n/a)50.31 (n/a)

input_length_2048-num_aie_columns_4-num_channels_2-tile_size_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8c138a8 — 2026-03-26 14:54:200.03 (n/a)0.03 (n/a)0.03 (n/a)0.01 (n/a)0.01 (n/a)842.90 (n/a)371.89 (n/a)265.35 (n/a)244.30 (n/a)214.86 (n/a)

input_length_2048-num_aie_columns_4-num_channels_2-tile_size_256-group_size_32

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8c138a8 — 2026-03-26 14:54:200.02 (n/a)0.02 (n/a)0.02 (n/a)0.02 (n/a)0.00 (n/a)315.20 (n/a)289.24 (n/a)286.20 (n/a)271.90 (n/a)15.83 (n/a)

input_length_2048-num_aie_columns_4-num_channels_2-tile_size_256-weighted_False

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8c138a8 — 2026-03-26 14:54:200.04 (n/a)0.03 (n/a)0.03 (n/a)0.01 (n/a)0.01 (n/a)567.60 (n/a)306.42 (n/a)243.90 (n/a)231.90 (n/a)146.16 (n/a)

input_length_2048-num_aie_columns_4-num_channels_2-tile_size_512-scalar_factor_3.0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8c138a8 — 2026-03-26 14:54:200.05 (n/a)0.04 (n/a)0.04 (n/a)0.03 (n/a)0.01 (n/a)417.80 (n/a)303.22 (n/a)296.20 (n/a)246.70 (n/a)67.89 (n/a)

input_length_2048-num_aie_columns_4-tile_size_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8c138a8 — 2026-03-26 14:54:200.05 (n/a)0.04 (n/a)0.04 (n/a)0.02 (n/a)0.01 (n/a)805.70 (n/a)341.81 (n/a)252.00 (n/a)224.20 (n/a)178.90 (n/a)

input_length_2048-num_aie_columns_8-num_channels_1-tile_size_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8c138a8 — 2026-03-26 14:54:200.04 (n/a)0.03 (n/a)0.03 (n/a)0.01 (n/a)0.01 (n/a)725.10 (n/a)315.54 (n/a)259.70 (n/a)231.30 (n/a)148.83 (n/a)

input_length_2048-num_aie_columns_8-num_channels_1-tile_size_256-group_size_32

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8c138a8 — 2026-03-26 14:54:200.02 (n/a)0.02 (n/a)0.02 (n/a)0.01 (n/a)0.00 (n/a)364.80 (n/a)301.72 (n/a)299.30 (n/a)259.70 (n/a)39.39 (n/a)

input_length_2048-num_aie_columns_8-num_channels_1-tile_size_256-weighted_False

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8c138a8 — 2026-03-26 14:54:200.04 (n/a)0.03 (n/a)0.03 (n/a)0.03 (n/a)0.00 (n/a)313.40 (n/a)251.34 (n/a)237.30 (n/a)232.60 (n/a)34.75 (n/a)

input_length_2048-num_aie_columns_8-num_channels_1-tile_size_256-weighted_True

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8c138a8 — 2026-03-26 14:54:200.03 (n/a)0.03 (n/a)0.03 (n/a)0.01 (n/a)0.01 (n/a)771.00 (n/a)364.80 (n/a)262.30 (n/a)260.00 (n/a)227.11 (n/a)

input_length_2048-num_aie_columns_8-num_channels_2-tile_size_128

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8c138a8 — 2026-03-26 14:54:200.03 (n/a)0.03 (n/a)0.03 (n/a)0.01 (n/a)0.01 (n/a)683.90 (n/a)312.01 (n/a)265.55 (n/a)251.20 (n/a)133.27 (n/a)

input_length_2048-num_aie_columns_8-num_channels_2-tile_size_128-group_size_32

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8c138a8 — 2026-03-26 14:54:200.02 (n/a)0.01 (n/a)0.02 (n/a)0.01 (n/a)0.00 (n/a)632.50 (n/a)382.44 (n/a)323.20 (n/a)315.50 (n/a)139.84 (n/a)

input_length_2048-num_aie_columns_8-num_channels_2-tile_size_128-weighted_False

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8c138a8 — 2026-03-26 14:54:200.03 (n/a)0.03 (n/a)0.03 (n/a)0.01 (n/a)0.01 (n/a)694.80 (n/a)356.86 (n/a)272.20 (n/a)270.30 (n/a)188.92 (n/a)

input_length_2048-num_aie_columns_8-num_channels_2-tile_size_256-scalar_factor_3.0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8c138a8 — 2026-03-26 14:54:200.05 (n/a)0.04 (n/a)0.05 (n/a)0.02 (n/a)0.01 (n/a)670.70 (n/a)354.26 (n/a)272.60 (n/a)258.10 (n/a)177.93 (n/a)

input_length_2048-num_aie_columns_8-tile_size_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8c138a8 — 2026-03-26 14:54:200.05 (n/a)0.04 (n/a)0.04 (n/a)0.01 (n/a)0.01 (n/a)661.10 (n/a)323.19 (n/a)264.00 (n/a)237.40 (n/a)129.94 (n/a)

input_length_2048-num_cores_1-num_channels_1-bypass_False-tile_size_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8c138a8 — 2026-03-26 14:54:200.03 (n/a)0.03 (n/a)0.03 (n/a)0.02 (n/a)0.01 (n/a)399.00 (n/a)292.80 (n/a)276.90 (n/a)237.30 (n/a)61.69 (n/a)

input_length_2048-num_cores_16-num_channels_2-bypass_False-tile_size_128

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8c138a8 — 2026-03-26 14:54:200.03 (n/a)0.03 (n/a)0.03 (n/a)0.01 (n/a)0.01 (n/a)688.40 (n/a)338.14 (n/a)250.90 (n/a)249.20 (n/a)195.81 (n/a)

input_length_2048-num_cores_2-num_channels_1-bypass_False-tile_size_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8c138a8 — 2026-03-26 14:54:200.03 (n/a)0.03 (n/a)0.03 (n/a)0.01 (n/a)0.01 (n/a)648.00 (n/a)326.46 (n/a)246.70 (n/a)243.20 (n/a)179.75 (n/a)

input_length_2048-num_cores_2-num_channels_2-bypass_False-tile_size_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8c138a8 — 2026-03-26 14:54:200.03 (n/a)0.03 (n/a)0.03 (n/a)0.02 (n/a)0.01 (n/a)418.70 (n/a)289.62 (n/a)255.60 (n/a)243.60 (n/a)74.18 (n/a)

input_length_2048-num_cores_4-num_channels_1-bypass_False-tile_size_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8c138a8 — 2026-03-26 14:54:200.04 (n/a)0.03 (n/a)0.03 (n/a)0.02 (n/a)0.00 (n/a)360.60 (n/a)273.08 (n/a)249.60 (n/a)234.00 (n/a)51.53 (n/a)

input_length_2048-num_cores_4-num_channels_2-bypass_False-tile_size_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8c138a8 — 2026-03-26 14:54:200.03 (n/a)0.03 (n/a)0.03 (n/a)0.02 (n/a)0.00 (n/a)392.30 (n/a)302.62 (n/a)302.40 (n/a)243.10 (n/a)56.30 (n/a)

input_length_2048-num_cores_8-num_channels_1-bypass_False-tile_size_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8c138a8 — 2026-03-26 14:54:200.04 (n/a)0.03 (n/a)0.03 (n/a)0.01 (n/a)0.01 (n/a)651.20 (n/a)326.04 (n/a)239.10 (n/a)230.00 (n/a)182.76 (n/a)

input_length_2048-num_cores_8-num_channels_2-bypass_False-tile_size_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8c138a8 — 2026-03-26 14:54:200.03 (n/a)0.03 (n/a)0.03 (n/a)0.01 (n/a)0.01 (n/a)631.30 (n/a)332.06 (n/a)258.60 (n/a)254.20 (n/a)167.29 (n/a)

input_length_32768-num_aie_columns_2-num_channels_2-tile_size_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8c138a8 — 2026-03-26 14:54:200.49 (n/a)0.43 (n/a)0.49 (n/a)0.30 (n/a)0.09 (n/a)441.30 (n/a)314.74 (n/a)268.70 (n/a)266.70 (n/a)75.74 (n/a)

input_length_32768-num_aie_columns_2-num_channels_2-tile_size_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8c138a8 — 2026-03-26 14:54:200.50 (n/a)0.42 (n/a)0.46 (n/a)0.28 (n/a)0.09 (n/a)469.70 (n/a)324.96 (n/a)287.40 (n/a)263.10 (n/a)83.92 (n/a)

input_length_32768-num_aie_columns_4-num_channels_4-tile_size_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8c138a8 — 2026-03-26 14:54:200.50 (n/a)0.38 (n/a)0.45 (n/a)0.19 (n/a)0.13 (n/a)675.60 (n/a)391.14 (n/a)290.00 (n/a)264.70 (n/a)174.76 (n/a)

layer_norm_1_cols_1_channels_2048_tile_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.07 (+35.61%)0.05 (+6.58%)0.04 (-0.65%)0.03 (-1.11%)0.01 (+98.33%)236.50 (+1.11%)182.90 (-3.12%)183.30 (+0.60%)118.50 (-26.26%)42.30 (+41.46%)
dc66da7 — 2026-03-25 19:23:450.05 (n/a)0.04 (n/a)0.04 (n/a)0.04 (n/a)0.01 (n/a)233.90 (n/a)188.80 (n/a)182.20 (n/a)160.70 (n/a)29.90 (n/a)

layer_norm_1_cols_2_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.07 (+9.47%)0.05 (+5.79%)0.05 (+5.48%)0.04 (-7.03%)0.01 (+51.12%)206.90 (+7.54%)163.06 (-2.60%)165.20 (-5.22%)113.20 (-8.71%)40.37 (+56.59%)
dc66da7 — 2026-03-25 19:23:450.07 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.01 (n/a)192.40 (n/a)167.42 (n/a)174.30 (n/a)124.00 (n/a)25.78 (n/a)

layer_norm_2_cols_1_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.06 (-0.97%)0.05 (+14.59%)0.05 (+47.43%)0.03 (+8.92%)0.01 (-0.31%)235.10 (-8.20%)183.06 (-13.03%)157.80 (-32.19%)148.10 (+1.02%)43.12 (-7.94%)
dc66da7 — 2026-03-25 19:23:450.06 (n/a)0.04 (n/a)0.04 (n/a)0.03 (n/a)0.01 (n/a)256.10 (n/a)210.48 (n/a)232.70 (n/a)146.60 (n/a)46.84 (n/a)

layer_norm_2_cols_2_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.07 (+14.65%)0.05 (+4.61%)0.05 (-7.99%)0.04 (+11.94%)0.01 (+28.58%)201.70 (-10.67%)168.58 (-3.65%)176.00 (+8.71%)122.10 (-12.79%)33.82 (+0.06%)
dc66da7 — 2026-03-25 19:23:450.06 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.01 (n/a)225.80 (n/a)174.96 (n/a)161.90 (n/a)140.00 (n/a)33.80 (n/a)

layer_norm_4_cols_1_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.06 (-31.05%)0.05 (-3.73%)0.05 (+9.03%)0.04 (+10.49%)0.01 (-63.42%)197.60 (-9.52%)167.50 (-4.83%)166.40 (-8.27%)129.00 (+45.11%)25.47 (-51.02%)
dc66da7 — 2026-03-25 19:23:450.09 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.02 (n/a)218.40 (n/a)176.00 (n/a)181.40 (n/a)88.90 (n/a)52.00 (n/a)

layer_norm_4_cols_2_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.05 (-1.52%)0.04 (-11.20%)0.04 (-25.45%)0.03 (-0.69%)0.01 (-9.35%)243.40 (+0.66%)202.20 (+12.10%)212.70 (+34.20%)156.90 (+1.55%)34.18 (-7.92%)
dc66da7 — 2026-03-25 19:23:450.05 (n/a)0.05 (n/a)0.05 (n/a)0.03 (n/a)0.01 (n/a)241.80 (n/a)180.38 (n/a)158.50 (n/a)154.50 (n/a)37.12 (n/a)

layer_norm_8_cols_1_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.05 (-6.07%)0.04 (+15.60%)0.04 (+21.55%)0.04 (+82.28%)0.01 (-55.28%)222.00 (-45.13%)190.18 (-21.85%)193.60 (-17.72%)153.10 (+6.47%)24.64 (-75.05%)
dc66da7 — 2026-03-25 19:23:450.06 (n/a)0.04 (n/a)0.03 (n/a)0.02 (n/a)0.01 (n/a)404.60 (n/a)243.34 (n/a)235.30 (n/a)143.80 (n/a)98.74 (n/a)

layer_norm_8_cols_2_channels_2048_tile_128

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.05 (+37.82%)0.04 (+27.22%)0.04 (+9.46%)0.03 (+34.60%)0.01 (+56.79%)251.50 (-25.72%)212.60 (-21.06%)226.70 (-8.66%)175.20 (-27.45%)33.04 (-18.34%)
dc66da7 — 2026-03-25 19:23:450.03 (n/a)0.03 (n/a)0.03 (n/a)0.02 (n/a)0.00 (n/a)338.60 (n/a)269.32 (n/a)248.20 (n/a)241.50 (n/a)40.46 (n/a)

matrix_vector_mul_128x128_32_1col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)
130b6ea — 2025-12-05 21:15:490.37 (+17.13%)0.30 (+24.87%)0.30 (+38.03%)0.23 (+12.47%)0.06 (+35.23%)
0a6c11c — 2025-12-04 00:39:100.32 (n/a)0.24 (n/a)0.22 (n/a)0.20 (n/a)0.04 (n/a)

matrix_vector_mul_128x128_32_1col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
331dcca — 2026-01-08 18:34:100.26 (-1.81%)0.24 (+4.06%)0.24 (+0.52%)0.21 (+16.97%)0.02 (-48.16%)0.26 (-1.81%)0.24 (+4.06%)0.24 (+0.52%)0.20 (+16.97%)0.02 (-48.16%)
a4b6ffe — 2026-01-05 19:24:050.26 (n/a)0.23 (n/a)0.24 (n/a)0.18 (n/a)0.04 (n/a)0.26 (n/a)0.23 (n/a)0.24 (n/a)0.17 (n/a)0.04 (n/a)

matrix_vector_mul_128x128_32tsi_128tso_1col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
8b6d0f4 — 2026-03-25 20:46:390.26 (+32.10%)0.19 (+11.55%)0.18 (+9.49%)0.13 (-2.22%)0.05 (+105.83%)0.25 (+32.10%)0.19 (+11.55%)0.18 (+9.49%)0.13 (-2.22%)0.05 (+105.83%)
dc66da7 — 2026-03-25 19:23:450.19 (n/a)0.17 (n/a)0.17 (n/a)0.14 (n/a)0.02 (n/a)0.19 (n/a)0.17 (n/a)0.17 (n/a)0.14 (n/a)0.02 (n/a)

matrix_vector_mul_128x128_32tsi_128tso_1col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
897d04e — 2026-03-06 22:12:250.37 (+5.44%)0.31 (-0.05%)0.34 (-1.99%)0.22 (-7.87%)0.06 (+15.13%)0.36 (+5.44%)0.31 (-0.05%)0.34 (-1.99%)0.21 (-7.87%)0.06 (+15.13%)
84d3478 — 2026-02-17 23:25:130.35 (n/a)0.31 (n/a)0.35 (n/a)0.23 (n/a)0.05 (n/a)0.35 (n/a)0.31 (n/a)0.34 (n/a)0.23 (n/a)0.05 (n/a)

matrix_vector_mul_2048x8192_1_1col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)
130b6ea — 2025-12-05 21:15:4913.57 (+3.47%)13.15 (+1.09%)13.10 (+0.16%)12.52 (-1.58%)0.37 (+153.19%)
0a6c11c — 2025-12-04 00:39:1013.11 (n/a)13.01 (n/a)13.07 (n/a)12.72 (n/a)0.15 (n/a)

matrix_vector_mul_2048x8192_1_1col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
331dcca — 2026-01-08 18:34:1013.43 (-0.07%)13.30 (+4.62%)13.27 (+1.29%)13.16 (+16.70%)0.12 (-86.69%)13.42 (-0.07%)13.29 (+4.62%)13.26 (+1.29%)13.16 (+16.70%)0.12 (-86.69%)
a4b6ffe — 2026-01-05 19:24:0513.44 (n/a)12.71 (n/a)13.10 (n/a)11.28 (n/a)0.87 (n/a)13.43 (n/a)12.70 (n/a)13.09 (n/a)11.27 (n/a)0.87 (n/a)

matrix_vector_mul_2048x8192_1_2col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)
130b6ea — 2025-12-05 21:15:4925.06 (+2.47%)23.71 (+1.81%)24.13 (+0.50%)20.90 (+5.41%)1.48 (-15.60%)
0a6c11c — 2025-12-04 00:39:1024.46 (n/a)23.29 (n/a)24.01 (n/a)19.83 (n/a)1.75 (n/a)

matrix_vector_mul_2048x8192_1_2col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
331dcca — 2026-01-08 18:34:1024.46 (-0.60%)21.07 (-8.03%)19.75 (-17.83%)18.08 (+1.18%)3.03 (+7.07%)24.44 (-0.60%)21.06 (-8.03%)19.74 (-17.83%)18.07 (+1.18%)3.03 (+7.07%)
a4b6ffe — 2026-01-05 19:24:0524.60 (n/a)22.91 (n/a)24.04 (n/a)17.87 (n/a)2.83 (n/a)24.59 (n/a)22.90 (n/a)24.02 (n/a)17.86 (n/a)2.83 (n/a)

matrix_vector_mul_2048x8192_1_4col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)
130b6ea — 2025-12-05 21:15:4942.05 (+1.54%)40.50 (+12.60%)40.49 (+2.36%)38.98 (+97.53%)0.97 (-88.09%)
0a6c11c — 2025-12-04 00:39:1041.41 (n/a)35.97 (n/a)39.55 (n/a)19.73 (n/a)8.15 (n/a)

matrix_vector_mul_2048x8192_1_4col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
331dcca — 2026-01-08 18:34:1042.58 (+1.51%)41.37 (+14.29%)41.25 (+4.89%)40.33 (+91.42%)0.92 (-89.18%)42.56 (+1.51%)41.34 (+14.29%)41.23 (+4.89%)40.30 (+91.42%)0.92 (-89.18%)
a4b6ffe — 2026-01-05 19:24:0541.95 (n/a)36.20 (n/a)39.33 (n/a)21.07 (n/a)8.54 (n/a)41.92 (n/a)36.17 (n/a)39.31 (n/a)21.05 (n/a)8.53 (n/a)

matrix_vector_mul_2048x8192_1_8col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)
130b6ea — 2025-12-05 21:15:4946.62 (+3.11%)44.82 (+0.17%)44.78 (+0.17%)42.60 (-3.80%)1.60 (+367.72%)
0a6c11c — 2025-12-04 00:39:1045.21 (n/a)44.75 (n/a)44.70 (n/a)44.29 (n/a)0.34 (n/a)

matrix_vector_mul_2048x8192_1_8col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
331dcca — 2026-01-08 18:34:1046.04 (+4.14%)45.06 (+3.47%)44.55 (+2.76%)44.30 (+2.84%)0.87 (+66.58%)46.01 (+4.14%)45.03 (+3.47%)44.52 (+2.76%)44.27 (+2.84%)0.87 (+66.58%)
a4b6ffe — 2026-01-05 19:24:0544.21 (n/a)43.55 (n/a)43.35 (n/a)43.07 (n/a)0.52 (n/a)44.18 (n/a)43.52 (n/a)43.32 (n/a)43.05 (n/a)0.52 (n/a)

matrix_vector_mul_2048x8192_1tsi_1024tso_2col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
8b6d0f4 — 2026-03-25 20:46:3925.44 (+1.95%)24.42 (-0.28%)24.79 (+0.16%)23.14 (-2.33%)1.05 (+106.47%)25.42 (+1.95%)24.41 (-0.28%)24.77 (+0.16%)23.12 (-2.33%)1.05 (+106.47%)
dc66da7 — 2026-03-25 19:23:4524.95 (n/a)24.49 (n/a)24.75 (n/a)23.69 (n/a)0.51 (n/a)24.94 (n/a)24.48 (n/a)24.73 (n/a)23.67 (n/a)0.51 (n/a)

matrix_vector_mul_2048x8192_1tsi_1024tso_2col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
897d04e — 2026-03-06 22:12:2524.47 (-3.32%)24.16 (-3.13%)24.20 (-3.09%)23.84 (-3.04%)0.25 (-14.36%)24.46 (-3.32%)24.15 (-3.13%)24.18 (-3.09%)23.83 (-3.04%)0.25 (-14.36%)
84d3478 — 2026-02-17 23:25:1325.31 (n/a)24.94 (n/a)24.97 (n/a)24.59 (n/a)0.29 (n/a)25.30 (n/a)24.93 (n/a)24.95 (n/a)24.58 (n/a)0.29 (n/a)

matrix_vector_mul_2048x8192_1tsi_2048tso_1col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
8b6d0f4 — 2026-03-25 20:46:3913.48 (-1.01%)13.15 (+1.31%)13.11 (-0.87%)12.95 (+6.77%)0.20 (-66.05%)13.47 (-1.01%)13.14 (+1.31%)13.11 (-0.87%)12.95 (+6.77%)0.20 (-66.05%)
dc66da7 — 2026-03-25 19:23:4513.62 (n/a)12.98 (n/a)13.23 (n/a)12.13 (n/a)0.58 (n/a)13.61 (n/a)12.97 (n/a)13.22 (n/a)12.13 (n/a)0.58 (n/a)

matrix_vector_mul_2048x8192_1tsi_2048tso_1col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
897d04e — 2026-03-06 22:12:2513.25 (-3.04%)13.02 (-2.92%)13.10 (-2.64%)12.56 (-4.14%)0.27 (+32.55%)13.24 (-3.04%)13.01 (-2.92%)13.09 (-2.64%)12.55 (-4.14%)0.27 (+32.55%)
84d3478 — 2026-02-17 23:25:1313.66 (n/a)13.41 (n/a)13.45 (n/a)13.10 (n/a)0.20 (n/a)13.65 (n/a)13.40 (n/a)13.44 (n/a)13.09 (n/a)0.20 (n/a)

matrix_vector_mul_2048x8192_1tsi_256tso_8col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
8b6d0f4 — 2026-03-25 20:46:3942.06 (-4.87%)40.85 (-4.53%)40.57 (-4.52%)39.78 (-4.49%)0.87 (-11.49%)42.04 (-4.87%)40.83 (-4.53%)40.55 (-4.52%)39.76 (-4.49%)0.87 (-11.49%)
dc66da7 — 2026-03-25 19:23:4544.21 (n/a)42.79 (n/a)42.49 (n/a)41.65 (n/a)0.99 (n/a)44.19 (n/a)42.76 (n/a)42.47 (n/a)41.63 (n/a)0.99 (n/a)

matrix_vector_mul_2048x8192_1tsi_256tso_8col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
897d04e — 2026-03-06 22:12:2547.71 (+3.51%)45.38 (+9.55%)45.22 (-1.27%)43.47 (+75.24%)1.52 (-83.67%)47.68 (+3.51%)45.35 (+9.55%)45.19 (-1.27%)43.45 (+75.24%)1.52 (-83.67%)
84d3478 — 2026-02-17 23:25:1346.09 (n/a)41.43 (n/a)45.80 (n/a)24.81 (n/a)9.31 (n/a)46.06 (n/a)41.40 (n/a)45.77 (n/a)24.79 (n/a)9.30 (n/a)

matrix_vector_mul_2048x8192_1tsi_512tso_4col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
8b6d0f4 — 2026-03-25 20:46:3940.94 (-0.67%)39.64 (+0.12%)39.96 (+1.02%)37.79 (+0.51%)1.25 (-21.17%)40.91 (-0.67%)39.61 (+0.12%)39.94 (+1.02%)37.77 (+0.51%)1.25 (-21.17%)
dc66da7 — 2026-03-25 19:23:4541.21 (n/a)39.59 (n/a)39.56 (n/a)37.60 (n/a)1.58 (n/a)41.19 (n/a)39.57 (n/a)39.53 (n/a)37.58 (n/a)1.58 (n/a)

matrix_vector_mul_2048x8192_1tsi_512tso_4col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
897d04e — 2026-03-06 22:12:2540.84 (-0.32%)39.53 (-1.25%)39.40 (-1.97%)38.87 (-0.69%)0.77 (-6.24%)40.81 (-0.32%)39.50 (-1.25%)39.37 (-1.97%)38.84 (-0.69%)0.77 (-6.24%)
84d3478 — 2026-02-17 23:25:1340.97 (n/a)40.03 (n/a)40.19 (n/a)39.14 (n/a)0.82 (n/a)40.95 (n/a)40.00 (n/a)40.17 (n/a)39.11 (n/a)0.82 (n/a)

matrix_vector_mul_8192x2048_4_1col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)
130b6ea — 2025-12-05 21:15:4913.29 (+1.63%)13.19 (+6.08%)13.19 (+1.17%)13.09 (+26.86%)0.08 (-92.94%)
0a6c11c — 2025-12-04 00:39:1013.08 (n/a)12.44 (n/a)13.03 (n/a)10.32 (n/a)1.07 (n/a)

matrix_vector_mul_8192x2048_4_1col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
331dcca — 2026-01-08 18:34:1013.68 (+3.42%)13.30 (+8.54%)13.38 (+7.25%)12.72 (+14.96%)0.35 (-66.09%)13.67 (+3.42%)13.29 (+8.54%)13.38 (+7.25%)12.71 (+14.96%)0.35 (-66.09%)
a4b6ffe — 2026-01-05 19:24:0513.23 (n/a)12.25 (n/a)12.48 (n/a)11.07 (n/a)1.04 (n/a)13.22 (n/a)12.24 (n/a)12.47 (n/a)11.06 (n/a)1.04 (n/a)

matrix_vector_mul_8192x2048_4_2col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)
130b6ea — 2025-12-05 21:15:4925.58 (+4.24%)24.23 (+6.59%)24.03 (+0.29%)23.58 (+37.29%)0.70 (-74.97%)
0a6c11c — 2025-12-04 00:39:1024.54 (n/a)22.73 (n/a)23.96 (n/a)17.18 (n/a)2.79 (n/a)

matrix_vector_mul_8192x2048_4_2col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
331dcca — 2026-01-08 18:34:1025.08 (+1.32%)24.68 (+13.42%)24.61 (+3.26%)24.47 (+48.32%)0.24 (-93.56%)25.07 (+1.32%)24.67 (+13.42%)24.59 (+3.26%)24.46 (+48.32%)0.24 (-93.56%)
a4b6ffe — 2026-01-05 19:24:0524.76 (n/a)21.76 (n/a)23.83 (n/a)16.50 (n/a)3.74 (n/a)24.74 (n/a)21.75 (n/a)23.82 (n/a)16.49 (n/a)3.73 (n/a)

matrix_vector_mul_8192x2048_4_4col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)
130b6ea — 2025-12-05 21:15:4940.94 (-2.19%)39.74 (-1.10%)39.76 (-0.60%)38.17 (-2.14%)1.04 (+4.39%)
0a6c11c — 2025-12-04 00:39:1041.86 (n/a)40.18 (n/a)40.00 (n/a)39.00 (n/a)1.00 (n/a)

matrix_vector_mul_8192x2048_4_4col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
331dcca — 2026-01-08 18:34:1040.74 (-1.09%)37.16 (-7.15%)40.25 (+1.47%)24.51 (-37.44%)7.07 (+736.13%)40.72 (-1.09%)37.13 (-7.15%)40.23 (+1.47%)24.50 (-37.44%)7.07 (+736.13%)
a4b6ffe — 2026-01-05 19:24:0541.19 (n/a)40.02 (n/a)39.67 (n/a)39.18 (n/a)0.85 (n/a)41.17 (n/a)39.99 (n/a)39.64 (n/a)39.16 (n/a)0.85 (n/a)

matrix_vector_mul_8192x2048_4_8col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)
130b6ea — 2025-12-05 21:15:4946.71 (-1.04%)44.91 (-2.34%)44.71 (-2.98%)43.80 (-1.44%)0.96 (+6.93%)
0a6c11c — 2025-12-04 00:39:1047.20 (n/a)45.98 (n/a)46.08 (n/a)44.44 (n/a)0.90 (n/a)

matrix_vector_mul_8192x2048_4_8col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
331dcca — 2026-01-08 18:34:1046.28 (+5.68%)44.89 (+14.59%)44.74 (+4.26%)44.20 (+62.80%)0.84 (-87.96%)46.25 (+5.68%)44.86 (+14.59%)44.71 (+4.26%)44.17 (+62.80%)0.84 (-87.96%)
a4b6ffe — 2026-01-05 19:24:0543.79 (n/a)39.17 (n/a)42.91 (n/a)27.15 (n/a)6.98 (n/a)43.77 (n/a)39.15 (n/a)42.88 (n/a)27.13 (n/a)6.98 (n/a)

matrix_vector_mul_8192x2048_4tsi_1024tso_1col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
8b6d0f4 — 2026-03-25 20:46:3913.35 (+0.83%)13.11 (+3.52%)13.10 (+3.94%)12.94 (+7.70%)0.15 (-69.46%)13.34 (+0.83%)13.10 (+3.52%)13.09 (+3.94%)12.94 (+7.70%)0.15 (-69.46%)
dc66da7 — 2026-03-25 19:23:4513.24 (n/a)12.66 (n/a)12.60 (n/a)12.02 (n/a)0.49 (n/a)13.23 (n/a)12.66 (n/a)12.59 (n/a)12.01 (n/a)0.49 (n/a)

matrix_vector_mul_8192x2048_4tsi_1024tso_1col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
897d04e — 2026-03-06 22:12:2513.34 (-0.79%)13.20 (-0.36%)13.18 (-1.55%)13.12 (+1.75%)0.08 (-65.93%)13.33 (-0.79%)13.19 (-0.36%)13.17 (-1.55%)13.11 (+1.75%)0.08 (-65.93%)
84d3478 — 2026-02-17 23:25:1313.45 (n/a)13.25 (n/a)13.39 (n/a)12.90 (n/a)0.25 (n/a)13.44 (n/a)13.24 (n/a)13.38 (n/a)12.89 (n/a)0.25 (n/a)

matrix_vector_mul_8192x2048_4tsi_1024tso_2col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
8b6d0f4 — 2026-03-25 20:46:3924.35 (-0.26%)24.07 (+0.89%)24.16 (+1.48%)23.64 (+2.17%)0.27 (-44.88%)24.34 (-0.26%)24.06 (+0.89%)24.15 (+1.48%)23.63 (+2.17%)0.27 (-44.88%)
dc66da7 — 2026-03-25 19:23:4524.41 (n/a)23.86 (n/a)23.81 (n/a)23.14 (n/a)0.49 (n/a)24.40 (n/a)23.84 (n/a)23.80 (n/a)23.12 (n/a)0.49 (n/a)

matrix_vector_mul_8192x2048_4tsi_1024tso_2col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
897d04e — 2026-03-06 22:12:2524.10 (-5.45%)23.07 (-1.85%)24.01 (-2.88%)19.26 (+7.47%)2.13 (-32.15%)24.08 (-5.45%)23.05 (-1.85%)23.99 (-2.88%)19.25 (+7.47%)2.13 (-32.15%)
84d3478 — 2026-02-17 23:25:1325.48 (n/a)23.50 (n/a)24.72 (n/a)17.92 (n/a)3.14 (n/a)25.47 (n/a)23.48 (n/a)24.71 (n/a)17.91 (n/a)3.14 (n/a)

matrix_vector_mul_8192x2048_4tsi_1024tso_4col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
8b6d0f4 — 2026-03-25 20:46:3940.84 (-1.16%)38.86 (-0.66%)38.49 (-1.61%)37.90 (+1.94%)1.14 (-23.22%)40.82 (-1.16%)38.84 (-0.66%)38.46 (-1.61%)37.87 (+1.94%)1.14 (-23.22%)
dc66da7 — 2026-03-25 19:23:4541.32 (n/a)39.12 (n/a)39.12 (n/a)37.18 (n/a)1.49 (n/a)41.30 (n/a)39.09 (n/a)39.09 (n/a)37.15 (n/a)1.49 (n/a)

matrix_vector_mul_8192x2048_4tsi_1024tso_4col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
897d04e — 2026-03-06 22:12:2540.80 (-0.16%)39.58 (-1.93%)39.75 (-1.71%)38.28 (-2.50%)1.10 (+67.33%)40.78 (-0.16%)39.55 (-1.93%)39.73 (-1.71%)38.26 (-2.50%)1.10 (+67.33%)
84d3478 — 2026-02-17 23:25:1340.87 (n/a)40.36 (n/a)40.45 (n/a)39.26 (n/a)0.66 (n/a)40.84 (n/a)40.33 (n/a)40.42 (n/a)39.24 (n/a)0.65 (n/a)

matrix_vector_mul_8192x2048_4tsi_1024tso_8col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
8b6d0f4 — 2026-03-25 20:46:3941.79 (-12.64%)40.98 (-5.83%)40.78 (-1.04%)40.36 (-0.60%)0.56 (-84.11%)41.76 (-12.64%)40.96 (-5.83%)40.76 (-1.04%)40.34 (-0.60%)0.56 (-84.11%)
dc66da7 — 2026-03-25 19:23:4547.83 (n/a)43.52 (n/a)41.21 (n/a)40.61 (n/a)3.53 (n/a)47.80 (n/a)43.50 (n/a)41.19 (n/a)40.58 (n/a)3.53 (n/a)

matrix_vector_mul_8192x2048_4tsi_1024tso_8col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
897d04e — 2026-03-06 22:12:2549.31 (+6.27%)46.12 (+1.67%)45.30 (-0.27%)44.62 (+1.87%)1.85 (+85.10%)49.28 (+6.27%)46.10 (+1.67%)45.28 (-0.27%)44.59 (+1.87%)1.85 (+85.10%)
84d3478 — 2026-02-17 23:25:1346.40 (n/a)45.37 (n/a)45.43 (n/a)43.80 (n/a)1.00 (n/a)46.37 (n/a)45.34 (n/a)45.40 (n/a)43.78 (n/a)1.00 (n/a)

mem_copy_16_cores_2_chans_2048_tile_128_False

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.05 (+7.16%)0.03 (-9.22%)0.03 (-18.40%)0.02 (-19.93%)0.01 (+42.24%)394.80 (+24.90%)264.56 (+14.35%)259.20 (+22.55%)179.80 (-6.65%)81.04 (+64.09%)
dc66da7 — 2026-03-25 19:23:450.04 (n/a)0.04 (n/a)0.04 (n/a)0.03 (n/a)0.01 (n/a)316.10 (n/a)231.36 (n/a)211.50 (n/a)192.60 (n/a)49.39 (n/a)

mem_copy_16_cores_2_chans_2048_tile_128_False0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.08 (-2.12%)0.06 (-16.44%)0.06 (-21.48%)0.03 (-23.84%)0.02 (+10.15%)247.60 (+31.28%)148.38 (+24.08%)128.30 (+27.41%)100.10 (+2.14%)58.70 (+50.73%)
84d3478 — 2026-02-17 23:25:130.08 (n/a)0.07 (n/a)0.08 (n/a)0.04 (n/a)0.02 (n/a)188.60 (n/a)119.58 (n/a)100.70 (n/a)98.00 (n/a)38.94 (n/a)

mem_copy_1_cols_1_channels_2048_tile_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
003b49d — 2025-11-14 19:30:540.05 (-16.99%)0.05 (-12.47%)0.05 (-7.32%)0.04 (-11.17%)0.00 (-29.61%)197.00 (+12.57%)171.00 (+13.70%)164.00 (+7.89%)153.00 (+20.47%)17.49 (-4.32%)
0999a9a — 2025-11-13 19:24:080.06 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.01 (n/a)175.00 (n/a)150.40 (n/a)152.00 (n/a)127.00 (n/a)18.28 (n/a)

mem_copy_1_cols_2_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
003b49d — 2025-11-14 19:30:540.06 (+5.56%)0.05 (-1.59%)0.05 (-1.39%)0.03 (-14.07%)0.01 (+25.03%)263.00 (+16.37%)166.00 (+4.53%)144.00 (+1.41%)126.00 (-5.26%)49.38 (+43.17%)
0999a9a — 2025-11-13 19:24:080.06 (n/a)0.05 (n/a)0.05 (n/a)0.03 (n/a)0.01 (n/a)226.00 (n/a)158.80 (n/a)142.00 (n/a)133.00 (n/a)34.49 (n/a)

mem_copy_1_cores_1_chans_2048_tile_2048_False

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.09 (+75.06%)0.05 (+7.82%)0.04 (-15.32%)0.03 (-13.78%)0.02 (+425.35%)249.10 (+15.97%)190.02 (+3.98%)211.70 (+18.07%)95.60 (-42.89%)62.62 (+236.69%)
dc66da7 — 2026-03-25 19:23:450.05 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.00 (n/a)214.80 (n/a)182.74 (n/a)179.30 (n/a)167.40 (n/a)18.60 (n/a)

mem_copy_1_cores_1_chans_2048_tile_2048_False0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.11 (-0.24%)0.08 (-17.53%)0.08 (-27.05%)0.06 (-15.43%)0.02 (+4.62%)148.50 (+18.23%)113.02 (+23.38%)103.60 (+37.04%)74.10 (+0.14%)31.69 (+33.33%)
84d3478 — 2026-02-17 23:25:130.11 (n/a)0.09 (n/a)0.11 (n/a)0.07 (n/a)0.02 (n/a)125.60 (n/a)91.60 (n/a)75.60 (n/a)74.00 (n/a)23.77 (n/a)

mem_copy_2_cols_1_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
003b49d — 2025-11-14 19:30:540.06 (-15.32%)0.05 (-17.09%)0.05 (-23.81%)0.04 (-15.54%)0.01 (-18.18%)193.00 (+18.40%)160.80 (+20.36%)168.00 (+31.25%)124.00 (+18.10%)22.92 (+11.39%)
0999a9a — 2025-11-13 19:24:080.07 (n/a)0.06 (n/a)0.06 (n/a)0.05 (n/a)0.01 (n/a)163.00 (n/a)133.60 (n/a)128.00 (n/a)105.00 (n/a)20.58 (n/a)

mem_copy_2_cols_2_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
003b49d — 2025-11-14 19:30:540.06 (-23.02%)0.05 (-6.98%)0.05 (-2.13%)0.04 (-15.31%)0.01 (-33.29%)196.00 (+18.07%)147.00 (+6.52%)141.00 (+2.17%)126.00 (+29.90%)25.27 (+8.13%)
0999a9a — 2025-11-13 19:24:080.08 (n/a)0.06 (n/a)0.06 (n/a)0.05 (n/a)0.01 (n/a)166.00 (n/a)138.00 (n/a)138.00 (n/a)97.00 (n/a)23.37 (n/a)

mem_copy_2_cores_1_chans_2048_tile_1024_False

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.07 (+13.70%)0.05 (-1.93%)0.05 (-7.07%)0.04 (-27.15%)0.02 (+157.23%)230.10 (+37.29%)166.58 (+9.82%)161.90 (+7.65%)109.70 (-12.03%)53.34 (+211.05%)
dc66da7 — 2026-03-25 19:23:450.07 (n/a)0.05 (n/a)0.05 (n/a)0.05 (n/a)0.01 (n/a)167.60 (n/a)151.68 (n/a)150.40 (n/a)124.70 (n/a)17.15 (n/a)

mem_copy_2_cores_1_chans_2048_tile_1024_False0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.11 (+27.03%)0.09 (+25.79%)0.11 (+27.09%)0.05 (+12.74%)0.03 (+43.93%)161.90 (-11.29%)94.34 (-18.35%)76.30 (-21.34%)74.00 (-21.36%)38.01 (+0.49%)
84d3478 — 2026-02-17 23:25:130.09 (n/a)0.08 (n/a)0.08 (n/a)0.04 (n/a)0.02 (n/a)182.50 (n/a)115.54 (n/a)97.00 (n/a)94.10 (n/a)37.82 (n/a)

mem_copy_2_cores_2_chans_2048_tile_1024_False

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.07 (-3.17%)0.05 (-7.71%)0.05 (-6.65%)0.04 (-20.60%)0.02 (+42.64%)223.40 (+25.93%)171.62 (+13.04%)172.10 (+7.09%)120.40 (+3.26%)49.35 (+82.52%)
dc66da7 — 2026-03-25 19:23:450.07 (n/a)0.06 (n/a)0.05 (n/a)0.05 (n/a)0.01 (n/a)177.40 (n/a)151.82 (n/a)160.70 (n/a)116.60 (n/a)27.04 (n/a)

mem_copy_2_cores_2_chans_2048_tile_1024_False0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.10 (+2.70%)0.08 (-5.38%)0.09 (+2.10%)0.04 (-43.30%)0.02 (+170.50%)201.60 (+76.38%)112.08 (+16.53%)90.80 (-2.05%)84.60 (-2.65%)50.14 (+375.75%)
84d3478 — 2026-02-17 23:25:130.09 (n/a)0.09 (n/a)0.09 (n/a)0.07 (n/a)0.01 (n/a)114.30 (n/a)96.18 (n/a)92.70 (n/a)86.90 (n/a)10.54 (n/a)

mem_copy_4_cols_1_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
003b49d — 2025-11-14 19:30:540.07 (+11.97%)0.05 (+3.09%)0.05 (+2.10%)0.03 (-11.35%)0.01 (+41.65%)229.00 (+12.81%)156.00 (-0.51%)143.00 (-2.05%)117.00 (-10.69%)38.21 (+48.71%)
0999a9a — 2025-11-13 19:24:080.06 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.01 (n/a)203.00 (n/a)156.80 (n/a)146.00 (n/a)131.00 (n/a)25.69 (n/a)

mem_copy_4_cols_2_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
003b49d — 2025-11-14 19:30:540.06 (+9.02%)0.05 (+5.59%)0.05 (+0.00%)0.04 (+3.63%)0.01 (+25.48%)193.00 (-3.50%)147.40 (-4.66%)147.00 (+0.00%)122.00 (-8.27%)25.24 (+7.48%)
0999a9a — 2025-11-13 19:24:080.06 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.01 (n/a)200.00 (n/a)154.60 (n/a)147.00 (n/a)133.00 (n/a)23.48 (n/a)

mem_copy_4_cores_1_chans_2048_tile_512_False

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.06 (-2.79%)0.05 (-5.43%)0.05 (+6.39%)0.03 (-12.50%)0.01 (+16.26%)242.20 (+14.25%)184.14 (+7.91%)167.80 (-6.05%)128.30 (+2.89%)46.84 (+42.55%)
dc66da7 — 2026-03-25 19:23:450.07 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.01 (n/a)212.00 (n/a)170.64 (n/a)178.60 (n/a)124.70 (n/a)32.86 (n/a)

mem_copy_4_cores_1_chans_2048_tile_512_False0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.09 (-5.87%)0.08 (-3.61%)0.08 (-0.79%)0.06 (-12.74%)0.01 (+22.04%)146.30 (+14.66%)108.02 (+4.96%)99.30 (+0.81%)95.10 (+6.26%)21.49 (+48.38%)
84d3478 — 2026-02-17 23:25:130.09 (n/a)0.08 (n/a)0.08 (n/a)0.06 (n/a)0.01 (n/a)127.60 (n/a)102.92 (n/a)98.50 (n/a)89.50 (n/a)14.49 (n/a)

mem_copy_4_cores_2_chans_2048_tile_512_False

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.05 (+7.78%)0.04 (-14.27%)0.04 (-10.58%)0.03 (-32.73%)0.01 (+109.81%)316.90 (+48.71%)226.08 (+22.64%)209.60 (+11.85%)150.20 (-7.23%)61.97 (+194.31%)
dc66da7 — 2026-03-25 19:23:450.05 (n/a)0.04 (n/a)0.04 (n/a)0.04 (n/a)0.01 (n/a)213.10 (n/a)184.34 (n/a)187.40 (n/a)161.90 (n/a)21.05 (n/a)

mem_copy_4_cores_2_chans_2048_tile_512_False0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.11 (+13.04%)0.09 (+17.19%)0.10 (+20.43%)0.06 (+17.23%)0.02 (+16.34%)142.80 (-14.70%)91.94 (-14.71%)79.00 (-17.02%)77.70 (-11.50%)28.50 (-14.82%)
84d3478 — 2026-02-17 23:25:130.09 (n/a)0.08 (n/a)0.09 (n/a)0.05 (n/a)0.02 (n/a)167.40 (n/a)107.80 (n/a)95.20 (n/a)87.80 (n/a)33.46 (n/a)

mem_copy_8_cols_1_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
003b49d — 2025-11-14 19:30:540.05 (-13.48%)0.05 (-17.79%)0.05 (-15.33%)0.03 (-25.09%)0.01 (+6.67%)267.00 (+33.50%)176.80 (+24.51%)150.00 (+18.11%)141.00 (+15.57%)47.41 (+61.41%)
0999a9a — 2025-11-13 19:24:080.06 (n/a)0.06 (n/a)0.06 (n/a)0.04 (n/a)0.01 (n/a)200.00 (n/a)142.00 (n/a)127.00 (n/a)122.00 (n/a)29.37 (n/a)

mem_copy_8_cols_2_channels_2048_tile_128

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
003b49d — 2025-11-14 19:30:540.05 (+13.38%)0.05 (+23.85%)0.05 (+37.91%)0.05 (+30.72%)0.00 (-56.85%)166.00 (-23.50%)153.40 (-20.60%)153.00 (-27.49%)142.00 (-11.80%)7.63 (-70.60%)
0999a9a — 2025-11-13 19:24:080.05 (n/a)0.04 (n/a)0.04 (n/a)0.04 (n/a)0.01 (n/a)217.00 (n/a)193.20 (n/a)211.00 (n/a)161.00 (n/a)25.96 (n/a)

mem_copy_8_cores_1_chans_2048_tile_256_False

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.06 (-5.71%)0.04 (+0.46%)0.04 (+2.21%)0.03 (-7.12%)0.01 (-7.04%)278.00 (+7.67%)209.24 (-0.41%)216.20 (-2.17%)141.90 (+6.05%)51.53 (+9.47%)
dc66da7 — 2026-03-25 19:23:450.06 (n/a)0.04 (n/a)0.04 (n/a)0.03 (n/a)0.01 (n/a)258.20 (n/a)210.10 (n/a)221.00 (n/a)133.80 (n/a)47.07 (n/a)

mem_copy_8_cores_1_chans_2048_tile_256_False0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.08 (-0.39%)0.07 (+10.07%)0.08 (+1.07%)0.06 (+38.38%)0.01 (-47.35%)136.70 (-27.75%)110.64 (-13.07%)104.70 (-1.13%)103.00 (+0.39%)14.59 (-60.78%)
84d3478 — 2026-02-17 23:25:130.08 (n/a)0.07 (n/a)0.08 (n/a)0.04 (n/a)0.02 (n/a)189.20 (n/a)127.28 (n/a)105.90 (n/a)102.60 (n/a)37.20 (n/a)

mem_copy_8_cores_2_chans_2048_tile_256_False

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.06 (+19.11%)0.04 (+1.88%)0.04 (-11.30%)0.03 (+24.20%)0.01 (+4.86%)262.00 (-19.48%)199.56 (-3.59%)200.50 (+12.70%)139.70 (-16.05%)45.93 (-31.63%)
dc66da7 — 2026-03-25 19:23:450.05 (n/a)0.04 (n/a)0.05 (n/a)0.03 (n/a)0.01 (n/a)325.40 (n/a)207.00 (n/a)177.90 (n/a)166.40 (n/a)67.19 (n/a)

mem_copy_8_cores_2_chans_2048_tile_256_False0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.08 (-17.85%)0.07 (-24.66%)0.08 (-18.40%)0.04 (-32.08%)0.02 (+21.78%)185.30 (+47.18%)127.72 (+37.87%)102.00 (+22.45%)100.40 (+21.70%)38.66 (+106.34%)
84d3478 — 2026-02-17 23:25:130.10 (n/a)0.09 (n/a)0.10 (n/a)0.07 (n/a)0.01 (n/a)125.90 (n/a)92.64 (n/a)83.30 (n/a)82.50 (n/a)18.74 (n/a)

mha

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.18 (+0.10%)0.18 (+0.09%)0.18 (+0.09%)0.18 (+0.08%)0.00 (+11.08%)47507.50 (-0.08%)47468.96 (-0.09%)47481.10 (-0.09%)47413.10 (-0.10%)37.32 (+10.96%)
dc66da7 — 2026-03-25 19:23:450.18 (n/a)0.18 (n/a)0.18 (n/a)0.18 (n/a)0.00 (n/a)47544.20 (n/a)47512.60 (n/a)47521.80 (n/a)47459.60 (n/a)33.63 (n/a)

mha0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
84d3478 — 2026-02-17 23:25:130.18 (-0.51%)0.18 (-0.36%)0.18 (-0.41%)0.18 (-0.21%)0.00 (-35.21%)47468.60 (+0.21%)47403.94 (+0.36%)47426.00 (+0.41%)47314.40 (+0.51%)68.83 (-34.73%)
1531781 — 2026-01-23 18:49:240.18 (n/a)0.18 (n/a)0.18 (n/a)0.18 (n/a)0.00 (n/a)47367.50 (n/a)47232.68 (n/a)47232.00 (n/a)47073.00 (n/a)105.45 (n/a)

mha_16384_64_1_8_0_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.18 (n/a)0.18 (n/a)0.18 (n/a)0.18 (n/a)0.00 (n/a)47533.70 (n/a)47380.56 (n/a)47337.90 (n/a)47268.00 (n/a)110.60 (n/a)

relu_1_cols_1_channels_2048_tile_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.07 (+14.06%)0.05 (+0.31%)0.05 (-3.71%)0.04 (-18.38%)0.01 (+129.69%)208.90 (+22.52%)161.96 (+2.56%)166.90 (+3.86%)120.40 (-12.31%)33.36 (+144.00%)
dc66da7 — 2026-03-25 19:23:450.06 (n/a)0.05 (n/a)0.05 (n/a)0.05 (n/a)0.00 (n/a)170.50 (n/a)157.92 (n/a)160.70 (n/a)137.30 (n/a)13.67 (n/a)

relu_2_cols_1_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.06 (-20.92%)0.05 (-6.86%)0.05 (-1.28%)0.03 (-20.49%)0.01 (-19.18%)264.20 (+25.81%)174.76 (+7.41%)149.00 (+1.29%)140.20 (+26.42%)52.29 (+24.11%)
dc66da7 — 2026-03-25 19:23:450.07 (n/a)0.05 (n/a)0.06 (n/a)0.04 (n/a)0.01 (n/a)210.00 (n/a)162.70 (n/a)147.10 (n/a)110.90 (n/a)42.13 (n/a)

relu_4_cols_1_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.05 (-8.73%)0.04 (-10.26%)0.04 (-4.04%)0.03 (-16.73%)0.01 (+5.34%)259.40 (+20.09%)206.80 (+12.91%)200.50 (+4.21%)150.20 (+9.56%)45.96 (+41.47%)
dc66da7 — 2026-03-25 19:23:450.06 (n/a)0.05 (n/a)0.04 (n/a)0.04 (n/a)0.01 (n/a)216.00 (n/a)183.16 (n/a)192.40 (n/a)137.10 (n/a)32.49 (n/a)

relu_8_cols_1_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.05 (+4.18%)0.04 (-0.33%)0.04 (-1.55%)0.02 (-20.26%)0.01 (+47.86%)369.50 (+25.38%)233.74 (+5.19%)215.30 (+1.56%)165.20 (-4.01%)80.83 (+78.85%)
dc66da7 — 2026-03-25 19:23:450.05 (n/a)0.04 (n/a)0.04 (n/a)0.03 (n/a)0.01 (n/a)294.70 (n/a)222.20 (n/a)212.00 (n/a)172.10 (n/a)45.19 (n/a)

rms_norm_1_cols_1_channels_2048_tile_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.06 (-18.06%)0.05 (-2.61%)0.05 (-1.31%)0.04 (+12.13%)0.01 (-43.39%)226.00 (-10.81%)172.66 (-1.55%)163.30 (+1.30%)140.00 (+22.06%)32.37 (-37.02%)
dc66da7 — 2026-03-25 19:23:450.07 (n/a)0.05 (n/a)0.05 (n/a)0.03 (n/a)0.01 (n/a)253.40 (n/a)175.38 (n/a)161.20 (n/a)114.70 (n/a)51.40 (n/a)

rms_norm_1_cols_2_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.08 (+16.34%)0.05 (+6.02%)0.05 (-3.17%)0.04 (+12.01%)0.02 (+27.81%)218.40 (-10.71%)162.92 (-3.80%)150.40 (+3.23%)108.00 (-14.01%)52.41 (+3.43%)
dc66da7 — 2026-03-25 19:23:450.07 (n/a)0.05 (n/a)0.06 (n/a)0.03 (n/a)0.01 (n/a)244.60 (n/a)169.36 (n/a)145.70 (n/a)125.60 (n/a)50.68 (n/a)

rms_norm_2_cols_1_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.06 (+9.52%)0.04 (-9.59%)0.04 (-17.77%)0.03 (-18.95%)0.01 (+55.38%)255.30 (+23.39%)199.88 (+13.52%)205.30 (+21.62%)137.80 (-8.68%)43.62 (+70.31%)
dc66da7 — 2026-03-25 19:23:450.05 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.01 (n/a)206.90 (n/a)176.08 (n/a)168.80 (n/a)150.90 (n/a)25.61 (n/a)

rms_norm_2_cols_2_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.06 (+6.05%)0.04 (+0.61%)0.04 (-6.97%)0.03 (-2.64%)0.01 (+44.42%)234.40 (+2.72%)195.80 (+1.42%)215.50 (+7.48%)141.90 (-5.65%)40.78 (+43.76%)
dc66da7 — 2026-03-25 19:23:450.05 (n/a)0.04 (n/a)0.04 (n/a)0.04 (n/a)0.01 (n/a)228.20 (n/a)193.06 (n/a)200.50 (n/a)150.40 (n/a)28.37 (n/a)

rms_norm_4_cols_1_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.06 (+10.73%)0.05 (-1.13%)0.05 (+5.17%)0.04 (-10.70%)0.01 (+73.44%)225.10 (+11.99%)181.84 (+3.00%)170.50 (-4.91%)140.00 (-9.68%)33.21 (+78.26%)
dc66da7 — 2026-03-25 19:23:450.05 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.00 (n/a)201.00 (n/a)176.54 (n/a)179.30 (n/a)155.00 (n/a)18.63 (n/a)

rms_norm_4_cols_2_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.06 (-0.51%)0.04 (-4.22%)0.04 (-6.40%)0.03 (+19.93%)0.01 (-12.26%)264.40 (-16.62%)208.96 (+1.64%)223.40 (+6.84%)140.70 (+0.50%)52.99 (-24.83%)
dc66da7 — 2026-03-25 19:23:450.06 (n/a)0.04 (n/a)0.04 (n/a)0.03 (n/a)0.01 (n/a)317.10 (n/a)205.58 (n/a)209.10 (n/a)140.00 (n/a)70.49 (n/a)

rms_norm_8_cols_1_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.06 (-8.19%)0.04 (-16.80%)0.03 (-25.89%)0.03 (-16.40%)0.01 (+5.02%)283.50 (+19.62%)220.72 (+22.74%)240.30 (+34.92%)128.00 (+8.94%)58.39 (+33.24%)
dc66da7 — 2026-03-25 19:23:450.07 (n/a)0.05 (n/a)0.05 (n/a)0.03 (n/a)0.01 (n/a)237.00 (n/a)179.82 (n/a)178.10 (n/a)117.50 (n/a)43.82 (n/a)

rms_norm_8_cols_2_channels_2048_tile_128

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.04 (+4.11%)0.04 (+2.57%)0.04 (-1.84%)0.03 (+8.94%)0.00 (+2.64%)250.80 (-8.20%)224.64 (-2.58%)233.20 (+1.88%)191.20 (-3.97%)25.83 (-9.64%)
dc66da7 — 2026-03-25 19:23:450.04 (n/a)0.04 (n/a)0.04 (n/a)0.03 (n/a)0.00 (n/a)273.20 (n/a)230.60 (n/a)228.90 (n/a)199.10 (n/a)28.59 (n/a)

rope_1_cols_2_channels_4096_tile_4096_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
1b2d230 — 2026-01-13 22:38:030.19 (-21.66%)0.17 (-8.53%)0.18 (+2.09%)0.14 (-6.71%)0.02 (-38.07%)174.10 (+7.20%)150.06 (+8.11%)139.60 (-2.04%)129.80 (+27.63%)20.36 (-10.65%)
d48746f — 2026-01-12 22:17:100.24 (n/a)0.18 (n/a)0.17 (n/a)0.15 (n/a)0.04 (n/a)162.40 (n/a)138.80 (n/a)142.50 (n/a)101.70 (n/a)22.79 (n/a)

rope_1c_32rows_512cols_32arows_0m

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.91 (+31.28%)0.72 (+30.97%)0.75 (+50.09%)0.52 (+23.14%)0.16 (+34.35%)189.50 (-18.81%)142.68 (-23.26%)131.80 (-33.40%)108.20 (-23.86%)33.21 (-13.83%)
dc66da7 — 2026-03-25 19:23:450.69 (n/a)0.55 (n/a)0.50 (n/a)0.42 (n/a)0.12 (n/a)233.40 (n/a)185.92 (n/a)197.90 (n/a)142.10 (n/a)38.54 (n/a)

rope_1c_32rows_512cols_8arows_0m

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.71 (+13.56%)0.53 (+16.33%)0.51 (+20.62%)0.43 (+13.26%)0.12 (+12.08%)172.60 (-11.71%)143.86 (-14.09%)145.70 (-17.07%)103.70 (-11.97%)28.08 (-12.19%)
dc66da7 — 2026-03-25 19:23:450.63 (n/a)0.46 (n/a)0.42 (n/a)0.38 (n/a)0.10 (n/a)195.50 (n/a)167.46 (n/a)175.70 (n/a)117.80 (n/a)31.98 (n/a)

rope_2_cols_2_channels_4096_tile_2048_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
1b2d230 — 2026-01-13 22:38:030.21 (-0.56%)0.17 (-10.62%)0.16 (-13.58%)0.13 (-13.33%)0.03 (+17.05%)183.70 (+15.39%)151.76 (+12.77%)155.20 (+15.73%)117.30 (+0.60%)24.24 (+35.73%)
d48746f — 2026-01-12 22:17:100.21 (n/a)0.19 (n/a)0.18 (n/a)0.15 (n/a)0.02 (n/a)159.20 (n/a)134.58 (n/a)134.10 (n/a)116.60 (n/a)17.86 (n/a)

rope_2c_32rows_512cols_32arows_0m

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.86 (+12.27%)0.68 (+4.35%)0.67 (-1.63%)0.56 (+6.41%)0.12 (+18.02%)174.80 (-6.02%)146.78 (-3.85%)146.60 (+1.66%)114.70 (-10.88%)23.97 (-1.25%)
dc66da7 — 2026-03-25 19:23:450.76 (n/a)0.66 (n/a)0.68 (n/a)0.53 (n/a)0.10 (n/a)186.00 (n/a)152.66 (n/a)144.20 (n/a)128.70 (n/a)24.27 (n/a)

rope_2c_32rows_512cols_8arows_0m

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.63 (+15.38%)0.42 (-5.03%)0.37 (-15.06%)0.26 (-23.68%)0.14 (+76.55%)289.00 (+31.07%)192.28 (+12.08%)197.90 (+17.73%)117.80 (-13.32%)64.38 (+98.12%)
dc66da7 — 2026-03-25 19:23:450.54 (n/a)0.44 (n/a)0.44 (n/a)0.33 (n/a)0.08 (n/a)220.50 (n/a)171.56 (n/a)168.10 (n/a)135.90 (n/a)32.49 (n/a)

rope_4_cols_2_channels_4096_tile_1024_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
1b2d230 — 2026-01-13 22:38:030.22 (+18.92%)0.17 (+6.24%)0.17 (+5.57%)0.14 (-0.50%)0.03 (+90.77%)171.10 (+0.47%)147.40 (-4.39%)148.10 (-5.31%)110.10 (-15.89%)23.59 (+60.67%)
d48746f — 2026-01-12 22:17:100.19 (n/a)0.16 (n/a)0.16 (n/a)0.14 (n/a)0.02 (n/a)170.30 (n/a)154.16 (n/a)156.40 (n/a)130.90 (n/a)14.69 (n/a)

rope_8_cols_2_channels_4096_tile_512_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
1b2d230 — 2026-01-13 22:38:030.17 (-1.30%)0.15 (+6.63%)0.15 (-8.78%)0.14 (+39.68%)0.01 (-67.35%)174.90 (-28.41%)163.42 (-10.36%)167.40 (+9.63%)146.50 (+1.31%)11.19 (-76.17%)
d48746f — 2026-01-12 22:17:100.17 (n/a)0.14 (n/a)0.16 (n/a)0.10 (n/a)0.03 (n/a)244.30 (n/a)182.30 (n/a)152.70 (n/a)144.60 (n/a)46.98 (n/a)

rope_8c_32rows_512cols_32arows_0m

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.60 (+0.58%)0.47 (-10.94%)0.47 (-14.67%)0.39 (-14.06%)0.09 (+57.47%)252.70 (+16.34%)213.46 (+14.31%)211.20 (+17.20%)163.30 (-0.61%)38.58 (+85.72%)
dc66da7 — 2026-03-25 19:23:450.60 (n/a)0.53 (n/a)0.55 (n/a)0.45 (n/a)0.06 (n/a)217.20 (n/a)186.74 (n/a)180.20 (n/a)164.30 (n/a)20.77 (n/a)

rope_8c_32rows_512cols_8arows_0m

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.58 (+33.33%)0.38 (-2.44%)0.32 (-20.46%)0.28 (-18.27%)0.13 (+203.34%)264.90 (+22.36%)209.28 (+9.84%)227.20 (+25.73%)126.60 (-25.00%)59.71 (+180.42%)
dc66da7 — 2026-03-25 19:23:450.44 (n/a)0.39 (n/a)0.41 (n/a)0.34 (n/a)0.04 (n/a)216.50 (n/a)190.54 (n/a)180.70 (n/a)168.80 (n/a)21.29 (n/a)

rows_32-cols_512-angle_rows_32-aie_columns_1-method_type_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8c138a8 — 2026-03-26 14:54:200.40 (n/a)0.35 (n/a)0.38 (n/a)0.23 (n/a)0.07 (n/a)420.60 (n/a)288.74 (n/a)256.00 (n/a)247.70 (n/a)74.07 (n/a)

rows_32-cols_512-angle_rows_32-aie_columns_2-method_type_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8c138a8 — 2026-03-26 14:54:200.39 (n/a)0.30 (n/a)0.34 (n/a)0.20 (n/a)0.08 (n/a)482.20 (n/a)348.02 (n/a)289.40 (n/a)249.50 (n/a)102.34 (n/a)

rows_32-cols_512-angle_rows_32-aie_columns_8-method_type_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8c138a8 — 2026-03-26 14:54:200.41 (n/a)0.28 (n/a)0.27 (n/a)0.13 (n/a)0.12 (n/a)742.30 (n/a)415.94 (n/a)364.80 (n/a)241.70 (n/a)204.69 (n/a)

rows_32-cols_512-angle_rows_8-aie_columns_1-method_type_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8c138a8 — 2026-03-26 14:54:200.30 (n/a)0.24 (n/a)0.24 (n/a)0.20 (n/a)0.04 (n/a)375.30 (n/a)306.66 (n/a)304.80 (n/a)244.40 (n/a)46.57 (n/a)

rows_32-cols_512-angle_rows_8-aie_columns_2-method_type_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8c138a8 — 2026-03-26 14:54:200.27 (n/a)0.24 (n/a)0.24 (n/a)0.20 (n/a)0.03 (n/a)368.50 (n/a)315.36 (n/a)313.20 (n/a)269.90 (n/a)35.05 (n/a)

rows_32-cols_512-angle_rows_8-aie_columns_8-method_type_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8c138a8 — 2026-03-26 14:54:200.27 (n/a)0.18 (n/a)0.23 (n/a)0.08 (n/a)0.09 (n/a)884.50 (n/a)504.40 (n/a)321.00 (n/a)273.30 (n/a)282.98 (n/a)

seq_len_16384-dim_64-num_heads_1-num_pipelines_8-num_kv_heads_0

No metrics available.

seq_len_256-embedding_dim_2048-hidden_dim_2048-prio_accuracy_False

No metrics available.

sigmoid_1_cols_1_channels_2048_tile_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.05 (-14.16%)0.05 (-6.29%)0.05 (-2.65%)0.04 (-7.71%)0.01 (-21.79%)225.80 (+8.35%)180.84 (+6.33%)170.90 (+2.70%)163.30 (+16.48%)25.41 (+0.55%)
dc66da7 — 2026-03-25 19:23:450.06 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.01 (n/a)208.40 (n/a)170.08 (n/a)166.40 (n/a)140.20 (n/a)25.28 (n/a)

sigmoid_2_cols_1_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.06 (+4.68%)0.05 (+2.94%)0.04 (-13.06%)0.04 (+15.35%)0.01 (+8.49%)232.90 (-13.32%)179.44 (-3.16%)197.20 (+15.05%)127.30 (-4.50%)46.58 (-13.93%)
dc66da7 — 2026-03-25 19:23:450.06 (n/a)0.05 (n/a)0.05 (n/a)0.03 (n/a)0.01 (n/a)268.70 (n/a)185.30 (n/a)171.40 (n/a)133.30 (n/a)54.12 (n/a)

sigmoid_4_cols_1_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.07 (+10.37%)0.05 (-2.49%)0.04 (-7.15%)0.04 (-7.47%)0.01 (+64.06%)210.80 (+8.10%)175.44 (+5.70%)183.30 (+7.70%)117.30 (-9.42%)39.46 (+67.78%)
dc66da7 — 2026-03-25 19:23:450.06 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.01 (n/a)195.00 (n/a)165.98 (n/a)170.20 (n/a)129.50 (n/a)23.52 (n/a)

sigmoid_8_cols_1_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.06 (+29.56%)0.05 (+22.81%)0.05 (+37.30%)0.03 (-8.45%)0.01 (+218.31%)242.70 (+9.23%)173.04 (-14.71%)150.20 (-27.16%)134.70 (-22.81%)46.84 (+169.71%)
dc66da7 — 2026-03-25 19:23:450.05 (n/a)0.04 (n/a)0.04 (n/a)0.04 (n/a)0.00 (n/a)222.20 (n/a)202.88 (n/a)206.20 (n/a)174.50 (n/a)17.37 (n/a)

silu_1_cols_1_channels_2048_tile_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.07 (+9.38%)0.05 (-4.51%)0.04 (-23.77%)0.04 (+15.02%)0.01 (+5.83%)207.90 (-13.05%)169.28 (+4.04%)188.60 (+31.15%)114.40 (-8.63%)38.46 (-17.21%)
dc66da7 — 2026-03-25 19:23:450.07 (n/a)0.05 (n/a)0.06 (n/a)0.03 (n/a)0.01 (n/a)239.10 (n/a)162.70 (n/a)143.80 (n/a)125.20 (n/a)46.46 (n/a)

silu_2_cols_1_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.06 (-4.63%)0.05 (-13.54%)0.05 (-12.20%)0.04 (-16.50%)0.01 (+19.45%)213.90 (+19.76%)180.30 (+16.99%)181.70 (+13.92%)134.00 (+4.85%)30.84 (+49.10%)
dc66da7 — 2026-03-25 19:23:450.06 (n/a)0.05 (n/a)0.05 (n/a)0.05 (n/a)0.01 (n/a)178.60 (n/a)154.12 (n/a)159.50 (n/a)127.80 (n/a)20.68 (n/a)

silu_4_cols_1_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.06 (+6.26%)0.05 (-8.81%)0.04 (-14.86%)0.04 (-17.09%)0.01 (+81.20%)214.80 (+20.61%)176.86 (+11.68%)182.90 (+17.47%)133.30 (-5.86%)30.25 (+101.37%)
dc66da7 — 2026-03-25 19:23:450.06 (n/a)0.05 (n/a)0.05 (n/a)0.05 (n/a)0.00 (n/a)178.10 (n/a)158.36 (n/a)155.70 (n/a)141.60 (n/a)15.02 (n/a)

silu_8_cols_1_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.06 (-6.11%)0.05 (-10.71%)0.04 (-10.38%)0.04 (-11.78%)0.01 (+1.86%)224.60 (+13.38%)185.30 (+12.62%)186.00 (+11.58%)140.40 (+6.53%)33.11 (+23.38%)
dc66da7 — 2026-03-25 19:23:450.06 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.01 (n/a)198.10 (n/a)164.54 (n/a)166.70 (n/a)131.80 (n/a)26.84 (n/a)

softmax_1_cols_2_channels_4096_tile_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.19 (-8.94%)0.17 (-8.66%)0.17 (-19.83%)0.14 (+4.32%)0.02 (-35.13%)115.80 (-4.14%)99.26 (+7.47%)98.60 (+24.65%)85.00 (+9.82%)13.65 (-30.86%)
84d3478 — 2026-02-17 23:25:130.21 (n/a)0.18 (n/a)0.21 (n/a)0.14 (n/a)0.04 (n/a)120.80 (n/a)92.36 (n/a)79.10 (n/a)77.40 (n/a)19.75 (n/a)

softmax_2_cols_2_channels_32768_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.64 (-42.09%)0.60 (-31.83%)0.59 (-30.19%)0.54 (-6.68%)0.05 (-76.96%)242.70 (+7.15%)221.12 (+40.50%)220.30 (+43.24%)203.40 (+72.67%)16.99 (-59.06%)
dc66da7 — 2026-03-25 19:23:451.11 (n/a)0.87 (n/a)0.85 (n/a)0.58 (n/a)0.20 (n/a)226.50 (n/a)157.38 (n/a)153.80 (n/a)117.80 (n/a)41.50 (n/a)

softmax_2_cols_2_channels_32768_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.68 (-34.93%)0.62 (-20.25%)0.60 (-17.49%)0.55 (-10.32%)0.05 (-71.29%)240.30 (+11.51%)214.32 (+20.87%)218.20 (+21.22%)193.80 (+53.69%)18.64 (-52.51%)
dc66da7 — 2026-03-25 19:23:451.04 (n/a)0.77 (n/a)0.73 (n/a)0.61 (n/a)0.18 (n/a)215.50 (n/a)177.32 (n/a)180.00 (n/a)126.10 (n/a)39.26 (n/a)

softmax_2_cols_2_channels_4096_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.20 (-0.87%)0.16 (+6.57%)0.17 (-9.55%)0.11 (+29.81%)0.03 (-44.34%)145.70 (-22.99%)104.68 (-15.00%)95.00 (+10.59%)83.80 (+0.84%)24.16 (-54.23%)
84d3478 — 2026-02-17 23:25:130.20 (n/a)0.15 (n/a)0.19 (n/a)0.09 (n/a)0.06 (n/a)189.20 (n/a)123.16 (n/a)85.90 (n/a)83.10 (n/a)52.80 (n/a)

softmax_2_cols_2_channels_4096_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.25 (+23.53%)0.18 (+9.42%)0.18 (-3.86%)0.11 (+2.99%)0.05 (+19.24%)144.10 (-2.90%)98.98 (-8.18%)92.10 (+4.07%)64.30 (-19.02%)29.58 (-7.13%)
84d3478 — 2026-02-17 23:25:130.21 (n/a)0.16 (n/a)0.19 (n/a)0.11 (n/a)0.04 (n/a)148.40 (n/a)107.80 (n/a)88.50 (n/a)79.40 (n/a)31.85 (n/a)

softmax_4_cols_4_channels_32768_tile_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.66 (+11.72%)0.50 (-0.51%)0.54 (-4.72%)0.35 (+6.44%)0.13 (+14.83%)377.70 (-6.04%)276.32 (+1.24%)242.20 (+4.94%)199.30 (-10.51%)76.50 (-0.31%)
dc66da7 — 2026-03-25 19:23:450.59 (n/a)0.51 (n/a)0.57 (n/a)0.33 (n/a)0.11 (n/a)402.00 (n/a)272.94 (n/a)230.80 (n/a)222.70 (n/a)76.73 (n/a)

swiglu

No metrics available.

swiglu_decode_1x2048x2048

No metrics available.

swiglu_decode_1x2048x2048_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.00 (-0.21%)0.00 (+7.74%)0.00 (+0.37%)0.00 (+8.84%)0.00 (-19.58%)5170.80 (-8.12%)3761.80 (-8.70%)3408.60 (-0.36%)3407.30 (+0.21%)787.66 (-22.98%)
84d3478 — 2026-02-17 23:25:130.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)5627.70 (n/a)4120.44 (n/a)3421.00 (n/a)3400.30 (n/a)1022.61 (n/a)

swiglu_prefill_256x2048x2048

No metrics available.

tanh_1_cols_1_channels_2048_tile_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.06 (-7.39%)0.04 (-16.97%)0.04 (-14.92%)0.03 (-28.80%)0.01 (+23.25%)296.40 (+40.47%)201.28 (+24.69%)188.60 (+17.51%)141.90 (+7.99%)59.45 (+90.30%)
dc66da7 — 2026-03-25 19:23:450.06 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.01 (n/a)211.00 (n/a)161.42 (n/a)160.50 (n/a)131.40 (n/a)31.24 (n/a)

tanh_2_cols_1_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.06 (-14.29%)0.05 (-6.63%)0.05 (+2.99%)0.03 (-30.73%)0.01 (+5.83%)293.30 (+44.41%)176.82 (+11.67%)159.70 (-2.92%)128.50 (+16.71%)67.26 (+86.47%)
dc66da7 — 2026-03-25 19:23:450.07 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.01 (n/a)203.10 (n/a)158.34 (n/a)164.50 (n/a)110.10 (n/a)36.07 (n/a)

tanh_4_cols_1_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.06 (-27.96%)0.05 (-4.84%)0.05 (+4.02%)0.04 (+16.65%)0.01 (-63.57%)187.60 (-14.30%)167.32 (-1.72%)171.90 (-3.86%)133.00 (+38.83%)20.69 (-55.09%)
dc66da7 — 2026-03-25 19:23:450.09 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.02 (n/a)218.90 (n/a)170.24 (n/a)178.80 (n/a)95.80 (n/a)46.06 (n/a)

tanh_8_cols_1_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.05 (-24.07%)0.04 (-15.48%)0.04 (-23.02%)0.03 (-3.09%)0.01 (-53.59%)262.30 (+3.19%)198.18 (+12.63%)183.30 (+29.91%)173.30 (+31.69%)36.55 (-34.15%)
dc66da7 — 2026-03-25 19:23:450.06 (n/a)0.05 (n/a)0.06 (n/a)0.03 (n/a)0.01 (n/a)254.20 (n/a)175.96 (n/a)141.10 (n/a)131.60 (n/a)55.51 (n/a)

transpose_2048_M_64_N_1_cols_1_channels_64_m_64_n_8_s

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:392.51 (-25.94%)2.10 (-20.30%)2.01 (-23.11%)1.76 (-13.49%)0.29 (-41.75%)298.70 (+15.60%)253.30 (+23.92%)261.10 (+30.09%)208.60 (+35.02%)33.97 (-9.04%)
dc66da7 — 2026-03-25 19:23:453.39 (n/a)2.64 (n/a)2.61 (n/a)2.03 (n/a)0.49 (n/a)258.40 (n/a)204.40 (n/a)200.70 (n/a)154.50 (n/a)37.34 (n/a)

transpose_2048_M_64_N_1_cols_1_channels_64_m_64_n_8_s0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:254.43 (+0.85%)3.73 (+3.83%)4.02 (-1.30%)2.97 (+16.77%)0.65 (-24.99%)176.70 (-14.35%)144.46 (-6.24%)130.40 (+1.32%)118.40 (-0.84%)26.60 (-35.16%)
84d3478 — 2026-02-17 23:25:134.39 (n/a)3.59 (n/a)4.07 (n/a)2.54 (n/a)0.87 (n/a)206.30 (n/a)154.08 (n/a)128.70 (n/a)119.40 (n/a)41.02 (n/a)

transpose_2048_M_64_N_1_cols_2_channels_64_m_64_n_8_s

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:392.72 (-10.04%)2.30 (-13.01%)2.26 (-13.66%)1.88 (-21.25%)0.34 (+46.90%)278.20 (+26.97%)231.60 (+16.34%)232.20 (+15.81%)192.40 (+11.15%)34.64 (+109.20%)
dc66da7 — 2026-03-25 19:23:453.03 (n/a)2.65 (n/a)2.61 (n/a)2.39 (n/a)0.23 (n/a)219.10 (n/a)199.08 (n/a)200.50 (n/a)173.10 (n/a)16.56 (n/a)

transpose_2048_M_64_N_1_cols_2_channels_64_m_64_n_8_s0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:255.27 (-14.18%)4.35 (-2.49%)4.45 (+2.15%)2.78 (-22.18%)0.98 (-3.38%)188.60 (+28.47%)126.98 (+4.20%)117.90 (-2.08%)99.40 (+16.53%)36.02 (+50.15%)
84d3478 — 2026-02-17 23:25:136.14 (n/a)4.46 (n/a)4.35 (n/a)3.57 (n/a)1.02 (n/a)146.80 (n/a)121.86 (n/a)120.40 (n/a)85.30 (n/a)23.99 (n/a)

weighted_rms_norm_1_cols_2_channels_2048_weights_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.09 (+6.38%)0.07 (-4.87%)0.07 (+1.68%)0.05 (-23.45%)0.01 (+101.20%)230.80 (+30.62%)177.68 (+7.40%)170.00 (-1.68%)138.30 (-5.98%)33.65 (+149.29%)
dc66da7 — 2026-03-25 19:23:450.08 (n/a)0.07 (n/a)0.07 (n/a)0.07 (n/a)0.01 (n/a)176.70 (n/a)165.44 (n/a)172.90 (n/a)147.10 (n/a)13.50 (n/a)

weighted_rms_norm_2_cols_2_channels_2048_weights_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.07 (-25.98%)0.06 (-18.74%)0.06 (-7.04%)0.04 (-20.93%)0.01 (-34.18%)237.00 (+26.47%)189.06 (+22.07%)179.50 (+7.55%)151.40 (+35.06%)34.27 (+14.91%)
dc66da7 — 2026-03-25 19:23:450.09 (n/a)0.07 (n/a)0.06 (n/a)0.05 (n/a)0.01 (n/a)187.40 (n/a)154.88 (n/a)166.90 (n/a)112.10 (n/a)29.83 (n/a)

weighted_rms_norm_4_cols_2_channels_2048_weights_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.06 (-13.39%)0.05 (-21.43%)0.04 (-31.53%)0.04 (-12.12%)0.01 (-0.80%)237.90 (+13.77%)205.32 (+28.15%)225.30 (+46.01%)147.80 (+15.47%)38.15 (+27.36%)
dc66da7 — 2026-03-25 19:23:450.07 (n/a)0.06 (n/a)0.06 (n/a)0.04 (n/a)0.01 (n/a)209.10 (n/a)160.22 (n/a)154.30 (n/a)128.00 (n/a)29.95 (n/a)

weighted_rms_norm_8_cols_2_channels_2048_weights_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.07 (+33.02%)0.04 (-11.48%)0.04 (-24.93%)0.02 (-32.99%)0.02 (+124.58%)371.50 (+49.26%)251.20 (+24.91%)247.70 (+33.17%)124.90 (-24.85%)88.76 (+136.07%)
dc66da7 — 2026-03-25 19:23:450.05 (n/a)0.04 (n/a)0.05 (n/a)0.03 (n/a)0.01 (n/a)248.90 (n/a)201.10 (n/a)186.00 (n/a)166.20 (n/a)37.60 (n/a)

@github-actions
Copy link
Contributor

📊 Test Results for Test Example Applications

14eee74 (2026_03_26_15_11_20)

IRONCLAD

Tested on 2026_03_26_15_11_20 at commit 14eee74.

Test Checks TTFT (mean)TPS (mean)
llama_3.2_1b_prompt_1024_tokens_1 ❌ 0/5 n/a n/a
llama_3.2_1b_prompt_1024_tokens_40 ❌ 0/5 n/a n/a
llama_3.2_1b_prompt_13_tokens_1 ❌ 0/5 n/a n/a
llama_3.2_1b_prompt_13_tokens_40 ❌ 0/5 n/a n/a
📈 Trends (vs main branch) for Test Example Applications

14eee74 (2026_03_26_15_11_20)

IRONCLAD Trends

llama_3.2_1b

Commit/Date Num Tokens (max)Num Tokens (mean)Num Tokens (median)Num Tokens (min)Num Tokens (stddev)TPS (max)TPS (mean)TPS (median)TPS (min)TPS (stddev)TTFT (max)TTFT (mean)TTFT (median)TTFT (min)TTFT (stddev)Total (max)Total (mean)Total (median)Total (min)Total (stddev)
130b6ea — 2025-12-05 21:33:1240.00 (+0.00%)40.00 (+0.00%)40.00 (+0.00%)40.00 (+0.00%)0.00 (n/a)4.71 (-0.42%)4.64 (-0.09%)4.64 (+0.65%)4.55 (-0.22%)0.05 (-17.66%)4.41 (-0.34%)4.39 (-0.19%)4.38 (-0.33%)4.37 (-0.15%)0.01 (-25.90%)12.96 (-0.00%)12.80 (+0.07%)12.80 (-0.23%)12.67 (+0.44%)0.09 (-21.12%)
0a6c11c — 2025-12-03 23:35:1540.00 (n/a)40.00 (n/a)40.00 (n/a)40.00 (n/a)0.00 (n/a)4.73 (n/a)4.64 (n/a)4.61 (n/a)4.56 (n/a)0.06 (n/a)4.42 (n/a)4.40 (n/a)4.40 (n/a)4.37 (n/a)0.02 (n/a)12.96 (n/a)12.79 (n/a)12.83 (n/a)12.62 (n/a)0.12 (n/a)

llama_3.2_1b_prompt_1024_tokens_1

Commit/Date TTFT (max)TTFT (mean)TTFT (median)TTFT (min)TTFT (stddev)
14eee74 — 2026-03-26 15:08:56n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)
8b6d0f4 — 2026-03-25 21:36:352.26 (n/a)2.25 (n/a)2.25 (n/a)2.23 (n/a)0.01 (n/a)

llama_3.2_1b_prompt_1024_tokens_40

Commit/Date TPS (max)TPS (mean)TPS (median)TPS (min)TPS (stddev)TTFT (max)TTFT (mean)TTFT (median)TTFT (min)TTFT (stddev)
14eee74 — 2026-03-26 15:08:56n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)
8b6d0f4 — 2026-03-25 21:36:354.30 (n/a)4.28 (n/a)4.28 (n/a)4.26 (n/a)0.02 (n/a)2.37 (n/a)2.27 (n/a)2.25 (n/a)2.23 (n/a)0.06 (n/a)

llama_3.2_1b_prompt_13_tokens_1

Commit/Date TTFT (max)TTFT (mean)TTFT (median)TTFT (min)TTFT (stddev)
14eee74 — 2026-03-26 15:08:56n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)
8b6d0f4 — 2026-03-25 21:36:352.19 (n/a)2.17 (n/a)2.17 (n/a)2.14 (n/a)0.02 (n/a)

llama_3.2_1b_prompt_13_tokens_40

Commit/Date TPS (max)TPS (mean)TPS (median)TPS (min)TPS (stddev)TTFT (max)TTFT (mean)TTFT (median)TTFT (min)TTFT (stddev)
14eee74 — 2026-03-26 15:08:56n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)
8b6d0f4 — 2026-03-25 21:36:354.30 (n/a)4.27 (n/a)4.27 (n/a)4.25 (n/a)0.02 (n/a)2.19 (n/a)2.17 (n/a)2.16 (n/a)2.15 (n/a)0.01 (n/a)

llama_3.2_1b_prompt_2048_tokens_1

Commit/Date Num_Tokens (max)Num_Tokens (mean)Num_Tokens (median)Num_Tokens (min)Num_Tokens (stddev)TPS (max)TPS (mean)TPS (median)TPS (min)TPS (stddev)TTFT (max)TTFT (mean)TTFT (median)TTFT (min)TTFT (stddev)
897d04e — 2026-03-06 22:56:071.00 (+0.00%)1.00 (+0.00%)1.00 (+0.00%)1.00 (+0.00%)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)2.68 (-1.06%)2.68 (-1.06%)2.68 (-1.06%)2.68 (-1.06%)0.00 (n/a)
84d3478 — 2026-02-17 23:16:231.00 (n/a)1.00 (n/a)1.00 (n/a)1.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)2.70 (n/a)2.70 (n/a)2.70 (n/a)2.70 (n/a)0.00 (n/a)

llama_3.2_1b_prompt_2048_tokens_40

Commit/Date Num_Tokens (max)Num_Tokens (mean)Num_Tokens (median)Num_Tokens (min)Num_Tokens (stddev)TPS (max)TPS (mean)TPS (median)TPS (min)TPS (stddev)TTFT (max)TTFT (mean)TTFT (median)TTFT (min)TTFT (stddev)
897d04e — 2026-03-06 22:56:0740.00 (+0.00%)40.00 (+0.00%)40.00 (+0.00%)40.00 (+0.00%)0.00 (n/a)4.00 (-1.72%)4.00 (-1.72%)4.00 (-1.72%)4.00 (-1.72%)0.00 (n/a)2.70 (-0.44%)2.70 (-0.44%)2.70 (-0.44%)2.70 (-0.44%)0.00 (n/a)
84d3478 — 2026-02-17 23:16:2340.00 (n/a)40.00 (n/a)40.00 (n/a)40.00 (n/a)0.00 (n/a)4.07 (n/a)4.07 (n/a)4.07 (n/a)4.07 (n/a)0.00 (n/a)2.71 (n/a)2.71 (n/a)2.71 (n/a)2.71 (n/a)0.00 (n/a)

@github-actions
Copy link
Contributor

📊 Test Results for Small Benchmark/Test Suite

14eee74 (2026_03_26_15_37_43)

IRONCLAD

Tested on 2026_03_26_15_37_43 at commit 14eee74.

Test Checks Latency (mean)Bandwidth (mean)Throughput (mean)
M_128-K_128-num_aie_columns_1-tile_size_input_32-tile_size_output_128 ✅ 5/5 n/a 0.12 0.12
M_1792-K_896-N_1152-num_aie_columns_8-b_col_maj_False-c_col_maj_True-m_64-k_32-n_48-trace_size_0 ✅ 5/5 2820.50 3.76 1478.18
M_192-K_384-N_64-num_aie_columns_4-b_col_maj_False-c_col_maj_False-m_48-k_96-n_16-trace_size_0 ✅ 5/5 514.88 0.48 20.46
M_192-K_384-N_64-num_aie_columns_4-b_col_maj_True-c_col_maj_True-m_48-k_96-n_16-trace_size_0 ✅ 5/5 534.08 0.51 21.61
M_2048-K_2048-N_2048-num_aie_columns_1-b_col_maj_False-c_col_maj_False-m_64-k_64-n_64-trace_size_0 ✅ 5/5 49945.60 0.50 343.97
M_2048-K_2048-N_2048-num_aie_columns_2-b_col_maj_True-c_col_maj_False-m_64-k_64-n_64-trace_size_0 ✅ 5/5 29385.26 0.86 584.68
M_2048-K_2048-N_2048-num_aie_columns_8-b_col_maj_True-c_col_maj_True-m_64-k_64-n_64-trace_size_0 ✅ 5/5 7687.44 3.29 2247.11
M_2048-K_8192-num_aie_columns_1-tile_size_input_1-tile_size_output_2048 ✅ 5/5 n/a 11.63 11.62
M_2048-K_8192-num_aie_columns_2-tile_size_input_1-tile_size_output_1024 ✅ 5/5 n/a 20.80 20.79
M_2048-K_8192-num_aie_columns_4-tile_size_input_1-tile_size_output_512 ✅ 5/5 n/a 32.02 32.00
M_2048-K_8192-num_aie_columns_8-tile_size_input_1-tile_size_output_256 ✅ 5/5 n/a 33.76 33.74
M_2048-N_64-aie_columns_1-channels_1-m_64-n_64-s_8 ✅ 5/5 343.16 1.54 n/a
M_2048-N_64-aie_columns_1-channels_2-m_64-n_64-s_8 ✅ 5/5 973.94 0.98 n/a
M_384-K_1536-N_1792-num_aie_columns_4-b_col_maj_True-c_col_maj_False-m_32-k_48-n_64-trace_size_0 ✅ 5/5 2097.50 3.84 1008.08
M_8192-K_2048-num_aie_columns_1-tile_size_input_4-tile_size_output_1024 ✅ 5/5 n/a 12.20 12.19
M_8192-K_2048-num_aie_columns_2-tile_size_input_4-tile_size_output_1024 ✅ 5/5 n/a 20.22 20.21
M_8192-K_2048-num_aie_columns_4-tile_size_input_4-tile_size_output_1024 ✅ 5/5 n/a 31.16 31.14
M_8192-K_2048-num_aie_columns_8-tile_size_input_4-tile_size_output_1024 ✅ 5/5 n/a 33.91 33.89
M_896-K_1792-N_640-num_aie_columns_8-b_col_maj_False-c_col_maj_True-m_32-k_64-n_80-trace_size_0 ✅ 5/5 1372.12 5.07 1565.74
embedding_dim_2048-hidden_dim_2048 ✅ 5/5 n/a n/a n/a
input_length_2048-num_aie_columns_1-num_channels_1-tile_size_2048 ✅ 25/25 283.01 0.03 n/a
input_length_2048-num_aie_columns_1-num_channels_1-tile_size_2048-group_size_32 ✅ 5/5 265.34 0.02 n/a
input_length_2048-num_aie_columns_1-num_channels_1-tile_size_2048-weighted_False ✅ 5/5 276.20 0.03 n/a
input_length_2048-num_aie_columns_1-num_channels_1-tile_size_2048-weighted_True ✅ 5/5 273.42 0.05 n/a
input_length_2048-num_aie_columns_1-num_channels_2-tile_size_1024 ✅ 10/10 264.64 0.03 n/a
input_length_2048-num_aie_columns_1-num_channels_2-tile_size_1024-group_size_32 ✅ 5/5 276.64 0.02 n/a
input_length_2048-num_aie_columns_1-num_channels_2-tile_size_1024-weighted_False ✅ 5/5 259.84 0.03 n/a
input_length_2048-num_aie_columns_1-num_channels_2-tile_size_2048-scalar_factor_3.0 ✅ 5/5 325.28 0.05 n/a
input_length_2048-num_aie_columns_1-tile_size_2048 ✅ 15/15 284.11 0.04 n/a
input_length_2048-num_aie_columns_2-num_channels_1-tile_size_1024 ✅ 25/25 295.97 0.03 n/a
input_length_2048-num_aie_columns_2-num_channels_1-tile_size_1024-group_size_32 ✅ 5/5 267.36 0.02 n/a
input_length_2048-num_aie_columns_2-num_channels_1-tile_size_1024-weighted_False ✅ 5/5 273.08 0.03 n/a
input_length_2048-num_aie_columns_2-num_channels_1-tile_size_1024-weighted_True ✅ 5/5 271.96 0.04 n/a
input_length_2048-num_aie_columns_2-num_channels_2-tile_size_1024-scalar_factor_3.0 ✅ 5/5 318.90 0.04 n/a
input_length_2048-num_aie_columns_2-num_channels_2-tile_size_512 ✅ 10/10 291.42 0.03 n/a
input_length_2048-num_aie_columns_2-num_channels_2-tile_size_512-group_size_32 ✅ 5/5 246.78 0.02 n/a
input_length_2048-num_aie_columns_2-num_channels_2-tile_size_512-weighted_False ✅ 5/5 260.96 0.03 n/a
input_length_2048-num_aie_columns_2-tile_size_1024 ✅ 15/15 330.51 0.04 n/a
input_length_2048-num_aie_columns_4-num_channels_1-tile_size_512 ✅ 25/25 302.31 0.03 n/a
input_length_2048-num_aie_columns_4-num_channels_1-tile_size_512-group_size_32 ✅ 5/5 348.26 0.02 n/a
input_length_2048-num_aie_columns_4-num_channels_1-tile_size_512-weighted_False ✅ 5/5 292.60 0.03 n/a
input_length_2048-num_aie_columns_4-num_channels_1-tile_size_512-weighted_True ✅ 5/5 310.72 0.03 n/a
input_length_2048-num_aie_columns_4-num_channels_2-tile_size_256 ✅ 10/10 340.30 0.03 n/a
input_length_2048-num_aie_columns_4-num_channels_2-tile_size_256-group_size_32 ✅ 5/5 308.20 0.02 n/a
input_length_2048-num_aie_columns_4-num_channels_2-tile_size_256-weighted_False ✅ 5/5 341.04 0.03 n/a
input_length_2048-num_aie_columns_4-num_channels_2-tile_size_512-scalar_factor_3.0 ✅ 5/5 300.20 0.05 n/a
input_length_2048-num_aie_columns_4-tile_size_512 ✅ 15/15 285.93 0.04 n/a
input_length_2048-num_aie_columns_8-num_channels_1-tile_size_256 ✅ 25/25 327.08 0.03 n/a
input_length_2048-num_aie_columns_8-num_channels_1-tile_size_256-group_size_32 ✅ 5/5 322.10 0.02 n/a
input_length_2048-num_aie_columns_8-num_channels_1-tile_size_256-weighted_False ✅ 5/5 293.64 0.03 n/a
input_length_2048-num_aie_columns_8-num_channels_1-tile_size_256-weighted_True ✅ 5/5 391.22 0.03 n/a
input_length_2048-num_aie_columns_8-num_channels_2-tile_size_128 ✅ 10/10 346.85 0.03 n/a
input_length_2048-num_aie_columns_8-num_channels_2-tile_size_128-group_size_32 ✅ 5/5 291.70 0.02 n/a
input_length_2048-num_aie_columns_8-num_channels_2-tile_size_128-weighted_False ✅ 5/5 341.48 0.03 n/a
input_length_2048-num_aie_columns_8-num_channels_2-tile_size_256-scalar_factor_3.0 ✅ 5/5 353.90 0.04 n/a
input_length_2048-num_aie_columns_8-tile_size_256 ✅ 15/15 305.93 0.04 n/a
input_length_2048-num_cores_1-num_channels_1-bypass_False-tile_size_2048 ✅ 5/5 272.38 0.03 n/a
input_length_2048-num_cores_16-num_channels_2-bypass_False-tile_size_128 ✅ 5/5 299.36 0.03 n/a
input_length_2048-num_cores_2-num_channels_1-bypass_False-tile_size_1024 ✅ 5/5 348.46 0.03 n/a
input_length_2048-num_cores_2-num_channels_2-bypass_False-tile_size_1024 ✅ 5/5 270.74 0.03 n/a
input_length_2048-num_cores_4-num_channels_1-bypass_False-tile_size_512 ✅ 5/5 390.94 0.03 n/a
input_length_2048-num_cores_4-num_channels_2-bypass_False-tile_size_512 ✅ 5/5 297.36 0.03 n/a
input_length_2048-num_cores_8-num_channels_1-bypass_False-tile_size_256 ✅ 5/5 410.64 0.03 n/a
input_length_2048-num_cores_8-num_channels_2-bypass_False-tile_size_256 ✅ 5/5 358.04 0.03 n/a
input_length_32768-num_aie_columns_2-num_channels_2-tile_size_1024 ✅ 5/5 371.62 0.37 n/a
input_length_32768-num_aie_columns_2-num_channels_2-tile_size_512 ✅ 5/5 442.58 0.32 n/a
input_length_32768-num_aie_columns_4-num_channels_4-tile_size_2048 ✅ 5/5 399.60 0.37 n/a
rows_32-cols_512-angle_rows_32-aie_columns_1-method_type_0 ✅ 5/5 388.78 0.28 n/a
rows_32-cols_512-angle_rows_32-aie_columns_2-method_type_0 ✅ 5/5 284.78 0.35 n/a
rows_32-cols_512-angle_rows_32-aie_columns_8-method_type_0 ✅ 5/5 368.90 0.32 n/a
rows_32-cols_512-angle_rows_8-aie_columns_1-method_type_0 ✅ 5/5 316.04 0.24 n/a
rows_32-cols_512-angle_rows_8-aie_columns_2-method_type_0 ✅ 5/5 321.92 0.24 n/a
rows_32-cols_512-angle_rows_8-aie_columns_8-method_type_0 ✅ 5/5 371.22 0.21 n/a
seq_len_16384-dim_64-num_heads_1-num_pipelines_8-num_kv_heads_0 ❌ 0/5 n/a n/a n/a
seq_len_256-embedding_dim_2048-hidden_dim_2048-prio_accuracy_False ✅ 5/5 n/a n/a n/a
📈 Trends (vs main branch) for Small Benchmark/Test Suite

14eee74 (2026_03_26_15_37_43)

IRONCLAD Trends

M_128-K_128-num_aie_columns_1-tile_size_input_32-tile_size_output_128

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
14eee74 — 2026-03-26 15:31:390.13 (n/a)0.12 (n/a)0.13 (n/a)0.09 (n/a)0.02 (n/a)0.13 (n/a)0.12 (n/a)0.12 (n/a)0.09 (n/a)0.02 (n/a)

M_1792-K_896-N_1152-num_aie_columns_8-b_col_maj_False-c_col_maj_True-m_64-k_32-n_48-trace_size_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
14eee74 — 2026-03-26 15:31:394.64 (n/a)3.76 (n/a)4.49 (n/a)1.89 (n/a)1.20 (n/a)4986.70 (n/a)2820.50 (n/a)2092.80 (n/a)2025.40 (n/a)1269.38 (n/a)1826.52 (n/a)1478.18 (n/a)1767.64 (n/a)741.85 (n/a)473.22 (n/a)

M_192-K_384-N_64-num_aie_columns_4-b_col_maj_False-c_col_maj_False-m_48-k_96-n_16-trace_size_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
14eee74 — 2026-03-26 15:31:390.78 (n/a)0.48 (n/a)0.48 (n/a)0.31 (n/a)0.19 (n/a)704.80 (n/a)514.88 (n/a)464.00 (n/a)282.40 (n/a)176.99 (n/a)33.41 (n/a)20.46 (n/a)20.34 (n/a)13.39 (n/a)8.09 (n/a)

M_192-K_384-N_64-num_aie_columns_4-b_col_maj_True-c_col_maj_True-m_48-k_96-n_16-trace_size_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
14eee74 — 2026-03-26 15:31:390.81 (n/a)0.51 (n/a)0.54 (n/a)0.25 (n/a)0.24 (n/a)896.70 (n/a)534.08 (n/a)406.30 (n/a)272.70 (n/a)271.61 (n/a)34.61 (n/a)21.61 (n/a)23.23 (n/a)10.52 (n/a)10.05 (n/a)

M_2048-K_2048-N_2048-num_aie_columns_1-b_col_maj_False-c_col_maj_False-m_64-k_64-n_64-trace_size_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
14eee74 — 2026-03-26 15:31:390.51 (n/a)0.50 (n/a)0.50 (n/a)0.50 (n/a)0.00 (n/a)50068.30 (n/a)49945.60 (n/a)50012.70 (n/a)49651.60 (n/a)167.60 (n/a)346.01 (n/a)343.97 (n/a)343.51 (n/a)343.13 (n/a)1.16 (n/a)

M_2048-K_2048-N_2048-num_aie_columns_2-b_col_maj_True-c_col_maj_False-m_64-k_64-n_64-trace_size_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
14eee74 — 2026-03-26 15:31:390.86 (n/a)0.86 (n/a)0.86 (n/a)0.85 (n/a)0.01 (n/a)29777.20 (n/a)29385.26 (n/a)29244.00 (n/a)29158.60 (n/a)255.08 (n/a)589.19 (n/a)584.68 (n/a)587.47 (n/a)576.95 (n/a)5.05 (n/a)

M_2048-K_2048-N_2048-num_aie_columns_8-b_col_maj_True-c_col_maj_True-m_64-k_64-n_64-trace_size_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
14eee74 — 2026-03-26 15:31:393.54 (n/a)3.29 (n/a)3.33 (n/a)2.90 (n/a)0.27 (n/a)8676.80 (n/a)7687.44 (n/a)7560.00 (n/a)7111.80 (n/a)650.27 (n/a)2415.67 (n/a)2247.11 (n/a)2272.47 (n/a)1979.97 (n/a)182.20 (n/a)

M_2048-K_8192-num_aie_columns_1-tile_size_input_1-tile_size_output_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
14eee74 — 2026-03-26 15:31:3912.42 (n/a)11.63 (n/a)12.32 (n/a)9.43 (n/a)1.28 (n/a)12.41 (n/a)11.62 (n/a)12.31 (n/a)9.42 (n/a)1.28 (n/a)

M_2048-K_8192-num_aie_columns_2-tile_size_input_1-tile_size_output_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
14eee74 — 2026-03-26 15:31:3921.79 (n/a)20.80 (n/a)21.66 (n/a)17.32 (n/a)1.95 (n/a)21.78 (n/a)20.79 (n/a)21.64 (n/a)17.31 (n/a)1.95 (n/a)

M_2048-K_8192-num_aie_columns_4-tile_size_input_1-tile_size_output_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
14eee74 — 2026-03-26 15:31:3934.64 (n/a)32.02 (n/a)33.20 (n/a)26.13 (n/a)3.39 (n/a)34.62 (n/a)32.00 (n/a)33.18 (n/a)26.12 (n/a)3.39 (n/a)

M_2048-K_8192-num_aie_columns_8-tile_size_input_1-tile_size_output_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
14eee74 — 2026-03-26 15:31:3936.28 (n/a)33.76 (n/a)35.15 (n/a)26.96 (n/a)3.84 (n/a)36.25 (n/a)33.74 (n/a)35.13 (n/a)26.95 (n/a)3.84 (n/a)

M_2048-N_64-aie_columns_1-channels_1-m_64-n_64-s_8

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
14eee74 — 2026-03-26 15:31:391.74 (n/a)1.54 (n/a)1.47 (n/a)1.39 (n/a)0.14 (n/a)376.70 (n/a)343.16 (n/a)356.70 (n/a)301.40 (n/a)30.59 (n/a)

M_2048-N_64-aie_columns_1-channels_2-m_64-n_64-s_8

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
14eee74 — 2026-03-26 15:31:391.80 (n/a)0.98 (n/a)0.74 (n/a)0.20 (n/a)0.68 (n/a)2671.40 (n/a)973.94 (n/a)708.20 (n/a)291.20 (n/a)979.66 (n/a)

M_384-K_1536-N_1792-num_aie_columns_4-b_col_maj_True-c_col_maj_False-m_32-k_48-n_64-trace_size_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
14eee74 — 2026-03-26 15:31:393.90 (n/a)3.84 (n/a)3.86 (n/a)3.73 (n/a)0.07 (n/a)2160.20 (n/a)2097.50 (n/a)2086.50 (n/a)2068.30 (n/a)36.31 (n/a)1022.08 (n/a)1008.08 (n/a)1013.14 (n/a)978.60 (n/a)17.12 (n/a)

M_8192-K_2048-num_aie_columns_1-tile_size_input_4-tile_size_output_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
14eee74 — 2026-03-26 15:31:3912.70 (n/a)12.20 (n/a)12.42 (n/a)11.07 (n/a)0.64 (n/a)12.69 (n/a)12.19 (n/a)12.41 (n/a)11.06 (n/a)0.64 (n/a)

M_8192-K_2048-num_aie_columns_2-tile_size_input_4-tile_size_output_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
14eee74 — 2026-03-26 15:31:3921.61 (n/a)20.22 (n/a)21.31 (n/a)15.63 (n/a)2.57 (n/a)21.59 (n/a)20.21 (n/a)21.30 (n/a)15.62 (n/a)2.57 (n/a)

M_8192-K_2048-num_aie_columns_4-tile_size_input_4-tile_size_output_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
14eee74 — 2026-03-26 15:31:3934.00 (n/a)31.16 (n/a)32.62 (n/a)24.11 (n/a)3.99 (n/a)33.98 (n/a)31.14 (n/a)32.60 (n/a)24.10 (n/a)3.99 (n/a)

M_8192-K_2048-num_aie_columns_8-tile_size_input_4-tile_size_output_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
14eee74 — 2026-03-26 15:31:3936.48 (n/a)33.91 (n/a)34.95 (n/a)28.22 (n/a)3.27 (n/a)36.45 (n/a)33.89 (n/a)34.93 (n/a)28.21 (n/a)3.27 (n/a)

M_896-K_1792-N_640-num_aie_columns_8-b_col_maj_False-c_col_maj_True-m_32-k_64-n_80-trace_size_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
14eee74 — 2026-03-26 15:31:395.92 (n/a)5.07 (n/a)5.54 (n/a)3.31 (n/a)1.04 (n/a)2006.80 (n/a)1372.12 (n/a)1200.70 (n/a)1124.00 (n/a)364.16 (n/a)1828.53 (n/a)1565.74 (n/a)1711.74 (n/a)1024.12 (n/a)322.14 (n/a)

axpy_1_cols_2_channels_2048_tile_2048_3.0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.08 (+17.48%)0.07 (+11.67%)0.07 (+26.71%)0.05 (+0.80%)0.01 (+108.36%)239.80 (-0.83%)190.62 (-7.91%)166.90 (-21.09%)155.40 (-14.90%)42.89 (+79.10%)
dc66da7 — 2026-03-25 19:23:450.07 (n/a)0.06 (n/a)0.06 (n/a)0.05 (n/a)0.01 (n/a)241.80 (n/a)207.00 (n/a)211.50 (n/a)182.60 (n/a)23.95 (n/a)

axpy_1_cols_2_channels_2048_tile_2048_3.0_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.16 (-3.31%)0.14 (-1.29%)0.15 (+16.18%)0.10 (-6.80%)0.03 (+3.89%)125.70 (+7.25%)93.56 (+2.12%)83.50 (-13.92%)75.00 (+3.45%)22.38 (+18.09%)
84d3478 — 2026-02-17 23:25:130.17 (n/a)0.14 (n/a)0.13 (n/a)0.10 (n/a)0.03 (n/a)117.20 (n/a)91.62 (n/a)97.00 (n/a)72.50 (n/a)18.95 (n/a)

axpy_2_cols_2_channels_2048_tile_1024_3.0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.09 (+14.45%)0.07 (-0.78%)0.07 (-0.40%)0.05 (-7.87%)0.01 (+68.50%)233.40 (+8.51%)188.20 (+2.47%)179.80 (+0.39%)143.50 (-12.66%)34.06 (+61.09%)
dc66da7 — 2026-03-25 19:23:450.07 (n/a)0.07 (n/a)0.07 (n/a)0.06 (n/a)0.01 (n/a)215.10 (n/a)183.66 (n/a)179.10 (n/a)164.30 (n/a)21.14 (n/a)

axpy_2_cols_2_channels_2048_tile_1024_3.0_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.13 (-0.73%)0.11 (-7.43%)0.11 (-8.63%)0.07 (-27.71%)0.02 (+50.39%)182.70 (+38.30%)120.86 (+11.97%)110.80 (+9.38%)96.10 (+0.73%)35.16 (+122.88%)
84d3478 — 2026-02-17 23:25:130.13 (n/a)0.12 (n/a)0.12 (n/a)0.09 (n/a)0.02 (n/a)132.10 (n/a)107.94 (n/a)101.30 (n/a)95.40 (n/a)15.78 (n/a)

axpy_4_cols_2_channels_2048_tile_512_3.0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.10 (+6.67%)0.06 (-1.56%)0.06 (-2.24%)0.05 (+3.60%)0.02 (+20.83%)251.50 (-3.49%)200.78 (+3.11%)191.70 (+2.29%)128.70 (-6.26%)50.37 (+11.30%)
dc66da7 — 2026-03-25 19:23:450.09 (n/a)0.07 (n/a)0.07 (n/a)0.05 (n/a)0.02 (n/a)260.60 (n/a)194.72 (n/a)187.40 (n/a)137.30 (n/a)45.26 (n/a)

axpy_4_cols_2_channels_2048_tile_512_3.0_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.13 (-10.91%)0.11 (-7.68%)0.12 (-0.73%)0.06 (-17.94%)0.03 (+0.99%)194.10 (+21.85%)124.52 (+10.29%)103.30 (+0.68%)97.20 (+12.24%)40.32 (+39.15%)
84d3478 — 2026-02-17 23:25:130.14 (n/a)0.11 (n/a)0.12 (n/a)0.08 (n/a)0.03 (n/a)159.30 (n/a)112.90 (n/a)102.60 (n/a)86.60 (n/a)28.97 (n/a)

axpy_8_cols_2_channels_2048_tile_256_3.0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.08 (+1.42%)0.06 (+9.86%)0.06 (+18.71%)0.03 (+13.51%)0.02 (-10.62%)376.00 (-11.90%)225.02 (-11.61%)196.20 (-15.76%)151.40 (-1.37%)87.15 (-17.99%)
dc66da7 — 2026-03-25 19:23:450.08 (n/a)0.05 (n/a)0.05 (n/a)0.03 (n/a)0.02 (n/a)426.80 (n/a)254.58 (n/a)232.90 (n/a)153.50 (n/a)106.27 (n/a)

axpy_8_cols_2_channels_2048_tile_256_3.0_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.13 (+0.01%)0.11 (+6.10%)0.11 (+0.22%)0.08 (-0.33%)0.02 (-9.83%)156.30 (+0.32%)115.26 (-6.40%)112.70 (-0.18%)95.50 (+0.00%)24.34 (-10.22%)
84d3478 — 2026-02-17 23:25:130.13 (n/a)0.10 (n/a)0.11 (n/a)0.08 (n/a)0.02 (n/a)155.80 (n/a)123.14 (n/a)112.90 (n/a)95.50 (n/a)27.11 (n/a)

dequant_1_cols_1_channels_2048_tile_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.04 (+17.85%)0.03 (+2.55%)0.03 (+1.65%)0.03 (-2.18%)0.00 (+114.37%)207.70 (+2.26%)178.64 (-0.96%)173.60 (-1.59%)141.60 (-15.16%)28.01 (+90.63%)
dc66da7 — 2026-03-25 19:23:450.03 (n/a)0.03 (n/a)0.03 (n/a)0.03 (n/a)0.00 (n/a)203.10 (n/a)180.38 (n/a)176.40 (n/a)166.90 (n/a)14.69 (n/a)

dequant_1_cols_1_channels_2048_tile_2048_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.05 (-13.00%)0.04 (-0.87%)0.04 (+1.16%)0.03 (-3.17%)0.01 (-31.33%)158.40 (+3.26%)128.12 (-0.19%)123.80 (-1.12%)110.40 (+14.88%)18.36 (-15.65%)
84d3478 — 2026-02-17 23:25:130.05 (n/a)0.04 (n/a)0.04 (n/a)0.03 (n/a)0.01 (n/a)153.40 (n/a)128.36 (n/a)125.20 (n/a)96.10 (n/a)21.77 (n/a)

dequant_1_cols_2_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.04 (+13.44%)0.03 (+8.77%)0.03 (+14.89%)0.02 (-9.15%)0.01 (+54.39%)252.70 (+10.06%)182.14 (-6.08%)174.50 (-12.97%)145.40 (-11.88%)42.38 (+56.31%)
dc66da7 — 2026-03-25 19:23:450.03 (n/a)0.03 (n/a)0.03 (n/a)0.02 (n/a)0.00 (n/a)229.60 (n/a)193.94 (n/a)200.50 (n/a)165.00 (n/a)27.11 (n/a)

dequant_1_cols_2_channels_2048_tile_1024_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.06 (-3.74%)0.05 (-12.63%)0.05 (-5.21%)0.03 (-33.47%)0.01 (+52.60%)162.70 (+50.37%)110.84 (+18.98%)98.80 (+5.44%)81.50 (+3.95%)31.02 (+149.23%)
84d3478 — 2026-02-17 23:25:130.07 (n/a)0.06 (n/a)0.06 (n/a)0.05 (n/a)0.01 (n/a)108.20 (n/a)93.16 (n/a)93.70 (n/a)78.40 (n/a)12.45 (n/a)

dequant_2_cols_1_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.05 (+21.89%)0.03 (+20.89%)0.03 (+21.41%)0.02 (+8.07%)0.01 (+32.97%)215.80 (-7.46%)168.56 (-16.36%)172.60 (-17.65%)113.20 (-17.97%)37.61 (+1.89%)
dc66da7 — 2026-03-25 19:23:450.04 (n/a)0.03 (n/a)0.03 (n/a)0.02 (n/a)0.01 (n/a)233.20 (n/a)201.52 (n/a)209.60 (n/a)138.00 (n/a)36.92 (n/a)

dequant_2_cols_1_channels_2048_tile_1024_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.06 (+8.90%)0.05 (+1.31%)0.05 (-1.85%)0.03 (+6.35%)0.01 (+7.11%)152.30 (-5.93%)115.08 (-1.20%)100.80 (+1.92%)88.10 (-8.23%)27.20 (-4.42%)
84d3478 — 2026-02-17 23:25:130.05 (n/a)0.05 (n/a)0.05 (n/a)0.03 (n/a)0.01 (n/a)161.90 (n/a)116.48 (n/a)98.90 (n/a)96.00 (n/a)28.46 (n/a)

dequant_2_cols_2_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.05 (+49.89%)0.03 (+24.89%)0.03 (+13.74%)0.02 (+15.73%)0.01 (+106.91%)219.80 (-13.60%)173.04 (-17.40%)185.70 (-12.07%)112.30 (-33.27%)41.15 (+16.74%)
dc66da7 — 2026-03-25 19:23:450.03 (n/a)0.03 (n/a)0.02 (n/a)0.02 (n/a)0.00 (n/a)254.40 (n/a)209.50 (n/a)211.20 (n/a)168.30 (n/a)35.25 (n/a)

dequant_2_cols_2_channels_2048_tile_512_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.07 (+27.96%)0.05 (+1.57%)0.04 (-18.06%)0.03 (+10.23%)0.01 (+22.79%)164.70 (-9.31%)119.12 (-1.39%)120.20 (+22.03%)75.60 (-21.82%)31.61 (-13.80%)
84d3478 — 2026-02-17 23:25:130.05 (n/a)0.05 (n/a)0.05 (n/a)0.03 (n/a)0.01 (n/a)181.60 (n/a)120.80 (n/a)98.50 (n/a)96.70 (n/a)36.67 (n/a)

dequant_4_cols_1_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.04 (+56.54%)0.03 (+15.78%)0.03 (+9.01%)0.02 (-5.58%)0.01 (+304.26%)243.40 (+5.92%)186.06 (-8.98%)182.60 (-8.29%)118.50 (-36.12%)46.36 (+164.07%)
dc66da7 — 2026-03-25 19:23:450.03 (n/a)0.03 (n/a)0.03 (n/a)0.02 (n/a)0.00 (n/a)229.80 (n/a)204.42 (n/a)199.10 (n/a)185.50 (n/a)17.56 (n/a)

dequant_4_cols_1_channels_2048_tile_512_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.05 (-2.55%)0.04 (-12.79%)0.05 (-16.25%)0.03 (-12.85%)0.01 (-1.20%)164.60 (+14.78%)122.14 (+15.14%)115.90 (+19.48%)97.50 (+2.63%)25.08 (+19.84%)
84d3478 — 2026-02-17 23:25:130.06 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.01 (n/a)143.40 (n/a)106.08 (n/a)97.00 (n/a)95.00 (n/a)20.93 (n/a)

dequant_4_cols_2_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.04 (+28.72%)0.03 (+2.89%)0.02 (-6.84%)0.02 (+6.27%)0.01 (+95.59%)243.20 (-5.88%)208.42 (-0.42%)222.90 (+7.32%)141.10 (-22.34%)41.08 (+38.16%)
dc66da7 — 2026-03-25 19:23:450.03 (n/a)0.03 (n/a)0.03 (n/a)0.02 (n/a)0.00 (n/a)258.40 (n/a)209.30 (n/a)207.70 (n/a)181.70 (n/a)29.74 (n/a)

dequant_4_cols_2_channels_2048_tile_256_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.06 (+20.83%)0.05 (+8.21%)0.04 (+1.42%)0.03 (+17.39%)0.01 (+14.84%)166.40 (-14.80%)122.34 (-7.97%)127.70 (-1.39%)84.60 (-17.22%)30.47 (-19.21%)
84d3478 — 2026-02-17 23:25:130.05 (n/a)0.04 (n/a)0.04 (n/a)0.03 (n/a)0.01 (n/a)195.30 (n/a)132.94 (n/a)129.50 (n/a)102.20 (n/a)37.71 (n/a)

dequant_8_cols_1_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.04 (-1.52%)0.03 (+2.17%)0.03 (+18.60%)0.02 (-11.90%)0.01 (+4.42%)268.50 (+13.53%)205.32 (-1.14%)191.00 (-15.67%)141.60 (+1.51%)49.39 (+23.46%)
dc66da7 — 2026-03-25 19:23:450.04 (n/a)0.03 (n/a)0.02 (n/a)0.02 (n/a)0.01 (n/a)236.50 (n/a)207.68 (n/a)226.50 (n/a)139.50 (n/a)40.00 (n/a)

dequant_8_cols_1_channels_2048_tile_256_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.05 (-25.19%)0.04 (-21.02%)0.04 (-19.70%)0.04 (-19.76%)0.00 (-37.92%)142.70 (+24.63%)126.44 (+26.24%)125.70 (+24.46%)115.50 (+33.68%)10.61 (+3.97%)
84d3478 — 2026-02-17 23:25:130.06 (n/a)0.05 (n/a)0.05 (n/a)0.05 (n/a)0.01 (n/a)114.50 (n/a)100.16 (n/a)101.00 (n/a)86.40 (n/a)10.20 (n/a)

dequant_8_cols_2_channels_2048_tile_128

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.02 (+5.64%)0.02 (+1.26%)0.02 (-1.25%)0.02 (-3.87%)0.00 (+124.04%)258.90 (+4.02%)239.08 (-0.87%)247.20 (+1.27%)215.50 (-5.36%)18.28 (+120.74%)
dc66da7 — 2026-03-25 19:23:450.02 (n/a)0.02 (n/a)0.02 (n/a)0.02 (n/a)0.00 (n/a)248.90 (n/a)241.18 (n/a)244.10 (n/a)227.70 (n/a)8.28 (n/a)

dequant_8_cols_2_channels_2048_tile_128_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.04 (-26.69%)0.04 (-11.93%)0.04 (-16.99%)0.03 (+27.28%)0.01 (-62.60%)193.10 (-21.41%)147.68 (+2.96%)136.10 (+20.44%)135.40 (+36.49%)25.41 (-58.96%)
84d3478 — 2026-02-17 23:25:130.05 (n/a)0.04 (n/a)0.05 (n/a)0.02 (n/a)0.01 (n/a)245.70 (n/a)143.44 (n/a)113.00 (n/a)99.20 (n/a)61.90 (n/a)

eltwise_add_1_cols_2_channels_2048_tile_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.07 (-32.66%)0.06 (-13.87%)0.06 (-9.19%)0.04 (-8.04%)0.01 (-49.50%)299.50 (+8.75%)223.84 (+11.90%)212.70 (+10.15%)187.60 (+48.42%)45.70 (-16.71%)
dc66da7 — 2026-03-25 19:23:450.10 (n/a)0.07 (n/a)0.06 (n/a)0.04 (n/a)0.02 (n/a)275.40 (n/a)200.04 (n/a)193.10 (n/a)126.40 (n/a)54.87 (n/a)

eltwise_add_2_cols_2_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.07 (-33.33%)0.06 (-18.97%)0.05 (-19.23%)0.05 (+18.13%)0.01 (-59.22%)270.80 (-15.37%)223.26 (+13.77%)229.40 (+23.80%)174.50 (+50.04%)37.51 (-50.17%)
dc66da7 — 2026-03-25 19:23:450.11 (n/a)0.07 (n/a)0.07 (n/a)0.04 (n/a)0.02 (n/a)320.00 (n/a)196.24 (n/a)185.30 (n/a)116.30 (n/a)75.29 (n/a)

eltwise_add_4_cols_2_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.07 (-8.97%)0.06 (-6.00%)0.06 (-2.02%)0.05 (-2.74%)0.01 (-14.89%)243.40 (+2.83%)208.60 (+6.01%)200.50 (+2.04%)178.10 (+9.87%)31.58 (-2.07%)
dc66da7 — 2026-03-25 19:23:450.08 (n/a)0.06 (n/a)0.06 (n/a)0.05 (n/a)0.01 (n/a)236.70 (n/a)196.78 (n/a)196.50 (n/a)162.10 (n/a)32.25 (n/a)

eltwise_add_8_cols_2_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.06 (-45.96%)0.05 (-24.77%)0.05 (-15.00%)0.05 (+4.78%)0.00 (-83.62%)259.60 (-4.56%)234.66 (+20.97%)224.10 (+17.64%)218.40 (+85.08%)18.77 (-71.47%)
dc66da7 — 2026-03-25 19:23:450.10 (n/a)0.07 (n/a)0.06 (n/a)0.05 (n/a)0.02 (n/a)272.00 (n/a)193.98 (n/a)190.50 (n/a)118.00 (n/a)65.77 (n/a)

eltwise_mul_1_cols_2_channels_2048_tile_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.11 (+3.13%)0.07 (+2.52%)0.06 (-1.87%)0.05 (-14.01%)0.02 (+21.62%)257.00 (+16.29%)184.10 (+0.67%)190.70 (+1.87%)114.40 (-3.05%)55.57 (+40.96%)
dc66da7 — 2026-03-25 19:23:450.10 (n/a)0.07 (n/a)0.07 (n/a)0.06 (n/a)0.02 (n/a)221.00 (n/a)182.88 (n/a)187.20 (n/a)118.00 (n/a)39.42 (n/a)

eltwise_mul_2_cols_2_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.09 (+2.43%)0.07 (+6.63%)0.08 (+30.25%)0.05 (-10.38%)0.02 (+9.28%)264.20 (+11.62%)182.10 (-4.96%)162.40 (-23.22%)137.30 (-2.42%)52.80 (+20.53%)
dc66da7 — 2026-03-25 19:23:450.09 (n/a)0.07 (n/a)0.06 (n/a)0.05 (n/a)0.02 (n/a)236.70 (n/a)191.60 (n/a)211.50 (n/a)140.70 (n/a)43.81 (n/a)

eltwise_mul_4_cols_2_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.10 (+22.79%)0.08 (+18.44%)0.09 (+34.73%)0.06 (-0.12%)0.01 (+158.99%)191.20 (+0.10%)154.44 (-13.56%)136.60 (-25.80%)126.60 (-18.53%)29.95 (+117.72%)
dc66da7 — 2026-03-25 19:23:450.08 (n/a)0.07 (n/a)0.07 (n/a)0.06 (n/a)0.01 (n/a)191.00 (n/a)178.66 (n/a)184.10 (n/a)155.40 (n/a)13.76 (n/a)

eltwise_mul_8_cols_2_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.08 (+34.81%)0.06 (+14.94%)0.06 (+6.87%)0.04 (-5.87%)0.01 (+158.26%)280.10 (+6.22%)210.00 (-9.80%)215.10 (-6.40%)150.70 (-25.80%)49.49 (+100.84%)
dc66da7 — 2026-03-25 19:23:450.06 (n/a)0.05 (n/a)0.05 (n/a)0.05 (n/a)0.01 (n/a)263.70 (n/a)232.82 (n/a)229.80 (n/a)203.10 (n/a)24.64 (n/a)

embedding_dim_2048-hidden_dim_2048

No metrics available.

gelu_1_cols_1_channels_2048_tile_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.05 (+9.18%)0.04 (-1.03%)0.04 (-2.51%)0.04 (-8.09%)0.01 (+114.58%)221.00 (+8.81%)195.42 (+1.97%)199.10 (+2.58%)163.60 (-8.40%)23.03 (+114.14%)
dc66da7 — 2026-03-25 19:23:450.05 (n/a)0.04 (n/a)0.04 (n/a)0.04 (n/a)0.00 (n/a)203.10 (n/a)191.64 (n/a)194.10 (n/a)178.60 (n/a)10.76 (n/a)

gelu_1_cols_2_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.04 (-26.60%)0.04 (-19.13%)0.04 (-18.64%)0.04 (-2.84%)0.00 (-66.68%)226.70 (+2.91%)199.16 (+20.50%)193.10 (+22.92%)186.40 (+36.16%)16.29 (-52.49%)
dc66da7 — 2026-03-25 19:23:450.06 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.01 (n/a)220.30 (n/a)165.28 (n/a)157.10 (n/a)136.90 (n/a)34.28 (n/a)

gelu_2_cols_1_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.06 (+2.08%)0.05 (+4.66%)0.04 (-1.75%)0.03 (-5.55%)0.01 (+12.71%)283.70 (+5.86%)191.18 (-3.05%)191.00 (+1.81%)137.60 (-1.99%)58.08 (+15.01%)
dc66da7 — 2026-03-25 19:23:450.06 (n/a)0.04 (n/a)0.04 (n/a)0.03 (n/a)0.01 (n/a)268.00 (n/a)197.20 (n/a)187.60 (n/a)140.40 (n/a)50.50 (n/a)

gelu_2_cols_2_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.07 (+2.20%)0.04 (-2.29%)0.04 (+1.04%)0.04 (+16.91%)0.01 (-4.76%)227.00 (-14.47%)194.56 (+0.69%)206.00 (-1.01%)119.40 (-2.21%)44.27 (-20.37%)
dc66da7 — 2026-03-25 19:23:450.07 (n/a)0.05 (n/a)0.04 (n/a)0.03 (n/a)0.01 (n/a)265.40 (n/a)193.22 (n/a)208.10 (n/a)122.10 (n/a)55.59 (n/a)

gelu_4_cols_1_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.07 (-9.75%)0.04 (-17.39%)0.04 (-20.48%)0.03 (+12.65%)0.01 (-29.41%)239.40 (-11.23%)196.08 (+15.06%)197.90 (+25.73%)124.70 (+10.84%)44.40 (-31.09%)
dc66da7 — 2026-03-25 19:23:450.07 (n/a)0.05 (n/a)0.05 (n/a)0.03 (n/a)0.02 (n/a)269.70 (n/a)170.42 (n/a)157.40 (n/a)112.50 (n/a)64.43 (n/a)

gelu_4_cols_2_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.04 (-6.29%)0.04 (+1.81%)0.04 (+0.94%)0.03 (-0.42%)0.00 (-22.18%)284.90 (+0.42%)233.36 (-2.39%)227.20 (-0.96%)204.80 (+6.72%)30.62 (-15.21%)
dc66da7 — 2026-03-25 19:23:450.04 (n/a)0.03 (n/a)0.04 (n/a)0.03 (n/a)0.01 (n/a)283.70 (n/a)239.08 (n/a)229.40 (n/a)191.90 (n/a)36.12 (n/a)

gelu_8_cols_1_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.05 (-18.24%)0.04 (-13.70%)0.04 (-4.91%)0.03 (-15.26%)0.01 (-28.46%)285.90 (+18.04%)223.38 (+14.20%)223.20 (+5.18%)156.90 (+22.29%)47.35 (+1.82%)
dc66da7 — 2026-03-25 19:23:450.06 (n/a)0.04 (n/a)0.04 (n/a)0.03 (n/a)0.01 (n/a)242.20 (n/a)195.60 (n/a)212.20 (n/a)128.30 (n/a)46.50 (n/a)

gelu_8_cols_2_channels_2048_tile_128

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.05 (+14.38%)0.04 (+10.49%)0.03 (+3.25%)0.03 (+13.03%)0.01 (+26.31%)305.70 (-11.52%)235.82 (-8.98%)241.80 (-3.12%)177.40 (-12.57%)49.18 (-6.09%)
dc66da7 — 2026-03-25 19:23:450.04 (n/a)0.03 (n/a)0.03 (n/a)0.02 (n/a)0.01 (n/a)345.50 (n/a)259.10 (n/a)249.60 (n/a)202.90 (n/a)52.37 (n/a)

gemm_1792x896x1152_64x32x48_8cols_ccolmaj

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
8b6d0f4 — 2026-03-25 20:46:394.89 (+9.71%)4.07 (+2.20%)4.10 (+2.03%)3.40 (+0.07%)0.55 (+44.15%)2767.30 (-0.07%)2344.32 (-1.52%)2291.20 (-1.99%)1922.40 (-8.85%)308.92 (+28.54%)1924.38 (+9.71%)1600.58 (+2.20%)1614.60 (+2.03%)1336.81 (+0.07%)215.84 (+44.15%)
dc66da7 — 2026-03-25 19:23:454.46 (n/a)3.98 (n/a)4.02 (n/a)3.40 (n/a)0.38 (n/a)2769.20 (n/a)2380.42 (n/a)2337.70 (n/a)2109.10 (n/a)240.33 (n/a)1754.05 (n/a)1566.13 (n/a)1582.49 (n/a)1335.88 (n/a)149.73 (n/a)

gemm_192x384x64_48x96x16_4cols

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
8b6d0f4 — 2026-03-25 20:46:391.35 (+17.71%)0.77 (-13.89%)0.65 (-34.22%)0.56 (+7.33%)0.32 (+33.14%)393.60 (-6.84%)315.66 (+18.48%)341.40 (+52.00%)164.30 (-15.05%)87.82 (-5.13%)57.45 (+17.71%)32.93 (-13.89%)27.64 (-34.22%)23.97 (+7.33%)13.81 (+33.14%)
dc66da7 — 2026-03-25 19:23:451.14 (n/a)0.90 (n/a)0.98 (n/a)0.52 (n/a)0.24 (n/a)422.50 (n/a)266.42 (n/a)224.60 (n/a)193.40 (n/a)92.57 (n/a)48.81 (n/a)38.25 (n/a)42.02 (n/a)22.34 (n/a)10.37 (n/a)

gemm_192x384x64_48x96x16_4cols_bcolmaj_ccolmaj

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
8b6d0f4 — 2026-03-25 20:46:391.18 (+31.18%)0.79 (-0.76%)0.72 (-17.59%)0.64 (+26.96%)0.22 (+35.66%)343.10 (-21.24%)294.50 (+0.95%)309.00 (+21.37%)188.10 (-23.78%)62.26 (-23.01%)50.17 (+31.18%)33.62 (-0.76%)30.54 (-17.59%)27.51 (+26.96%)9.40 (+35.66%)
dc66da7 — 2026-03-25 19:23:450.90 (n/a)0.79 (n/a)0.87 (n/a)0.51 (n/a)0.16 (n/a)435.60 (n/a)291.72 (n/a)254.60 (n/a)246.80 (n/a)80.87 (n/a)38.24 (n/a)33.88 (n/a)37.06 (n/a)21.67 (n/a)6.93 (n/a)

gemm_2048x2048x2048_64x64x32_8_cols_0_bcolmaj_0_ccolmaj_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)GFLOP/s (max)GFLOP/s (mean)GFLOP/s (median)GFLOP/s (min)GFLOP/s (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
3a053d3 — 2025-11-26 01:12:163.72 (-0.11%)3.55 (-2.75%)3.56 (-2.75%)3.36 (-5.33%)0.12 (+76.10%)2725.23 (-0.11%)2602.52 (-2.75%)2612.11 (-2.75%)2459.89 (-5.33%)86.27 (+76.10%)6984.00 (+5.63%)6608.60 (+2.91%)6577.00 (+2.83%)6304.00 (+0.11%)221.87 (+87.15%)
9e98049 — 2025-11-25 20:50:553.72 (n/a)3.65 (n/a)3.66 (n/a)3.54 (n/a)0.07 (n/a)2728.26 (n/a)2676.06 (n/a)2686.03 (n/a)2598.29 (n/a)48.99 (n/a)6612.00 (n/a)6422.00 (n/a)6396.00 (n/a)6297.00 (n/a)118.55 (n/a)

gemm_2048x2048x2048_64x64x32_8_cols_0_bcolmaj_1_ccolmaj_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)GFLOP/s (max)GFLOP/s (mean)GFLOP/s (median)GFLOP/s (min)GFLOP/s (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
3a053d3 — 2025-11-26 01:12:163.40 (-3.14%)3.38 (-0.82%)3.38 (-0.82%)3.35 (+4.93%)0.02 (-84.51%)2493.81 (-3.14%)2475.27 (-0.82%)2479.06 (-0.82%)2453.92 (+4.93%)13.33 (-84.51%)7001.00 (-4.70%)6940.80 (+0.71%)6930.00 (+0.83%)6889.00 (+3.24%)37.43 (-84.76%)
9e98049 — 2025-11-25 20:50:553.51 (n/a)3.40 (n/a)3.41 (n/a)3.19 (n/a)0.12 (n/a)2574.53 (n/a)2495.72 (n/a)2499.62 (n/a)2338.67 (n/a)86.02 (n/a)7346.00 (n/a)6892.20 (n/a)6873.00 (n/a)6673.00 (n/a)245.57 (n/a)

gemm_2048x2048x2048_64x64x32_8_cols_1_bcolmaj_0_ccolmaj_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)GFLOP/s (max)GFLOP/s (mean)GFLOP/s (median)GFLOP/s (min)GFLOP/s (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
3a053d3 — 2025-11-26 01:12:163.02 (+0.52%)2.87 (-2.05%)2.86 (-2.24%)2.70 (-6.16%)0.11 (+131.66%)2216.19 (+0.52%)2104.44 (-2.05%)2094.08 (-2.24%)1980.84 (-6.16%)80.25 (+131.67%)8673.00 (+6.56%)8175.60 (+2.21%)8204.00 (+2.29%)7752.00 (-0.51%)314.03 (+145.36%)
9e98049 — 2025-11-25 20:50:553.01 (n/a)2.93 (n/a)2.92 (n/a)2.88 (n/a)0.05 (n/a)2204.81 (n/a)2148.41 (n/a)2142.13 (n/a)2110.81 (n/a)34.64 (n/a)8139.00 (n/a)7998.60 (n/a)8020.00 (n/a)7792.00 (n/a)127.99 (n/a)

gemm_2048x2048x2048_64x64x64_1cols

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
8b6d0f4 — 2026-03-25 20:46:390.51 (+0.13%)0.51 (+0.07%)0.51 (+0.04%)0.51 (+0.11%)0.00 (+8.62%)49296.10 (-0.11%)49206.24 (-0.07%)49283.30 (-0.04%)48878.90 (-0.13%)183.15 (+8.40%)351.48 (+0.13%)349.14 (+0.07%)348.59 (+0.04%)348.50 (+0.11%)1.31 (+8.62%)
dc66da7 — 2026-03-25 19:23:450.51 (n/a)0.51 (n/a)0.51 (n/a)0.51 (n/a)0.00 (n/a)49348.10 (n/a)49241.40 (n/a)49302.10 (n/a)48941.60 (n/a)168.96 (n/a)351.03 (n/a)348.89 (n/a)348.46 (n/a)348.14 (n/a)1.20 (n/a)

gemm_2048x2048x2048_64x64x64_2_cols_0_bcolmaj_0_ccolmaj_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
130b6ea — 2025-12-05 21:15:491.04 (+0.94%)1.03 (+0.23%)1.03 (+0.09%)1.02 (+0.07%)0.00 (+197.51%)24576.00 (-0.07%)24485.82 (-0.23%)24530.40 (-0.09%)24266.40 (-0.94%)111.26 (+194.35%)
0a6c11c — 2025-12-04 00:39:101.03 (n/a)1.03 (n/a)1.03 (n/a)1.02 (n/a)0.00 (n/a)24594.30 (n/a)24542.14 (n/a)24551.90 (n/a)24495.50 (n/a)37.80 (n/a)

gemm_2048x2048x2048_64x64x64_2_cols_0_bcolmaj_0_ccolmaj_0_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
a4b6ffe — 2026-01-05 19:24:051.04 (+0.28%)1.03 (-0.15%)1.03 (+0.05%)1.02 (-0.52%)0.01 (+42.72%)24719.80 (+0.52%)24512.42 (+0.16%)24529.20 (-0.05%)24092.60 (-0.28%)252.92 (+43.15%)713.08 (+0.28%)700.92 (-0.15%)700.38 (+0.05%)694.98 (-0.52%)7.30 (+42.71%)
50ad37c — 2025-12-29 20:35:531.04 (n/a)1.03 (n/a)1.03 (n/a)1.02 (n/a)0.01 (n/a)24592.30 (n/a)24473.70 (n/a)24540.90 (n/a)24160.80 (n/a)176.68 (n/a)711.06 (n/a)702.00 (n/a)700.05 (n/a)698.59 (n/a)5.12 (n/a)

gemm_2048x2048x2048_64x64x64_2_cols_0_bcolmaj_1_ccolmaj_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
130b6ea — 2025-12-05 21:15:490.99 (+1.69%)0.97 (+0.29%)0.97 (-0.04%)0.97 (-0.15%)0.01 (+434.92%)26025.30 (+0.15%)25839.34 (-0.29%)25916.70 (+0.04%)25446.00 (-1.67%)202.05 (+425.44%)
0a6c11c — 2025-12-04 00:39:100.97 (n/a)0.97 (n/a)0.97 (n/a)0.97 (n/a)0.00 (n/a)25986.90 (n/a)25913.52 (n/a)25905.70 (n/a)25877.00 (n/a)38.45 (n/a)

gemm_2048x2048x2048_64x64x64_2_cols_0_bcolmaj_1_ccolmaj_0_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
a4b6ffe — 2026-01-05 19:24:050.97 (-0.01%)0.97 (+0.03%)0.97 (+0.04%)0.97 (+0.00%)0.00 (-6.43%)25927.90 (-0.00%)25903.20 (-0.03%)25898.60 (-0.04%)25892.80 (+0.01%)14.05 (-6.56%)663.50 (-0.01%)663.23 (+0.03%)663.35 (+0.04%)662.60 (+0.00%)0.36 (-6.43%)
50ad37c — 2025-12-29 20:35:530.97 (n/a)0.97 (n/a)0.97 (n/a)0.97 (n/a)0.00 (n/a)25928.10 (n/a)25910.14 (n/a)25908.10 (n/a)25891.20 (n/a)15.04 (n/a)663.54 (n/a)663.06 (n/a)663.11 (n/a)662.60 (n/a)0.38 (n/a)

gemm_2048x2048x2048_64x64x64_2_cols_1_bcolmaj_0_ccolmaj_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
130b6ea — 2025-12-05 21:15:490.92 (-1.02%)0.92 (-0.66%)0.92 (-0.10%)0.91 (-0.78%)0.01 (-4.54%)27751.40 (+0.79%)27480.74 (+0.66%)27354.80 (+0.10%)27320.50 (+1.03%)182.69 (-2.84%)
0a6c11c — 2025-12-04 00:39:100.93 (n/a)0.92 (n/a)0.92 (n/a)0.91 (n/a)0.01 (n/a)27534.80 (n/a)27299.32 (n/a)27326.70 (n/a)27042.30 (n/a)188.04 (n/a)

gemm_2048x2048x2048_64x64x64_2_cols_1_bcolmaj_0_ccolmaj_0_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
a4b6ffe — 2026-01-05 19:24:050.92 (+0.11%)0.91 (-0.12%)0.91 (-0.21%)0.91 (-0.24%)0.00 (+30.34%)27723.50 (+0.24%)27548.94 (+0.12%)27522.90 (+0.21%)27403.60 (-0.11%)124.26 (+30.58%)626.92 (+0.11%)623.62 (-0.12%)624.20 (-0.21%)619.69 (-0.24%)2.81 (+30.34%)
50ad37c — 2025-12-29 20:35:530.92 (n/a)0.91 (n/a)0.92 (n/a)0.91 (n/a)0.00 (n/a)27657.10 (n/a)27515.58 (n/a)27464.70 (n/a)27434.60 (n/a)95.16 (n/a)626.21 (n/a)624.37 (n/a)625.52 (n/a)621.17 (n/a)2.16 (n/a)

gemm_2048x2048x2048_64x64x64_2cols_bcolmaj

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
8b6d0f4 — 2026-03-25 20:46:390.88 (+0.21%)0.87 (-0.11%)0.87 (-0.26%)0.87 (+0.03%)0.00 (+24.60%)29019.10 (-0.03%)28862.30 (+0.12%)28862.50 (+0.26%)28665.10 (-0.21%)148.73 (+24.28%)599.33 (+0.21%)595.25 (-0.11%)595.23 (-0.26%)592.02 (+0.03%)3.07 (+24.61%)
dc66da7 — 2026-03-25 19:23:450.88 (n/a)0.87 (n/a)0.87 (n/a)0.87 (n/a)0.00 (n/a)29027.70 (n/a)28829.10 (n/a)28788.30 (n/a)28724.70 (n/a)119.67 (n/a)598.09 (n/a)595.93 (n/a)596.77 (n/a)591.84 (n/a)2.46 (n/a)

gemm_2048x2048x2048_64x64x64_8_cols_0_bcolmaj_0_ccolmaj_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)GFLOP/s (max)GFLOP/s (mean)GFLOP/s (median)GFLOP/s (min)GFLOP/s (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
31ab24e — 2025-11-22 02:01:563.65 (-1.06%)3.59 (+0.83%)3.63 (-0.26%)3.45 (+2.66%)0.07 (-44.71%)2678.08 (-1.06%)2632.13 (+0.83%)2658.60 (-0.26%)2528.31 (+2.66%)53.86 (-44.71%)6795.00 (-2.59%)6529.80 (-0.92%)6462.00 (+0.26%)6415.00 (+1.07%)137.14 (-45.27%)
d9a64c5 — 2025-11-21 20:49:353.69 (n/a)3.56 (n/a)3.64 (n/a)3.36 (n/a)0.13 (n/a)2706.77 (n/a)2610.50 (n/a)2665.61 (n/a)2462.71 (n/a)97.41 (n/a)6976.00 (n/a)6590.40 (n/a)6445.00 (n/a)6347.00 (n/a)250.58 (n/a)

gemm_2048x2048x2048_64x64x64_8_cols_0_bcolmaj_0_ccolmaj_0_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
d48746f — 2026-01-12 22:17:103.98 (+0.20%)3.84 (-2.95%)3.76 (-5.27%)3.72 (-4.97%)0.13 (+473.97%)6760.00 (+5.23%)6559.50 (+3.13%)6697.10 (+5.56%)6321.00 (-0.20%)218.68 (+499.40%)2717.91 (+0.20%)2621.44 (-2.95%)2565.27 (-5.27%)2541.40 (-4.97%)88.37 (+473.97%)
331dcca — 2026-01-08 18:34:103.97 (n/a)3.96 (n/a)3.97 (n/a)3.92 (n/a)0.02 (n/a)6423.80 (n/a)6360.32 (n/a)6344.30 (n/a)6333.90 (n/a)36.48 (n/a)2712.39 (n/a)2701.17 (n/a)2707.91 (n/a)2674.40 (n/a)15.40 (n/a)

gemm_2048x2048x2048_64x64x64_8_cols_0_bcolmaj_1_ccolmaj_0_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
d48746f — 2026-01-12 22:17:103.75 (+0.28%)3.70 (+1.25%)3.71 (-0.21%)3.65 (+4.77%)0.04 (-66.90%)6901.80 (-4.56%)6795.46 (-1.30%)6790.20 (+0.21%)6715.90 (-0.28%)67.05 (-68.40%)2558.08 (+0.28%)2528.33 (+1.25%)2530.10 (-0.21%)2489.18 (+4.77%)24.81 (-66.90%)
331dcca — 2026-01-08 18:34:103.74 (n/a)3.66 (n/a)3.71 (n/a)3.48 (n/a)0.11 (n/a)7231.20 (n/a)6885.18 (n/a)6775.70 (n/a)6734.90 (n/a)212.21 (n/a)2550.87 (n/a)2497.04 (n/a)2535.50 (n/a)2375.78 (n/a)74.96 (n/a)

gemm_2048x2048x2048_64x64x64_8_cols_1_bcolmaj_0_ccolmaj_0_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
d48746f — 2026-01-12 22:17:103.62 (-0.77%)3.54 (+0.14%)3.60 (-0.04%)3.44 (+2.17%)0.09 (-25.77%)7307.40 (-2.13%)7103.46 (-0.18%)6982.50 (+0.04%)6948.20 (+0.77%)182.95 (-26.55%)2472.55 (-0.77%)2419.80 (+0.14%)2460.41 (-0.04%)2351.04 (+2.17%)61.76 (-25.77%)
331dcca — 2026-01-08 18:34:103.65 (n/a)3.54 (n/a)3.61 (n/a)3.37 (n/a)0.12 (n/a)7466.20 (n/a)7116.40 (n/a)6979.60 (n/a)6895.10 (n/a)249.09 (n/a)2491.61 (n/a)2416.45 (n/a)2461.45 (n/a)2301.00 (n/a)83.20 (n/a)

gemm_2048x2048x2048_64x64x64_8cols_bcolmaj_ccolmaj

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
8b6d0f4 — 2026-03-25 20:46:393.47 (-0.81%)3.41 (+0.82%)3.42 (-0.23%)3.35 (+4.63%)0.05 (-63.03%)7522.10 (-4.43%)7390.72 (-0.92%)7367.60 (+0.23%)7255.80 (+0.82%)107.14 (-64.21%)2367.74 (-0.81%)2324.90 (+0.82%)2331.81 (-0.23%)2283.92 (+4.63%)33.69 (-63.03%)
dc66da7 — 2026-03-25 19:23:453.50 (n/a)3.38 (n/a)3.42 (n/a)3.20 (n/a)0.13 (n/a)7870.70 (n/a)7459.22 (n/a)7350.70 (n/a)7196.70 (n/a)299.40 (n/a)2387.20 (n/a)2306.10 (n/a)2337.18 (n/a)2182.77 (n/a)91.13 (n/a)

gemm_384x1536x1792_32x48x64_4cols_bcolmaj

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
8b6d0f4 — 2026-03-25 20:46:394.16 (+12.80%)3.50 (+5.22%)3.50 (+2.51%)2.86 (-0.71%)0.63 (+68.68%)2814.80 (+0.72%)2362.04 (-3.41%)2304.60 (-2.45%)1936.00 (-11.34%)430.86 (+52.19%)1091.93 (+12.80%)919.16 (+5.22%)917.28 (+2.51%)751.01 (-0.71%)166.31 (+68.68%)
dc66da7 — 2026-03-25 19:23:453.69 (n/a)3.33 (n/a)3.41 (n/a)2.88 (n/a)0.38 (n/a)2794.70 (n/a)2445.42 (n/a)2362.50 (n/a)2183.70 (n/a)283.10 (n/a)968.06 (n/a)873.57 (n/a)894.79 (n/a)756.39 (n/a)98.60 (n/a)

gemm_896x1792x640_32x64x80_8cols_ccolmaj

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
8b6d0f4 — 2026-03-25 20:46:396.32 (+6.48%)4.74 (+3.95%)4.60 (+2.04%)3.49 (+3.11%)1.02 (+11.51%)1906.60 (-3.02%)1453.08 (-3.45%)1446.20 (-2.01%)1052.10 (-6.09%)304.01 (+0.60%)1953.36 (+6.48%)1465.54 (+3.95%)1421.06 (+2.04%)1077.93 (+3.11%)313.81 (+11.51%)
dc66da7 — 2026-03-25 19:23:455.94 (n/a)4.56 (n/a)4.51 (n/a)3.38 (n/a)0.91 (n/a)1966.00 (n/a)1504.94 (n/a)1475.80 (n/a)1120.30 (n/a)302.18 (n/a)1834.47 (n/a)1409.81 (n/a)1392.60 (n/a)1045.38 (n/a)281.42 (n/a)

input_length_2048-num_aie_columns_1-num_channels_1-tile_size_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
14eee74 — 2026-03-26 15:31:390.04 (n/a)0.03 (n/a)0.03 (n/a)0.01 (n/a)0.01 (n/a)631.90 (n/a)283.01 (n/a)245.40 (n/a)221.80 (n/a)91.35 (n/a)

input_length_2048-num_aie_columns_1-num_channels_1-tile_size_2048-group_size_32

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
14eee74 — 2026-03-26 15:31:390.02 (n/a)0.02 (n/a)0.02 (n/a)0.02 (n/a)0.00 (n/a)338.20 (n/a)265.34 (n/a)246.90 (n/a)244.90 (n/a)40.80 (n/a)

input_length_2048-num_aie_columns_1-num_channels_1-tile_size_2048-weighted_False

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
14eee74 — 2026-03-26 15:31:390.03 (n/a)0.03 (n/a)0.03 (n/a)0.02 (n/a)0.00 (n/a)365.80 (n/a)276.20 (n/a)249.90 (n/a)246.70 (n/a)50.91 (n/a)

input_length_2048-num_aie_columns_1-num_channels_1-tile_size_2048-weighted_True

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
14eee74 — 2026-03-26 15:31:390.05 (n/a)0.05 (n/a)0.05 (n/a)0.03 (n/a)0.01 (n/a)356.40 (n/a)273.42 (n/a)257.30 (n/a)246.10 (n/a)46.71 (n/a)

input_length_2048-num_aie_columns_1-num_channels_2-tile_size_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
14eee74 — 2026-03-26 15:31:390.04 (n/a)0.03 (n/a)0.03 (n/a)0.02 (n/a)0.00 (n/a)424.30 (n/a)264.64 (n/a)245.95 (n/a)231.70 (n/a)58.18 (n/a)

input_length_2048-num_aie_columns_1-num_channels_2-tile_size_1024-group_size_32

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
14eee74 — 2026-03-26 15:31:390.02 (n/a)0.02 (n/a)0.02 (n/a)0.02 (n/a)0.00 (n/a)332.80 (n/a)276.64 (n/a)250.70 (n/a)240.60 (n/a)43.43 (n/a)

input_length_2048-num_aie_columns_1-num_channels_2-tile_size_1024-weighted_False

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
14eee74 — 2026-03-26 15:31:390.03 (n/a)0.03 (n/a)0.03 (n/a)0.03 (n/a)0.00 (n/a)319.60 (n/a)259.84 (n/a)247.30 (n/a)236.40 (n/a)33.77 (n/a)

input_length_2048-num_aie_columns_1-num_channels_2-tile_size_2048-scalar_factor_3.0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
14eee74 — 2026-03-26 15:31:390.05 (n/a)0.05 (n/a)0.05 (n/a)0.02 (n/a)0.02 (n/a)691.10 (n/a)325.28 (n/a)236.80 (n/a)225.70 (n/a)204.62 (n/a)

input_length_2048-num_aie_columns_1-tile_size_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
14eee74 — 2026-03-26 15:31:390.05 (n/a)0.04 (n/a)0.04 (n/a)0.02 (n/a)0.01 (n/a)452.70 (n/a)284.11 (n/a)244.30 (n/a)225.90 (n/a)69.75 (n/a)

input_length_2048-num_aie_columns_2-num_channels_1-tile_size_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
14eee74 — 2026-03-26 15:31:390.04 (n/a)0.03 (n/a)0.03 (n/a)0.01 (n/a)0.01 (n/a)708.60 (n/a)295.97 (n/a)249.00 (n/a)218.20 (n/a)118.10 (n/a)

input_length_2048-num_aie_columns_2-num_channels_1-tile_size_1024-group_size_32

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
14eee74 — 2026-03-26 15:31:390.02 (n/a)0.02 (n/a)0.02 (n/a)0.02 (n/a)0.00 (n/a)324.10 (n/a)267.36 (n/a)254.40 (n/a)248.90 (n/a)31.97 (n/a)

input_length_2048-num_aie_columns_2-num_channels_1-tile_size_1024-weighted_False

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
14eee74 — 2026-03-26 15:31:390.03 (n/a)0.03 (n/a)0.03 (n/a)0.02 (n/a)0.00 (n/a)342.30 (n/a)273.08 (n/a)248.80 (n/a)238.70 (n/a)43.04 (n/a)

input_length_2048-num_aie_columns_2-num_channels_1-tile_size_1024-weighted_True

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
14eee74 — 2026-03-26 15:31:390.04 (n/a)0.04 (n/a)0.04 (n/a)0.03 (n/a)0.01 (n/a)359.50 (n/a)271.96 (n/a)255.20 (n/a)242.80 (n/a)49.29 (n/a)

input_length_2048-num_aie_columns_2-num_channels_2-tile_size_1024-scalar_factor_3.0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
14eee74 — 2026-03-26 15:31:390.05 (n/a)0.04 (n/a)0.05 (n/a)0.02 (n/a)0.01 (n/a)587.90 (n/a)318.90 (n/a)251.50 (n/a)247.10 (n/a)150.42 (n/a)

input_length_2048-num_aie_columns_2-num_channels_2-tile_size_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
14eee74 — 2026-03-26 15:31:390.04 (n/a)0.03 (n/a)0.03 (n/a)0.01 (n/a)0.01 (n/a)606.10 (n/a)291.42 (n/a)241.95 (n/a)229.70 (n/a)121.21 (n/a)

input_length_2048-num_aie_columns_2-num_channels_2-tile_size_512-group_size_32

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
14eee74 — 2026-03-26 15:31:390.02 (n/a)0.02 (n/a)0.02 (n/a)0.02 (n/a)0.00 (n/a)317.90 (n/a)246.78 (n/a)229.60 (n/a)227.20 (n/a)39.78 (n/a)

input_length_2048-num_aie_columns_2-num_channels_2-tile_size_512-weighted_False

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
14eee74 — 2026-03-26 15:31:390.04 (n/a)0.03 (n/a)0.03 (n/a)0.02 (n/a)0.01 (n/a)351.60 (n/a)260.96 (n/a)238.70 (n/a)231.60 (n/a)51.24 (n/a)

input_length_2048-num_aie_columns_2-tile_size_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
14eee74 — 2026-03-26 15:31:390.06 (n/a)0.04 (n/a)0.04 (n/a)0.01 (n/a)0.02 (n/a)714.90 (n/a)330.51 (n/a)241.20 (n/a)221.90 (n/a)184.14 (n/a)

input_length_2048-num_aie_columns_4-num_channels_1-tile_size_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
14eee74 — 2026-03-26 15:31:390.04 (n/a)0.03 (n/a)0.03 (n/a)0.01 (n/a)0.01 (n/a)642.10 (n/a)302.31 (n/a)252.10 (n/a)229.60 (n/a)105.21 (n/a)

input_length_2048-num_aie_columns_4-num_channels_1-tile_size_512-group_size_32

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
14eee74 — 2026-03-26 15:31:390.02 (n/a)0.02 (n/a)0.02 (n/a)0.01 (n/a)0.01 (n/a)745.20 (n/a)348.26 (n/a)252.30 (n/a)244.00 (n/a)221.93 (n/a)

input_length_2048-num_aie_columns_4-num_channels_1-tile_size_512-weighted_False

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
14eee74 — 2026-03-26 15:31:390.03 (n/a)0.03 (n/a)0.03 (n/a)0.02 (n/a)0.01 (n/a)468.30 (n/a)292.60 (n/a)247.20 (n/a)246.90 (n/a)98.26 (n/a)

input_length_2048-num_aie_columns_4-num_channels_1-tile_size_512-weighted_True

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
14eee74 — 2026-03-26 15:31:390.04 (n/a)0.03 (n/a)0.04 (n/a)0.01 (n/a)0.01 (n/a)632.50 (n/a)310.72 (n/a)231.60 (n/a)225.80 (n/a)179.90 (n/a)

input_length_2048-num_aie_columns_4-num_channels_2-tile_size_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
14eee74 — 2026-03-26 15:31:390.03 (n/a)0.03 (n/a)0.03 (n/a)0.01 (n/a)0.01 (n/a)694.80 (n/a)340.30 (n/a)257.80 (n/a)240.40 (n/a)181.80 (n/a)

input_length_2048-num_aie_columns_4-num_channels_2-tile_size_256-group_size_32

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
14eee74 — 2026-03-26 15:31:390.02 (n/a)0.02 (n/a)0.02 (n/a)0.01 (n/a)0.01 (n/a)599.80 (n/a)308.20 (n/a)234.80 (n/a)234.00 (n/a)163.02 (n/a)

input_length_2048-num_aie_columns_4-num_channels_2-tile_size_256-weighted_False

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
14eee74 — 2026-03-26 15:31:390.03 (n/a)0.03 (n/a)0.03 (n/a)0.01 (n/a)0.01 (n/a)676.40 (n/a)341.04 (n/a)258.30 (n/a)255.20 (n/a)187.48 (n/a)

input_length_2048-num_aie_columns_4-num_channels_2-tile_size_512-scalar_factor_3.0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
14eee74 — 2026-03-26 15:31:390.05 (n/a)0.05 (n/a)0.05 (n/a)0.02 (n/a)0.01 (n/a)580.70 (n/a)300.20 (n/a)230.30 (n/a)224.00 (n/a)156.88 (n/a)

input_length_2048-num_aie_columns_4-tile_size_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
14eee74 — 2026-03-26 15:31:390.05 (n/a)0.04 (n/a)0.04 (n/a)0.02 (n/a)0.01 (n/a)655.00 (n/a)285.93 (n/a)245.10 (n/a)225.00 (n/a)107.95 (n/a)

input_length_2048-num_aie_columns_8-num_channels_1-tile_size_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
14eee74 — 2026-03-26 15:31:390.04 (n/a)0.03 (n/a)0.03 (n/a)0.01 (n/a)0.01 (n/a)969.40 (n/a)327.08 (n/a)257.10 (n/a)230.60 (n/a)181.03 (n/a)

input_length_2048-num_aie_columns_8-num_channels_1-tile_size_256-group_size_32

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
14eee74 — 2026-03-26 15:31:390.02 (n/a)0.02 (n/a)0.02 (n/a)0.01 (n/a)0.01 (n/a)572.00 (n/a)322.10 (n/a)255.80 (n/a)248.60 (n/a)140.15 (n/a)

input_length_2048-num_aie_columns_8-num_channels_1-tile_size_256-weighted_False

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
14eee74 — 2026-03-26 15:31:390.03 (n/a)0.03 (n/a)0.03 (n/a)0.02 (n/a)0.01 (n/a)498.50 (n/a)293.64 (n/a)242.00 (n/a)239.60 (n/a)114.58 (n/a)

input_length_2048-num_aie_columns_8-num_channels_1-tile_size_256-weighted_True

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
14eee74 — 2026-03-26 15:31:390.03 (n/a)0.03 (n/a)0.03 (n/a)0.01 (n/a)0.01 (n/a)920.00 (n/a)391.22 (n/a)259.80 (n/a)257.90 (n/a)295.60 (n/a)

input_length_2048-num_aie_columns_8-num_channels_2-tile_size_128

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
14eee74 — 2026-03-26 15:31:390.03 (n/a)0.03 (n/a)0.03 (n/a)0.01 (n/a)0.01 (n/a)680.50 (n/a)346.85 (n/a)272.25 (n/a)250.80 (n/a)160.42 (n/a)

input_length_2048-num_aie_columns_8-num_channels_2-tile_size_128-group_size_32

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
14eee74 — 2026-03-26 15:31:390.02 (n/a)0.02 (n/a)0.02 (n/a)0.02 (n/a)0.00 (n/a)317.60 (n/a)291.70 (n/a)278.90 (n/a)269.10 (n/a)22.60 (n/a)

input_length_2048-num_aie_columns_8-num_channels_2-tile_size_128-weighted_False

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
14eee74 — 2026-03-26 15:31:390.03 (n/a)0.03 (n/a)0.03 (n/a)0.01 (n/a)0.01 (n/a)691.70 (n/a)341.48 (n/a)254.40 (n/a)251.70 (n/a)195.79 (n/a)

input_length_2048-num_aie_columns_8-num_channels_2-tile_size_256-scalar_factor_3.0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
14eee74 — 2026-03-26 15:31:390.05 (n/a)0.04 (n/a)0.05 (n/a)0.02 (n/a)0.01 (n/a)685.40 (n/a)353.90 (n/a)261.10 (n/a)258.70 (n/a)186.38 (n/a)

input_length_2048-num_aie_columns_8-tile_size_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
14eee74 — 2026-03-26 15:31:390.05 (n/a)0.04 (n/a)0.04 (n/a)0.01 (n/a)0.01 (n/a)929.30 (n/a)305.93 (n/a)257.50 (n/a)231.40 (n/a)175.80 (n/a)

input_length_2048-num_cores_1-num_channels_1-bypass_False-tile_size_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
14eee74 — 2026-03-26 15:31:390.03 (n/a)0.03 (n/a)0.03 (n/a)0.02 (n/a)0.00 (n/a)343.50 (n/a)272.38 (n/a)257.90 (n/a)243.40 (n/a)41.06 (n/a)

input_length_2048-num_cores_16-num_channels_2-bypass_False-tile_size_128

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
14eee74 — 2026-03-26 15:31:390.03 (n/a)0.03 (n/a)0.03 (n/a)0.02 (n/a)0.00 (n/a)410.60 (n/a)299.36 (n/a)272.20 (n/a)267.50 (n/a)62.30 (n/a)

input_length_2048-num_cores_2-num_channels_1-bypass_False-tile_size_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
14eee74 — 2026-03-26 15:31:390.04 (n/a)0.03 (n/a)0.04 (n/a)0.01 (n/a)0.01 (n/a)706.30 (n/a)348.46 (n/a)228.70 (n/a)221.60 (n/a)209.03 (n/a)

input_length_2048-num_cores_2-num_channels_2-bypass_False-tile_size_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
14eee74 — 2026-03-26 15:31:390.04 (n/a)0.03 (n/a)0.04 (n/a)0.02 (n/a)0.01 (n/a)409.10 (n/a)270.74 (n/a)227.70 (n/a)223.20 (n/a)79.41 (n/a)

input_length_2048-num_cores_4-num_channels_1-bypass_False-tile_size_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
14eee74 — 2026-03-26 15:31:390.04 (n/a)0.03 (n/a)0.03 (n/a)0.01 (n/a)0.01 (n/a)834.60 (n/a)390.94 (n/a)257.30 (n/a)231.50 (n/a)256.94 (n/a)

input_length_2048-num_cores_4-num_channels_2-bypass_False-tile_size_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
14eee74 — 2026-03-26 15:31:390.03 (n/a)0.03 (n/a)0.03 (n/a)0.02 (n/a)0.00 (n/a)395.40 (n/a)297.36 (n/a)277.50 (n/a)245.00 (n/a)59.52 (n/a)

input_length_2048-num_cores_8-num_channels_1-bypass_False-tile_size_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
14eee74 — 2026-03-26 15:31:390.03 (n/a)0.03 (n/a)0.03 (n/a)0.01 (n/a)0.01 (n/a)1010.20 (n/a)410.64 (n/a)259.40 (n/a)255.10 (n/a)335.22 (n/a)

input_length_2048-num_cores_8-num_channels_2-bypass_False-tile_size_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
14eee74 — 2026-03-26 15:31:390.03 (n/a)0.03 (n/a)0.03 (n/a)0.01 (n/a)0.01 (n/a)727.30 (n/a)358.04 (n/a)261.80 (n/a)255.60 (n/a)206.81 (n/a)

input_length_32768-num_aie_columns_2-num_channels_2-tile_size_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
14eee74 — 2026-03-26 15:31:390.54 (n/a)0.37 (n/a)0.31 (n/a)0.30 (n/a)0.10 (n/a)430.80 (n/a)371.62 (n/a)422.70 (n/a)242.90 (n/a)82.87 (n/a)

input_length_32768-num_aie_columns_2-num_channels_2-tile_size_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
14eee74 — 2026-03-26 15:31:390.46 (n/a)0.32 (n/a)0.32 (n/a)0.18 (n/a)0.10 (n/a)722.10 (n/a)442.58 (n/a)414.70 (n/a)286.50 (n/a)166.03 (n/a)

input_length_32768-num_aie_columns_4-num_channels_4-tile_size_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
14eee74 — 2026-03-26 15:31:390.47 (n/a)0.37 (n/a)0.40 (n/a)0.18 (n/a)0.12 (n/a)719.40 (n/a)399.60 (n/a)325.80 (n/a)281.00 (n/a)184.19 (n/a)

layer_norm_1_cols_1_channels_2048_tile_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.07 (+35.61%)0.05 (+6.58%)0.04 (-0.65%)0.03 (-1.11%)0.01 (+98.33%)236.50 (+1.11%)182.90 (-3.12%)183.30 (+0.60%)118.50 (-26.26%)42.30 (+41.46%)
dc66da7 — 2026-03-25 19:23:450.05 (n/a)0.04 (n/a)0.04 (n/a)0.04 (n/a)0.01 (n/a)233.90 (n/a)188.80 (n/a)182.20 (n/a)160.70 (n/a)29.90 (n/a)

layer_norm_1_cols_2_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.07 (+9.47%)0.05 (+5.79%)0.05 (+5.48%)0.04 (-7.03%)0.01 (+51.12%)206.90 (+7.54%)163.06 (-2.60%)165.20 (-5.22%)113.20 (-8.71%)40.37 (+56.59%)
dc66da7 — 2026-03-25 19:23:450.07 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.01 (n/a)192.40 (n/a)167.42 (n/a)174.30 (n/a)124.00 (n/a)25.78 (n/a)

layer_norm_2_cols_1_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.06 (-0.97%)0.05 (+14.59%)0.05 (+47.43%)0.03 (+8.92%)0.01 (-0.31%)235.10 (-8.20%)183.06 (-13.03%)157.80 (-32.19%)148.10 (+1.02%)43.12 (-7.94%)
dc66da7 — 2026-03-25 19:23:450.06 (n/a)0.04 (n/a)0.04 (n/a)0.03 (n/a)0.01 (n/a)256.10 (n/a)210.48 (n/a)232.70 (n/a)146.60 (n/a)46.84 (n/a)

layer_norm_2_cols_2_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.07 (+14.65%)0.05 (+4.61%)0.05 (-7.99%)0.04 (+11.94%)0.01 (+28.58%)201.70 (-10.67%)168.58 (-3.65%)176.00 (+8.71%)122.10 (-12.79%)33.82 (+0.06%)
dc66da7 — 2026-03-25 19:23:450.06 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.01 (n/a)225.80 (n/a)174.96 (n/a)161.90 (n/a)140.00 (n/a)33.80 (n/a)

layer_norm_4_cols_1_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.06 (-31.05%)0.05 (-3.73%)0.05 (+9.03%)0.04 (+10.49%)0.01 (-63.42%)197.60 (-9.52%)167.50 (-4.83%)166.40 (-8.27%)129.00 (+45.11%)25.47 (-51.02%)
dc66da7 — 2026-03-25 19:23:450.09 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.02 (n/a)218.40 (n/a)176.00 (n/a)181.40 (n/a)88.90 (n/a)52.00 (n/a)

layer_norm_4_cols_2_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.05 (-1.52%)0.04 (-11.20%)0.04 (-25.45%)0.03 (-0.69%)0.01 (-9.35%)243.40 (+0.66%)202.20 (+12.10%)212.70 (+34.20%)156.90 (+1.55%)34.18 (-7.92%)
dc66da7 — 2026-03-25 19:23:450.05 (n/a)0.05 (n/a)0.05 (n/a)0.03 (n/a)0.01 (n/a)241.80 (n/a)180.38 (n/a)158.50 (n/a)154.50 (n/a)37.12 (n/a)

layer_norm_8_cols_1_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.05 (-6.07%)0.04 (+15.60%)0.04 (+21.55%)0.04 (+82.28%)0.01 (-55.28%)222.00 (-45.13%)190.18 (-21.85%)193.60 (-17.72%)153.10 (+6.47%)24.64 (-75.05%)
dc66da7 — 2026-03-25 19:23:450.06 (n/a)0.04 (n/a)0.03 (n/a)0.02 (n/a)0.01 (n/a)404.60 (n/a)243.34 (n/a)235.30 (n/a)143.80 (n/a)98.74 (n/a)

layer_norm_8_cols_2_channels_2048_tile_128

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.05 (+37.82%)0.04 (+27.22%)0.04 (+9.46%)0.03 (+34.60%)0.01 (+56.79%)251.50 (-25.72%)212.60 (-21.06%)226.70 (-8.66%)175.20 (-27.45%)33.04 (-18.34%)
dc66da7 — 2026-03-25 19:23:450.03 (n/a)0.03 (n/a)0.03 (n/a)0.02 (n/a)0.00 (n/a)338.60 (n/a)269.32 (n/a)248.20 (n/a)241.50 (n/a)40.46 (n/a)

matrix_vector_mul_128x128_32_1col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)
130b6ea — 2025-12-05 21:15:490.37 (+17.13%)0.30 (+24.87%)0.30 (+38.03%)0.23 (+12.47%)0.06 (+35.23%)
0a6c11c — 2025-12-04 00:39:100.32 (n/a)0.24 (n/a)0.22 (n/a)0.20 (n/a)0.04 (n/a)

matrix_vector_mul_128x128_32_1col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
331dcca — 2026-01-08 18:34:100.26 (-1.81%)0.24 (+4.06%)0.24 (+0.52%)0.21 (+16.97%)0.02 (-48.16%)0.26 (-1.81%)0.24 (+4.06%)0.24 (+0.52%)0.20 (+16.97%)0.02 (-48.16%)
a4b6ffe — 2026-01-05 19:24:050.26 (n/a)0.23 (n/a)0.24 (n/a)0.18 (n/a)0.04 (n/a)0.26 (n/a)0.23 (n/a)0.24 (n/a)0.17 (n/a)0.04 (n/a)

matrix_vector_mul_128x128_32tsi_128tso_1col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
8b6d0f4 — 2026-03-25 20:46:390.26 (+32.10%)0.19 (+11.55%)0.18 (+9.49%)0.13 (-2.22%)0.05 (+105.83%)0.25 (+32.10%)0.19 (+11.55%)0.18 (+9.49%)0.13 (-2.22%)0.05 (+105.83%)
dc66da7 — 2026-03-25 19:23:450.19 (n/a)0.17 (n/a)0.17 (n/a)0.14 (n/a)0.02 (n/a)0.19 (n/a)0.17 (n/a)0.17 (n/a)0.14 (n/a)0.02 (n/a)

matrix_vector_mul_128x128_32tsi_128tso_1col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
897d04e — 2026-03-06 22:12:250.37 (+5.44%)0.31 (-0.05%)0.34 (-1.99%)0.22 (-7.87%)0.06 (+15.13%)0.36 (+5.44%)0.31 (-0.05%)0.34 (-1.99%)0.21 (-7.87%)0.06 (+15.13%)
84d3478 — 2026-02-17 23:25:130.35 (n/a)0.31 (n/a)0.35 (n/a)0.23 (n/a)0.05 (n/a)0.35 (n/a)0.31 (n/a)0.34 (n/a)0.23 (n/a)0.05 (n/a)

matrix_vector_mul_2048x8192_1_1col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)
130b6ea — 2025-12-05 21:15:4913.57 (+3.47%)13.15 (+1.09%)13.10 (+0.16%)12.52 (-1.58%)0.37 (+153.19%)
0a6c11c — 2025-12-04 00:39:1013.11 (n/a)13.01 (n/a)13.07 (n/a)12.72 (n/a)0.15 (n/a)

matrix_vector_mul_2048x8192_1_1col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
331dcca — 2026-01-08 18:34:1013.43 (-0.07%)13.30 (+4.62%)13.27 (+1.29%)13.16 (+16.70%)0.12 (-86.69%)13.42 (-0.07%)13.29 (+4.62%)13.26 (+1.29%)13.16 (+16.70%)0.12 (-86.69%)
a4b6ffe — 2026-01-05 19:24:0513.44 (n/a)12.71 (n/a)13.10 (n/a)11.28 (n/a)0.87 (n/a)13.43 (n/a)12.70 (n/a)13.09 (n/a)11.27 (n/a)0.87 (n/a)

matrix_vector_mul_2048x8192_1_2col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)
130b6ea — 2025-12-05 21:15:4925.06 (+2.47%)23.71 (+1.81%)24.13 (+0.50%)20.90 (+5.41%)1.48 (-15.60%)
0a6c11c — 2025-12-04 00:39:1024.46 (n/a)23.29 (n/a)24.01 (n/a)19.83 (n/a)1.75 (n/a)

matrix_vector_mul_2048x8192_1_2col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
331dcca — 2026-01-08 18:34:1024.46 (-0.60%)21.07 (-8.03%)19.75 (-17.83%)18.08 (+1.18%)3.03 (+7.07%)24.44 (-0.60%)21.06 (-8.03%)19.74 (-17.83%)18.07 (+1.18%)3.03 (+7.07%)
a4b6ffe — 2026-01-05 19:24:0524.60 (n/a)22.91 (n/a)24.04 (n/a)17.87 (n/a)2.83 (n/a)24.59 (n/a)22.90 (n/a)24.02 (n/a)17.86 (n/a)2.83 (n/a)

matrix_vector_mul_2048x8192_1_4col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)
130b6ea — 2025-12-05 21:15:4942.05 (+1.54%)40.50 (+12.60%)40.49 (+2.36%)38.98 (+97.53%)0.97 (-88.09%)
0a6c11c — 2025-12-04 00:39:1041.41 (n/a)35.97 (n/a)39.55 (n/a)19.73 (n/a)8.15 (n/a)

matrix_vector_mul_2048x8192_1_4col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
331dcca — 2026-01-08 18:34:1042.58 (+1.51%)41.37 (+14.29%)41.25 (+4.89%)40.33 (+91.42%)0.92 (-89.18%)42.56 (+1.51%)41.34 (+14.29%)41.23 (+4.89%)40.30 (+91.42%)0.92 (-89.18%)
a4b6ffe — 2026-01-05 19:24:0541.95 (n/a)36.20 (n/a)39.33 (n/a)21.07 (n/a)8.54 (n/a)41.92 (n/a)36.17 (n/a)39.31 (n/a)21.05 (n/a)8.53 (n/a)

matrix_vector_mul_2048x8192_1_8col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)
130b6ea — 2025-12-05 21:15:4946.62 (+3.11%)44.82 (+0.17%)44.78 (+0.17%)42.60 (-3.80%)1.60 (+367.72%)
0a6c11c — 2025-12-04 00:39:1045.21 (n/a)44.75 (n/a)44.70 (n/a)44.29 (n/a)0.34 (n/a)

matrix_vector_mul_2048x8192_1_8col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
331dcca — 2026-01-08 18:34:1046.04 (+4.14%)45.06 (+3.47%)44.55 (+2.76%)44.30 (+2.84%)0.87 (+66.58%)46.01 (+4.14%)45.03 (+3.47%)44.52 (+2.76%)44.27 (+2.84%)0.87 (+66.58%)
a4b6ffe — 2026-01-05 19:24:0544.21 (n/a)43.55 (n/a)43.35 (n/a)43.07 (n/a)0.52 (n/a)44.18 (n/a)43.52 (n/a)43.32 (n/a)43.05 (n/a)0.52 (n/a)

matrix_vector_mul_2048x8192_1tsi_1024tso_2col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
8b6d0f4 — 2026-03-25 20:46:3925.44 (+1.95%)24.42 (-0.28%)24.79 (+0.16%)23.14 (-2.33%)1.05 (+106.47%)25.42 (+1.95%)24.41 (-0.28%)24.77 (+0.16%)23.12 (-2.33%)1.05 (+106.47%)
dc66da7 — 2026-03-25 19:23:4524.95 (n/a)24.49 (n/a)24.75 (n/a)23.69 (n/a)0.51 (n/a)24.94 (n/a)24.48 (n/a)24.73 (n/a)23.67 (n/a)0.51 (n/a)

matrix_vector_mul_2048x8192_1tsi_1024tso_2col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
897d04e — 2026-03-06 22:12:2524.47 (-3.32%)24.16 (-3.13%)24.20 (-3.09%)23.84 (-3.04%)0.25 (-14.36%)24.46 (-3.32%)24.15 (-3.13%)24.18 (-3.09%)23.83 (-3.04%)0.25 (-14.36%)
84d3478 — 2026-02-17 23:25:1325.31 (n/a)24.94 (n/a)24.97 (n/a)24.59 (n/a)0.29 (n/a)25.30 (n/a)24.93 (n/a)24.95 (n/a)24.58 (n/a)0.29 (n/a)

matrix_vector_mul_2048x8192_1tsi_2048tso_1col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
8b6d0f4 — 2026-03-25 20:46:3913.48 (-1.01%)13.15 (+1.31%)13.11 (-0.87%)12.95 (+6.77%)0.20 (-66.05%)13.47 (-1.01%)13.14 (+1.31%)13.11 (-0.87%)12.95 (+6.77%)0.20 (-66.05%)
dc66da7 — 2026-03-25 19:23:4513.62 (n/a)12.98 (n/a)13.23 (n/a)12.13 (n/a)0.58 (n/a)13.61 (n/a)12.97 (n/a)13.22 (n/a)12.13 (n/a)0.58 (n/a)

matrix_vector_mul_2048x8192_1tsi_2048tso_1col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
897d04e — 2026-03-06 22:12:2513.25 (-3.04%)13.02 (-2.92%)13.10 (-2.64%)12.56 (-4.14%)0.27 (+32.55%)13.24 (-3.04%)13.01 (-2.92%)13.09 (-2.64%)12.55 (-4.14%)0.27 (+32.55%)
84d3478 — 2026-02-17 23:25:1313.66 (n/a)13.41 (n/a)13.45 (n/a)13.10 (n/a)0.20 (n/a)13.65 (n/a)13.40 (n/a)13.44 (n/a)13.09 (n/a)0.20 (n/a)

matrix_vector_mul_2048x8192_1tsi_256tso_8col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
8b6d0f4 — 2026-03-25 20:46:3942.06 (-4.87%)40.85 (-4.53%)40.57 (-4.52%)39.78 (-4.49%)0.87 (-11.49%)42.04 (-4.87%)40.83 (-4.53%)40.55 (-4.52%)39.76 (-4.49%)0.87 (-11.49%)
dc66da7 — 2026-03-25 19:23:4544.21 (n/a)42.79 (n/a)42.49 (n/a)41.65 (n/a)0.99 (n/a)44.19 (n/a)42.76 (n/a)42.47 (n/a)41.63 (n/a)0.99 (n/a)

matrix_vector_mul_2048x8192_1tsi_256tso_8col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
897d04e — 2026-03-06 22:12:2547.71 (+3.51%)45.38 (+9.55%)45.22 (-1.27%)43.47 (+75.24%)1.52 (-83.67%)47.68 (+3.51%)45.35 (+9.55%)45.19 (-1.27%)43.45 (+75.24%)1.52 (-83.67%)
84d3478 — 2026-02-17 23:25:1346.09 (n/a)41.43 (n/a)45.80 (n/a)24.81 (n/a)9.31 (n/a)46.06 (n/a)41.40 (n/a)45.77 (n/a)24.79 (n/a)9.30 (n/a)

matrix_vector_mul_2048x8192_1tsi_512tso_4col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
8b6d0f4 — 2026-03-25 20:46:3940.94 (-0.67%)39.64 (+0.12%)39.96 (+1.02%)37.79 (+0.51%)1.25 (-21.17%)40.91 (-0.67%)39.61 (+0.12%)39.94 (+1.02%)37.77 (+0.51%)1.25 (-21.17%)
dc66da7 — 2026-03-25 19:23:4541.21 (n/a)39.59 (n/a)39.56 (n/a)37.60 (n/a)1.58 (n/a)41.19 (n/a)39.57 (n/a)39.53 (n/a)37.58 (n/a)1.58 (n/a)

matrix_vector_mul_2048x8192_1tsi_512tso_4col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
897d04e — 2026-03-06 22:12:2540.84 (-0.32%)39.53 (-1.25%)39.40 (-1.97%)38.87 (-0.69%)0.77 (-6.24%)40.81 (-0.32%)39.50 (-1.25%)39.37 (-1.97%)38.84 (-0.69%)0.77 (-6.24%)
84d3478 — 2026-02-17 23:25:1340.97 (n/a)40.03 (n/a)40.19 (n/a)39.14 (n/a)0.82 (n/a)40.95 (n/a)40.00 (n/a)40.17 (n/a)39.11 (n/a)0.82 (n/a)

matrix_vector_mul_8192x2048_4_1col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)
130b6ea — 2025-12-05 21:15:4913.29 (+1.63%)13.19 (+6.08%)13.19 (+1.17%)13.09 (+26.86%)0.08 (-92.94%)
0a6c11c — 2025-12-04 00:39:1013.08 (n/a)12.44 (n/a)13.03 (n/a)10.32 (n/a)1.07 (n/a)

matrix_vector_mul_8192x2048_4_1col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
331dcca — 2026-01-08 18:34:1013.68 (+3.42%)13.30 (+8.54%)13.38 (+7.25%)12.72 (+14.96%)0.35 (-66.09%)13.67 (+3.42%)13.29 (+8.54%)13.38 (+7.25%)12.71 (+14.96%)0.35 (-66.09%)
a4b6ffe — 2026-01-05 19:24:0513.23 (n/a)12.25 (n/a)12.48 (n/a)11.07 (n/a)1.04 (n/a)13.22 (n/a)12.24 (n/a)12.47 (n/a)11.06 (n/a)1.04 (n/a)

matrix_vector_mul_8192x2048_4_2col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)
130b6ea — 2025-12-05 21:15:4925.58 (+4.24%)24.23 (+6.59%)24.03 (+0.29%)23.58 (+37.29%)0.70 (-74.97%)
0a6c11c — 2025-12-04 00:39:1024.54 (n/a)22.73 (n/a)23.96 (n/a)17.18 (n/a)2.79 (n/a)

matrix_vector_mul_8192x2048_4_2col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
331dcca — 2026-01-08 18:34:1025.08 (+1.32%)24.68 (+13.42%)24.61 (+3.26%)24.47 (+48.32%)0.24 (-93.56%)25.07 (+1.32%)24.67 (+13.42%)24.59 (+3.26%)24.46 (+48.32%)0.24 (-93.56%)
a4b6ffe — 2026-01-05 19:24:0524.76 (n/a)21.76 (n/a)23.83 (n/a)16.50 (n/a)3.74 (n/a)24.74 (n/a)21.75 (n/a)23.82 (n/a)16.49 (n/a)3.73 (n/a)

matrix_vector_mul_8192x2048_4_4col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)
130b6ea — 2025-12-05 21:15:4940.94 (-2.19%)39.74 (-1.10%)39.76 (-0.60%)38.17 (-2.14%)1.04 (+4.39%)
0a6c11c — 2025-12-04 00:39:1041.86 (n/a)40.18 (n/a)40.00 (n/a)39.00 (n/a)1.00 (n/a)

matrix_vector_mul_8192x2048_4_4col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
331dcca — 2026-01-08 18:34:1040.74 (-1.09%)37.16 (-7.15%)40.25 (+1.47%)24.51 (-37.44%)7.07 (+736.13%)40.72 (-1.09%)37.13 (-7.15%)40.23 (+1.47%)24.50 (-37.44%)7.07 (+736.13%)
a4b6ffe — 2026-01-05 19:24:0541.19 (n/a)40.02 (n/a)39.67 (n/a)39.18 (n/a)0.85 (n/a)41.17 (n/a)39.99 (n/a)39.64 (n/a)39.16 (n/a)0.85 (n/a)

matrix_vector_mul_8192x2048_4_8col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)
130b6ea — 2025-12-05 21:15:4946.71 (-1.04%)44.91 (-2.34%)44.71 (-2.98%)43.80 (-1.44%)0.96 (+6.93%)
0a6c11c — 2025-12-04 00:39:1047.20 (n/a)45.98 (n/a)46.08 (n/a)44.44 (n/a)0.90 (n/a)

matrix_vector_mul_8192x2048_4_8col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
331dcca — 2026-01-08 18:34:1046.28 (+5.68%)44.89 (+14.59%)44.74 (+4.26%)44.20 (+62.80%)0.84 (-87.96%)46.25 (+5.68%)44.86 (+14.59%)44.71 (+4.26%)44.17 (+62.80%)0.84 (-87.96%)
a4b6ffe — 2026-01-05 19:24:0543.79 (n/a)39.17 (n/a)42.91 (n/a)27.15 (n/a)6.98 (n/a)43.77 (n/a)39.15 (n/a)42.88 (n/a)27.13 (n/a)6.98 (n/a)

matrix_vector_mul_8192x2048_4tsi_1024tso_1col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
8b6d0f4 — 2026-03-25 20:46:3913.35 (+0.83%)13.11 (+3.52%)13.10 (+3.94%)12.94 (+7.70%)0.15 (-69.46%)13.34 (+0.83%)13.10 (+3.52%)13.09 (+3.94%)12.94 (+7.70%)0.15 (-69.46%)
dc66da7 — 2026-03-25 19:23:4513.24 (n/a)12.66 (n/a)12.60 (n/a)12.02 (n/a)0.49 (n/a)13.23 (n/a)12.66 (n/a)12.59 (n/a)12.01 (n/a)0.49 (n/a)

matrix_vector_mul_8192x2048_4tsi_1024tso_1col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
897d04e — 2026-03-06 22:12:2513.34 (-0.79%)13.20 (-0.36%)13.18 (-1.55%)13.12 (+1.75%)0.08 (-65.93%)13.33 (-0.79%)13.19 (-0.36%)13.17 (-1.55%)13.11 (+1.75%)0.08 (-65.93%)
84d3478 — 2026-02-17 23:25:1313.45 (n/a)13.25 (n/a)13.39 (n/a)12.90 (n/a)0.25 (n/a)13.44 (n/a)13.24 (n/a)13.38 (n/a)12.89 (n/a)0.25 (n/a)

matrix_vector_mul_8192x2048_4tsi_1024tso_2col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
8b6d0f4 — 2026-03-25 20:46:3924.35 (-0.26%)24.07 (+0.89%)24.16 (+1.48%)23.64 (+2.17%)0.27 (-44.88%)24.34 (-0.26%)24.06 (+0.89%)24.15 (+1.48%)23.63 (+2.17%)0.27 (-44.88%)
dc66da7 — 2026-03-25 19:23:4524.41 (n/a)23.86 (n/a)23.81 (n/a)23.14 (n/a)0.49 (n/a)24.40 (n/a)23.84 (n/a)23.80 (n/a)23.12 (n/a)0.49 (n/a)

matrix_vector_mul_8192x2048_4tsi_1024tso_2col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
897d04e — 2026-03-06 22:12:2524.10 (-5.45%)23.07 (-1.85%)24.01 (-2.88%)19.26 (+7.47%)2.13 (-32.15%)24.08 (-5.45%)23.05 (-1.85%)23.99 (-2.88%)19.25 (+7.47%)2.13 (-32.15%)
84d3478 — 2026-02-17 23:25:1325.48 (n/a)23.50 (n/a)24.72 (n/a)17.92 (n/a)3.14 (n/a)25.47 (n/a)23.48 (n/a)24.71 (n/a)17.91 (n/a)3.14 (n/a)

matrix_vector_mul_8192x2048_4tsi_1024tso_4col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
8b6d0f4 — 2026-03-25 20:46:3940.84 (-1.16%)38.86 (-0.66%)38.49 (-1.61%)37.90 (+1.94%)1.14 (-23.22%)40.82 (-1.16%)38.84 (-0.66%)38.46 (-1.61%)37.87 (+1.94%)1.14 (-23.22%)
dc66da7 — 2026-03-25 19:23:4541.32 (n/a)39.12 (n/a)39.12 (n/a)37.18 (n/a)1.49 (n/a)41.30 (n/a)39.09 (n/a)39.09 (n/a)37.15 (n/a)1.49 (n/a)

matrix_vector_mul_8192x2048_4tsi_1024tso_4col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
897d04e — 2026-03-06 22:12:2540.80 (-0.16%)39.58 (-1.93%)39.75 (-1.71%)38.28 (-2.50%)1.10 (+67.33%)40.78 (-0.16%)39.55 (-1.93%)39.73 (-1.71%)38.26 (-2.50%)1.10 (+67.33%)
84d3478 — 2026-02-17 23:25:1340.87 (n/a)40.36 (n/a)40.45 (n/a)39.26 (n/a)0.66 (n/a)40.84 (n/a)40.33 (n/a)40.42 (n/a)39.24 (n/a)0.65 (n/a)

matrix_vector_mul_8192x2048_4tsi_1024tso_8col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
8b6d0f4 — 2026-03-25 20:46:3941.79 (-12.64%)40.98 (-5.83%)40.78 (-1.04%)40.36 (-0.60%)0.56 (-84.11%)41.76 (-12.64%)40.96 (-5.83%)40.76 (-1.04%)40.34 (-0.60%)0.56 (-84.11%)
dc66da7 — 2026-03-25 19:23:4547.83 (n/a)43.52 (n/a)41.21 (n/a)40.61 (n/a)3.53 (n/a)47.80 (n/a)43.50 (n/a)41.19 (n/a)40.58 (n/a)3.53 (n/a)

matrix_vector_mul_8192x2048_4tsi_1024tso_8col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
897d04e — 2026-03-06 22:12:2549.31 (+6.27%)46.12 (+1.67%)45.30 (-0.27%)44.62 (+1.87%)1.85 (+85.10%)49.28 (+6.27%)46.10 (+1.67%)45.28 (-0.27%)44.59 (+1.87%)1.85 (+85.10%)
84d3478 — 2026-02-17 23:25:1346.40 (n/a)45.37 (n/a)45.43 (n/a)43.80 (n/a)1.00 (n/a)46.37 (n/a)45.34 (n/a)45.40 (n/a)43.78 (n/a)1.00 (n/a)

mem_copy_16_cores_2_chans_2048_tile_128_False

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.05 (+7.16%)0.03 (-9.22%)0.03 (-18.40%)0.02 (-19.93%)0.01 (+42.24%)394.80 (+24.90%)264.56 (+14.35%)259.20 (+22.55%)179.80 (-6.65%)81.04 (+64.09%)
dc66da7 — 2026-03-25 19:23:450.04 (n/a)0.04 (n/a)0.04 (n/a)0.03 (n/a)0.01 (n/a)316.10 (n/a)231.36 (n/a)211.50 (n/a)192.60 (n/a)49.39 (n/a)

mem_copy_16_cores_2_chans_2048_tile_128_False0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.08 (-2.12%)0.06 (-16.44%)0.06 (-21.48%)0.03 (-23.84%)0.02 (+10.15%)247.60 (+31.28%)148.38 (+24.08%)128.30 (+27.41%)100.10 (+2.14%)58.70 (+50.73%)
84d3478 — 2026-02-17 23:25:130.08 (n/a)0.07 (n/a)0.08 (n/a)0.04 (n/a)0.02 (n/a)188.60 (n/a)119.58 (n/a)100.70 (n/a)98.00 (n/a)38.94 (n/a)

mem_copy_1_cols_1_channels_2048_tile_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
003b49d — 2025-11-14 19:30:540.05 (-16.99%)0.05 (-12.47%)0.05 (-7.32%)0.04 (-11.17%)0.00 (-29.61%)197.00 (+12.57%)171.00 (+13.70%)164.00 (+7.89%)153.00 (+20.47%)17.49 (-4.32%)
0999a9a — 2025-11-13 19:24:080.06 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.01 (n/a)175.00 (n/a)150.40 (n/a)152.00 (n/a)127.00 (n/a)18.28 (n/a)

mem_copy_1_cols_2_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
003b49d — 2025-11-14 19:30:540.06 (+5.56%)0.05 (-1.59%)0.05 (-1.39%)0.03 (-14.07%)0.01 (+25.03%)263.00 (+16.37%)166.00 (+4.53%)144.00 (+1.41%)126.00 (-5.26%)49.38 (+43.17%)
0999a9a — 2025-11-13 19:24:080.06 (n/a)0.05 (n/a)0.05 (n/a)0.03 (n/a)0.01 (n/a)226.00 (n/a)158.80 (n/a)142.00 (n/a)133.00 (n/a)34.49 (n/a)

mem_copy_1_cores_1_chans_2048_tile_2048_False

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.09 (+75.06%)0.05 (+7.82%)0.04 (-15.32%)0.03 (-13.78%)0.02 (+425.35%)249.10 (+15.97%)190.02 (+3.98%)211.70 (+18.07%)95.60 (-42.89%)62.62 (+236.69%)
dc66da7 — 2026-03-25 19:23:450.05 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.00 (n/a)214.80 (n/a)182.74 (n/a)179.30 (n/a)167.40 (n/a)18.60 (n/a)

mem_copy_1_cores_1_chans_2048_tile_2048_False0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.11 (-0.24%)0.08 (-17.53%)0.08 (-27.05%)0.06 (-15.43%)0.02 (+4.62%)148.50 (+18.23%)113.02 (+23.38%)103.60 (+37.04%)74.10 (+0.14%)31.69 (+33.33%)
84d3478 — 2026-02-17 23:25:130.11 (n/a)0.09 (n/a)0.11 (n/a)0.07 (n/a)0.02 (n/a)125.60 (n/a)91.60 (n/a)75.60 (n/a)74.00 (n/a)23.77 (n/a)

mem_copy_2_cols_1_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
003b49d — 2025-11-14 19:30:540.06 (-15.32%)0.05 (-17.09%)0.05 (-23.81%)0.04 (-15.54%)0.01 (-18.18%)193.00 (+18.40%)160.80 (+20.36%)168.00 (+31.25%)124.00 (+18.10%)22.92 (+11.39%)
0999a9a — 2025-11-13 19:24:080.07 (n/a)0.06 (n/a)0.06 (n/a)0.05 (n/a)0.01 (n/a)163.00 (n/a)133.60 (n/a)128.00 (n/a)105.00 (n/a)20.58 (n/a)

mem_copy_2_cols_2_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
003b49d — 2025-11-14 19:30:540.06 (-23.02%)0.05 (-6.98%)0.05 (-2.13%)0.04 (-15.31%)0.01 (-33.29%)196.00 (+18.07%)147.00 (+6.52%)141.00 (+2.17%)126.00 (+29.90%)25.27 (+8.13%)
0999a9a — 2025-11-13 19:24:080.08 (n/a)0.06 (n/a)0.06 (n/a)0.05 (n/a)0.01 (n/a)166.00 (n/a)138.00 (n/a)138.00 (n/a)97.00 (n/a)23.37 (n/a)

mem_copy_2_cores_1_chans_2048_tile_1024_False

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.07 (+13.70%)0.05 (-1.93%)0.05 (-7.07%)0.04 (-27.15%)0.02 (+157.23%)230.10 (+37.29%)166.58 (+9.82%)161.90 (+7.65%)109.70 (-12.03%)53.34 (+211.05%)
dc66da7 — 2026-03-25 19:23:450.07 (n/a)0.05 (n/a)0.05 (n/a)0.05 (n/a)0.01 (n/a)167.60 (n/a)151.68 (n/a)150.40 (n/a)124.70 (n/a)17.15 (n/a)

mem_copy_2_cores_1_chans_2048_tile_1024_False0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.11 (+27.03%)0.09 (+25.79%)0.11 (+27.09%)0.05 (+12.74%)0.03 (+43.93%)161.90 (-11.29%)94.34 (-18.35%)76.30 (-21.34%)74.00 (-21.36%)38.01 (+0.49%)
84d3478 — 2026-02-17 23:25:130.09 (n/a)0.08 (n/a)0.08 (n/a)0.04 (n/a)0.02 (n/a)182.50 (n/a)115.54 (n/a)97.00 (n/a)94.10 (n/a)37.82 (n/a)

mem_copy_2_cores_2_chans_2048_tile_1024_False

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.07 (-3.17%)0.05 (-7.71%)0.05 (-6.65%)0.04 (-20.60%)0.02 (+42.64%)223.40 (+25.93%)171.62 (+13.04%)172.10 (+7.09%)120.40 (+3.26%)49.35 (+82.52%)
dc66da7 — 2026-03-25 19:23:450.07 (n/a)0.06 (n/a)0.05 (n/a)0.05 (n/a)0.01 (n/a)177.40 (n/a)151.82 (n/a)160.70 (n/a)116.60 (n/a)27.04 (n/a)

mem_copy_2_cores_2_chans_2048_tile_1024_False0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.10 (+2.70%)0.08 (-5.38%)0.09 (+2.10%)0.04 (-43.30%)0.02 (+170.50%)201.60 (+76.38%)112.08 (+16.53%)90.80 (-2.05%)84.60 (-2.65%)50.14 (+375.75%)
84d3478 — 2026-02-17 23:25:130.09 (n/a)0.09 (n/a)0.09 (n/a)0.07 (n/a)0.01 (n/a)114.30 (n/a)96.18 (n/a)92.70 (n/a)86.90 (n/a)10.54 (n/a)

mem_copy_4_cols_1_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
003b49d — 2025-11-14 19:30:540.07 (+11.97%)0.05 (+3.09%)0.05 (+2.10%)0.03 (-11.35%)0.01 (+41.65%)229.00 (+12.81%)156.00 (-0.51%)143.00 (-2.05%)117.00 (-10.69%)38.21 (+48.71%)
0999a9a — 2025-11-13 19:24:080.06 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.01 (n/a)203.00 (n/a)156.80 (n/a)146.00 (n/a)131.00 (n/a)25.69 (n/a)

mem_copy_4_cols_2_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
003b49d — 2025-11-14 19:30:540.06 (+9.02%)0.05 (+5.59%)0.05 (+0.00%)0.04 (+3.63%)0.01 (+25.48%)193.00 (-3.50%)147.40 (-4.66%)147.00 (+0.00%)122.00 (-8.27%)25.24 (+7.48%)
0999a9a — 2025-11-13 19:24:080.06 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.01 (n/a)200.00 (n/a)154.60 (n/a)147.00 (n/a)133.00 (n/a)23.48 (n/a)

mem_copy_4_cores_1_chans_2048_tile_512_False

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.06 (-2.79%)0.05 (-5.43%)0.05 (+6.39%)0.03 (-12.50%)0.01 (+16.26%)242.20 (+14.25%)184.14 (+7.91%)167.80 (-6.05%)128.30 (+2.89%)46.84 (+42.55%)
dc66da7 — 2026-03-25 19:23:450.07 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.01 (n/a)212.00 (n/a)170.64 (n/a)178.60 (n/a)124.70 (n/a)32.86 (n/a)

mem_copy_4_cores_1_chans_2048_tile_512_False0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.09 (-5.87%)0.08 (-3.61%)0.08 (-0.79%)0.06 (-12.74%)0.01 (+22.04%)146.30 (+14.66%)108.02 (+4.96%)99.30 (+0.81%)95.10 (+6.26%)21.49 (+48.38%)
84d3478 — 2026-02-17 23:25:130.09 (n/a)0.08 (n/a)0.08 (n/a)0.06 (n/a)0.01 (n/a)127.60 (n/a)102.92 (n/a)98.50 (n/a)89.50 (n/a)14.49 (n/a)

mem_copy_4_cores_2_chans_2048_tile_512_False

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.05 (+7.78%)0.04 (-14.27%)0.04 (-10.58%)0.03 (-32.73%)0.01 (+109.81%)316.90 (+48.71%)226.08 (+22.64%)209.60 (+11.85%)150.20 (-7.23%)61.97 (+194.31%)
dc66da7 — 2026-03-25 19:23:450.05 (n/a)0.04 (n/a)0.04 (n/a)0.04 (n/a)0.01 (n/a)213.10 (n/a)184.34 (n/a)187.40 (n/a)161.90 (n/a)21.05 (n/a)

mem_copy_4_cores_2_chans_2048_tile_512_False0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.11 (+13.04%)0.09 (+17.19%)0.10 (+20.43%)0.06 (+17.23%)0.02 (+16.34%)142.80 (-14.70%)91.94 (-14.71%)79.00 (-17.02%)77.70 (-11.50%)28.50 (-14.82%)
84d3478 — 2026-02-17 23:25:130.09 (n/a)0.08 (n/a)0.09 (n/a)0.05 (n/a)0.02 (n/a)167.40 (n/a)107.80 (n/a)95.20 (n/a)87.80 (n/a)33.46 (n/a)

mem_copy_8_cols_1_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
003b49d — 2025-11-14 19:30:540.05 (-13.48%)0.05 (-17.79%)0.05 (-15.33%)0.03 (-25.09%)0.01 (+6.67%)267.00 (+33.50%)176.80 (+24.51%)150.00 (+18.11%)141.00 (+15.57%)47.41 (+61.41%)
0999a9a — 2025-11-13 19:24:080.06 (n/a)0.06 (n/a)0.06 (n/a)0.04 (n/a)0.01 (n/a)200.00 (n/a)142.00 (n/a)127.00 (n/a)122.00 (n/a)29.37 (n/a)

mem_copy_8_cols_2_channels_2048_tile_128

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
003b49d — 2025-11-14 19:30:540.05 (+13.38%)0.05 (+23.85%)0.05 (+37.91%)0.05 (+30.72%)0.00 (-56.85%)166.00 (-23.50%)153.40 (-20.60%)153.00 (-27.49%)142.00 (-11.80%)7.63 (-70.60%)
0999a9a — 2025-11-13 19:24:080.05 (n/a)0.04 (n/a)0.04 (n/a)0.04 (n/a)0.01 (n/a)217.00 (n/a)193.20 (n/a)211.00 (n/a)161.00 (n/a)25.96 (n/a)

mem_copy_8_cores_1_chans_2048_tile_256_False

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.06 (-5.71%)0.04 (+0.46%)0.04 (+2.21%)0.03 (-7.12%)0.01 (-7.04%)278.00 (+7.67%)209.24 (-0.41%)216.20 (-2.17%)141.90 (+6.05%)51.53 (+9.47%)
dc66da7 — 2026-03-25 19:23:450.06 (n/a)0.04 (n/a)0.04 (n/a)0.03 (n/a)0.01 (n/a)258.20 (n/a)210.10 (n/a)221.00 (n/a)133.80 (n/a)47.07 (n/a)

mem_copy_8_cores_1_chans_2048_tile_256_False0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.08 (-0.39%)0.07 (+10.07%)0.08 (+1.07%)0.06 (+38.38%)0.01 (-47.35%)136.70 (-27.75%)110.64 (-13.07%)104.70 (-1.13%)103.00 (+0.39%)14.59 (-60.78%)
84d3478 — 2026-02-17 23:25:130.08 (n/a)0.07 (n/a)0.08 (n/a)0.04 (n/a)0.02 (n/a)189.20 (n/a)127.28 (n/a)105.90 (n/a)102.60 (n/a)37.20 (n/a)

mem_copy_8_cores_2_chans_2048_tile_256_False

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.06 (+19.11%)0.04 (+1.88%)0.04 (-11.30%)0.03 (+24.20%)0.01 (+4.86%)262.00 (-19.48%)199.56 (-3.59%)200.50 (+12.70%)139.70 (-16.05%)45.93 (-31.63%)
dc66da7 — 2026-03-25 19:23:450.05 (n/a)0.04 (n/a)0.05 (n/a)0.03 (n/a)0.01 (n/a)325.40 (n/a)207.00 (n/a)177.90 (n/a)166.40 (n/a)67.19 (n/a)

mem_copy_8_cores_2_chans_2048_tile_256_False0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.08 (-17.85%)0.07 (-24.66%)0.08 (-18.40%)0.04 (-32.08%)0.02 (+21.78%)185.30 (+47.18%)127.72 (+37.87%)102.00 (+22.45%)100.40 (+21.70%)38.66 (+106.34%)
84d3478 — 2026-02-17 23:25:130.10 (n/a)0.09 (n/a)0.10 (n/a)0.07 (n/a)0.01 (n/a)125.90 (n/a)92.64 (n/a)83.30 (n/a)82.50 (n/a)18.74 (n/a)

mha

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.18 (+0.10%)0.18 (+0.09%)0.18 (+0.09%)0.18 (+0.08%)0.00 (+11.08%)47507.50 (-0.08%)47468.96 (-0.09%)47481.10 (-0.09%)47413.10 (-0.10%)37.32 (+10.96%)
dc66da7 — 2026-03-25 19:23:450.18 (n/a)0.18 (n/a)0.18 (n/a)0.18 (n/a)0.00 (n/a)47544.20 (n/a)47512.60 (n/a)47521.80 (n/a)47459.60 (n/a)33.63 (n/a)

mha0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
84d3478 — 2026-02-17 23:25:130.18 (-0.51%)0.18 (-0.36%)0.18 (-0.41%)0.18 (-0.21%)0.00 (-35.21%)47468.60 (+0.21%)47403.94 (+0.36%)47426.00 (+0.41%)47314.40 (+0.51%)68.83 (-34.73%)
1531781 — 2026-01-23 18:49:240.18 (n/a)0.18 (n/a)0.18 (n/a)0.18 (n/a)0.00 (n/a)47367.50 (n/a)47232.68 (n/a)47232.00 (n/a)47073.00 (n/a)105.45 (n/a)

mha_16384_64_1_8_0_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.18 (n/a)0.18 (n/a)0.18 (n/a)0.18 (n/a)0.00 (n/a)47533.70 (n/a)47380.56 (n/a)47337.90 (n/a)47268.00 (n/a)110.60 (n/a)

relu_1_cols_1_channels_2048_tile_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.07 (+14.06%)0.05 (+0.31%)0.05 (-3.71%)0.04 (-18.38%)0.01 (+129.69%)208.90 (+22.52%)161.96 (+2.56%)166.90 (+3.86%)120.40 (-12.31%)33.36 (+144.00%)
dc66da7 — 2026-03-25 19:23:450.06 (n/a)0.05 (n/a)0.05 (n/a)0.05 (n/a)0.00 (n/a)170.50 (n/a)157.92 (n/a)160.70 (n/a)137.30 (n/a)13.67 (n/a)

relu_2_cols_1_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.06 (-20.92%)0.05 (-6.86%)0.05 (-1.28%)0.03 (-20.49%)0.01 (-19.18%)264.20 (+25.81%)174.76 (+7.41%)149.00 (+1.29%)140.20 (+26.42%)52.29 (+24.11%)
dc66da7 — 2026-03-25 19:23:450.07 (n/a)0.05 (n/a)0.06 (n/a)0.04 (n/a)0.01 (n/a)210.00 (n/a)162.70 (n/a)147.10 (n/a)110.90 (n/a)42.13 (n/a)

relu_4_cols_1_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.05 (-8.73%)0.04 (-10.26%)0.04 (-4.04%)0.03 (-16.73%)0.01 (+5.34%)259.40 (+20.09%)206.80 (+12.91%)200.50 (+4.21%)150.20 (+9.56%)45.96 (+41.47%)
dc66da7 — 2026-03-25 19:23:450.06 (n/a)0.05 (n/a)0.04 (n/a)0.04 (n/a)0.01 (n/a)216.00 (n/a)183.16 (n/a)192.40 (n/a)137.10 (n/a)32.49 (n/a)

relu_8_cols_1_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.05 (+4.18%)0.04 (-0.33%)0.04 (-1.55%)0.02 (-20.26%)0.01 (+47.86%)369.50 (+25.38%)233.74 (+5.19%)215.30 (+1.56%)165.20 (-4.01%)80.83 (+78.85%)
dc66da7 — 2026-03-25 19:23:450.05 (n/a)0.04 (n/a)0.04 (n/a)0.03 (n/a)0.01 (n/a)294.70 (n/a)222.20 (n/a)212.00 (n/a)172.10 (n/a)45.19 (n/a)

rms_norm_1_cols_1_channels_2048_tile_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.06 (-18.06%)0.05 (-2.61%)0.05 (-1.31%)0.04 (+12.13%)0.01 (-43.39%)226.00 (-10.81%)172.66 (-1.55%)163.30 (+1.30%)140.00 (+22.06%)32.37 (-37.02%)
dc66da7 — 2026-03-25 19:23:450.07 (n/a)0.05 (n/a)0.05 (n/a)0.03 (n/a)0.01 (n/a)253.40 (n/a)175.38 (n/a)161.20 (n/a)114.70 (n/a)51.40 (n/a)

rms_norm_1_cols_2_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.08 (+16.34%)0.05 (+6.02%)0.05 (-3.17%)0.04 (+12.01%)0.02 (+27.81%)218.40 (-10.71%)162.92 (-3.80%)150.40 (+3.23%)108.00 (-14.01%)52.41 (+3.43%)
dc66da7 — 2026-03-25 19:23:450.07 (n/a)0.05 (n/a)0.06 (n/a)0.03 (n/a)0.01 (n/a)244.60 (n/a)169.36 (n/a)145.70 (n/a)125.60 (n/a)50.68 (n/a)

rms_norm_2_cols_1_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.06 (+9.52%)0.04 (-9.59%)0.04 (-17.77%)0.03 (-18.95%)0.01 (+55.38%)255.30 (+23.39%)199.88 (+13.52%)205.30 (+21.62%)137.80 (-8.68%)43.62 (+70.31%)
dc66da7 — 2026-03-25 19:23:450.05 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.01 (n/a)206.90 (n/a)176.08 (n/a)168.80 (n/a)150.90 (n/a)25.61 (n/a)

rms_norm_2_cols_2_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.06 (+6.05%)0.04 (+0.61%)0.04 (-6.97%)0.03 (-2.64%)0.01 (+44.42%)234.40 (+2.72%)195.80 (+1.42%)215.50 (+7.48%)141.90 (-5.65%)40.78 (+43.76%)
dc66da7 — 2026-03-25 19:23:450.05 (n/a)0.04 (n/a)0.04 (n/a)0.04 (n/a)0.01 (n/a)228.20 (n/a)193.06 (n/a)200.50 (n/a)150.40 (n/a)28.37 (n/a)

rms_norm_4_cols_1_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.06 (+10.73%)0.05 (-1.13%)0.05 (+5.17%)0.04 (-10.70%)0.01 (+73.44%)225.10 (+11.99%)181.84 (+3.00%)170.50 (-4.91%)140.00 (-9.68%)33.21 (+78.26%)
dc66da7 — 2026-03-25 19:23:450.05 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.00 (n/a)201.00 (n/a)176.54 (n/a)179.30 (n/a)155.00 (n/a)18.63 (n/a)

rms_norm_4_cols_2_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.06 (-0.51%)0.04 (-4.22%)0.04 (-6.40%)0.03 (+19.93%)0.01 (-12.26%)264.40 (-16.62%)208.96 (+1.64%)223.40 (+6.84%)140.70 (+0.50%)52.99 (-24.83%)
dc66da7 — 2026-03-25 19:23:450.06 (n/a)0.04 (n/a)0.04 (n/a)0.03 (n/a)0.01 (n/a)317.10 (n/a)205.58 (n/a)209.10 (n/a)140.00 (n/a)70.49 (n/a)

rms_norm_8_cols_1_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.06 (-8.19%)0.04 (-16.80%)0.03 (-25.89%)0.03 (-16.40%)0.01 (+5.02%)283.50 (+19.62%)220.72 (+22.74%)240.30 (+34.92%)128.00 (+8.94%)58.39 (+33.24%)
dc66da7 — 2026-03-25 19:23:450.07 (n/a)0.05 (n/a)0.05 (n/a)0.03 (n/a)0.01 (n/a)237.00 (n/a)179.82 (n/a)178.10 (n/a)117.50 (n/a)43.82 (n/a)

rms_norm_8_cols_2_channels_2048_tile_128

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.04 (+4.11%)0.04 (+2.57%)0.04 (-1.84%)0.03 (+8.94%)0.00 (+2.64%)250.80 (-8.20%)224.64 (-2.58%)233.20 (+1.88%)191.20 (-3.97%)25.83 (-9.64%)
dc66da7 — 2026-03-25 19:23:450.04 (n/a)0.04 (n/a)0.04 (n/a)0.03 (n/a)0.00 (n/a)273.20 (n/a)230.60 (n/a)228.90 (n/a)199.10 (n/a)28.59 (n/a)

rope_1_cols_2_channels_4096_tile_4096_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
1b2d230 — 2026-01-13 22:38:030.19 (-21.66%)0.17 (-8.53%)0.18 (+2.09%)0.14 (-6.71%)0.02 (-38.07%)174.10 (+7.20%)150.06 (+8.11%)139.60 (-2.04%)129.80 (+27.63%)20.36 (-10.65%)
d48746f — 2026-01-12 22:17:100.24 (n/a)0.18 (n/a)0.17 (n/a)0.15 (n/a)0.04 (n/a)162.40 (n/a)138.80 (n/a)142.50 (n/a)101.70 (n/a)22.79 (n/a)

rope_1c_32rows_512cols_32arows_0m

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.91 (+31.28%)0.72 (+30.97%)0.75 (+50.09%)0.52 (+23.14%)0.16 (+34.35%)189.50 (-18.81%)142.68 (-23.26%)131.80 (-33.40%)108.20 (-23.86%)33.21 (-13.83%)
dc66da7 — 2026-03-25 19:23:450.69 (n/a)0.55 (n/a)0.50 (n/a)0.42 (n/a)0.12 (n/a)233.40 (n/a)185.92 (n/a)197.90 (n/a)142.10 (n/a)38.54 (n/a)

rope_1c_32rows_512cols_8arows_0m

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.71 (+13.56%)0.53 (+16.33%)0.51 (+20.62%)0.43 (+13.26%)0.12 (+12.08%)172.60 (-11.71%)143.86 (-14.09%)145.70 (-17.07%)103.70 (-11.97%)28.08 (-12.19%)
dc66da7 — 2026-03-25 19:23:450.63 (n/a)0.46 (n/a)0.42 (n/a)0.38 (n/a)0.10 (n/a)195.50 (n/a)167.46 (n/a)175.70 (n/a)117.80 (n/a)31.98 (n/a)

rope_2_cols_2_channels_4096_tile_2048_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
1b2d230 — 2026-01-13 22:38:030.21 (-0.56%)0.17 (-10.62%)0.16 (-13.58%)0.13 (-13.33%)0.03 (+17.05%)183.70 (+15.39%)151.76 (+12.77%)155.20 (+15.73%)117.30 (+0.60%)24.24 (+35.73%)
d48746f — 2026-01-12 22:17:100.21 (n/a)0.19 (n/a)0.18 (n/a)0.15 (n/a)0.02 (n/a)159.20 (n/a)134.58 (n/a)134.10 (n/a)116.60 (n/a)17.86 (n/a)

rope_2c_32rows_512cols_32arows_0m

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.86 (+12.27%)0.68 (+4.35%)0.67 (-1.63%)0.56 (+6.41%)0.12 (+18.02%)174.80 (-6.02%)146.78 (-3.85%)146.60 (+1.66%)114.70 (-10.88%)23.97 (-1.25%)
dc66da7 — 2026-03-25 19:23:450.76 (n/a)0.66 (n/a)0.68 (n/a)0.53 (n/a)0.10 (n/a)186.00 (n/a)152.66 (n/a)144.20 (n/a)128.70 (n/a)24.27 (n/a)

rope_2c_32rows_512cols_8arows_0m

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.63 (+15.38%)0.42 (-5.03%)0.37 (-15.06%)0.26 (-23.68%)0.14 (+76.55%)289.00 (+31.07%)192.28 (+12.08%)197.90 (+17.73%)117.80 (-13.32%)64.38 (+98.12%)
dc66da7 — 2026-03-25 19:23:450.54 (n/a)0.44 (n/a)0.44 (n/a)0.33 (n/a)0.08 (n/a)220.50 (n/a)171.56 (n/a)168.10 (n/a)135.90 (n/a)32.49 (n/a)

rope_4_cols_2_channels_4096_tile_1024_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
1b2d230 — 2026-01-13 22:38:030.22 (+18.92%)0.17 (+6.24%)0.17 (+5.57%)0.14 (-0.50%)0.03 (+90.77%)171.10 (+0.47%)147.40 (-4.39%)148.10 (-5.31%)110.10 (-15.89%)23.59 (+60.67%)
d48746f — 2026-01-12 22:17:100.19 (n/a)0.16 (n/a)0.16 (n/a)0.14 (n/a)0.02 (n/a)170.30 (n/a)154.16 (n/a)156.40 (n/a)130.90 (n/a)14.69 (n/a)

rope_8_cols_2_channels_4096_tile_512_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
1b2d230 — 2026-01-13 22:38:030.17 (-1.30%)0.15 (+6.63%)0.15 (-8.78%)0.14 (+39.68%)0.01 (-67.35%)174.90 (-28.41%)163.42 (-10.36%)167.40 (+9.63%)146.50 (+1.31%)11.19 (-76.17%)
d48746f — 2026-01-12 22:17:100.17 (n/a)0.14 (n/a)0.16 (n/a)0.10 (n/a)0.03 (n/a)244.30 (n/a)182.30 (n/a)152.70 (n/a)144.60 (n/a)46.98 (n/a)

rope_8c_32rows_512cols_32arows_0m

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.60 (+0.58%)0.47 (-10.94%)0.47 (-14.67%)0.39 (-14.06%)0.09 (+57.47%)252.70 (+16.34%)213.46 (+14.31%)211.20 (+17.20%)163.30 (-0.61%)38.58 (+85.72%)
dc66da7 — 2026-03-25 19:23:450.60 (n/a)0.53 (n/a)0.55 (n/a)0.45 (n/a)0.06 (n/a)217.20 (n/a)186.74 (n/a)180.20 (n/a)164.30 (n/a)20.77 (n/a)

rope_8c_32rows_512cols_8arows_0m

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.58 (+33.33%)0.38 (-2.44%)0.32 (-20.46%)0.28 (-18.27%)0.13 (+203.34%)264.90 (+22.36%)209.28 (+9.84%)227.20 (+25.73%)126.60 (-25.00%)59.71 (+180.42%)
dc66da7 — 2026-03-25 19:23:450.44 (n/a)0.39 (n/a)0.41 (n/a)0.34 (n/a)0.04 (n/a)216.50 (n/a)190.54 (n/a)180.70 (n/a)168.80 (n/a)21.29 (n/a)

rows_32-cols_512-angle_rows_32-aie_columns_1-method_type_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
14eee74 — 2026-03-26 15:31:390.38 (n/a)0.28 (n/a)0.25 (n/a)0.17 (n/a)0.09 (n/a)567.20 (n/a)388.78 (n/a)390.80 (n/a)259.70 (n/a)126.72 (n/a)

rows_32-cols_512-angle_rows_32-aie_columns_2-method_type_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
14eee74 — 2026-03-26 15:31:390.42 (n/a)0.35 (n/a)0.37 (n/a)0.27 (n/a)0.06 (n/a)359.40 (n/a)284.78 (n/a)264.10 (n/a)236.60 (n/a)52.58 (n/a)

rows_32-cols_512-angle_rows_32-aie_columns_8-method_type_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
14eee74 — 2026-03-26 15:31:390.40 (n/a)0.32 (n/a)0.36 (n/a)0.13 (n/a)0.11 (n/a)755.20 (n/a)368.90 (n/a)271.00 (n/a)246.90 (n/a)218.13 (n/a)

rows_32-cols_512-angle_rows_8-aie_columns_1-method_type_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
14eee74 — 2026-03-26 15:31:390.27 (n/a)0.24 (n/a)0.25 (n/a)0.18 (n/a)0.04 (n/a)418.40 (n/a)316.04 (n/a)293.50 (n/a)275.40 (n/a)57.94 (n/a)

rows_32-cols_512-angle_rows_8-aie_columns_2-method_type_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
14eee74 — 2026-03-26 15:31:390.26 (n/a)0.24 (n/a)0.26 (n/a)0.16 (n/a)0.04 (n/a)449.30 (n/a)321.92 (n/a)286.10 (n/a)278.40 (n/a)72.31 (n/a)

rows_32-cols_512-angle_rows_8-aie_columns_8-method_type_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
14eee74 — 2026-03-26 15:31:390.24 (n/a)0.21 (n/a)0.23 (n/a)0.12 (n/a)0.05 (n/a)600.80 (n/a)371.22 (n/a)317.20 (n/a)302.30 (n/a)128.53 (n/a)

seq_len_16384-dim_64-num_heads_1-num_pipelines_8-num_kv_heads_0

No metrics available.

seq_len_256-embedding_dim_2048-hidden_dim_2048-prio_accuracy_False

No metrics available.

sigmoid_1_cols_1_channels_2048_tile_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.05 (-14.16%)0.05 (-6.29%)0.05 (-2.65%)0.04 (-7.71%)0.01 (-21.79%)225.80 (+8.35%)180.84 (+6.33%)170.90 (+2.70%)163.30 (+16.48%)25.41 (+0.55%)
dc66da7 — 2026-03-25 19:23:450.06 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.01 (n/a)208.40 (n/a)170.08 (n/a)166.40 (n/a)140.20 (n/a)25.28 (n/a)

sigmoid_2_cols_1_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.06 (+4.68%)0.05 (+2.94%)0.04 (-13.06%)0.04 (+15.35%)0.01 (+8.49%)232.90 (-13.32%)179.44 (-3.16%)197.20 (+15.05%)127.30 (-4.50%)46.58 (-13.93%)
dc66da7 — 2026-03-25 19:23:450.06 (n/a)0.05 (n/a)0.05 (n/a)0.03 (n/a)0.01 (n/a)268.70 (n/a)185.30 (n/a)171.40 (n/a)133.30 (n/a)54.12 (n/a)

sigmoid_4_cols_1_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.07 (+10.37%)0.05 (-2.49%)0.04 (-7.15%)0.04 (-7.47%)0.01 (+64.06%)210.80 (+8.10%)175.44 (+5.70%)183.30 (+7.70%)117.30 (-9.42%)39.46 (+67.78%)
dc66da7 — 2026-03-25 19:23:450.06 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.01 (n/a)195.00 (n/a)165.98 (n/a)170.20 (n/a)129.50 (n/a)23.52 (n/a)

sigmoid_8_cols_1_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.06 (+29.56%)0.05 (+22.81%)0.05 (+37.30%)0.03 (-8.45%)0.01 (+218.31%)242.70 (+9.23%)173.04 (-14.71%)150.20 (-27.16%)134.70 (-22.81%)46.84 (+169.71%)
dc66da7 — 2026-03-25 19:23:450.05 (n/a)0.04 (n/a)0.04 (n/a)0.04 (n/a)0.00 (n/a)222.20 (n/a)202.88 (n/a)206.20 (n/a)174.50 (n/a)17.37 (n/a)

silu_1_cols_1_channels_2048_tile_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.07 (+9.38%)0.05 (-4.51%)0.04 (-23.77%)0.04 (+15.02%)0.01 (+5.83%)207.90 (-13.05%)169.28 (+4.04%)188.60 (+31.15%)114.40 (-8.63%)38.46 (-17.21%)
dc66da7 — 2026-03-25 19:23:450.07 (n/a)0.05 (n/a)0.06 (n/a)0.03 (n/a)0.01 (n/a)239.10 (n/a)162.70 (n/a)143.80 (n/a)125.20 (n/a)46.46 (n/a)

silu_2_cols_1_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.06 (-4.63%)0.05 (-13.54%)0.05 (-12.20%)0.04 (-16.50%)0.01 (+19.45%)213.90 (+19.76%)180.30 (+16.99%)181.70 (+13.92%)134.00 (+4.85%)30.84 (+49.10%)
dc66da7 — 2026-03-25 19:23:450.06 (n/a)0.05 (n/a)0.05 (n/a)0.05 (n/a)0.01 (n/a)178.60 (n/a)154.12 (n/a)159.50 (n/a)127.80 (n/a)20.68 (n/a)

silu_4_cols_1_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.06 (+6.26%)0.05 (-8.81%)0.04 (-14.86%)0.04 (-17.09%)0.01 (+81.20%)214.80 (+20.61%)176.86 (+11.68%)182.90 (+17.47%)133.30 (-5.86%)30.25 (+101.37%)
dc66da7 — 2026-03-25 19:23:450.06 (n/a)0.05 (n/a)0.05 (n/a)0.05 (n/a)0.00 (n/a)178.10 (n/a)158.36 (n/a)155.70 (n/a)141.60 (n/a)15.02 (n/a)

silu_8_cols_1_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.06 (-6.11%)0.05 (-10.71%)0.04 (-10.38%)0.04 (-11.78%)0.01 (+1.86%)224.60 (+13.38%)185.30 (+12.62%)186.00 (+11.58%)140.40 (+6.53%)33.11 (+23.38%)
dc66da7 — 2026-03-25 19:23:450.06 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.01 (n/a)198.10 (n/a)164.54 (n/a)166.70 (n/a)131.80 (n/a)26.84 (n/a)

softmax_1_cols_2_channels_4096_tile_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.19 (-8.94%)0.17 (-8.66%)0.17 (-19.83%)0.14 (+4.32%)0.02 (-35.13%)115.80 (-4.14%)99.26 (+7.47%)98.60 (+24.65%)85.00 (+9.82%)13.65 (-30.86%)
84d3478 — 2026-02-17 23:25:130.21 (n/a)0.18 (n/a)0.21 (n/a)0.14 (n/a)0.04 (n/a)120.80 (n/a)92.36 (n/a)79.10 (n/a)77.40 (n/a)19.75 (n/a)

softmax_2_cols_2_channels_32768_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.64 (-42.09%)0.60 (-31.83%)0.59 (-30.19%)0.54 (-6.68%)0.05 (-76.96%)242.70 (+7.15%)221.12 (+40.50%)220.30 (+43.24%)203.40 (+72.67%)16.99 (-59.06%)
dc66da7 — 2026-03-25 19:23:451.11 (n/a)0.87 (n/a)0.85 (n/a)0.58 (n/a)0.20 (n/a)226.50 (n/a)157.38 (n/a)153.80 (n/a)117.80 (n/a)41.50 (n/a)

softmax_2_cols_2_channels_32768_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.68 (-34.93%)0.62 (-20.25%)0.60 (-17.49%)0.55 (-10.32%)0.05 (-71.29%)240.30 (+11.51%)214.32 (+20.87%)218.20 (+21.22%)193.80 (+53.69%)18.64 (-52.51%)
dc66da7 — 2026-03-25 19:23:451.04 (n/a)0.77 (n/a)0.73 (n/a)0.61 (n/a)0.18 (n/a)215.50 (n/a)177.32 (n/a)180.00 (n/a)126.10 (n/a)39.26 (n/a)

softmax_2_cols_2_channels_4096_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.20 (-0.87%)0.16 (+6.57%)0.17 (-9.55%)0.11 (+29.81%)0.03 (-44.34%)145.70 (-22.99%)104.68 (-15.00%)95.00 (+10.59%)83.80 (+0.84%)24.16 (-54.23%)
84d3478 — 2026-02-17 23:25:130.20 (n/a)0.15 (n/a)0.19 (n/a)0.09 (n/a)0.06 (n/a)189.20 (n/a)123.16 (n/a)85.90 (n/a)83.10 (n/a)52.80 (n/a)

softmax_2_cols_2_channels_4096_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.25 (+23.53%)0.18 (+9.42%)0.18 (-3.86%)0.11 (+2.99%)0.05 (+19.24%)144.10 (-2.90%)98.98 (-8.18%)92.10 (+4.07%)64.30 (-19.02%)29.58 (-7.13%)
84d3478 — 2026-02-17 23:25:130.21 (n/a)0.16 (n/a)0.19 (n/a)0.11 (n/a)0.04 (n/a)148.40 (n/a)107.80 (n/a)88.50 (n/a)79.40 (n/a)31.85 (n/a)

softmax_4_cols_4_channels_32768_tile_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.66 (+11.72%)0.50 (-0.51%)0.54 (-4.72%)0.35 (+6.44%)0.13 (+14.83%)377.70 (-6.04%)276.32 (+1.24%)242.20 (+4.94%)199.30 (-10.51%)76.50 (-0.31%)
dc66da7 — 2026-03-25 19:23:450.59 (n/a)0.51 (n/a)0.57 (n/a)0.33 (n/a)0.11 (n/a)402.00 (n/a)272.94 (n/a)230.80 (n/a)222.70 (n/a)76.73 (n/a)

swiglu

No metrics available.

swiglu_decode_1x2048x2048

No metrics available.

swiglu_decode_1x2048x2048_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.00 (-0.21%)0.00 (+7.74%)0.00 (+0.37%)0.00 (+8.84%)0.00 (-19.58%)5170.80 (-8.12%)3761.80 (-8.70%)3408.60 (-0.36%)3407.30 (+0.21%)787.66 (-22.98%)
84d3478 — 2026-02-17 23:25:130.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)5627.70 (n/a)4120.44 (n/a)3421.00 (n/a)3400.30 (n/a)1022.61 (n/a)

swiglu_prefill_256x2048x2048

No metrics available.

tanh_1_cols_1_channels_2048_tile_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.06 (-7.39%)0.04 (-16.97%)0.04 (-14.92%)0.03 (-28.80%)0.01 (+23.25%)296.40 (+40.47%)201.28 (+24.69%)188.60 (+17.51%)141.90 (+7.99%)59.45 (+90.30%)
dc66da7 — 2026-03-25 19:23:450.06 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.01 (n/a)211.00 (n/a)161.42 (n/a)160.50 (n/a)131.40 (n/a)31.24 (n/a)

tanh_2_cols_1_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.06 (-14.29%)0.05 (-6.63%)0.05 (+2.99%)0.03 (-30.73%)0.01 (+5.83%)293.30 (+44.41%)176.82 (+11.67%)159.70 (-2.92%)128.50 (+16.71%)67.26 (+86.47%)
dc66da7 — 2026-03-25 19:23:450.07 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.01 (n/a)203.10 (n/a)158.34 (n/a)164.50 (n/a)110.10 (n/a)36.07 (n/a)

tanh_4_cols_1_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.06 (-27.96%)0.05 (-4.84%)0.05 (+4.02%)0.04 (+16.65%)0.01 (-63.57%)187.60 (-14.30%)167.32 (-1.72%)171.90 (-3.86%)133.00 (+38.83%)20.69 (-55.09%)
dc66da7 — 2026-03-25 19:23:450.09 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.02 (n/a)218.90 (n/a)170.24 (n/a)178.80 (n/a)95.80 (n/a)46.06 (n/a)

tanh_8_cols_1_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.05 (-24.07%)0.04 (-15.48%)0.04 (-23.02%)0.03 (-3.09%)0.01 (-53.59%)262.30 (+3.19%)198.18 (+12.63%)183.30 (+29.91%)173.30 (+31.69%)36.55 (-34.15%)
dc66da7 — 2026-03-25 19:23:450.06 (n/a)0.05 (n/a)0.06 (n/a)0.03 (n/a)0.01 (n/a)254.20 (n/a)175.96 (n/a)141.10 (n/a)131.60 (n/a)55.51 (n/a)

transpose_2048_M_64_N_1_cols_1_channels_64_m_64_n_8_s

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:392.51 (-25.94%)2.10 (-20.30%)2.01 (-23.11%)1.76 (-13.49%)0.29 (-41.75%)298.70 (+15.60%)253.30 (+23.92%)261.10 (+30.09%)208.60 (+35.02%)33.97 (-9.04%)
dc66da7 — 2026-03-25 19:23:453.39 (n/a)2.64 (n/a)2.61 (n/a)2.03 (n/a)0.49 (n/a)258.40 (n/a)204.40 (n/a)200.70 (n/a)154.50 (n/a)37.34 (n/a)

transpose_2048_M_64_N_1_cols_1_channels_64_m_64_n_8_s0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:254.43 (+0.85%)3.73 (+3.83%)4.02 (-1.30%)2.97 (+16.77%)0.65 (-24.99%)176.70 (-14.35%)144.46 (-6.24%)130.40 (+1.32%)118.40 (-0.84%)26.60 (-35.16%)
84d3478 — 2026-02-17 23:25:134.39 (n/a)3.59 (n/a)4.07 (n/a)2.54 (n/a)0.87 (n/a)206.30 (n/a)154.08 (n/a)128.70 (n/a)119.40 (n/a)41.02 (n/a)

transpose_2048_M_64_N_1_cols_2_channels_64_m_64_n_8_s

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:392.72 (-10.04%)2.30 (-13.01%)2.26 (-13.66%)1.88 (-21.25%)0.34 (+46.90%)278.20 (+26.97%)231.60 (+16.34%)232.20 (+15.81%)192.40 (+11.15%)34.64 (+109.20%)
dc66da7 — 2026-03-25 19:23:453.03 (n/a)2.65 (n/a)2.61 (n/a)2.39 (n/a)0.23 (n/a)219.10 (n/a)199.08 (n/a)200.50 (n/a)173.10 (n/a)16.56 (n/a)

transpose_2048_M_64_N_1_cols_2_channels_64_m_64_n_8_s0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:255.27 (-14.18%)4.35 (-2.49%)4.45 (+2.15%)2.78 (-22.18%)0.98 (-3.38%)188.60 (+28.47%)126.98 (+4.20%)117.90 (-2.08%)99.40 (+16.53%)36.02 (+50.15%)
84d3478 — 2026-02-17 23:25:136.14 (n/a)4.46 (n/a)4.35 (n/a)3.57 (n/a)1.02 (n/a)146.80 (n/a)121.86 (n/a)120.40 (n/a)85.30 (n/a)23.99 (n/a)

weighted_rms_norm_1_cols_2_channels_2048_weights_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.09 (+6.38%)0.07 (-4.87%)0.07 (+1.68%)0.05 (-23.45%)0.01 (+101.20%)230.80 (+30.62%)177.68 (+7.40%)170.00 (-1.68%)138.30 (-5.98%)33.65 (+149.29%)
dc66da7 — 2026-03-25 19:23:450.08 (n/a)0.07 (n/a)0.07 (n/a)0.07 (n/a)0.01 (n/a)176.70 (n/a)165.44 (n/a)172.90 (n/a)147.10 (n/a)13.50 (n/a)

weighted_rms_norm_2_cols_2_channels_2048_weights_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.07 (-25.98%)0.06 (-18.74%)0.06 (-7.04%)0.04 (-20.93%)0.01 (-34.18%)237.00 (+26.47%)189.06 (+22.07%)179.50 (+7.55%)151.40 (+35.06%)34.27 (+14.91%)
dc66da7 — 2026-03-25 19:23:450.09 (n/a)0.07 (n/a)0.06 (n/a)0.05 (n/a)0.01 (n/a)187.40 (n/a)154.88 (n/a)166.90 (n/a)112.10 (n/a)29.83 (n/a)

weighted_rms_norm_4_cols_2_channels_2048_weights_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.06 (-13.39%)0.05 (-21.43%)0.04 (-31.53%)0.04 (-12.12%)0.01 (-0.80%)237.90 (+13.77%)205.32 (+28.15%)225.30 (+46.01%)147.80 (+15.47%)38.15 (+27.36%)
dc66da7 — 2026-03-25 19:23:450.07 (n/a)0.06 (n/a)0.06 (n/a)0.04 (n/a)0.01 (n/a)209.10 (n/a)160.22 (n/a)154.30 (n/a)128.00 (n/a)29.95 (n/a)

weighted_rms_norm_8_cols_2_channels_2048_weights_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.07 (+33.02%)0.04 (-11.48%)0.04 (-24.93%)0.02 (-32.99%)0.02 (+124.58%)371.50 (+49.26%)251.20 (+24.91%)247.70 (+33.17%)124.90 (-24.85%)88.76 (+136.07%)
dc66da7 — 2026-03-25 19:23:450.05 (n/a)0.04 (n/a)0.05 (n/a)0.03 (n/a)0.01 (n/a)248.90 (n/a)201.10 (n/a)186.00 (n/a)166.20 (n/a)37.60 (n/a)

- mha.cc: include mm.cc once (col-major via -DB_COL_MAJ flag) and define
  zero_bf16_rowmaj and matmul_bf16_bf16_rowmaj directly in the same TU
  using the matmul_vectorized_2x2_mmul template with b_row_maj=true,
  eliminating transitive cross-object symbol references
- op.py: replace mha_mm.o, mha_mm_rowmaj.o, mha_mha.o with single mha.o
  compiled with mm col-major flags and all three source dependencies
- design.py: update all Kernel references from mha_mm.o/mha_mha.o to mha.o
- compilation/base.py: relax PeanoCompilationRule to allow multiple source
  dependencies per KernelObjectArtifact (>= 1 instead of == 1)

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
@github-actions
Copy link
Contributor

📊 Test Results for Small Benchmark/Test Suite

8728b32 (2026_03_26_21_31_35)

IRONCLAD

Tested on 2026_03_26_21_31_35 at commit 8728b32.

Test Checks Latency (mean)Bandwidth (mean)Throughput (mean)
M_128-K_128-num_aie_columns_1-tile_size_input_32-tile_size_output_128 ✅ 5/5 n/a 0.09 0.09
M_1792-K_896-N_1152-num_aie_columns_8-b_col_maj_False-c_col_maj_True-m_64-k_32-n_48-trace_size_0 ✅ 5/5 2418.36 3.97 1560.94
M_192-K_384-N_64-num_aie_columns_4-b_col_maj_False-c_col_maj_False-m_48-k_96-n_16-trace_size_0 ✅ 5/5 660.30 0.38 16.40
M_192-K_384-N_64-num_aie_columns_4-b_col_maj_True-c_col_maj_True-m_48-k_96-n_16-trace_size_0 ✅ 5/5 753.00 0.34 14.45
M_2048-K_2048-N_2048-num_aie_columns_1-b_col_maj_False-c_col_maj_False-m_64-k_64-n_64-trace_size_0 ✅ 5/5 50029.12 0.50 343.40
M_2048-K_2048-N_2048-num_aie_columns_2-b_col_maj_True-c_col_maj_False-m_64-k_64-n_64-trace_size_0 ✅ 5/5 29459.48 0.85 583.18
M_2048-K_2048-N_2048-num_aie_columns_8-b_col_maj_True-c_col_maj_True-m_64-k_64-n_64-trace_size_0 ✅ 5/5 8031.32 3.14 2143.81
M_2048-K_8192-num_aie_columns_1-tile_size_input_1-tile_size_output_2048 ✅ 5/5 n/a 11.91 11.90
M_2048-K_8192-num_aie_columns_2-tile_size_input_1-tile_size_output_1024 ✅ 5/5 n/a 20.64 20.62
M_2048-K_8192-num_aie_columns_4-tile_size_input_1-tile_size_output_512 ✅ 5/5 n/a 32.70 32.68
M_2048-K_8192-num_aie_columns_8-tile_size_input_1-tile_size_output_256 ✅ 5/5 n/a 32.02 32.00
M_2048-N_64-aie_columns_1-channels_1-m_64-n_64-s_8 ✅ 5/5 493.52 1.27 n/a
M_2048-N_64-aie_columns_1-channels_2-m_64-n_64-s_8 ✅ 5/5 358.24 1.56 n/a
M_384-K_1536-N_1792-num_aie_columns_4-b_col_maj_True-c_col_maj_False-m_32-k_48-n_64-trace_size_0 ✅ 5/5 2429.88 3.43 899.73
M_8192-K_2048-num_aie_columns_1-tile_size_input_4-tile_size_output_1024 ✅ 5/5 n/a 11.73 11.72
M_8192-K_2048-num_aie_columns_2-tile_size_input_4-tile_size_output_1024 ✅ 5/5 n/a 19.62 19.61
M_8192-K_2048-num_aie_columns_4-tile_size_input_4-tile_size_output_1024 ✅ 5/5 n/a 31.71 31.69
M_8192-K_2048-num_aie_columns_8-tile_size_input_4-tile_size_output_1024 ✅ 5/5 n/a 31.39 31.37
M_896-K_1792-N_640-num_aie_columns_8-b_col_maj_False-c_col_maj_True-m_32-k_64-n_80-trace_size_0 ✅ 5/5 1441.34 4.75 1467.03
embedding_dim_2048-hidden_dim_2048 ✅ 5/5 n/a n/a n/a
input_length_2048-num_aie_columns_1-num_channels_1-tile_size_2048 ✅ 25/25 281.38 0.03 n/a
input_length_2048-num_aie_columns_1-num_channels_1-tile_size_2048-group_size_32 ✅ 5/5 341.86 0.02 n/a
input_length_2048-num_aie_columns_1-num_channels_1-tile_size_2048-weighted_False ✅ 5/5 265.36 0.03 n/a
input_length_2048-num_aie_columns_1-num_channels_1-tile_size_2048-weighted_True ✅ 5/5 255.24 0.05 n/a
input_length_2048-num_aie_columns_1-num_channels_2-tile_size_1024 ✅ 10/10 291.86 0.03 n/a
input_length_2048-num_aie_columns_1-num_channels_2-tile_size_1024-group_size_32 ✅ 5/5 398.10 0.02 n/a
input_length_2048-num_aie_columns_1-num_channels_2-tile_size_1024-weighted_False ✅ 5/5 311.52 0.03 n/a
input_length_2048-num_aie_columns_1-num_channels_2-tile_size_2048-scalar_factor_3.0 ✅ 5/5 275.26 0.05 n/a
input_length_2048-num_aie_columns_1-tile_size_2048 ✅ 15/15 273.21 0.04 n/a
input_length_2048-num_aie_columns_2-num_channels_1-tile_size_1024 ✅ 25/25 273.55 0.03 n/a
input_length_2048-num_aie_columns_2-num_channels_1-tile_size_1024-group_size_32 ✅ 5/5 353.84 0.02 n/a
input_length_2048-num_aie_columns_2-num_channels_1-tile_size_1024-weighted_False ✅ 5/5 378.60 0.03 n/a
input_length_2048-num_aie_columns_2-num_channels_1-tile_size_1024-weighted_True ✅ 5/5 286.40 0.04 n/a
input_length_2048-num_aie_columns_2-num_channels_2-tile_size_1024-scalar_factor_3.0 ✅ 5/5 261.16 0.05 n/a
input_length_2048-num_aie_columns_2-num_channels_2-tile_size_512 ✅ 10/10 340.40 0.03 n/a
input_length_2048-num_aie_columns_2-num_channels_2-tile_size_512-group_size_32 ✅ 5/5 316.66 0.02 n/a
input_length_2048-num_aie_columns_2-num_channels_2-tile_size_512-weighted_False ✅ 5/5 305.74 0.03 n/a
input_length_2048-num_aie_columns_2-tile_size_1024 ✅ 15/15 289.59 0.04 n/a
input_length_2048-num_aie_columns_4-num_channels_1-tile_size_512 ✅ 25/25 381.44 0.03 n/a
input_length_2048-num_aie_columns_4-num_channels_1-tile_size_512-group_size_32 ✅ 5/5 324.90 0.02 n/a
input_length_2048-num_aie_columns_4-num_channels_1-tile_size_512-weighted_False ✅ 5/5 263.98 0.03 n/a
input_length_2048-num_aie_columns_4-num_channels_1-tile_size_512-weighted_True ✅ 5/5 308.88 0.03 n/a
input_length_2048-num_aie_columns_4-num_channels_2-tile_size_256 ✅ 10/10 314.42 0.03 n/a
input_length_2048-num_aie_columns_4-num_channels_2-tile_size_256-group_size_32 ✅ 5/5 244.60 0.02 n/a
input_length_2048-num_aie_columns_4-num_channels_2-tile_size_256-weighted_False ✅ 5/5 279.16 0.03 n/a
input_length_2048-num_aie_columns_4-num_channels_2-tile_size_512-scalar_factor_3.0 ✅ 5/5 236.92 0.05 n/a
input_length_2048-num_aie_columns_4-tile_size_512 ✅ 15/15 284.31 0.04 n/a
input_length_2048-num_aie_columns_8-num_channels_1-tile_size_256 ✅ 25/25 327.56 0.03 n/a
input_length_2048-num_aie_columns_8-num_channels_1-tile_size_256-group_size_32 ✅ 5/5 276.06 0.02 n/a
input_length_2048-num_aie_columns_8-num_channels_1-tile_size_256-weighted_False ✅ 5/5 333.12 0.03 n/a
input_length_2048-num_aie_columns_8-num_channels_1-tile_size_256-weighted_True ✅ 5/5 358.14 0.03 n/a
input_length_2048-num_aie_columns_8-num_channels_2-tile_size_128 ✅ 10/10 347.36 0.03 n/a
input_length_2048-num_aie_columns_8-num_channels_2-tile_size_128-group_size_32 ✅ 5/5 342.56 0.02 n/a
input_length_2048-num_aie_columns_8-num_channels_2-tile_size_128-weighted_False ✅ 5/5 357.72 0.03 n/a
input_length_2048-num_aie_columns_8-num_channels_2-tile_size_256-scalar_factor_3.0 ✅ 5/5 271.78 0.05 n/a
input_length_2048-num_aie_columns_8-tile_size_256 ✅ 15/15 311.51 0.04 n/a
input_length_2048-num_cores_1-num_channels_1-bypass_False-tile_size_2048 ✅ 5/5 280.80 0.03 n/a
input_length_2048-num_cores_16-num_channels_2-bypass_False-tile_size_128 ✅ 5/5 353.86 0.03 n/a
input_length_2048-num_cores_2-num_channels_1-bypass_False-tile_size_1024 ✅ 5/5 285.84 0.03 n/a
input_length_2048-num_cores_2-num_channels_2-bypass_False-tile_size_1024 ✅ 5/5 265.54 0.03 n/a
input_length_2048-num_cores_4-num_channels_1-bypass_False-tile_size_512 ✅ 5/5 282.36 0.03 n/a
input_length_2048-num_cores_4-num_channels_2-bypass_False-tile_size_512 ✅ 5/5 262.30 0.03 n/a
input_length_2048-num_cores_8-num_channels_1-bypass_False-tile_size_256 ✅ 5/5 316.38 0.03 n/a
input_length_2048-num_cores_8-num_channels_2-bypass_False-tile_size_256 ✅ 5/5 336.16 0.03 n/a
input_length_32768-num_aie_columns_2-num_channels_2-tile_size_1024 ✅ 5/5 320.08 0.47 n/a
input_length_32768-num_aie_columns_2-num_channels_2-tile_size_512 ✅ 5/5 275.54 0.49 n/a
input_length_32768-num_aie_columns_4-num_channels_4-tile_size_2048 ✅ 5/5 524.02 0.35 n/a
rows_32-cols_512-angle_rows_32-aie_columns_1-method_type_0 ✅ 5/5 467.46 0.26 n/a
rows_32-cols_512-angle_rows_32-aie_columns_2-method_type_0 ✅ 5/5 398.42 0.31 n/a
rows_32-cols_512-angle_rows_32-aie_columns_8-method_type_0 ✅ 5/5 354.14 0.32 n/a
rows_32-cols_512-angle_rows_8-aie_columns_1-method_type_0 ✅ 5/5 381.12 0.21 n/a
rows_32-cols_512-angle_rows_8-aie_columns_2-method_type_0 ✅ 5/5 385.04 0.22 n/a
rows_32-cols_512-angle_rows_8-aie_columns_8-method_type_0 ✅ 5/5 345.98 0.23 n/a
seq_len_16384-dim_64-num_heads_1-num_pipelines_8-num_kv_heads_0 ✅ 5/5 41367.72 0.20 n/a
seq_len_256-embedding_dim_2048-hidden_dim_2048-prio_accuracy_False ✅ 5/5 n/a n/a n/a
📈 Trends (vs main branch) for Small Benchmark/Test Suite

8728b32 (2026_03_26_21_31_35)

IRONCLAD Trends

M_128-K_128-num_aie_columns_1-tile_size_input_32-tile_size_output_128

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
8728b32 — 2026-03-26 21:24:130.13 (n/a)0.09 (n/a)0.08 (n/a)0.04 (n/a)0.04 (n/a)0.13 (n/a)0.09 (n/a)0.08 (n/a)0.04 (n/a)0.04 (n/a)

M_1792-K_896-N_1152-num_aie_columns_8-b_col_maj_False-c_col_maj_True-m_64-k_32-n_48-trace_size_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
8728b32 — 2026-03-26 21:24:134.60 (n/a)3.97 (n/a)4.06 (n/a)3.16 (n/a)0.61 (n/a)2978.80 (n/a)2418.36 (n/a)2315.50 (n/a)2046.00 (n/a)393.67 (n/a)1808.09 (n/a)1560.94 (n/a)1597.66 (n/a)1241.89 (n/a)241.02 (n/a)

M_192-K_384-N_64-num_aie_columns_4-b_col_maj_False-c_col_maj_False-m_48-k_96-n_16-trace_size_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
8728b32 — 2026-03-26 21:24:130.58 (n/a)0.38 (n/a)0.30 (n/a)0.23 (n/a)0.16 (n/a)942.80 (n/a)660.30 (n/a)728.50 (n/a)382.80 (n/a)253.44 (n/a)24.65 (n/a)16.40 (n/a)12.96 (n/a)10.01 (n/a)6.94 (n/a)

M_192-K_384-N_64-num_aie_columns_4-b_col_maj_True-c_col_maj_True-m_48-k_96-n_16-trace_size_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
8728b32 — 2026-03-26 21:24:130.64 (n/a)0.34 (n/a)0.28 (n/a)0.21 (n/a)0.17 (n/a)1053.60 (n/a)753.00 (n/a)789.90 (n/a)344.70 (n/a)256.20 (n/a)27.38 (n/a)14.45 (n/a)11.95 (n/a)8.96 (n/a)7.35 (n/a)

M_2048-K_2048-N_2048-num_aie_columns_1-b_col_maj_False-c_col_maj_False-m_64-k_64-n_64-trace_size_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
8728b32 — 2026-03-26 21:24:130.50 (n/a)0.50 (n/a)0.50 (n/a)0.50 (n/a)0.00 (n/a)50046.60 (n/a)50029.12 (n/a)50035.50 (n/a)50003.80 (n/a)16.53 (n/a)343.57 (n/a)343.40 (n/a)343.35 (n/a)343.28 (n/a)0.11 (n/a)

M_2048-K_2048-N_2048-num_aie_columns_2-b_col_maj_True-c_col_maj_False-m_64-k_64-n_64-trace_size_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
8728b32 — 2026-03-26 21:24:130.86 (n/a)0.85 (n/a)0.85 (n/a)0.85 (n/a)0.00 (n/a)29726.60 (n/a)29459.48 (n/a)29450.60 (n/a)29297.50 (n/a)167.89 (n/a)586.39 (n/a)583.18 (n/a)583.35 (n/a)577.93 (n/a)3.31 (n/a)

M_2048-K_2048-N_2048-num_aie_columns_8-b_col_maj_True-c_col_maj_True-m_64-k_64-n_64-trace_size_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
8728b32 — 2026-03-26 21:24:133.40 (n/a)3.14 (n/a)3.13 (n/a)2.95 (n/a)0.17 (n/a)8543.90 (n/a)8031.32 (n/a)8046.00 (n/a)7406.90 (n/a)415.96 (n/a)2319.44 (n/a)2143.81 (n/a)2135.20 (n/a)2010.77 (n/a)113.57 (n/a)

M_2048-K_8192-num_aie_columns_1-tile_size_input_1-tile_size_output_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
8728b32 — 2026-03-26 21:24:1312.42 (n/a)11.91 (n/a)12.15 (n/a)11.01 (n/a)0.60 (n/a)12.41 (n/a)11.90 (n/a)12.14 (n/a)11.00 (n/a)0.60 (n/a)

M_2048-K_8192-num_aie_columns_2-tile_size_input_1-tile_size_output_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
8728b32 — 2026-03-26 21:24:1321.81 (n/a)20.64 (n/a)21.16 (n/a)17.71 (n/a)1.66 (n/a)21.80 (n/a)20.62 (n/a)21.15 (n/a)17.70 (n/a)1.66 (n/a)

M_2048-K_8192-num_aie_columns_4-tile_size_input_1-tile_size_output_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
8728b32 — 2026-03-26 21:24:1335.42 (n/a)32.70 (n/a)34.10 (n/a)26.26 (n/a)3.70 (n/a)35.40 (n/a)32.68 (n/a)34.08 (n/a)26.25 (n/a)3.70 (n/a)

M_2048-K_8192-num_aie_columns_8-tile_size_input_1-tile_size_output_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
8728b32 — 2026-03-26 21:24:1335.16 (n/a)32.02 (n/a)34.41 (n/a)27.25 (n/a)3.95 (n/a)35.13 (n/a)32.00 (n/a)34.39 (n/a)27.23 (n/a)3.95 (n/a)

M_2048-N_64-aie_columns_1-channels_1-m_64-n_64-s_8

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8728b32 — 2026-03-26 21:24:132.02 (n/a)1.27 (n/a)1.25 (n/a)0.68 (n/a)0.58 (n/a)768.20 (n/a)493.52 (n/a)420.90 (n/a)259.10 (n/a)229.89 (n/a)

M_2048-N_64-aie_columns_1-channels_2-m_64-n_64-s_8

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8728b32 — 2026-03-26 21:24:132.07 (n/a)1.56 (n/a)1.33 (n/a)1.17 (n/a)0.44 (n/a)449.10 (n/a)358.24 (n/a)395.30 (n/a)253.80 (n/a)94.12 (n/a)

M_384-K_1536-N_1792-num_aie_columns_4-b_col_maj_True-c_col_maj_False-m_32-k_48-n_64-trace_size_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
8728b32 — 2026-03-26 21:24:133.89 (n/a)3.43 (n/a)3.80 (n/a)2.40 (n/a)0.64 (n/a)3353.00 (n/a)2429.88 (n/a)2120.30 (n/a)2071.30 (n/a)545.72 (n/a)1020.58 (n/a)899.73 (n/a)997.01 (n/a)630.45 (n/a)166.70 (n/a)

M_8192-K_2048-num_aie_columns_1-tile_size_input_4-tile_size_output_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
8728b32 — 2026-03-26 21:24:1312.44 (n/a)11.73 (n/a)12.29 (n/a)9.32 (n/a)1.35 (n/a)12.43 (n/a)11.72 (n/a)12.28 (n/a)9.31 (n/a)1.35 (n/a)

M_8192-K_2048-num_aie_columns_2-tile_size_input_4-tile_size_output_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
8728b32 — 2026-03-26 21:24:1321.41 (n/a)19.62 (n/a)21.18 (n/a)13.02 (n/a)3.69 (n/a)21.39 (n/a)19.61 (n/a)21.16 (n/a)13.02 (n/a)3.69 (n/a)

M_8192-K_2048-num_aie_columns_4-tile_size_input_4-tile_size_output_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
8728b32 — 2026-03-26 21:24:1333.79 (n/a)31.71 (n/a)32.62 (n/a)26.37 (n/a)3.05 (n/a)33.77 (n/a)31.69 (n/a)32.60 (n/a)26.35 (n/a)3.05 (n/a)

M_8192-K_2048-num_aie_columns_8-tile_size_input_4-tile_size_output_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
8728b32 — 2026-03-26 21:24:1336.12 (n/a)31.39 (n/a)35.26 (n/a)24.93 (n/a)5.90 (n/a)36.10 (n/a)31.37 (n/a)35.24 (n/a)24.91 (n/a)5.90 (n/a)

M_896-K_1792-N_640-num_aie_columns_8-b_col_maj_False-c_col_maj_True-m_32-k_64-n_80-trace_size_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
8728b32 — 2026-03-26 21:24:135.77 (n/a)4.75 (n/a)4.47 (n/a)3.83 (n/a)0.90 (n/a)1737.20 (n/a)1441.34 (n/a)1487.80 (n/a)1153.30 (n/a)266.47 (n/a)1782.02 (n/a)1467.03 (n/a)1381.39 (n/a)1183.07 (n/a)279.00 (n/a)

axpy_1_cols_2_channels_2048_tile_2048_3.0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.08 (+17.48%)0.07 (+11.67%)0.07 (+26.71%)0.05 (+0.80%)0.01 (+108.36%)239.80 (-0.83%)190.62 (-7.91%)166.90 (-21.09%)155.40 (-14.90%)42.89 (+79.10%)
dc66da7 — 2026-03-25 19:23:450.07 (n/a)0.06 (n/a)0.06 (n/a)0.05 (n/a)0.01 (n/a)241.80 (n/a)207.00 (n/a)211.50 (n/a)182.60 (n/a)23.95 (n/a)

axpy_1_cols_2_channels_2048_tile_2048_3.0_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.16 (-3.31%)0.14 (-1.29%)0.15 (+16.18%)0.10 (-6.80%)0.03 (+3.89%)125.70 (+7.25%)93.56 (+2.12%)83.50 (-13.92%)75.00 (+3.45%)22.38 (+18.09%)
84d3478 — 2026-02-17 23:25:130.17 (n/a)0.14 (n/a)0.13 (n/a)0.10 (n/a)0.03 (n/a)117.20 (n/a)91.62 (n/a)97.00 (n/a)72.50 (n/a)18.95 (n/a)

axpy_2_cols_2_channels_2048_tile_1024_3.0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.09 (+14.45%)0.07 (-0.78%)0.07 (-0.40%)0.05 (-7.87%)0.01 (+68.50%)233.40 (+8.51%)188.20 (+2.47%)179.80 (+0.39%)143.50 (-12.66%)34.06 (+61.09%)
dc66da7 — 2026-03-25 19:23:450.07 (n/a)0.07 (n/a)0.07 (n/a)0.06 (n/a)0.01 (n/a)215.10 (n/a)183.66 (n/a)179.10 (n/a)164.30 (n/a)21.14 (n/a)

axpy_2_cols_2_channels_2048_tile_1024_3.0_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.13 (-0.73%)0.11 (-7.43%)0.11 (-8.63%)0.07 (-27.71%)0.02 (+50.39%)182.70 (+38.30%)120.86 (+11.97%)110.80 (+9.38%)96.10 (+0.73%)35.16 (+122.88%)
84d3478 — 2026-02-17 23:25:130.13 (n/a)0.12 (n/a)0.12 (n/a)0.09 (n/a)0.02 (n/a)132.10 (n/a)107.94 (n/a)101.30 (n/a)95.40 (n/a)15.78 (n/a)

axpy_4_cols_2_channels_2048_tile_512_3.0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.10 (+6.67%)0.06 (-1.56%)0.06 (-2.24%)0.05 (+3.60%)0.02 (+20.83%)251.50 (-3.49%)200.78 (+3.11%)191.70 (+2.29%)128.70 (-6.26%)50.37 (+11.30%)
dc66da7 — 2026-03-25 19:23:450.09 (n/a)0.07 (n/a)0.07 (n/a)0.05 (n/a)0.02 (n/a)260.60 (n/a)194.72 (n/a)187.40 (n/a)137.30 (n/a)45.26 (n/a)

axpy_4_cols_2_channels_2048_tile_512_3.0_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.13 (-10.91%)0.11 (-7.68%)0.12 (-0.73%)0.06 (-17.94%)0.03 (+0.99%)194.10 (+21.85%)124.52 (+10.29%)103.30 (+0.68%)97.20 (+12.24%)40.32 (+39.15%)
84d3478 — 2026-02-17 23:25:130.14 (n/a)0.11 (n/a)0.12 (n/a)0.08 (n/a)0.03 (n/a)159.30 (n/a)112.90 (n/a)102.60 (n/a)86.60 (n/a)28.97 (n/a)

axpy_8_cols_2_channels_2048_tile_256_3.0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.08 (+1.42%)0.06 (+9.86%)0.06 (+18.71%)0.03 (+13.51%)0.02 (-10.62%)376.00 (-11.90%)225.02 (-11.61%)196.20 (-15.76%)151.40 (-1.37%)87.15 (-17.99%)
dc66da7 — 2026-03-25 19:23:450.08 (n/a)0.05 (n/a)0.05 (n/a)0.03 (n/a)0.02 (n/a)426.80 (n/a)254.58 (n/a)232.90 (n/a)153.50 (n/a)106.27 (n/a)

axpy_8_cols_2_channels_2048_tile_256_3.0_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.13 (+0.01%)0.11 (+6.10%)0.11 (+0.22%)0.08 (-0.33%)0.02 (-9.83%)156.30 (+0.32%)115.26 (-6.40%)112.70 (-0.18%)95.50 (+0.00%)24.34 (-10.22%)
84d3478 — 2026-02-17 23:25:130.13 (n/a)0.10 (n/a)0.11 (n/a)0.08 (n/a)0.02 (n/a)155.80 (n/a)123.14 (n/a)112.90 (n/a)95.50 (n/a)27.11 (n/a)

dequant_1_cols_1_channels_2048_tile_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.04 (+17.85%)0.03 (+2.55%)0.03 (+1.65%)0.03 (-2.18%)0.00 (+114.37%)207.70 (+2.26%)178.64 (-0.96%)173.60 (-1.59%)141.60 (-15.16%)28.01 (+90.63%)
dc66da7 — 2026-03-25 19:23:450.03 (n/a)0.03 (n/a)0.03 (n/a)0.03 (n/a)0.00 (n/a)203.10 (n/a)180.38 (n/a)176.40 (n/a)166.90 (n/a)14.69 (n/a)

dequant_1_cols_1_channels_2048_tile_2048_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.05 (-13.00%)0.04 (-0.87%)0.04 (+1.16%)0.03 (-3.17%)0.01 (-31.33%)158.40 (+3.26%)128.12 (-0.19%)123.80 (-1.12%)110.40 (+14.88%)18.36 (-15.65%)
84d3478 — 2026-02-17 23:25:130.05 (n/a)0.04 (n/a)0.04 (n/a)0.03 (n/a)0.01 (n/a)153.40 (n/a)128.36 (n/a)125.20 (n/a)96.10 (n/a)21.77 (n/a)

dequant_1_cols_2_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.04 (+13.44%)0.03 (+8.77%)0.03 (+14.89%)0.02 (-9.15%)0.01 (+54.39%)252.70 (+10.06%)182.14 (-6.08%)174.50 (-12.97%)145.40 (-11.88%)42.38 (+56.31%)
dc66da7 — 2026-03-25 19:23:450.03 (n/a)0.03 (n/a)0.03 (n/a)0.02 (n/a)0.00 (n/a)229.60 (n/a)193.94 (n/a)200.50 (n/a)165.00 (n/a)27.11 (n/a)

dequant_1_cols_2_channels_2048_tile_1024_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.06 (-3.74%)0.05 (-12.63%)0.05 (-5.21%)0.03 (-33.47%)0.01 (+52.60%)162.70 (+50.37%)110.84 (+18.98%)98.80 (+5.44%)81.50 (+3.95%)31.02 (+149.23%)
84d3478 — 2026-02-17 23:25:130.07 (n/a)0.06 (n/a)0.06 (n/a)0.05 (n/a)0.01 (n/a)108.20 (n/a)93.16 (n/a)93.70 (n/a)78.40 (n/a)12.45 (n/a)

dequant_2_cols_1_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.05 (+21.89%)0.03 (+20.89%)0.03 (+21.41%)0.02 (+8.07%)0.01 (+32.97%)215.80 (-7.46%)168.56 (-16.36%)172.60 (-17.65%)113.20 (-17.97%)37.61 (+1.89%)
dc66da7 — 2026-03-25 19:23:450.04 (n/a)0.03 (n/a)0.03 (n/a)0.02 (n/a)0.01 (n/a)233.20 (n/a)201.52 (n/a)209.60 (n/a)138.00 (n/a)36.92 (n/a)

dequant_2_cols_1_channels_2048_tile_1024_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.06 (+8.90%)0.05 (+1.31%)0.05 (-1.85%)0.03 (+6.35%)0.01 (+7.11%)152.30 (-5.93%)115.08 (-1.20%)100.80 (+1.92%)88.10 (-8.23%)27.20 (-4.42%)
84d3478 — 2026-02-17 23:25:130.05 (n/a)0.05 (n/a)0.05 (n/a)0.03 (n/a)0.01 (n/a)161.90 (n/a)116.48 (n/a)98.90 (n/a)96.00 (n/a)28.46 (n/a)

dequant_2_cols_2_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.05 (+49.89%)0.03 (+24.89%)0.03 (+13.74%)0.02 (+15.73%)0.01 (+106.91%)219.80 (-13.60%)173.04 (-17.40%)185.70 (-12.07%)112.30 (-33.27%)41.15 (+16.74%)
dc66da7 — 2026-03-25 19:23:450.03 (n/a)0.03 (n/a)0.02 (n/a)0.02 (n/a)0.00 (n/a)254.40 (n/a)209.50 (n/a)211.20 (n/a)168.30 (n/a)35.25 (n/a)

dequant_2_cols_2_channels_2048_tile_512_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.07 (+27.96%)0.05 (+1.57%)0.04 (-18.06%)0.03 (+10.23%)0.01 (+22.79%)164.70 (-9.31%)119.12 (-1.39%)120.20 (+22.03%)75.60 (-21.82%)31.61 (-13.80%)
84d3478 — 2026-02-17 23:25:130.05 (n/a)0.05 (n/a)0.05 (n/a)0.03 (n/a)0.01 (n/a)181.60 (n/a)120.80 (n/a)98.50 (n/a)96.70 (n/a)36.67 (n/a)

dequant_4_cols_1_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.04 (+56.54%)0.03 (+15.78%)0.03 (+9.01%)0.02 (-5.58%)0.01 (+304.26%)243.40 (+5.92%)186.06 (-8.98%)182.60 (-8.29%)118.50 (-36.12%)46.36 (+164.07%)
dc66da7 — 2026-03-25 19:23:450.03 (n/a)0.03 (n/a)0.03 (n/a)0.02 (n/a)0.00 (n/a)229.80 (n/a)204.42 (n/a)199.10 (n/a)185.50 (n/a)17.56 (n/a)

dequant_4_cols_1_channels_2048_tile_512_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.05 (-2.55%)0.04 (-12.79%)0.05 (-16.25%)0.03 (-12.85%)0.01 (-1.20%)164.60 (+14.78%)122.14 (+15.14%)115.90 (+19.48%)97.50 (+2.63%)25.08 (+19.84%)
84d3478 — 2026-02-17 23:25:130.06 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.01 (n/a)143.40 (n/a)106.08 (n/a)97.00 (n/a)95.00 (n/a)20.93 (n/a)

dequant_4_cols_2_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.04 (+28.72%)0.03 (+2.89%)0.02 (-6.84%)0.02 (+6.27%)0.01 (+95.59%)243.20 (-5.88%)208.42 (-0.42%)222.90 (+7.32%)141.10 (-22.34%)41.08 (+38.16%)
dc66da7 — 2026-03-25 19:23:450.03 (n/a)0.03 (n/a)0.03 (n/a)0.02 (n/a)0.00 (n/a)258.40 (n/a)209.30 (n/a)207.70 (n/a)181.70 (n/a)29.74 (n/a)

dequant_4_cols_2_channels_2048_tile_256_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.06 (+20.83%)0.05 (+8.21%)0.04 (+1.42%)0.03 (+17.39%)0.01 (+14.84%)166.40 (-14.80%)122.34 (-7.97%)127.70 (-1.39%)84.60 (-17.22%)30.47 (-19.21%)
84d3478 — 2026-02-17 23:25:130.05 (n/a)0.04 (n/a)0.04 (n/a)0.03 (n/a)0.01 (n/a)195.30 (n/a)132.94 (n/a)129.50 (n/a)102.20 (n/a)37.71 (n/a)

dequant_8_cols_1_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.04 (-1.52%)0.03 (+2.17%)0.03 (+18.60%)0.02 (-11.90%)0.01 (+4.42%)268.50 (+13.53%)205.32 (-1.14%)191.00 (-15.67%)141.60 (+1.51%)49.39 (+23.46%)
dc66da7 — 2026-03-25 19:23:450.04 (n/a)0.03 (n/a)0.02 (n/a)0.02 (n/a)0.01 (n/a)236.50 (n/a)207.68 (n/a)226.50 (n/a)139.50 (n/a)40.00 (n/a)

dequant_8_cols_1_channels_2048_tile_256_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.05 (-25.19%)0.04 (-21.02%)0.04 (-19.70%)0.04 (-19.76%)0.00 (-37.92%)142.70 (+24.63%)126.44 (+26.24%)125.70 (+24.46%)115.50 (+33.68%)10.61 (+3.97%)
84d3478 — 2026-02-17 23:25:130.06 (n/a)0.05 (n/a)0.05 (n/a)0.05 (n/a)0.01 (n/a)114.50 (n/a)100.16 (n/a)101.00 (n/a)86.40 (n/a)10.20 (n/a)

dequant_8_cols_2_channels_2048_tile_128

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.02 (+5.64%)0.02 (+1.26%)0.02 (-1.25%)0.02 (-3.87%)0.00 (+124.04%)258.90 (+4.02%)239.08 (-0.87%)247.20 (+1.27%)215.50 (-5.36%)18.28 (+120.74%)
dc66da7 — 2026-03-25 19:23:450.02 (n/a)0.02 (n/a)0.02 (n/a)0.02 (n/a)0.00 (n/a)248.90 (n/a)241.18 (n/a)244.10 (n/a)227.70 (n/a)8.28 (n/a)

dequant_8_cols_2_channels_2048_tile_128_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.04 (-26.69%)0.04 (-11.93%)0.04 (-16.99%)0.03 (+27.28%)0.01 (-62.60%)193.10 (-21.41%)147.68 (+2.96%)136.10 (+20.44%)135.40 (+36.49%)25.41 (-58.96%)
84d3478 — 2026-02-17 23:25:130.05 (n/a)0.04 (n/a)0.05 (n/a)0.02 (n/a)0.01 (n/a)245.70 (n/a)143.44 (n/a)113.00 (n/a)99.20 (n/a)61.90 (n/a)

eltwise_add_1_cols_2_channels_2048_tile_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.07 (-32.66%)0.06 (-13.87%)0.06 (-9.19%)0.04 (-8.04%)0.01 (-49.50%)299.50 (+8.75%)223.84 (+11.90%)212.70 (+10.15%)187.60 (+48.42%)45.70 (-16.71%)
dc66da7 — 2026-03-25 19:23:450.10 (n/a)0.07 (n/a)0.06 (n/a)0.04 (n/a)0.02 (n/a)275.40 (n/a)200.04 (n/a)193.10 (n/a)126.40 (n/a)54.87 (n/a)

eltwise_add_2_cols_2_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.07 (-33.33%)0.06 (-18.97%)0.05 (-19.23%)0.05 (+18.13%)0.01 (-59.22%)270.80 (-15.37%)223.26 (+13.77%)229.40 (+23.80%)174.50 (+50.04%)37.51 (-50.17%)
dc66da7 — 2026-03-25 19:23:450.11 (n/a)0.07 (n/a)0.07 (n/a)0.04 (n/a)0.02 (n/a)320.00 (n/a)196.24 (n/a)185.30 (n/a)116.30 (n/a)75.29 (n/a)

eltwise_add_4_cols_2_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.07 (-8.97%)0.06 (-6.00%)0.06 (-2.02%)0.05 (-2.74%)0.01 (-14.89%)243.40 (+2.83%)208.60 (+6.01%)200.50 (+2.04%)178.10 (+9.87%)31.58 (-2.07%)
dc66da7 — 2026-03-25 19:23:450.08 (n/a)0.06 (n/a)0.06 (n/a)0.05 (n/a)0.01 (n/a)236.70 (n/a)196.78 (n/a)196.50 (n/a)162.10 (n/a)32.25 (n/a)

eltwise_add_8_cols_2_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.06 (-45.96%)0.05 (-24.77%)0.05 (-15.00%)0.05 (+4.78%)0.00 (-83.62%)259.60 (-4.56%)234.66 (+20.97%)224.10 (+17.64%)218.40 (+85.08%)18.77 (-71.47%)
dc66da7 — 2026-03-25 19:23:450.10 (n/a)0.07 (n/a)0.06 (n/a)0.05 (n/a)0.02 (n/a)272.00 (n/a)193.98 (n/a)190.50 (n/a)118.00 (n/a)65.77 (n/a)

eltwise_mul_1_cols_2_channels_2048_tile_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.11 (+3.13%)0.07 (+2.52%)0.06 (-1.87%)0.05 (-14.01%)0.02 (+21.62%)257.00 (+16.29%)184.10 (+0.67%)190.70 (+1.87%)114.40 (-3.05%)55.57 (+40.96%)
dc66da7 — 2026-03-25 19:23:450.10 (n/a)0.07 (n/a)0.07 (n/a)0.06 (n/a)0.02 (n/a)221.00 (n/a)182.88 (n/a)187.20 (n/a)118.00 (n/a)39.42 (n/a)

eltwise_mul_2_cols_2_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.09 (+2.43%)0.07 (+6.63%)0.08 (+30.25%)0.05 (-10.38%)0.02 (+9.28%)264.20 (+11.62%)182.10 (-4.96%)162.40 (-23.22%)137.30 (-2.42%)52.80 (+20.53%)
dc66da7 — 2026-03-25 19:23:450.09 (n/a)0.07 (n/a)0.06 (n/a)0.05 (n/a)0.02 (n/a)236.70 (n/a)191.60 (n/a)211.50 (n/a)140.70 (n/a)43.81 (n/a)

eltwise_mul_4_cols_2_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.10 (+22.79%)0.08 (+18.44%)0.09 (+34.73%)0.06 (-0.12%)0.01 (+158.99%)191.20 (+0.10%)154.44 (-13.56%)136.60 (-25.80%)126.60 (-18.53%)29.95 (+117.72%)
dc66da7 — 2026-03-25 19:23:450.08 (n/a)0.07 (n/a)0.07 (n/a)0.06 (n/a)0.01 (n/a)191.00 (n/a)178.66 (n/a)184.10 (n/a)155.40 (n/a)13.76 (n/a)

eltwise_mul_8_cols_2_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.08 (+34.81%)0.06 (+14.94%)0.06 (+6.87%)0.04 (-5.87%)0.01 (+158.26%)280.10 (+6.22%)210.00 (-9.80%)215.10 (-6.40%)150.70 (-25.80%)49.49 (+100.84%)
dc66da7 — 2026-03-25 19:23:450.06 (n/a)0.05 (n/a)0.05 (n/a)0.05 (n/a)0.01 (n/a)263.70 (n/a)232.82 (n/a)229.80 (n/a)203.10 (n/a)24.64 (n/a)

embedding_dim_2048-hidden_dim_2048

No metrics available.

gelu_1_cols_1_channels_2048_tile_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.05 (+9.18%)0.04 (-1.03%)0.04 (-2.51%)0.04 (-8.09%)0.01 (+114.58%)221.00 (+8.81%)195.42 (+1.97%)199.10 (+2.58%)163.60 (-8.40%)23.03 (+114.14%)
dc66da7 — 2026-03-25 19:23:450.05 (n/a)0.04 (n/a)0.04 (n/a)0.04 (n/a)0.00 (n/a)203.10 (n/a)191.64 (n/a)194.10 (n/a)178.60 (n/a)10.76 (n/a)

gelu_1_cols_2_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.04 (-26.60%)0.04 (-19.13%)0.04 (-18.64%)0.04 (-2.84%)0.00 (-66.68%)226.70 (+2.91%)199.16 (+20.50%)193.10 (+22.92%)186.40 (+36.16%)16.29 (-52.49%)
dc66da7 — 2026-03-25 19:23:450.06 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.01 (n/a)220.30 (n/a)165.28 (n/a)157.10 (n/a)136.90 (n/a)34.28 (n/a)

gelu_2_cols_1_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.06 (+2.08%)0.05 (+4.66%)0.04 (-1.75%)0.03 (-5.55%)0.01 (+12.71%)283.70 (+5.86%)191.18 (-3.05%)191.00 (+1.81%)137.60 (-1.99%)58.08 (+15.01%)
dc66da7 — 2026-03-25 19:23:450.06 (n/a)0.04 (n/a)0.04 (n/a)0.03 (n/a)0.01 (n/a)268.00 (n/a)197.20 (n/a)187.60 (n/a)140.40 (n/a)50.50 (n/a)

gelu_2_cols_2_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.07 (+2.20%)0.04 (-2.29%)0.04 (+1.04%)0.04 (+16.91%)0.01 (-4.76%)227.00 (-14.47%)194.56 (+0.69%)206.00 (-1.01%)119.40 (-2.21%)44.27 (-20.37%)
dc66da7 — 2026-03-25 19:23:450.07 (n/a)0.05 (n/a)0.04 (n/a)0.03 (n/a)0.01 (n/a)265.40 (n/a)193.22 (n/a)208.10 (n/a)122.10 (n/a)55.59 (n/a)

gelu_4_cols_1_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.07 (-9.75%)0.04 (-17.39%)0.04 (-20.48%)0.03 (+12.65%)0.01 (-29.41%)239.40 (-11.23%)196.08 (+15.06%)197.90 (+25.73%)124.70 (+10.84%)44.40 (-31.09%)
dc66da7 — 2026-03-25 19:23:450.07 (n/a)0.05 (n/a)0.05 (n/a)0.03 (n/a)0.02 (n/a)269.70 (n/a)170.42 (n/a)157.40 (n/a)112.50 (n/a)64.43 (n/a)

gelu_4_cols_2_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.04 (-6.29%)0.04 (+1.81%)0.04 (+0.94%)0.03 (-0.42%)0.00 (-22.18%)284.90 (+0.42%)233.36 (-2.39%)227.20 (-0.96%)204.80 (+6.72%)30.62 (-15.21%)
dc66da7 — 2026-03-25 19:23:450.04 (n/a)0.03 (n/a)0.04 (n/a)0.03 (n/a)0.01 (n/a)283.70 (n/a)239.08 (n/a)229.40 (n/a)191.90 (n/a)36.12 (n/a)

gelu_8_cols_1_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.05 (-18.24%)0.04 (-13.70%)0.04 (-4.91%)0.03 (-15.26%)0.01 (-28.46%)285.90 (+18.04%)223.38 (+14.20%)223.20 (+5.18%)156.90 (+22.29%)47.35 (+1.82%)
dc66da7 — 2026-03-25 19:23:450.06 (n/a)0.04 (n/a)0.04 (n/a)0.03 (n/a)0.01 (n/a)242.20 (n/a)195.60 (n/a)212.20 (n/a)128.30 (n/a)46.50 (n/a)

gelu_8_cols_2_channels_2048_tile_128

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.05 (+14.38%)0.04 (+10.49%)0.03 (+3.25%)0.03 (+13.03%)0.01 (+26.31%)305.70 (-11.52%)235.82 (-8.98%)241.80 (-3.12%)177.40 (-12.57%)49.18 (-6.09%)
dc66da7 — 2026-03-25 19:23:450.04 (n/a)0.03 (n/a)0.03 (n/a)0.02 (n/a)0.01 (n/a)345.50 (n/a)259.10 (n/a)249.60 (n/a)202.90 (n/a)52.37 (n/a)

gemm_1792x896x1152_64x32x48_8cols_ccolmaj

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
8b6d0f4 — 2026-03-25 20:46:394.89 (+9.71%)4.07 (+2.20%)4.10 (+2.03%)3.40 (+0.07%)0.55 (+44.15%)2767.30 (-0.07%)2344.32 (-1.52%)2291.20 (-1.99%)1922.40 (-8.85%)308.92 (+28.54%)1924.38 (+9.71%)1600.58 (+2.20%)1614.60 (+2.03%)1336.81 (+0.07%)215.84 (+44.15%)
dc66da7 — 2026-03-25 19:23:454.46 (n/a)3.98 (n/a)4.02 (n/a)3.40 (n/a)0.38 (n/a)2769.20 (n/a)2380.42 (n/a)2337.70 (n/a)2109.10 (n/a)240.33 (n/a)1754.05 (n/a)1566.13 (n/a)1582.49 (n/a)1335.88 (n/a)149.73 (n/a)

gemm_192x384x64_48x96x16_4cols

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
8b6d0f4 — 2026-03-25 20:46:391.35 (+17.71%)0.77 (-13.89%)0.65 (-34.22%)0.56 (+7.33%)0.32 (+33.14%)393.60 (-6.84%)315.66 (+18.48%)341.40 (+52.00%)164.30 (-15.05%)87.82 (-5.13%)57.45 (+17.71%)32.93 (-13.89%)27.64 (-34.22%)23.97 (+7.33%)13.81 (+33.14%)
dc66da7 — 2026-03-25 19:23:451.14 (n/a)0.90 (n/a)0.98 (n/a)0.52 (n/a)0.24 (n/a)422.50 (n/a)266.42 (n/a)224.60 (n/a)193.40 (n/a)92.57 (n/a)48.81 (n/a)38.25 (n/a)42.02 (n/a)22.34 (n/a)10.37 (n/a)

gemm_192x384x64_48x96x16_4cols_bcolmaj_ccolmaj

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
8b6d0f4 — 2026-03-25 20:46:391.18 (+31.18%)0.79 (-0.76%)0.72 (-17.59%)0.64 (+26.96%)0.22 (+35.66%)343.10 (-21.24%)294.50 (+0.95%)309.00 (+21.37%)188.10 (-23.78%)62.26 (-23.01%)50.17 (+31.18%)33.62 (-0.76%)30.54 (-17.59%)27.51 (+26.96%)9.40 (+35.66%)
dc66da7 — 2026-03-25 19:23:450.90 (n/a)0.79 (n/a)0.87 (n/a)0.51 (n/a)0.16 (n/a)435.60 (n/a)291.72 (n/a)254.60 (n/a)246.80 (n/a)80.87 (n/a)38.24 (n/a)33.88 (n/a)37.06 (n/a)21.67 (n/a)6.93 (n/a)

gemm_2048x2048x2048_64x64x32_8_cols_0_bcolmaj_0_ccolmaj_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)GFLOP/s (max)GFLOP/s (mean)GFLOP/s (median)GFLOP/s (min)GFLOP/s (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
3a053d3 — 2025-11-26 01:12:163.72 (-0.11%)3.55 (-2.75%)3.56 (-2.75%)3.36 (-5.33%)0.12 (+76.10%)2725.23 (-0.11%)2602.52 (-2.75%)2612.11 (-2.75%)2459.89 (-5.33%)86.27 (+76.10%)6984.00 (+5.63%)6608.60 (+2.91%)6577.00 (+2.83%)6304.00 (+0.11%)221.87 (+87.15%)
9e98049 — 2025-11-25 20:50:553.72 (n/a)3.65 (n/a)3.66 (n/a)3.54 (n/a)0.07 (n/a)2728.26 (n/a)2676.06 (n/a)2686.03 (n/a)2598.29 (n/a)48.99 (n/a)6612.00 (n/a)6422.00 (n/a)6396.00 (n/a)6297.00 (n/a)118.55 (n/a)

gemm_2048x2048x2048_64x64x32_8_cols_0_bcolmaj_1_ccolmaj_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)GFLOP/s (max)GFLOP/s (mean)GFLOP/s (median)GFLOP/s (min)GFLOP/s (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
3a053d3 — 2025-11-26 01:12:163.40 (-3.14%)3.38 (-0.82%)3.38 (-0.82%)3.35 (+4.93%)0.02 (-84.51%)2493.81 (-3.14%)2475.27 (-0.82%)2479.06 (-0.82%)2453.92 (+4.93%)13.33 (-84.51%)7001.00 (-4.70%)6940.80 (+0.71%)6930.00 (+0.83%)6889.00 (+3.24%)37.43 (-84.76%)
9e98049 — 2025-11-25 20:50:553.51 (n/a)3.40 (n/a)3.41 (n/a)3.19 (n/a)0.12 (n/a)2574.53 (n/a)2495.72 (n/a)2499.62 (n/a)2338.67 (n/a)86.02 (n/a)7346.00 (n/a)6892.20 (n/a)6873.00 (n/a)6673.00 (n/a)245.57 (n/a)

gemm_2048x2048x2048_64x64x32_8_cols_1_bcolmaj_0_ccolmaj_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)GFLOP/s (max)GFLOP/s (mean)GFLOP/s (median)GFLOP/s (min)GFLOP/s (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
3a053d3 — 2025-11-26 01:12:163.02 (+0.52%)2.87 (-2.05%)2.86 (-2.24%)2.70 (-6.16%)0.11 (+131.66%)2216.19 (+0.52%)2104.44 (-2.05%)2094.08 (-2.24%)1980.84 (-6.16%)80.25 (+131.67%)8673.00 (+6.56%)8175.60 (+2.21%)8204.00 (+2.29%)7752.00 (-0.51%)314.03 (+145.36%)
9e98049 — 2025-11-25 20:50:553.01 (n/a)2.93 (n/a)2.92 (n/a)2.88 (n/a)0.05 (n/a)2204.81 (n/a)2148.41 (n/a)2142.13 (n/a)2110.81 (n/a)34.64 (n/a)8139.00 (n/a)7998.60 (n/a)8020.00 (n/a)7792.00 (n/a)127.99 (n/a)

gemm_2048x2048x2048_64x64x64_1cols

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
8b6d0f4 — 2026-03-25 20:46:390.51 (+0.13%)0.51 (+0.07%)0.51 (+0.04%)0.51 (+0.11%)0.00 (+8.62%)49296.10 (-0.11%)49206.24 (-0.07%)49283.30 (-0.04%)48878.90 (-0.13%)183.15 (+8.40%)351.48 (+0.13%)349.14 (+0.07%)348.59 (+0.04%)348.50 (+0.11%)1.31 (+8.62%)
dc66da7 — 2026-03-25 19:23:450.51 (n/a)0.51 (n/a)0.51 (n/a)0.51 (n/a)0.00 (n/a)49348.10 (n/a)49241.40 (n/a)49302.10 (n/a)48941.60 (n/a)168.96 (n/a)351.03 (n/a)348.89 (n/a)348.46 (n/a)348.14 (n/a)1.20 (n/a)

gemm_2048x2048x2048_64x64x64_2_cols_0_bcolmaj_0_ccolmaj_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
130b6ea — 2025-12-05 21:15:491.04 (+0.94%)1.03 (+0.23%)1.03 (+0.09%)1.02 (+0.07%)0.00 (+197.51%)24576.00 (-0.07%)24485.82 (-0.23%)24530.40 (-0.09%)24266.40 (-0.94%)111.26 (+194.35%)
0a6c11c — 2025-12-04 00:39:101.03 (n/a)1.03 (n/a)1.03 (n/a)1.02 (n/a)0.00 (n/a)24594.30 (n/a)24542.14 (n/a)24551.90 (n/a)24495.50 (n/a)37.80 (n/a)

gemm_2048x2048x2048_64x64x64_2_cols_0_bcolmaj_0_ccolmaj_0_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
a4b6ffe — 2026-01-05 19:24:051.04 (+0.28%)1.03 (-0.15%)1.03 (+0.05%)1.02 (-0.52%)0.01 (+42.72%)24719.80 (+0.52%)24512.42 (+0.16%)24529.20 (-0.05%)24092.60 (-0.28%)252.92 (+43.15%)713.08 (+0.28%)700.92 (-0.15%)700.38 (+0.05%)694.98 (-0.52%)7.30 (+42.71%)
50ad37c — 2025-12-29 20:35:531.04 (n/a)1.03 (n/a)1.03 (n/a)1.02 (n/a)0.01 (n/a)24592.30 (n/a)24473.70 (n/a)24540.90 (n/a)24160.80 (n/a)176.68 (n/a)711.06 (n/a)702.00 (n/a)700.05 (n/a)698.59 (n/a)5.12 (n/a)

gemm_2048x2048x2048_64x64x64_2_cols_0_bcolmaj_1_ccolmaj_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
130b6ea — 2025-12-05 21:15:490.99 (+1.69%)0.97 (+0.29%)0.97 (-0.04%)0.97 (-0.15%)0.01 (+434.92%)26025.30 (+0.15%)25839.34 (-0.29%)25916.70 (+0.04%)25446.00 (-1.67%)202.05 (+425.44%)
0a6c11c — 2025-12-04 00:39:100.97 (n/a)0.97 (n/a)0.97 (n/a)0.97 (n/a)0.00 (n/a)25986.90 (n/a)25913.52 (n/a)25905.70 (n/a)25877.00 (n/a)38.45 (n/a)

gemm_2048x2048x2048_64x64x64_2_cols_0_bcolmaj_1_ccolmaj_0_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
a4b6ffe — 2026-01-05 19:24:050.97 (-0.01%)0.97 (+0.03%)0.97 (+0.04%)0.97 (+0.00%)0.00 (-6.43%)25927.90 (-0.00%)25903.20 (-0.03%)25898.60 (-0.04%)25892.80 (+0.01%)14.05 (-6.56%)663.50 (-0.01%)663.23 (+0.03%)663.35 (+0.04%)662.60 (+0.00%)0.36 (-6.43%)
50ad37c — 2025-12-29 20:35:530.97 (n/a)0.97 (n/a)0.97 (n/a)0.97 (n/a)0.00 (n/a)25928.10 (n/a)25910.14 (n/a)25908.10 (n/a)25891.20 (n/a)15.04 (n/a)663.54 (n/a)663.06 (n/a)663.11 (n/a)662.60 (n/a)0.38 (n/a)

gemm_2048x2048x2048_64x64x64_2_cols_1_bcolmaj_0_ccolmaj_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
130b6ea — 2025-12-05 21:15:490.92 (-1.02%)0.92 (-0.66%)0.92 (-0.10%)0.91 (-0.78%)0.01 (-4.54%)27751.40 (+0.79%)27480.74 (+0.66%)27354.80 (+0.10%)27320.50 (+1.03%)182.69 (-2.84%)
0a6c11c — 2025-12-04 00:39:100.93 (n/a)0.92 (n/a)0.92 (n/a)0.91 (n/a)0.01 (n/a)27534.80 (n/a)27299.32 (n/a)27326.70 (n/a)27042.30 (n/a)188.04 (n/a)

gemm_2048x2048x2048_64x64x64_2_cols_1_bcolmaj_0_ccolmaj_0_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
a4b6ffe — 2026-01-05 19:24:050.92 (+0.11%)0.91 (-0.12%)0.91 (-0.21%)0.91 (-0.24%)0.00 (+30.34%)27723.50 (+0.24%)27548.94 (+0.12%)27522.90 (+0.21%)27403.60 (-0.11%)124.26 (+30.58%)626.92 (+0.11%)623.62 (-0.12%)624.20 (-0.21%)619.69 (-0.24%)2.81 (+30.34%)
50ad37c — 2025-12-29 20:35:530.92 (n/a)0.91 (n/a)0.92 (n/a)0.91 (n/a)0.00 (n/a)27657.10 (n/a)27515.58 (n/a)27464.70 (n/a)27434.60 (n/a)95.16 (n/a)626.21 (n/a)624.37 (n/a)625.52 (n/a)621.17 (n/a)2.16 (n/a)

gemm_2048x2048x2048_64x64x64_2cols_bcolmaj

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
8b6d0f4 — 2026-03-25 20:46:390.88 (+0.21%)0.87 (-0.11%)0.87 (-0.26%)0.87 (+0.03%)0.00 (+24.60%)29019.10 (-0.03%)28862.30 (+0.12%)28862.50 (+0.26%)28665.10 (-0.21%)148.73 (+24.28%)599.33 (+0.21%)595.25 (-0.11%)595.23 (-0.26%)592.02 (+0.03%)3.07 (+24.61%)
dc66da7 — 2026-03-25 19:23:450.88 (n/a)0.87 (n/a)0.87 (n/a)0.87 (n/a)0.00 (n/a)29027.70 (n/a)28829.10 (n/a)28788.30 (n/a)28724.70 (n/a)119.67 (n/a)598.09 (n/a)595.93 (n/a)596.77 (n/a)591.84 (n/a)2.46 (n/a)

gemm_2048x2048x2048_64x64x64_8_cols_0_bcolmaj_0_ccolmaj_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)GFLOP/s (max)GFLOP/s (mean)GFLOP/s (median)GFLOP/s (min)GFLOP/s (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
31ab24e — 2025-11-22 02:01:563.65 (-1.06%)3.59 (+0.83%)3.63 (-0.26%)3.45 (+2.66%)0.07 (-44.71%)2678.08 (-1.06%)2632.13 (+0.83%)2658.60 (-0.26%)2528.31 (+2.66%)53.86 (-44.71%)6795.00 (-2.59%)6529.80 (-0.92%)6462.00 (+0.26%)6415.00 (+1.07%)137.14 (-45.27%)
d9a64c5 — 2025-11-21 20:49:353.69 (n/a)3.56 (n/a)3.64 (n/a)3.36 (n/a)0.13 (n/a)2706.77 (n/a)2610.50 (n/a)2665.61 (n/a)2462.71 (n/a)97.41 (n/a)6976.00 (n/a)6590.40 (n/a)6445.00 (n/a)6347.00 (n/a)250.58 (n/a)

gemm_2048x2048x2048_64x64x64_8_cols_0_bcolmaj_0_ccolmaj_0_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
d48746f — 2026-01-12 22:17:103.98 (+0.20%)3.84 (-2.95%)3.76 (-5.27%)3.72 (-4.97%)0.13 (+473.97%)6760.00 (+5.23%)6559.50 (+3.13%)6697.10 (+5.56%)6321.00 (-0.20%)218.68 (+499.40%)2717.91 (+0.20%)2621.44 (-2.95%)2565.27 (-5.27%)2541.40 (-4.97%)88.37 (+473.97%)
331dcca — 2026-01-08 18:34:103.97 (n/a)3.96 (n/a)3.97 (n/a)3.92 (n/a)0.02 (n/a)6423.80 (n/a)6360.32 (n/a)6344.30 (n/a)6333.90 (n/a)36.48 (n/a)2712.39 (n/a)2701.17 (n/a)2707.91 (n/a)2674.40 (n/a)15.40 (n/a)

gemm_2048x2048x2048_64x64x64_8_cols_0_bcolmaj_1_ccolmaj_0_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
d48746f — 2026-01-12 22:17:103.75 (+0.28%)3.70 (+1.25%)3.71 (-0.21%)3.65 (+4.77%)0.04 (-66.90%)6901.80 (-4.56%)6795.46 (-1.30%)6790.20 (+0.21%)6715.90 (-0.28%)67.05 (-68.40%)2558.08 (+0.28%)2528.33 (+1.25%)2530.10 (-0.21%)2489.18 (+4.77%)24.81 (-66.90%)
331dcca — 2026-01-08 18:34:103.74 (n/a)3.66 (n/a)3.71 (n/a)3.48 (n/a)0.11 (n/a)7231.20 (n/a)6885.18 (n/a)6775.70 (n/a)6734.90 (n/a)212.21 (n/a)2550.87 (n/a)2497.04 (n/a)2535.50 (n/a)2375.78 (n/a)74.96 (n/a)

gemm_2048x2048x2048_64x64x64_8_cols_1_bcolmaj_0_ccolmaj_0_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
d48746f — 2026-01-12 22:17:103.62 (-0.77%)3.54 (+0.14%)3.60 (-0.04%)3.44 (+2.17%)0.09 (-25.77%)7307.40 (-2.13%)7103.46 (-0.18%)6982.50 (+0.04%)6948.20 (+0.77%)182.95 (-26.55%)2472.55 (-0.77%)2419.80 (+0.14%)2460.41 (-0.04%)2351.04 (+2.17%)61.76 (-25.77%)
331dcca — 2026-01-08 18:34:103.65 (n/a)3.54 (n/a)3.61 (n/a)3.37 (n/a)0.12 (n/a)7466.20 (n/a)7116.40 (n/a)6979.60 (n/a)6895.10 (n/a)249.09 (n/a)2491.61 (n/a)2416.45 (n/a)2461.45 (n/a)2301.00 (n/a)83.20 (n/a)

gemm_2048x2048x2048_64x64x64_8cols_bcolmaj_ccolmaj

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
8b6d0f4 — 2026-03-25 20:46:393.47 (-0.81%)3.41 (+0.82%)3.42 (-0.23%)3.35 (+4.63%)0.05 (-63.03%)7522.10 (-4.43%)7390.72 (-0.92%)7367.60 (+0.23%)7255.80 (+0.82%)107.14 (-64.21%)2367.74 (-0.81%)2324.90 (+0.82%)2331.81 (-0.23%)2283.92 (+4.63%)33.69 (-63.03%)
dc66da7 — 2026-03-25 19:23:453.50 (n/a)3.38 (n/a)3.42 (n/a)3.20 (n/a)0.13 (n/a)7870.70 (n/a)7459.22 (n/a)7350.70 (n/a)7196.70 (n/a)299.40 (n/a)2387.20 (n/a)2306.10 (n/a)2337.18 (n/a)2182.77 (n/a)91.13 (n/a)

gemm_384x1536x1792_32x48x64_4cols_bcolmaj

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
8b6d0f4 — 2026-03-25 20:46:394.16 (+12.80%)3.50 (+5.22%)3.50 (+2.51%)2.86 (-0.71%)0.63 (+68.68%)2814.80 (+0.72%)2362.04 (-3.41%)2304.60 (-2.45%)1936.00 (-11.34%)430.86 (+52.19%)1091.93 (+12.80%)919.16 (+5.22%)917.28 (+2.51%)751.01 (-0.71%)166.31 (+68.68%)
dc66da7 — 2026-03-25 19:23:453.69 (n/a)3.33 (n/a)3.41 (n/a)2.88 (n/a)0.38 (n/a)2794.70 (n/a)2445.42 (n/a)2362.50 (n/a)2183.70 (n/a)283.10 (n/a)968.06 (n/a)873.57 (n/a)894.79 (n/a)756.39 (n/a)98.60 (n/a)

gemm_896x1792x640_32x64x80_8cols_ccolmaj

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
8b6d0f4 — 2026-03-25 20:46:396.32 (+6.48%)4.74 (+3.95%)4.60 (+2.04%)3.49 (+3.11%)1.02 (+11.51%)1906.60 (-3.02%)1453.08 (-3.45%)1446.20 (-2.01%)1052.10 (-6.09%)304.01 (+0.60%)1953.36 (+6.48%)1465.54 (+3.95%)1421.06 (+2.04%)1077.93 (+3.11%)313.81 (+11.51%)
dc66da7 — 2026-03-25 19:23:455.94 (n/a)4.56 (n/a)4.51 (n/a)3.38 (n/a)0.91 (n/a)1966.00 (n/a)1504.94 (n/a)1475.80 (n/a)1120.30 (n/a)302.18 (n/a)1834.47 (n/a)1409.81 (n/a)1392.60 (n/a)1045.38 (n/a)281.42 (n/a)

input_length_2048-num_aie_columns_1-num_channels_1-tile_size_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8728b32 — 2026-03-26 21:24:130.04 (n/a)0.03 (n/a)0.03 (n/a)0.02 (n/a)0.01 (n/a)484.50 (n/a)281.38 (n/a)252.90 (n/a)227.90 (n/a)65.69 (n/a)

input_length_2048-num_aie_columns_1-num_channels_1-tile_size_2048-group_size_32

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8728b32 — 2026-03-26 21:24:130.02 (n/a)0.02 (n/a)0.02 (n/a)0.01 (n/a)0.01 (n/a)662.40 (n/a)341.86 (n/a)270.60 (n/a)241.70 (n/a)180.27 (n/a)

input_length_2048-num_aie_columns_1-num_channels_1-tile_size_2048-weighted_False

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8728b32 — 2026-03-26 21:24:130.04 (n/a)0.03 (n/a)0.03 (n/a)0.02 (n/a)0.01 (n/a)375.30 (n/a)265.36 (n/a)245.60 (n/a)226.40 (n/a)62.42 (n/a)

input_length_2048-num_aie_columns_1-num_channels_1-tile_size_2048-weighted_True

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8728b32 — 2026-03-26 21:24:130.05 (n/a)0.05 (n/a)0.05 (n/a)0.03 (n/a)0.01 (n/a)354.90 (n/a)255.24 (n/a)231.00 (n/a)226.40 (n/a)55.77 (n/a)

input_length_2048-num_aie_columns_1-num_channels_2-tile_size_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8728b32 — 2026-03-26 21:24:130.03 (n/a)0.03 (n/a)0.03 (n/a)0.02 (n/a)0.01 (n/a)447.00 (n/a)291.86 (n/a)259.85 (n/a)244.60 (n/a)67.71 (n/a)

input_length_2048-num_aie_columns_1-num_channels_2-tile_size_1024-group_size_32

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8728b32 — 2026-03-26 21:24:130.02 (n/a)0.02 (n/a)0.02 (n/a)0.01 (n/a)0.01 (n/a)709.10 (n/a)398.10 (n/a)255.80 (n/a)221.40 (n/a)229.57 (n/a)

input_length_2048-num_aie_columns_1-num_channels_2-tile_size_1024-weighted_False

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8728b32 — 2026-03-26 21:24:130.03 (n/a)0.03 (n/a)0.03 (n/a)0.01 (n/a)0.01 (n/a)574.60 (n/a)311.52 (n/a)246.80 (n/a)239.60 (n/a)147.14 (n/a)

input_length_2048-num_aie_columns_1-num_channels_2-tile_size_2048-scalar_factor_3.0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8728b32 — 2026-03-26 21:24:130.06 (n/a)0.05 (n/a)0.05 (n/a)0.03 (n/a)0.01 (n/a)419.10 (n/a)275.26 (n/a)240.60 (n/a)219.80 (n/a)81.97 (n/a)

input_length_2048-num_aie_columns_1-tile_size_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8728b32 — 2026-03-26 21:24:130.06 (n/a)0.04 (n/a)0.05 (n/a)0.02 (n/a)0.01 (n/a)562.00 (n/a)273.21 (n/a)245.10 (n/a)218.00 (n/a)89.64 (n/a)

input_length_2048-num_aie_columns_2-num_channels_1-tile_size_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8728b32 — 2026-03-26 21:24:130.04 (n/a)0.03 (n/a)0.03 (n/a)0.02 (n/a)0.00 (n/a)389.70 (n/a)273.55 (n/a)251.80 (n/a)219.30 (n/a)54.74 (n/a)

input_length_2048-num_aie_columns_2-num_channels_1-tile_size_1024-group_size_32

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8728b32 — 2026-03-26 21:24:130.02 (n/a)0.02 (n/a)0.02 (n/a)0.01 (n/a)0.01 (n/a)647.00 (n/a)353.84 (n/a)241.30 (n/a)235.70 (n/a)178.88 (n/a)

input_length_2048-num_aie_columns_2-num_channels_1-tile_size_1024-weighted_False

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8728b32 — 2026-03-26 21:24:130.04 (n/a)0.03 (n/a)0.03 (n/a)0.01 (n/a)0.01 (n/a)929.60 (n/a)378.60 (n/a)243.50 (n/a)229.30 (n/a)308.13 (n/a)

input_length_2048-num_aie_columns_2-num_channels_1-tile_size_1024-weighted_True

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8728b32 — 2026-03-26 21:24:130.04 (n/a)0.04 (n/a)0.04 (n/a)0.02 (n/a)0.01 (n/a)427.80 (n/a)286.40 (n/a)249.40 (n/a)248.00 (n/a)79.16 (n/a)

input_length_2048-num_aie_columns_2-num_channels_2-tile_size_1024-scalar_factor_3.0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8728b32 — 2026-03-26 21:24:130.05 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.00 (n/a)316.20 (n/a)261.16 (n/a)248.10 (n/a)246.00 (n/a)30.79 (n/a)

input_length_2048-num_aie_columns_2-num_channels_2-tile_size_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8728b32 — 2026-03-26 21:24:130.04 (n/a)0.03 (n/a)0.03 (n/a)0.01 (n/a)0.01 (n/a)713.60 (n/a)340.40 (n/a)250.70 (n/a)225.70 (n/a)188.01 (n/a)

input_length_2048-num_aie_columns_2-num_channels_2-tile_size_512-group_size_32

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8728b32 — 2026-03-26 21:24:130.02 (n/a)0.02 (n/a)0.02 (n/a)0.01 (n/a)0.00 (n/a)535.20 (n/a)316.66 (n/a)270.10 (n/a)242.60 (n/a)123.26 (n/a)

input_length_2048-num_aie_columns_2-num_channels_2-tile_size_512-weighted_False

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8728b32 — 2026-03-26 21:24:130.03 (n/a)0.03 (n/a)0.03 (n/a)0.02 (n/a)0.01 (n/a)439.50 (n/a)305.74 (n/a)293.30 (n/a)246.20 (n/a)78.25 (n/a)

input_length_2048-num_aie_columns_2-tile_size_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8728b32 — 2026-03-26 21:24:130.05 (n/a)0.04 (n/a)0.04 (n/a)0.02 (n/a)0.01 (n/a)603.60 (n/a)289.59 (n/a)253.40 (n/a)224.50 (n/a)99.13 (n/a)

input_length_2048-num_aie_columns_4-num_channels_1-tile_size_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8728b32 — 2026-03-26 21:24:130.04 (n/a)0.03 (n/a)0.03 (n/a)0.00 (n/a)0.01 (n/a)2075.50 (n/a)381.44 (n/a)255.60 (n/a)224.20 (n/a)371.33 (n/a)

input_length_2048-num_aie_columns_4-num_channels_1-tile_size_512-group_size_32

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8728b32 — 2026-03-26 21:24:130.02 (n/a)0.02 (n/a)0.02 (n/a)0.01 (n/a)0.01 (n/a)632.10 (n/a)324.90 (n/a)251.60 (n/a)239.60 (n/a)171.84 (n/a)

input_length_2048-num_aie_columns_4-num_channels_1-tile_size_512-weighted_False

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8728b32 — 2026-03-26 21:24:130.04 (n/a)0.03 (n/a)0.04 (n/a)0.02 (n/a)0.01 (n/a)400.00 (n/a)263.98 (n/a)231.10 (n/a)226.90 (n/a)76.06 (n/a)

input_length_2048-num_aie_columns_4-num_channels_1-tile_size_512-weighted_True

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8728b32 — 2026-03-26 21:24:130.04 (n/a)0.03 (n/a)0.04 (n/a)0.02 (n/a)0.01 (n/a)438.00 (n/a)308.88 (n/a)248.50 (n/a)231.30 (n/a)93.55 (n/a)

input_length_2048-num_aie_columns_4-num_channels_2-tile_size_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8728b32 — 2026-03-26 21:24:130.03 (n/a)0.03 (n/a)0.03 (n/a)0.01 (n/a)0.01 (n/a)676.10 (n/a)314.42 (n/a)265.90 (n/a)253.50 (n/a)129.47 (n/a)

input_length_2048-num_aie_columns_4-num_channels_2-tile_size_256-group_size_32

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8728b32 — 2026-03-26 21:24:130.02 (n/a)0.02 (n/a)0.02 (n/a)0.02 (n/a)0.00 (n/a)308.90 (n/a)244.60 (n/a)232.70 (n/a)214.40 (n/a)36.87 (n/a)

input_length_2048-num_aie_columns_4-num_channels_2-tile_size_256-weighted_False

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8728b32 — 2026-03-26 21:24:130.03 (n/a)0.03 (n/a)0.03 (n/a)0.02 (n/a)0.00 (n/a)366.70 (n/a)279.16 (n/a)257.30 (n/a)255.30 (n/a)48.96 (n/a)

input_length_2048-num_aie_columns_4-num_channels_2-tile_size_512-scalar_factor_3.0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8728b32 — 2026-03-26 21:24:130.06 (n/a)0.05 (n/a)0.06 (n/a)0.04 (n/a)0.01 (n/a)294.90 (n/a)236.92 (n/a)222.50 (n/a)218.90 (n/a)32.53 (n/a)

input_length_2048-num_aie_columns_4-tile_size_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8728b32 — 2026-03-26 21:24:130.05 (n/a)0.04 (n/a)0.04 (n/a)0.02 (n/a)0.01 (n/a)623.70 (n/a)284.31 (n/a)241.70 (n/a)224.70 (n/a)105.35 (n/a)

input_length_2048-num_aie_columns_8-num_channels_1-tile_size_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8728b32 — 2026-03-26 21:24:130.04 (n/a)0.03 (n/a)0.03 (n/a)0.01 (n/a)0.01 (n/a)685.00 (n/a)327.56 (n/a)259.40 (n/a)229.30 (n/a)145.26 (n/a)

input_length_2048-num_aie_columns_8-num_channels_1-tile_size_256-group_size_32

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8728b32 — 2026-03-26 21:24:130.02 (n/a)0.02 (n/a)0.02 (n/a)0.02 (n/a)0.00 (n/a)331.70 (n/a)276.06 (n/a)255.50 (n/a)236.90 (n/a)39.14 (n/a)

input_length_2048-num_aie_columns_8-num_channels_1-tile_size_256-weighted_False

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8728b32 — 2026-03-26 21:24:130.03 (n/a)0.03 (n/a)0.03 (n/a)0.01 (n/a)0.01 (n/a)693.80 (n/a)333.12 (n/a)249.50 (n/a)234.10 (n/a)201.80 (n/a)

input_length_2048-num_aie_columns_8-num_channels_1-tile_size_256-weighted_True

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8728b32 — 2026-03-26 21:24:130.03 (n/a)0.03 (n/a)0.03 (n/a)0.01 (n/a)0.01 (n/a)681.20 (n/a)358.14 (n/a)296.80 (n/a)251.50 (n/a)182.07 (n/a)

input_length_2048-num_aie_columns_8-num_channels_2-tile_size_128

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8728b32 — 2026-03-26 21:24:130.03 (n/a)0.03 (n/a)0.03 (n/a)0.01 (n/a)0.01 (n/a)661.20 (n/a)347.36 (n/a)276.70 (n/a)253.70 (n/a)162.34 (n/a)

input_length_2048-num_aie_columns_8-num_channels_2-tile_size_128-group_size_32

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8728b32 — 2026-03-26 21:24:130.02 (n/a)0.02 (n/a)0.02 (n/a)0.01 (n/a)0.01 (n/a)650.40 (n/a)342.56 (n/a)268.30 (n/a)223.20 (n/a)174.60 (n/a)

input_length_2048-num_aie_columns_8-num_channels_2-tile_size_128-weighted_False

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8728b32 — 2026-03-26 21:24:130.03 (n/a)0.03 (n/a)0.03 (n/a)0.01 (n/a)0.01 (n/a)689.50 (n/a)357.72 (n/a)276.50 (n/a)269.70 (n/a)185.51 (n/a)

input_length_2048-num_aie_columns_8-num_channels_2-tile_size_256-scalar_factor_3.0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8728b32 — 2026-03-26 21:24:130.05 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.01 (n/a)333.90 (n/a)271.78 (n/a)257.30 (n/a)251.40 (n/a)35.09 (n/a)

input_length_2048-num_aie_columns_8-tile_size_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8728b32 — 2026-03-26 21:24:130.05 (n/a)0.04 (n/a)0.04 (n/a)0.02 (n/a)0.01 (n/a)642.30 (n/a)311.51 (n/a)265.60 (n/a)253.20 (n/a)109.46 (n/a)

input_length_2048-num_cores_1-num_channels_1-bypass_False-tile_size_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8728b32 — 2026-03-26 21:24:130.04 (n/a)0.03 (n/a)0.03 (n/a)0.02 (n/a)0.01 (n/a)380.50 (n/a)280.80 (n/a)281.70 (n/a)227.60 (n/a)61.57 (n/a)

input_length_2048-num_cores_16-num_channels_2-bypass_False-tile_size_128

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8728b32 — 2026-03-26 21:24:130.03 (n/a)0.03 (n/a)0.03 (n/a)0.01 (n/a)0.01 (n/a)673.40 (n/a)353.86 (n/a)274.60 (n/a)269.50 (n/a)178.65 (n/a)

input_length_2048-num_cores_2-num_channels_1-bypass_False-tile_size_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8728b32 — 2026-03-26 21:24:130.03 (n/a)0.03 (n/a)0.03 (n/a)0.02 (n/a)0.01 (n/a)437.20 (n/a)285.84 (n/a)251.70 (n/a)238.30 (n/a)84.92 (n/a)

input_length_2048-num_cores_2-num_channels_2-bypass_False-tile_size_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8728b32 — 2026-03-26 21:24:130.04 (n/a)0.03 (n/a)0.04 (n/a)0.02 (n/a)0.01 (n/a)376.80 (n/a)265.54 (n/a)233.70 (n/a)221.70 (n/a)64.64 (n/a)

input_length_2048-num_cores_4-num_channels_1-bypass_False-tile_size_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8728b32 — 2026-03-26 21:24:130.03 (n/a)0.03 (n/a)0.03 (n/a)0.02 (n/a)0.01 (n/a)431.20 (n/a)282.36 (n/a)251.10 (n/a)236.70 (n/a)83.44 (n/a)

input_length_2048-num_cores_4-num_channels_2-bypass_False-tile_size_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8728b32 — 2026-03-26 21:24:130.03 (n/a)0.03 (n/a)0.03 (n/a)0.03 (n/a)0.00 (n/a)319.10 (n/a)262.30 (n/a)247.80 (n/a)246.00 (n/a)31.87 (n/a)

input_length_2048-num_cores_8-num_channels_1-bypass_False-tile_size_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8728b32 — 2026-03-26 21:24:130.04 (n/a)0.03 (n/a)0.03 (n/a)0.01 (n/a)0.01 (n/a)638.90 (n/a)316.38 (n/a)235.40 (n/a)232.40 (n/a)180.31 (n/a)

input_length_2048-num_cores_8-num_channels_2-bypass_False-tile_size_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8728b32 — 2026-03-26 21:24:130.03 (n/a)0.03 (n/a)0.03 (n/a)0.01 (n/a)0.01 (n/a)680.20 (n/a)336.16 (n/a)254.20 (n/a)240.40 (n/a)192.41 (n/a)

input_length_32768-num_aie_columns_2-num_channels_2-tile_size_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8728b32 — 2026-03-26 21:24:130.54 (n/a)0.47 (n/a)0.54 (n/a)0.22 (n/a)0.14 (n/a)605.70 (n/a)320.08 (n/a)243.80 (n/a)241.70 (n/a)160.03 (n/a)

input_length_32768-num_aie_columns_2-num_channels_2-tile_size_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8728b32 — 2026-03-26 21:24:130.57 (n/a)0.49 (n/a)0.54 (n/a)0.35 (n/a)0.09 (n/a)372.30 (n/a)275.54 (n/a)244.70 (n/a)231.20 (n/a)59.56 (n/a)

input_length_32768-num_aie_columns_4-num_channels_4-tile_size_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8728b32 — 2026-03-26 21:24:130.52 (n/a)0.35 (n/a)0.46 (n/a)0.13 (n/a)0.19 (n/a)993.20 (n/a)524.02 (n/a)283.80 (n/a)251.40 (n/a)357.89 (n/a)

layer_norm_1_cols_1_channels_2048_tile_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.07 (+35.61%)0.05 (+6.58%)0.04 (-0.65%)0.03 (-1.11%)0.01 (+98.33%)236.50 (+1.11%)182.90 (-3.12%)183.30 (+0.60%)118.50 (-26.26%)42.30 (+41.46%)
dc66da7 — 2026-03-25 19:23:450.05 (n/a)0.04 (n/a)0.04 (n/a)0.04 (n/a)0.01 (n/a)233.90 (n/a)188.80 (n/a)182.20 (n/a)160.70 (n/a)29.90 (n/a)

layer_norm_1_cols_2_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.07 (+9.47%)0.05 (+5.79%)0.05 (+5.48%)0.04 (-7.03%)0.01 (+51.12%)206.90 (+7.54%)163.06 (-2.60%)165.20 (-5.22%)113.20 (-8.71%)40.37 (+56.59%)
dc66da7 — 2026-03-25 19:23:450.07 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.01 (n/a)192.40 (n/a)167.42 (n/a)174.30 (n/a)124.00 (n/a)25.78 (n/a)

layer_norm_2_cols_1_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.06 (-0.97%)0.05 (+14.59%)0.05 (+47.43%)0.03 (+8.92%)0.01 (-0.31%)235.10 (-8.20%)183.06 (-13.03%)157.80 (-32.19%)148.10 (+1.02%)43.12 (-7.94%)
dc66da7 — 2026-03-25 19:23:450.06 (n/a)0.04 (n/a)0.04 (n/a)0.03 (n/a)0.01 (n/a)256.10 (n/a)210.48 (n/a)232.70 (n/a)146.60 (n/a)46.84 (n/a)

layer_norm_2_cols_2_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.07 (+14.65%)0.05 (+4.61%)0.05 (-7.99%)0.04 (+11.94%)0.01 (+28.58%)201.70 (-10.67%)168.58 (-3.65%)176.00 (+8.71%)122.10 (-12.79%)33.82 (+0.06%)
dc66da7 — 2026-03-25 19:23:450.06 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.01 (n/a)225.80 (n/a)174.96 (n/a)161.90 (n/a)140.00 (n/a)33.80 (n/a)

layer_norm_4_cols_1_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.06 (-31.05%)0.05 (-3.73%)0.05 (+9.03%)0.04 (+10.49%)0.01 (-63.42%)197.60 (-9.52%)167.50 (-4.83%)166.40 (-8.27%)129.00 (+45.11%)25.47 (-51.02%)
dc66da7 — 2026-03-25 19:23:450.09 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.02 (n/a)218.40 (n/a)176.00 (n/a)181.40 (n/a)88.90 (n/a)52.00 (n/a)

layer_norm_4_cols_2_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.05 (-1.52%)0.04 (-11.20%)0.04 (-25.45%)0.03 (-0.69%)0.01 (-9.35%)243.40 (+0.66%)202.20 (+12.10%)212.70 (+34.20%)156.90 (+1.55%)34.18 (-7.92%)
dc66da7 — 2026-03-25 19:23:450.05 (n/a)0.05 (n/a)0.05 (n/a)0.03 (n/a)0.01 (n/a)241.80 (n/a)180.38 (n/a)158.50 (n/a)154.50 (n/a)37.12 (n/a)

layer_norm_8_cols_1_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.05 (-6.07%)0.04 (+15.60%)0.04 (+21.55%)0.04 (+82.28%)0.01 (-55.28%)222.00 (-45.13%)190.18 (-21.85%)193.60 (-17.72%)153.10 (+6.47%)24.64 (-75.05%)
dc66da7 — 2026-03-25 19:23:450.06 (n/a)0.04 (n/a)0.03 (n/a)0.02 (n/a)0.01 (n/a)404.60 (n/a)243.34 (n/a)235.30 (n/a)143.80 (n/a)98.74 (n/a)

layer_norm_8_cols_2_channels_2048_tile_128

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.05 (+37.82%)0.04 (+27.22%)0.04 (+9.46%)0.03 (+34.60%)0.01 (+56.79%)251.50 (-25.72%)212.60 (-21.06%)226.70 (-8.66%)175.20 (-27.45%)33.04 (-18.34%)
dc66da7 — 2026-03-25 19:23:450.03 (n/a)0.03 (n/a)0.03 (n/a)0.02 (n/a)0.00 (n/a)338.60 (n/a)269.32 (n/a)248.20 (n/a)241.50 (n/a)40.46 (n/a)

matrix_vector_mul_128x128_32_1col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)
130b6ea — 2025-12-05 21:15:490.37 (+17.13%)0.30 (+24.87%)0.30 (+38.03%)0.23 (+12.47%)0.06 (+35.23%)
0a6c11c — 2025-12-04 00:39:100.32 (n/a)0.24 (n/a)0.22 (n/a)0.20 (n/a)0.04 (n/a)

matrix_vector_mul_128x128_32_1col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
331dcca — 2026-01-08 18:34:100.26 (-1.81%)0.24 (+4.06%)0.24 (+0.52%)0.21 (+16.97%)0.02 (-48.16%)0.26 (-1.81%)0.24 (+4.06%)0.24 (+0.52%)0.20 (+16.97%)0.02 (-48.16%)
a4b6ffe — 2026-01-05 19:24:050.26 (n/a)0.23 (n/a)0.24 (n/a)0.18 (n/a)0.04 (n/a)0.26 (n/a)0.23 (n/a)0.24 (n/a)0.17 (n/a)0.04 (n/a)

matrix_vector_mul_128x128_32tsi_128tso_1col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
8b6d0f4 — 2026-03-25 20:46:390.26 (+32.10%)0.19 (+11.55%)0.18 (+9.49%)0.13 (-2.22%)0.05 (+105.83%)0.25 (+32.10%)0.19 (+11.55%)0.18 (+9.49%)0.13 (-2.22%)0.05 (+105.83%)
dc66da7 — 2026-03-25 19:23:450.19 (n/a)0.17 (n/a)0.17 (n/a)0.14 (n/a)0.02 (n/a)0.19 (n/a)0.17 (n/a)0.17 (n/a)0.14 (n/a)0.02 (n/a)

matrix_vector_mul_128x128_32tsi_128tso_1col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
897d04e — 2026-03-06 22:12:250.37 (+5.44%)0.31 (-0.05%)0.34 (-1.99%)0.22 (-7.87%)0.06 (+15.13%)0.36 (+5.44%)0.31 (-0.05%)0.34 (-1.99%)0.21 (-7.87%)0.06 (+15.13%)
84d3478 — 2026-02-17 23:25:130.35 (n/a)0.31 (n/a)0.35 (n/a)0.23 (n/a)0.05 (n/a)0.35 (n/a)0.31 (n/a)0.34 (n/a)0.23 (n/a)0.05 (n/a)

matrix_vector_mul_2048x8192_1_1col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)
130b6ea — 2025-12-05 21:15:4913.57 (+3.47%)13.15 (+1.09%)13.10 (+0.16%)12.52 (-1.58%)0.37 (+153.19%)
0a6c11c — 2025-12-04 00:39:1013.11 (n/a)13.01 (n/a)13.07 (n/a)12.72 (n/a)0.15 (n/a)

matrix_vector_mul_2048x8192_1_1col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
331dcca — 2026-01-08 18:34:1013.43 (-0.07%)13.30 (+4.62%)13.27 (+1.29%)13.16 (+16.70%)0.12 (-86.69%)13.42 (-0.07%)13.29 (+4.62%)13.26 (+1.29%)13.16 (+16.70%)0.12 (-86.69%)
a4b6ffe — 2026-01-05 19:24:0513.44 (n/a)12.71 (n/a)13.10 (n/a)11.28 (n/a)0.87 (n/a)13.43 (n/a)12.70 (n/a)13.09 (n/a)11.27 (n/a)0.87 (n/a)

matrix_vector_mul_2048x8192_1_2col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)
130b6ea — 2025-12-05 21:15:4925.06 (+2.47%)23.71 (+1.81%)24.13 (+0.50%)20.90 (+5.41%)1.48 (-15.60%)
0a6c11c — 2025-12-04 00:39:1024.46 (n/a)23.29 (n/a)24.01 (n/a)19.83 (n/a)1.75 (n/a)

matrix_vector_mul_2048x8192_1_2col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
331dcca — 2026-01-08 18:34:1024.46 (-0.60%)21.07 (-8.03%)19.75 (-17.83%)18.08 (+1.18%)3.03 (+7.07%)24.44 (-0.60%)21.06 (-8.03%)19.74 (-17.83%)18.07 (+1.18%)3.03 (+7.07%)
a4b6ffe — 2026-01-05 19:24:0524.60 (n/a)22.91 (n/a)24.04 (n/a)17.87 (n/a)2.83 (n/a)24.59 (n/a)22.90 (n/a)24.02 (n/a)17.86 (n/a)2.83 (n/a)

matrix_vector_mul_2048x8192_1_4col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)
130b6ea — 2025-12-05 21:15:4942.05 (+1.54%)40.50 (+12.60%)40.49 (+2.36%)38.98 (+97.53%)0.97 (-88.09%)
0a6c11c — 2025-12-04 00:39:1041.41 (n/a)35.97 (n/a)39.55 (n/a)19.73 (n/a)8.15 (n/a)

matrix_vector_mul_2048x8192_1_4col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
331dcca — 2026-01-08 18:34:1042.58 (+1.51%)41.37 (+14.29%)41.25 (+4.89%)40.33 (+91.42%)0.92 (-89.18%)42.56 (+1.51%)41.34 (+14.29%)41.23 (+4.89%)40.30 (+91.42%)0.92 (-89.18%)
a4b6ffe — 2026-01-05 19:24:0541.95 (n/a)36.20 (n/a)39.33 (n/a)21.07 (n/a)8.54 (n/a)41.92 (n/a)36.17 (n/a)39.31 (n/a)21.05 (n/a)8.53 (n/a)

matrix_vector_mul_2048x8192_1_8col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)
130b6ea — 2025-12-05 21:15:4946.62 (+3.11%)44.82 (+0.17%)44.78 (+0.17%)42.60 (-3.80%)1.60 (+367.72%)
0a6c11c — 2025-12-04 00:39:1045.21 (n/a)44.75 (n/a)44.70 (n/a)44.29 (n/a)0.34 (n/a)

matrix_vector_mul_2048x8192_1_8col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
331dcca — 2026-01-08 18:34:1046.04 (+4.14%)45.06 (+3.47%)44.55 (+2.76%)44.30 (+2.84%)0.87 (+66.58%)46.01 (+4.14%)45.03 (+3.47%)44.52 (+2.76%)44.27 (+2.84%)0.87 (+66.58%)
a4b6ffe — 2026-01-05 19:24:0544.21 (n/a)43.55 (n/a)43.35 (n/a)43.07 (n/a)0.52 (n/a)44.18 (n/a)43.52 (n/a)43.32 (n/a)43.05 (n/a)0.52 (n/a)

matrix_vector_mul_2048x8192_1tsi_1024tso_2col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
8b6d0f4 — 2026-03-25 20:46:3925.44 (+1.95%)24.42 (-0.28%)24.79 (+0.16%)23.14 (-2.33%)1.05 (+106.47%)25.42 (+1.95%)24.41 (-0.28%)24.77 (+0.16%)23.12 (-2.33%)1.05 (+106.47%)
dc66da7 — 2026-03-25 19:23:4524.95 (n/a)24.49 (n/a)24.75 (n/a)23.69 (n/a)0.51 (n/a)24.94 (n/a)24.48 (n/a)24.73 (n/a)23.67 (n/a)0.51 (n/a)

matrix_vector_mul_2048x8192_1tsi_1024tso_2col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
897d04e — 2026-03-06 22:12:2524.47 (-3.32%)24.16 (-3.13%)24.20 (-3.09%)23.84 (-3.04%)0.25 (-14.36%)24.46 (-3.32%)24.15 (-3.13%)24.18 (-3.09%)23.83 (-3.04%)0.25 (-14.36%)
84d3478 — 2026-02-17 23:25:1325.31 (n/a)24.94 (n/a)24.97 (n/a)24.59 (n/a)0.29 (n/a)25.30 (n/a)24.93 (n/a)24.95 (n/a)24.58 (n/a)0.29 (n/a)

matrix_vector_mul_2048x8192_1tsi_2048tso_1col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
8b6d0f4 — 2026-03-25 20:46:3913.48 (-1.01%)13.15 (+1.31%)13.11 (-0.87%)12.95 (+6.77%)0.20 (-66.05%)13.47 (-1.01%)13.14 (+1.31%)13.11 (-0.87%)12.95 (+6.77%)0.20 (-66.05%)
dc66da7 — 2026-03-25 19:23:4513.62 (n/a)12.98 (n/a)13.23 (n/a)12.13 (n/a)0.58 (n/a)13.61 (n/a)12.97 (n/a)13.22 (n/a)12.13 (n/a)0.58 (n/a)

matrix_vector_mul_2048x8192_1tsi_2048tso_1col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
897d04e — 2026-03-06 22:12:2513.25 (-3.04%)13.02 (-2.92%)13.10 (-2.64%)12.56 (-4.14%)0.27 (+32.55%)13.24 (-3.04%)13.01 (-2.92%)13.09 (-2.64%)12.55 (-4.14%)0.27 (+32.55%)
84d3478 — 2026-02-17 23:25:1313.66 (n/a)13.41 (n/a)13.45 (n/a)13.10 (n/a)0.20 (n/a)13.65 (n/a)13.40 (n/a)13.44 (n/a)13.09 (n/a)0.20 (n/a)

matrix_vector_mul_2048x8192_1tsi_256tso_8col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
8b6d0f4 — 2026-03-25 20:46:3942.06 (-4.87%)40.85 (-4.53%)40.57 (-4.52%)39.78 (-4.49%)0.87 (-11.49%)42.04 (-4.87%)40.83 (-4.53%)40.55 (-4.52%)39.76 (-4.49%)0.87 (-11.49%)
dc66da7 — 2026-03-25 19:23:4544.21 (n/a)42.79 (n/a)42.49 (n/a)41.65 (n/a)0.99 (n/a)44.19 (n/a)42.76 (n/a)42.47 (n/a)41.63 (n/a)0.99 (n/a)

matrix_vector_mul_2048x8192_1tsi_256tso_8col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
897d04e — 2026-03-06 22:12:2547.71 (+3.51%)45.38 (+9.55%)45.22 (-1.27%)43.47 (+75.24%)1.52 (-83.67%)47.68 (+3.51%)45.35 (+9.55%)45.19 (-1.27%)43.45 (+75.24%)1.52 (-83.67%)
84d3478 — 2026-02-17 23:25:1346.09 (n/a)41.43 (n/a)45.80 (n/a)24.81 (n/a)9.31 (n/a)46.06 (n/a)41.40 (n/a)45.77 (n/a)24.79 (n/a)9.30 (n/a)

matrix_vector_mul_2048x8192_1tsi_512tso_4col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
8b6d0f4 — 2026-03-25 20:46:3940.94 (-0.67%)39.64 (+0.12%)39.96 (+1.02%)37.79 (+0.51%)1.25 (-21.17%)40.91 (-0.67%)39.61 (+0.12%)39.94 (+1.02%)37.77 (+0.51%)1.25 (-21.17%)
dc66da7 — 2026-03-25 19:23:4541.21 (n/a)39.59 (n/a)39.56 (n/a)37.60 (n/a)1.58 (n/a)41.19 (n/a)39.57 (n/a)39.53 (n/a)37.58 (n/a)1.58 (n/a)

matrix_vector_mul_2048x8192_1tsi_512tso_4col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
897d04e — 2026-03-06 22:12:2540.84 (-0.32%)39.53 (-1.25%)39.40 (-1.97%)38.87 (-0.69%)0.77 (-6.24%)40.81 (-0.32%)39.50 (-1.25%)39.37 (-1.97%)38.84 (-0.69%)0.77 (-6.24%)
84d3478 — 2026-02-17 23:25:1340.97 (n/a)40.03 (n/a)40.19 (n/a)39.14 (n/a)0.82 (n/a)40.95 (n/a)40.00 (n/a)40.17 (n/a)39.11 (n/a)0.82 (n/a)

matrix_vector_mul_8192x2048_4_1col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)
130b6ea — 2025-12-05 21:15:4913.29 (+1.63%)13.19 (+6.08%)13.19 (+1.17%)13.09 (+26.86%)0.08 (-92.94%)
0a6c11c — 2025-12-04 00:39:1013.08 (n/a)12.44 (n/a)13.03 (n/a)10.32 (n/a)1.07 (n/a)

matrix_vector_mul_8192x2048_4_1col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
331dcca — 2026-01-08 18:34:1013.68 (+3.42%)13.30 (+8.54%)13.38 (+7.25%)12.72 (+14.96%)0.35 (-66.09%)13.67 (+3.42%)13.29 (+8.54%)13.38 (+7.25%)12.71 (+14.96%)0.35 (-66.09%)
a4b6ffe — 2026-01-05 19:24:0513.23 (n/a)12.25 (n/a)12.48 (n/a)11.07 (n/a)1.04 (n/a)13.22 (n/a)12.24 (n/a)12.47 (n/a)11.06 (n/a)1.04 (n/a)

matrix_vector_mul_8192x2048_4_2col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)
130b6ea — 2025-12-05 21:15:4925.58 (+4.24%)24.23 (+6.59%)24.03 (+0.29%)23.58 (+37.29%)0.70 (-74.97%)
0a6c11c — 2025-12-04 00:39:1024.54 (n/a)22.73 (n/a)23.96 (n/a)17.18 (n/a)2.79 (n/a)

matrix_vector_mul_8192x2048_4_2col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
331dcca — 2026-01-08 18:34:1025.08 (+1.32%)24.68 (+13.42%)24.61 (+3.26%)24.47 (+48.32%)0.24 (-93.56%)25.07 (+1.32%)24.67 (+13.42%)24.59 (+3.26%)24.46 (+48.32%)0.24 (-93.56%)
a4b6ffe — 2026-01-05 19:24:0524.76 (n/a)21.76 (n/a)23.83 (n/a)16.50 (n/a)3.74 (n/a)24.74 (n/a)21.75 (n/a)23.82 (n/a)16.49 (n/a)3.73 (n/a)

matrix_vector_mul_8192x2048_4_4col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)
130b6ea — 2025-12-05 21:15:4940.94 (-2.19%)39.74 (-1.10%)39.76 (-0.60%)38.17 (-2.14%)1.04 (+4.39%)
0a6c11c — 2025-12-04 00:39:1041.86 (n/a)40.18 (n/a)40.00 (n/a)39.00 (n/a)1.00 (n/a)

matrix_vector_mul_8192x2048_4_4col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
331dcca — 2026-01-08 18:34:1040.74 (-1.09%)37.16 (-7.15%)40.25 (+1.47%)24.51 (-37.44%)7.07 (+736.13%)40.72 (-1.09%)37.13 (-7.15%)40.23 (+1.47%)24.50 (-37.44%)7.07 (+736.13%)
a4b6ffe — 2026-01-05 19:24:0541.19 (n/a)40.02 (n/a)39.67 (n/a)39.18 (n/a)0.85 (n/a)41.17 (n/a)39.99 (n/a)39.64 (n/a)39.16 (n/a)0.85 (n/a)

matrix_vector_mul_8192x2048_4_8col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)
130b6ea — 2025-12-05 21:15:4946.71 (-1.04%)44.91 (-2.34%)44.71 (-2.98%)43.80 (-1.44%)0.96 (+6.93%)
0a6c11c — 2025-12-04 00:39:1047.20 (n/a)45.98 (n/a)46.08 (n/a)44.44 (n/a)0.90 (n/a)

matrix_vector_mul_8192x2048_4_8col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
331dcca — 2026-01-08 18:34:1046.28 (+5.68%)44.89 (+14.59%)44.74 (+4.26%)44.20 (+62.80%)0.84 (-87.96%)46.25 (+5.68%)44.86 (+14.59%)44.71 (+4.26%)44.17 (+62.80%)0.84 (-87.96%)
a4b6ffe — 2026-01-05 19:24:0543.79 (n/a)39.17 (n/a)42.91 (n/a)27.15 (n/a)6.98 (n/a)43.77 (n/a)39.15 (n/a)42.88 (n/a)27.13 (n/a)6.98 (n/a)

matrix_vector_mul_8192x2048_4tsi_1024tso_1col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
8b6d0f4 — 2026-03-25 20:46:3913.35 (+0.83%)13.11 (+3.52%)13.10 (+3.94%)12.94 (+7.70%)0.15 (-69.46%)13.34 (+0.83%)13.10 (+3.52%)13.09 (+3.94%)12.94 (+7.70%)0.15 (-69.46%)
dc66da7 — 2026-03-25 19:23:4513.24 (n/a)12.66 (n/a)12.60 (n/a)12.02 (n/a)0.49 (n/a)13.23 (n/a)12.66 (n/a)12.59 (n/a)12.01 (n/a)0.49 (n/a)

matrix_vector_mul_8192x2048_4tsi_1024tso_1col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
897d04e — 2026-03-06 22:12:2513.34 (-0.79%)13.20 (-0.36%)13.18 (-1.55%)13.12 (+1.75%)0.08 (-65.93%)13.33 (-0.79%)13.19 (-0.36%)13.17 (-1.55%)13.11 (+1.75%)0.08 (-65.93%)
84d3478 — 2026-02-17 23:25:1313.45 (n/a)13.25 (n/a)13.39 (n/a)12.90 (n/a)0.25 (n/a)13.44 (n/a)13.24 (n/a)13.38 (n/a)12.89 (n/a)0.25 (n/a)

matrix_vector_mul_8192x2048_4tsi_1024tso_2col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
8b6d0f4 — 2026-03-25 20:46:3924.35 (-0.26%)24.07 (+0.89%)24.16 (+1.48%)23.64 (+2.17%)0.27 (-44.88%)24.34 (-0.26%)24.06 (+0.89%)24.15 (+1.48%)23.63 (+2.17%)0.27 (-44.88%)
dc66da7 — 2026-03-25 19:23:4524.41 (n/a)23.86 (n/a)23.81 (n/a)23.14 (n/a)0.49 (n/a)24.40 (n/a)23.84 (n/a)23.80 (n/a)23.12 (n/a)0.49 (n/a)

matrix_vector_mul_8192x2048_4tsi_1024tso_2col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
897d04e — 2026-03-06 22:12:2524.10 (-5.45%)23.07 (-1.85%)24.01 (-2.88%)19.26 (+7.47%)2.13 (-32.15%)24.08 (-5.45%)23.05 (-1.85%)23.99 (-2.88%)19.25 (+7.47%)2.13 (-32.15%)
84d3478 — 2026-02-17 23:25:1325.48 (n/a)23.50 (n/a)24.72 (n/a)17.92 (n/a)3.14 (n/a)25.47 (n/a)23.48 (n/a)24.71 (n/a)17.91 (n/a)3.14 (n/a)

matrix_vector_mul_8192x2048_4tsi_1024tso_4col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
8b6d0f4 — 2026-03-25 20:46:3940.84 (-1.16%)38.86 (-0.66%)38.49 (-1.61%)37.90 (+1.94%)1.14 (-23.22%)40.82 (-1.16%)38.84 (-0.66%)38.46 (-1.61%)37.87 (+1.94%)1.14 (-23.22%)
dc66da7 — 2026-03-25 19:23:4541.32 (n/a)39.12 (n/a)39.12 (n/a)37.18 (n/a)1.49 (n/a)41.30 (n/a)39.09 (n/a)39.09 (n/a)37.15 (n/a)1.49 (n/a)

matrix_vector_mul_8192x2048_4tsi_1024tso_4col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
897d04e — 2026-03-06 22:12:2540.80 (-0.16%)39.58 (-1.93%)39.75 (-1.71%)38.28 (-2.50%)1.10 (+67.33%)40.78 (-0.16%)39.55 (-1.93%)39.73 (-1.71%)38.26 (-2.50%)1.10 (+67.33%)
84d3478 — 2026-02-17 23:25:1340.87 (n/a)40.36 (n/a)40.45 (n/a)39.26 (n/a)0.66 (n/a)40.84 (n/a)40.33 (n/a)40.42 (n/a)39.24 (n/a)0.65 (n/a)

matrix_vector_mul_8192x2048_4tsi_1024tso_8col

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
8b6d0f4 — 2026-03-25 20:46:3941.79 (-12.64%)40.98 (-5.83%)40.78 (-1.04%)40.36 (-0.60%)0.56 (-84.11%)41.76 (-12.64%)40.96 (-5.83%)40.76 (-1.04%)40.34 (-0.60%)0.56 (-84.11%)
dc66da7 — 2026-03-25 19:23:4547.83 (n/a)43.52 (n/a)41.21 (n/a)40.61 (n/a)3.53 (n/a)47.80 (n/a)43.50 (n/a)41.19 (n/a)40.58 (n/a)3.53 (n/a)

matrix_vector_mul_8192x2048_4tsi_1024tso_8col0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Throughput (max)Throughput (mean)Throughput (median)Throughput (min)Throughput (stddev)
897d04e — 2026-03-06 22:12:2549.31 (+6.27%)46.12 (+1.67%)45.30 (-0.27%)44.62 (+1.87%)1.85 (+85.10%)49.28 (+6.27%)46.10 (+1.67%)45.28 (-0.27%)44.59 (+1.87%)1.85 (+85.10%)
84d3478 — 2026-02-17 23:25:1346.40 (n/a)45.37 (n/a)45.43 (n/a)43.80 (n/a)1.00 (n/a)46.37 (n/a)45.34 (n/a)45.40 (n/a)43.78 (n/a)1.00 (n/a)

mem_copy_16_cores_2_chans_2048_tile_128_False

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.05 (+7.16%)0.03 (-9.22%)0.03 (-18.40%)0.02 (-19.93%)0.01 (+42.24%)394.80 (+24.90%)264.56 (+14.35%)259.20 (+22.55%)179.80 (-6.65%)81.04 (+64.09%)
dc66da7 — 2026-03-25 19:23:450.04 (n/a)0.04 (n/a)0.04 (n/a)0.03 (n/a)0.01 (n/a)316.10 (n/a)231.36 (n/a)211.50 (n/a)192.60 (n/a)49.39 (n/a)

mem_copy_16_cores_2_chans_2048_tile_128_False0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.08 (-2.12%)0.06 (-16.44%)0.06 (-21.48%)0.03 (-23.84%)0.02 (+10.15%)247.60 (+31.28%)148.38 (+24.08%)128.30 (+27.41%)100.10 (+2.14%)58.70 (+50.73%)
84d3478 — 2026-02-17 23:25:130.08 (n/a)0.07 (n/a)0.08 (n/a)0.04 (n/a)0.02 (n/a)188.60 (n/a)119.58 (n/a)100.70 (n/a)98.00 (n/a)38.94 (n/a)

mem_copy_1_cols_1_channels_2048_tile_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
003b49d — 2025-11-14 19:30:540.05 (-16.99%)0.05 (-12.47%)0.05 (-7.32%)0.04 (-11.17%)0.00 (-29.61%)197.00 (+12.57%)171.00 (+13.70%)164.00 (+7.89%)153.00 (+20.47%)17.49 (-4.32%)
0999a9a — 2025-11-13 19:24:080.06 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.01 (n/a)175.00 (n/a)150.40 (n/a)152.00 (n/a)127.00 (n/a)18.28 (n/a)

mem_copy_1_cols_2_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
003b49d — 2025-11-14 19:30:540.06 (+5.56%)0.05 (-1.59%)0.05 (-1.39%)0.03 (-14.07%)0.01 (+25.03%)263.00 (+16.37%)166.00 (+4.53%)144.00 (+1.41%)126.00 (-5.26%)49.38 (+43.17%)
0999a9a — 2025-11-13 19:24:080.06 (n/a)0.05 (n/a)0.05 (n/a)0.03 (n/a)0.01 (n/a)226.00 (n/a)158.80 (n/a)142.00 (n/a)133.00 (n/a)34.49 (n/a)

mem_copy_1_cores_1_chans_2048_tile_2048_False

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.09 (+75.06%)0.05 (+7.82%)0.04 (-15.32%)0.03 (-13.78%)0.02 (+425.35%)249.10 (+15.97%)190.02 (+3.98%)211.70 (+18.07%)95.60 (-42.89%)62.62 (+236.69%)
dc66da7 — 2026-03-25 19:23:450.05 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.00 (n/a)214.80 (n/a)182.74 (n/a)179.30 (n/a)167.40 (n/a)18.60 (n/a)

mem_copy_1_cores_1_chans_2048_tile_2048_False0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.11 (-0.24%)0.08 (-17.53%)0.08 (-27.05%)0.06 (-15.43%)0.02 (+4.62%)148.50 (+18.23%)113.02 (+23.38%)103.60 (+37.04%)74.10 (+0.14%)31.69 (+33.33%)
84d3478 — 2026-02-17 23:25:130.11 (n/a)0.09 (n/a)0.11 (n/a)0.07 (n/a)0.02 (n/a)125.60 (n/a)91.60 (n/a)75.60 (n/a)74.00 (n/a)23.77 (n/a)

mem_copy_2_cols_1_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
003b49d — 2025-11-14 19:30:540.06 (-15.32%)0.05 (-17.09%)0.05 (-23.81%)0.04 (-15.54%)0.01 (-18.18%)193.00 (+18.40%)160.80 (+20.36%)168.00 (+31.25%)124.00 (+18.10%)22.92 (+11.39%)
0999a9a — 2025-11-13 19:24:080.07 (n/a)0.06 (n/a)0.06 (n/a)0.05 (n/a)0.01 (n/a)163.00 (n/a)133.60 (n/a)128.00 (n/a)105.00 (n/a)20.58 (n/a)

mem_copy_2_cols_2_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
003b49d — 2025-11-14 19:30:540.06 (-23.02%)0.05 (-6.98%)0.05 (-2.13%)0.04 (-15.31%)0.01 (-33.29%)196.00 (+18.07%)147.00 (+6.52%)141.00 (+2.17%)126.00 (+29.90%)25.27 (+8.13%)
0999a9a — 2025-11-13 19:24:080.08 (n/a)0.06 (n/a)0.06 (n/a)0.05 (n/a)0.01 (n/a)166.00 (n/a)138.00 (n/a)138.00 (n/a)97.00 (n/a)23.37 (n/a)

mem_copy_2_cores_1_chans_2048_tile_1024_False

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.07 (+13.70%)0.05 (-1.93%)0.05 (-7.07%)0.04 (-27.15%)0.02 (+157.23%)230.10 (+37.29%)166.58 (+9.82%)161.90 (+7.65%)109.70 (-12.03%)53.34 (+211.05%)
dc66da7 — 2026-03-25 19:23:450.07 (n/a)0.05 (n/a)0.05 (n/a)0.05 (n/a)0.01 (n/a)167.60 (n/a)151.68 (n/a)150.40 (n/a)124.70 (n/a)17.15 (n/a)

mem_copy_2_cores_1_chans_2048_tile_1024_False0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.11 (+27.03%)0.09 (+25.79%)0.11 (+27.09%)0.05 (+12.74%)0.03 (+43.93%)161.90 (-11.29%)94.34 (-18.35%)76.30 (-21.34%)74.00 (-21.36%)38.01 (+0.49%)
84d3478 — 2026-02-17 23:25:130.09 (n/a)0.08 (n/a)0.08 (n/a)0.04 (n/a)0.02 (n/a)182.50 (n/a)115.54 (n/a)97.00 (n/a)94.10 (n/a)37.82 (n/a)

mem_copy_2_cores_2_chans_2048_tile_1024_False

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.07 (-3.17%)0.05 (-7.71%)0.05 (-6.65%)0.04 (-20.60%)0.02 (+42.64%)223.40 (+25.93%)171.62 (+13.04%)172.10 (+7.09%)120.40 (+3.26%)49.35 (+82.52%)
dc66da7 — 2026-03-25 19:23:450.07 (n/a)0.06 (n/a)0.05 (n/a)0.05 (n/a)0.01 (n/a)177.40 (n/a)151.82 (n/a)160.70 (n/a)116.60 (n/a)27.04 (n/a)

mem_copy_2_cores_2_chans_2048_tile_1024_False0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.10 (+2.70%)0.08 (-5.38%)0.09 (+2.10%)0.04 (-43.30%)0.02 (+170.50%)201.60 (+76.38%)112.08 (+16.53%)90.80 (-2.05%)84.60 (-2.65%)50.14 (+375.75%)
84d3478 — 2026-02-17 23:25:130.09 (n/a)0.09 (n/a)0.09 (n/a)0.07 (n/a)0.01 (n/a)114.30 (n/a)96.18 (n/a)92.70 (n/a)86.90 (n/a)10.54 (n/a)

mem_copy_4_cols_1_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
003b49d — 2025-11-14 19:30:540.07 (+11.97%)0.05 (+3.09%)0.05 (+2.10%)0.03 (-11.35%)0.01 (+41.65%)229.00 (+12.81%)156.00 (-0.51%)143.00 (-2.05%)117.00 (-10.69%)38.21 (+48.71%)
0999a9a — 2025-11-13 19:24:080.06 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.01 (n/a)203.00 (n/a)156.80 (n/a)146.00 (n/a)131.00 (n/a)25.69 (n/a)

mem_copy_4_cols_2_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
003b49d — 2025-11-14 19:30:540.06 (+9.02%)0.05 (+5.59%)0.05 (+0.00%)0.04 (+3.63%)0.01 (+25.48%)193.00 (-3.50%)147.40 (-4.66%)147.00 (+0.00%)122.00 (-8.27%)25.24 (+7.48%)
0999a9a — 2025-11-13 19:24:080.06 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.01 (n/a)200.00 (n/a)154.60 (n/a)147.00 (n/a)133.00 (n/a)23.48 (n/a)

mem_copy_4_cores_1_chans_2048_tile_512_False

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.06 (-2.79%)0.05 (-5.43%)0.05 (+6.39%)0.03 (-12.50%)0.01 (+16.26%)242.20 (+14.25%)184.14 (+7.91%)167.80 (-6.05%)128.30 (+2.89%)46.84 (+42.55%)
dc66da7 — 2026-03-25 19:23:450.07 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.01 (n/a)212.00 (n/a)170.64 (n/a)178.60 (n/a)124.70 (n/a)32.86 (n/a)

mem_copy_4_cores_1_chans_2048_tile_512_False0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.09 (-5.87%)0.08 (-3.61%)0.08 (-0.79%)0.06 (-12.74%)0.01 (+22.04%)146.30 (+14.66%)108.02 (+4.96%)99.30 (+0.81%)95.10 (+6.26%)21.49 (+48.38%)
84d3478 — 2026-02-17 23:25:130.09 (n/a)0.08 (n/a)0.08 (n/a)0.06 (n/a)0.01 (n/a)127.60 (n/a)102.92 (n/a)98.50 (n/a)89.50 (n/a)14.49 (n/a)

mem_copy_4_cores_2_chans_2048_tile_512_False

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.05 (+7.78%)0.04 (-14.27%)0.04 (-10.58%)0.03 (-32.73%)0.01 (+109.81%)316.90 (+48.71%)226.08 (+22.64%)209.60 (+11.85%)150.20 (-7.23%)61.97 (+194.31%)
dc66da7 — 2026-03-25 19:23:450.05 (n/a)0.04 (n/a)0.04 (n/a)0.04 (n/a)0.01 (n/a)213.10 (n/a)184.34 (n/a)187.40 (n/a)161.90 (n/a)21.05 (n/a)

mem_copy_4_cores_2_chans_2048_tile_512_False0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.11 (+13.04%)0.09 (+17.19%)0.10 (+20.43%)0.06 (+17.23%)0.02 (+16.34%)142.80 (-14.70%)91.94 (-14.71%)79.00 (-17.02%)77.70 (-11.50%)28.50 (-14.82%)
84d3478 — 2026-02-17 23:25:130.09 (n/a)0.08 (n/a)0.09 (n/a)0.05 (n/a)0.02 (n/a)167.40 (n/a)107.80 (n/a)95.20 (n/a)87.80 (n/a)33.46 (n/a)

mem_copy_8_cols_1_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
003b49d — 2025-11-14 19:30:540.05 (-13.48%)0.05 (-17.79%)0.05 (-15.33%)0.03 (-25.09%)0.01 (+6.67%)267.00 (+33.50%)176.80 (+24.51%)150.00 (+18.11%)141.00 (+15.57%)47.41 (+61.41%)
0999a9a — 2025-11-13 19:24:080.06 (n/a)0.06 (n/a)0.06 (n/a)0.04 (n/a)0.01 (n/a)200.00 (n/a)142.00 (n/a)127.00 (n/a)122.00 (n/a)29.37 (n/a)

mem_copy_8_cols_2_channels_2048_tile_128

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
003b49d — 2025-11-14 19:30:540.05 (+13.38%)0.05 (+23.85%)0.05 (+37.91%)0.05 (+30.72%)0.00 (-56.85%)166.00 (-23.50%)153.40 (-20.60%)153.00 (-27.49%)142.00 (-11.80%)7.63 (-70.60%)
0999a9a — 2025-11-13 19:24:080.05 (n/a)0.04 (n/a)0.04 (n/a)0.04 (n/a)0.01 (n/a)217.00 (n/a)193.20 (n/a)211.00 (n/a)161.00 (n/a)25.96 (n/a)

mem_copy_8_cores_1_chans_2048_tile_256_False

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.06 (-5.71%)0.04 (+0.46%)0.04 (+2.21%)0.03 (-7.12%)0.01 (-7.04%)278.00 (+7.67%)209.24 (-0.41%)216.20 (-2.17%)141.90 (+6.05%)51.53 (+9.47%)
dc66da7 — 2026-03-25 19:23:450.06 (n/a)0.04 (n/a)0.04 (n/a)0.03 (n/a)0.01 (n/a)258.20 (n/a)210.10 (n/a)221.00 (n/a)133.80 (n/a)47.07 (n/a)

mem_copy_8_cores_1_chans_2048_tile_256_False0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.08 (-0.39%)0.07 (+10.07%)0.08 (+1.07%)0.06 (+38.38%)0.01 (-47.35%)136.70 (-27.75%)110.64 (-13.07%)104.70 (-1.13%)103.00 (+0.39%)14.59 (-60.78%)
84d3478 — 2026-02-17 23:25:130.08 (n/a)0.07 (n/a)0.08 (n/a)0.04 (n/a)0.02 (n/a)189.20 (n/a)127.28 (n/a)105.90 (n/a)102.60 (n/a)37.20 (n/a)

mem_copy_8_cores_2_chans_2048_tile_256_False

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.06 (+19.11%)0.04 (+1.88%)0.04 (-11.30%)0.03 (+24.20%)0.01 (+4.86%)262.00 (-19.48%)199.56 (-3.59%)200.50 (+12.70%)139.70 (-16.05%)45.93 (-31.63%)
dc66da7 — 2026-03-25 19:23:450.05 (n/a)0.04 (n/a)0.05 (n/a)0.03 (n/a)0.01 (n/a)325.40 (n/a)207.00 (n/a)177.90 (n/a)166.40 (n/a)67.19 (n/a)

mem_copy_8_cores_2_chans_2048_tile_256_False0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.08 (-17.85%)0.07 (-24.66%)0.08 (-18.40%)0.04 (-32.08%)0.02 (+21.78%)185.30 (+47.18%)127.72 (+37.87%)102.00 (+22.45%)100.40 (+21.70%)38.66 (+106.34%)
84d3478 — 2026-02-17 23:25:130.10 (n/a)0.09 (n/a)0.10 (n/a)0.07 (n/a)0.01 (n/a)125.90 (n/a)92.64 (n/a)83.30 (n/a)82.50 (n/a)18.74 (n/a)

mha

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.18 (+0.10%)0.18 (+0.09%)0.18 (+0.09%)0.18 (+0.08%)0.00 (+11.08%)47507.50 (-0.08%)47468.96 (-0.09%)47481.10 (-0.09%)47413.10 (-0.10%)37.32 (+10.96%)
dc66da7 — 2026-03-25 19:23:450.18 (n/a)0.18 (n/a)0.18 (n/a)0.18 (n/a)0.00 (n/a)47544.20 (n/a)47512.60 (n/a)47521.80 (n/a)47459.60 (n/a)33.63 (n/a)

mha0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
84d3478 — 2026-02-17 23:25:130.18 (-0.51%)0.18 (-0.36%)0.18 (-0.41%)0.18 (-0.21%)0.00 (-35.21%)47468.60 (+0.21%)47403.94 (+0.36%)47426.00 (+0.41%)47314.40 (+0.51%)68.83 (-34.73%)
1531781 — 2026-01-23 18:49:240.18 (n/a)0.18 (n/a)0.18 (n/a)0.18 (n/a)0.00 (n/a)47367.50 (n/a)47232.68 (n/a)47232.00 (n/a)47073.00 (n/a)105.45 (n/a)

mha_16384_64_1_8_0_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.18 (n/a)0.18 (n/a)0.18 (n/a)0.18 (n/a)0.00 (n/a)47533.70 (n/a)47380.56 (n/a)47337.90 (n/a)47268.00 (n/a)110.60 (n/a)

relu_1_cols_1_channels_2048_tile_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.07 (+14.06%)0.05 (+0.31%)0.05 (-3.71%)0.04 (-18.38%)0.01 (+129.69%)208.90 (+22.52%)161.96 (+2.56%)166.90 (+3.86%)120.40 (-12.31%)33.36 (+144.00%)
dc66da7 — 2026-03-25 19:23:450.06 (n/a)0.05 (n/a)0.05 (n/a)0.05 (n/a)0.00 (n/a)170.50 (n/a)157.92 (n/a)160.70 (n/a)137.30 (n/a)13.67 (n/a)

relu_2_cols_1_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.06 (-20.92%)0.05 (-6.86%)0.05 (-1.28%)0.03 (-20.49%)0.01 (-19.18%)264.20 (+25.81%)174.76 (+7.41%)149.00 (+1.29%)140.20 (+26.42%)52.29 (+24.11%)
dc66da7 — 2026-03-25 19:23:450.07 (n/a)0.05 (n/a)0.06 (n/a)0.04 (n/a)0.01 (n/a)210.00 (n/a)162.70 (n/a)147.10 (n/a)110.90 (n/a)42.13 (n/a)

relu_4_cols_1_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.05 (-8.73%)0.04 (-10.26%)0.04 (-4.04%)0.03 (-16.73%)0.01 (+5.34%)259.40 (+20.09%)206.80 (+12.91%)200.50 (+4.21%)150.20 (+9.56%)45.96 (+41.47%)
dc66da7 — 2026-03-25 19:23:450.06 (n/a)0.05 (n/a)0.04 (n/a)0.04 (n/a)0.01 (n/a)216.00 (n/a)183.16 (n/a)192.40 (n/a)137.10 (n/a)32.49 (n/a)

relu_8_cols_1_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.05 (+4.18%)0.04 (-0.33%)0.04 (-1.55%)0.02 (-20.26%)0.01 (+47.86%)369.50 (+25.38%)233.74 (+5.19%)215.30 (+1.56%)165.20 (-4.01%)80.83 (+78.85%)
dc66da7 — 2026-03-25 19:23:450.05 (n/a)0.04 (n/a)0.04 (n/a)0.03 (n/a)0.01 (n/a)294.70 (n/a)222.20 (n/a)212.00 (n/a)172.10 (n/a)45.19 (n/a)

rms_norm_1_cols_1_channels_2048_tile_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.06 (-18.06%)0.05 (-2.61%)0.05 (-1.31%)0.04 (+12.13%)0.01 (-43.39%)226.00 (-10.81%)172.66 (-1.55%)163.30 (+1.30%)140.00 (+22.06%)32.37 (-37.02%)
dc66da7 — 2026-03-25 19:23:450.07 (n/a)0.05 (n/a)0.05 (n/a)0.03 (n/a)0.01 (n/a)253.40 (n/a)175.38 (n/a)161.20 (n/a)114.70 (n/a)51.40 (n/a)

rms_norm_1_cols_2_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.08 (+16.34%)0.05 (+6.02%)0.05 (-3.17%)0.04 (+12.01%)0.02 (+27.81%)218.40 (-10.71%)162.92 (-3.80%)150.40 (+3.23%)108.00 (-14.01%)52.41 (+3.43%)
dc66da7 — 2026-03-25 19:23:450.07 (n/a)0.05 (n/a)0.06 (n/a)0.03 (n/a)0.01 (n/a)244.60 (n/a)169.36 (n/a)145.70 (n/a)125.60 (n/a)50.68 (n/a)

rms_norm_2_cols_1_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.06 (+9.52%)0.04 (-9.59%)0.04 (-17.77%)0.03 (-18.95%)0.01 (+55.38%)255.30 (+23.39%)199.88 (+13.52%)205.30 (+21.62%)137.80 (-8.68%)43.62 (+70.31%)
dc66da7 — 2026-03-25 19:23:450.05 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.01 (n/a)206.90 (n/a)176.08 (n/a)168.80 (n/a)150.90 (n/a)25.61 (n/a)

rms_norm_2_cols_2_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.06 (+6.05%)0.04 (+0.61%)0.04 (-6.97%)0.03 (-2.64%)0.01 (+44.42%)234.40 (+2.72%)195.80 (+1.42%)215.50 (+7.48%)141.90 (-5.65%)40.78 (+43.76%)
dc66da7 — 2026-03-25 19:23:450.05 (n/a)0.04 (n/a)0.04 (n/a)0.04 (n/a)0.01 (n/a)228.20 (n/a)193.06 (n/a)200.50 (n/a)150.40 (n/a)28.37 (n/a)

rms_norm_4_cols_1_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.06 (+10.73%)0.05 (-1.13%)0.05 (+5.17%)0.04 (-10.70%)0.01 (+73.44%)225.10 (+11.99%)181.84 (+3.00%)170.50 (-4.91%)140.00 (-9.68%)33.21 (+78.26%)
dc66da7 — 2026-03-25 19:23:450.05 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.00 (n/a)201.00 (n/a)176.54 (n/a)179.30 (n/a)155.00 (n/a)18.63 (n/a)

rms_norm_4_cols_2_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.06 (-0.51%)0.04 (-4.22%)0.04 (-6.40%)0.03 (+19.93%)0.01 (-12.26%)264.40 (-16.62%)208.96 (+1.64%)223.40 (+6.84%)140.70 (+0.50%)52.99 (-24.83%)
dc66da7 — 2026-03-25 19:23:450.06 (n/a)0.04 (n/a)0.04 (n/a)0.03 (n/a)0.01 (n/a)317.10 (n/a)205.58 (n/a)209.10 (n/a)140.00 (n/a)70.49 (n/a)

rms_norm_8_cols_1_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.06 (-8.19%)0.04 (-16.80%)0.03 (-25.89%)0.03 (-16.40%)0.01 (+5.02%)283.50 (+19.62%)220.72 (+22.74%)240.30 (+34.92%)128.00 (+8.94%)58.39 (+33.24%)
dc66da7 — 2026-03-25 19:23:450.07 (n/a)0.05 (n/a)0.05 (n/a)0.03 (n/a)0.01 (n/a)237.00 (n/a)179.82 (n/a)178.10 (n/a)117.50 (n/a)43.82 (n/a)

rms_norm_8_cols_2_channels_2048_tile_128

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.04 (+4.11%)0.04 (+2.57%)0.04 (-1.84%)0.03 (+8.94%)0.00 (+2.64%)250.80 (-8.20%)224.64 (-2.58%)233.20 (+1.88%)191.20 (-3.97%)25.83 (-9.64%)
dc66da7 — 2026-03-25 19:23:450.04 (n/a)0.04 (n/a)0.04 (n/a)0.03 (n/a)0.00 (n/a)273.20 (n/a)230.60 (n/a)228.90 (n/a)199.10 (n/a)28.59 (n/a)

rope_1_cols_2_channels_4096_tile_4096_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
1b2d230 — 2026-01-13 22:38:030.19 (-21.66%)0.17 (-8.53%)0.18 (+2.09%)0.14 (-6.71%)0.02 (-38.07%)174.10 (+7.20%)150.06 (+8.11%)139.60 (-2.04%)129.80 (+27.63%)20.36 (-10.65%)
d48746f — 2026-01-12 22:17:100.24 (n/a)0.18 (n/a)0.17 (n/a)0.15 (n/a)0.04 (n/a)162.40 (n/a)138.80 (n/a)142.50 (n/a)101.70 (n/a)22.79 (n/a)

rope_1c_32rows_512cols_32arows_0m

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.91 (+31.28%)0.72 (+30.97%)0.75 (+50.09%)0.52 (+23.14%)0.16 (+34.35%)189.50 (-18.81%)142.68 (-23.26%)131.80 (-33.40%)108.20 (-23.86%)33.21 (-13.83%)
dc66da7 — 2026-03-25 19:23:450.69 (n/a)0.55 (n/a)0.50 (n/a)0.42 (n/a)0.12 (n/a)233.40 (n/a)185.92 (n/a)197.90 (n/a)142.10 (n/a)38.54 (n/a)

rope_1c_32rows_512cols_8arows_0m

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.71 (+13.56%)0.53 (+16.33%)0.51 (+20.62%)0.43 (+13.26%)0.12 (+12.08%)172.60 (-11.71%)143.86 (-14.09%)145.70 (-17.07%)103.70 (-11.97%)28.08 (-12.19%)
dc66da7 — 2026-03-25 19:23:450.63 (n/a)0.46 (n/a)0.42 (n/a)0.38 (n/a)0.10 (n/a)195.50 (n/a)167.46 (n/a)175.70 (n/a)117.80 (n/a)31.98 (n/a)

rope_2_cols_2_channels_4096_tile_2048_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
1b2d230 — 2026-01-13 22:38:030.21 (-0.56%)0.17 (-10.62%)0.16 (-13.58%)0.13 (-13.33%)0.03 (+17.05%)183.70 (+15.39%)151.76 (+12.77%)155.20 (+15.73%)117.30 (+0.60%)24.24 (+35.73%)
d48746f — 2026-01-12 22:17:100.21 (n/a)0.19 (n/a)0.18 (n/a)0.15 (n/a)0.02 (n/a)159.20 (n/a)134.58 (n/a)134.10 (n/a)116.60 (n/a)17.86 (n/a)

rope_2c_32rows_512cols_32arows_0m

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.86 (+12.27%)0.68 (+4.35%)0.67 (-1.63%)0.56 (+6.41%)0.12 (+18.02%)174.80 (-6.02%)146.78 (-3.85%)146.60 (+1.66%)114.70 (-10.88%)23.97 (-1.25%)
dc66da7 — 2026-03-25 19:23:450.76 (n/a)0.66 (n/a)0.68 (n/a)0.53 (n/a)0.10 (n/a)186.00 (n/a)152.66 (n/a)144.20 (n/a)128.70 (n/a)24.27 (n/a)

rope_2c_32rows_512cols_8arows_0m

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.63 (+15.38%)0.42 (-5.03%)0.37 (-15.06%)0.26 (-23.68%)0.14 (+76.55%)289.00 (+31.07%)192.28 (+12.08%)197.90 (+17.73%)117.80 (-13.32%)64.38 (+98.12%)
dc66da7 — 2026-03-25 19:23:450.54 (n/a)0.44 (n/a)0.44 (n/a)0.33 (n/a)0.08 (n/a)220.50 (n/a)171.56 (n/a)168.10 (n/a)135.90 (n/a)32.49 (n/a)

rope_4_cols_2_channels_4096_tile_1024_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
1b2d230 — 2026-01-13 22:38:030.22 (+18.92%)0.17 (+6.24%)0.17 (+5.57%)0.14 (-0.50%)0.03 (+90.77%)171.10 (+0.47%)147.40 (-4.39%)148.10 (-5.31%)110.10 (-15.89%)23.59 (+60.67%)
d48746f — 2026-01-12 22:17:100.19 (n/a)0.16 (n/a)0.16 (n/a)0.14 (n/a)0.02 (n/a)170.30 (n/a)154.16 (n/a)156.40 (n/a)130.90 (n/a)14.69 (n/a)

rope_8_cols_2_channels_4096_tile_512_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
1b2d230 — 2026-01-13 22:38:030.17 (-1.30%)0.15 (+6.63%)0.15 (-8.78%)0.14 (+39.68%)0.01 (-67.35%)174.90 (-28.41%)163.42 (-10.36%)167.40 (+9.63%)146.50 (+1.31%)11.19 (-76.17%)
d48746f — 2026-01-12 22:17:100.17 (n/a)0.14 (n/a)0.16 (n/a)0.10 (n/a)0.03 (n/a)244.30 (n/a)182.30 (n/a)152.70 (n/a)144.60 (n/a)46.98 (n/a)

rope_8c_32rows_512cols_32arows_0m

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.60 (+0.58%)0.47 (-10.94%)0.47 (-14.67%)0.39 (-14.06%)0.09 (+57.47%)252.70 (+16.34%)213.46 (+14.31%)211.20 (+17.20%)163.30 (-0.61%)38.58 (+85.72%)
dc66da7 — 2026-03-25 19:23:450.60 (n/a)0.53 (n/a)0.55 (n/a)0.45 (n/a)0.06 (n/a)217.20 (n/a)186.74 (n/a)180.20 (n/a)164.30 (n/a)20.77 (n/a)

rope_8c_32rows_512cols_8arows_0m

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.58 (+33.33%)0.38 (-2.44%)0.32 (-20.46%)0.28 (-18.27%)0.13 (+203.34%)264.90 (+22.36%)209.28 (+9.84%)227.20 (+25.73%)126.60 (-25.00%)59.71 (+180.42%)
dc66da7 — 2026-03-25 19:23:450.44 (n/a)0.39 (n/a)0.41 (n/a)0.34 (n/a)0.04 (n/a)216.50 (n/a)190.54 (n/a)180.70 (n/a)168.80 (n/a)21.29 (n/a)

rows_32-cols_512-angle_rows_32-aie_columns_1-method_type_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8728b32 — 2026-03-26 21:24:130.38 (n/a)0.26 (n/a)0.33 (n/a)0.11 (n/a)0.12 (n/a)891.10 (n/a)467.46 (n/a)301.80 (n/a)261.60 (n/a)273.55 (n/a)

rows_32-cols_512-angle_rows_32-aie_columns_2-method_type_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8728b32 — 2026-03-26 21:24:130.40 (n/a)0.31 (n/a)0.35 (n/a)0.11 (n/a)0.11 (n/a)877.70 (n/a)398.42 (n/a)281.60 (n/a)244.20 (n/a)269.45 (n/a)

rows_32-cols_512-angle_rows_32-aie_columns_8-method_type_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8728b32 — 2026-03-26 21:24:130.40 (n/a)0.32 (n/a)0.34 (n/a)0.15 (n/a)0.10 (n/a)668.00 (n/a)354.14 (n/a)285.00 (n/a)246.50 (n/a)177.17 (n/a)

rows_32-cols_512-angle_rows_8-aie_columns_1-method_type_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8728b32 — 2026-03-26 21:24:130.26 (n/a)0.21 (n/a)0.25 (n/a)0.12 (n/a)0.06 (n/a)621.80 (n/a)381.12 (n/a)294.70 (n/a)282.20 (n/a)145.31 (n/a)

rows_32-cols_512-angle_rows_8-aie_columns_2-method_type_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8728b32 — 2026-03-26 21:24:130.27 (n/a)0.22 (n/a)0.25 (n/a)0.10 (n/a)0.07 (n/a)773.90 (n/a)385.04 (n/a)293.30 (n/a)270.00 (n/a)217.64 (n/a)

rows_32-cols_512-angle_rows_8-aie_columns_8-method_type_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8728b32 — 2026-03-26 21:24:130.29 (n/a)0.23 (n/a)0.25 (n/a)0.12 (n/a)0.06 (n/a)591.00 (n/a)345.98 (n/a)296.00 (n/a)256.60 (n/a)138.69 (n/a)

seq_len_16384-dim_64-num_heads_1-num_pipelines_8-num_kv_heads_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8728b32 — 2026-03-26 21:24:130.20 (n/a)0.20 (n/a)0.20 (n/a)0.20 (n/a)0.00 (n/a)41402.60 (n/a)41367.72 (n/a)41376.40 (n/a)41324.10 (n/a)35.92 (n/a)

seq_len_256-embedding_dim_2048-hidden_dim_2048-prio_accuracy_False

No metrics available.

sigmoid_1_cols_1_channels_2048_tile_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.05 (-14.16%)0.05 (-6.29%)0.05 (-2.65%)0.04 (-7.71%)0.01 (-21.79%)225.80 (+8.35%)180.84 (+6.33%)170.90 (+2.70%)163.30 (+16.48%)25.41 (+0.55%)
dc66da7 — 2026-03-25 19:23:450.06 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.01 (n/a)208.40 (n/a)170.08 (n/a)166.40 (n/a)140.20 (n/a)25.28 (n/a)

sigmoid_2_cols_1_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.06 (+4.68%)0.05 (+2.94%)0.04 (-13.06%)0.04 (+15.35%)0.01 (+8.49%)232.90 (-13.32%)179.44 (-3.16%)197.20 (+15.05%)127.30 (-4.50%)46.58 (-13.93%)
dc66da7 — 2026-03-25 19:23:450.06 (n/a)0.05 (n/a)0.05 (n/a)0.03 (n/a)0.01 (n/a)268.70 (n/a)185.30 (n/a)171.40 (n/a)133.30 (n/a)54.12 (n/a)

sigmoid_4_cols_1_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.07 (+10.37%)0.05 (-2.49%)0.04 (-7.15%)0.04 (-7.47%)0.01 (+64.06%)210.80 (+8.10%)175.44 (+5.70%)183.30 (+7.70%)117.30 (-9.42%)39.46 (+67.78%)
dc66da7 — 2026-03-25 19:23:450.06 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.01 (n/a)195.00 (n/a)165.98 (n/a)170.20 (n/a)129.50 (n/a)23.52 (n/a)

sigmoid_8_cols_1_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.06 (+29.56%)0.05 (+22.81%)0.05 (+37.30%)0.03 (-8.45%)0.01 (+218.31%)242.70 (+9.23%)173.04 (-14.71%)150.20 (-27.16%)134.70 (-22.81%)46.84 (+169.71%)
dc66da7 — 2026-03-25 19:23:450.05 (n/a)0.04 (n/a)0.04 (n/a)0.04 (n/a)0.00 (n/a)222.20 (n/a)202.88 (n/a)206.20 (n/a)174.50 (n/a)17.37 (n/a)

silu_1_cols_1_channels_2048_tile_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.07 (+9.38%)0.05 (-4.51%)0.04 (-23.77%)0.04 (+15.02%)0.01 (+5.83%)207.90 (-13.05%)169.28 (+4.04%)188.60 (+31.15%)114.40 (-8.63%)38.46 (-17.21%)
dc66da7 — 2026-03-25 19:23:450.07 (n/a)0.05 (n/a)0.06 (n/a)0.03 (n/a)0.01 (n/a)239.10 (n/a)162.70 (n/a)143.80 (n/a)125.20 (n/a)46.46 (n/a)

silu_2_cols_1_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.06 (-4.63%)0.05 (-13.54%)0.05 (-12.20%)0.04 (-16.50%)0.01 (+19.45%)213.90 (+19.76%)180.30 (+16.99%)181.70 (+13.92%)134.00 (+4.85%)30.84 (+49.10%)
dc66da7 — 2026-03-25 19:23:450.06 (n/a)0.05 (n/a)0.05 (n/a)0.05 (n/a)0.01 (n/a)178.60 (n/a)154.12 (n/a)159.50 (n/a)127.80 (n/a)20.68 (n/a)

silu_4_cols_1_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.06 (+6.26%)0.05 (-8.81%)0.04 (-14.86%)0.04 (-17.09%)0.01 (+81.20%)214.80 (+20.61%)176.86 (+11.68%)182.90 (+17.47%)133.30 (-5.86%)30.25 (+101.37%)
dc66da7 — 2026-03-25 19:23:450.06 (n/a)0.05 (n/a)0.05 (n/a)0.05 (n/a)0.00 (n/a)178.10 (n/a)158.36 (n/a)155.70 (n/a)141.60 (n/a)15.02 (n/a)

silu_8_cols_1_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.06 (-6.11%)0.05 (-10.71%)0.04 (-10.38%)0.04 (-11.78%)0.01 (+1.86%)224.60 (+13.38%)185.30 (+12.62%)186.00 (+11.58%)140.40 (+6.53%)33.11 (+23.38%)
dc66da7 — 2026-03-25 19:23:450.06 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.01 (n/a)198.10 (n/a)164.54 (n/a)166.70 (n/a)131.80 (n/a)26.84 (n/a)

softmax_1_cols_2_channels_4096_tile_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.19 (-8.94%)0.17 (-8.66%)0.17 (-19.83%)0.14 (+4.32%)0.02 (-35.13%)115.80 (-4.14%)99.26 (+7.47%)98.60 (+24.65%)85.00 (+9.82%)13.65 (-30.86%)
84d3478 — 2026-02-17 23:25:130.21 (n/a)0.18 (n/a)0.21 (n/a)0.14 (n/a)0.04 (n/a)120.80 (n/a)92.36 (n/a)79.10 (n/a)77.40 (n/a)19.75 (n/a)

softmax_2_cols_2_channels_32768_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.64 (-42.09%)0.60 (-31.83%)0.59 (-30.19%)0.54 (-6.68%)0.05 (-76.96%)242.70 (+7.15%)221.12 (+40.50%)220.30 (+43.24%)203.40 (+72.67%)16.99 (-59.06%)
dc66da7 — 2026-03-25 19:23:451.11 (n/a)0.87 (n/a)0.85 (n/a)0.58 (n/a)0.20 (n/a)226.50 (n/a)157.38 (n/a)153.80 (n/a)117.80 (n/a)41.50 (n/a)

softmax_2_cols_2_channels_32768_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.68 (-34.93%)0.62 (-20.25%)0.60 (-17.49%)0.55 (-10.32%)0.05 (-71.29%)240.30 (+11.51%)214.32 (+20.87%)218.20 (+21.22%)193.80 (+53.69%)18.64 (-52.51%)
dc66da7 — 2026-03-25 19:23:451.04 (n/a)0.77 (n/a)0.73 (n/a)0.61 (n/a)0.18 (n/a)215.50 (n/a)177.32 (n/a)180.00 (n/a)126.10 (n/a)39.26 (n/a)

softmax_2_cols_2_channels_4096_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.20 (-0.87%)0.16 (+6.57%)0.17 (-9.55%)0.11 (+29.81%)0.03 (-44.34%)145.70 (-22.99%)104.68 (-15.00%)95.00 (+10.59%)83.80 (+0.84%)24.16 (-54.23%)
84d3478 — 2026-02-17 23:25:130.20 (n/a)0.15 (n/a)0.19 (n/a)0.09 (n/a)0.06 (n/a)189.20 (n/a)123.16 (n/a)85.90 (n/a)83.10 (n/a)52.80 (n/a)

softmax_2_cols_2_channels_4096_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.25 (+23.53%)0.18 (+9.42%)0.18 (-3.86%)0.11 (+2.99%)0.05 (+19.24%)144.10 (-2.90%)98.98 (-8.18%)92.10 (+4.07%)64.30 (-19.02%)29.58 (-7.13%)
84d3478 — 2026-02-17 23:25:130.21 (n/a)0.16 (n/a)0.19 (n/a)0.11 (n/a)0.04 (n/a)148.40 (n/a)107.80 (n/a)88.50 (n/a)79.40 (n/a)31.85 (n/a)

softmax_4_cols_4_channels_32768_tile_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.66 (+11.72%)0.50 (-0.51%)0.54 (-4.72%)0.35 (+6.44%)0.13 (+14.83%)377.70 (-6.04%)276.32 (+1.24%)242.20 (+4.94%)199.30 (-10.51%)76.50 (-0.31%)
dc66da7 — 2026-03-25 19:23:450.59 (n/a)0.51 (n/a)0.57 (n/a)0.33 (n/a)0.11 (n/a)402.00 (n/a)272.94 (n/a)230.80 (n/a)222.70 (n/a)76.73 (n/a)

swiglu

No metrics available.

swiglu_decode_1x2048x2048

No metrics available.

swiglu_decode_1x2048x2048_0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:250.00 (-0.21%)0.00 (+7.74%)0.00 (+0.37%)0.00 (+8.84%)0.00 (-19.58%)5170.80 (-8.12%)3761.80 (-8.70%)3408.60 (-0.36%)3407.30 (+0.21%)787.66 (-22.98%)
84d3478 — 2026-02-17 23:25:130.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)5627.70 (n/a)4120.44 (n/a)3421.00 (n/a)3400.30 (n/a)1022.61 (n/a)

swiglu_prefill_256x2048x2048

No metrics available.

tanh_1_cols_1_channels_2048_tile_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.06 (-7.39%)0.04 (-16.97%)0.04 (-14.92%)0.03 (-28.80%)0.01 (+23.25%)296.40 (+40.47%)201.28 (+24.69%)188.60 (+17.51%)141.90 (+7.99%)59.45 (+90.30%)
dc66da7 — 2026-03-25 19:23:450.06 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.01 (n/a)211.00 (n/a)161.42 (n/a)160.50 (n/a)131.40 (n/a)31.24 (n/a)

tanh_2_cols_1_channels_2048_tile_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.06 (-14.29%)0.05 (-6.63%)0.05 (+2.99%)0.03 (-30.73%)0.01 (+5.83%)293.30 (+44.41%)176.82 (+11.67%)159.70 (-2.92%)128.50 (+16.71%)67.26 (+86.47%)
dc66da7 — 2026-03-25 19:23:450.07 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.01 (n/a)203.10 (n/a)158.34 (n/a)164.50 (n/a)110.10 (n/a)36.07 (n/a)

tanh_4_cols_1_channels_2048_tile_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.06 (-27.96%)0.05 (-4.84%)0.05 (+4.02%)0.04 (+16.65%)0.01 (-63.57%)187.60 (-14.30%)167.32 (-1.72%)171.90 (-3.86%)133.00 (+38.83%)20.69 (-55.09%)
dc66da7 — 2026-03-25 19:23:450.09 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.02 (n/a)218.90 (n/a)170.24 (n/a)178.80 (n/a)95.80 (n/a)46.06 (n/a)

tanh_8_cols_1_channels_2048_tile_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.05 (-24.07%)0.04 (-15.48%)0.04 (-23.02%)0.03 (-3.09%)0.01 (-53.59%)262.30 (+3.19%)198.18 (+12.63%)183.30 (+29.91%)173.30 (+31.69%)36.55 (-34.15%)
dc66da7 — 2026-03-25 19:23:450.06 (n/a)0.05 (n/a)0.06 (n/a)0.03 (n/a)0.01 (n/a)254.20 (n/a)175.96 (n/a)141.10 (n/a)131.60 (n/a)55.51 (n/a)

transpose_2048_M_64_N_1_cols_1_channels_64_m_64_n_8_s

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:392.51 (-25.94%)2.10 (-20.30%)2.01 (-23.11%)1.76 (-13.49%)0.29 (-41.75%)298.70 (+15.60%)253.30 (+23.92%)261.10 (+30.09%)208.60 (+35.02%)33.97 (-9.04%)
dc66da7 — 2026-03-25 19:23:453.39 (n/a)2.64 (n/a)2.61 (n/a)2.03 (n/a)0.49 (n/a)258.40 (n/a)204.40 (n/a)200.70 (n/a)154.50 (n/a)37.34 (n/a)

transpose_2048_M_64_N_1_cols_1_channels_64_m_64_n_8_s0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:254.43 (+0.85%)3.73 (+3.83%)4.02 (-1.30%)2.97 (+16.77%)0.65 (-24.99%)176.70 (-14.35%)144.46 (-6.24%)130.40 (+1.32%)118.40 (-0.84%)26.60 (-35.16%)
84d3478 — 2026-02-17 23:25:134.39 (n/a)3.59 (n/a)4.07 (n/a)2.54 (n/a)0.87 (n/a)206.30 (n/a)154.08 (n/a)128.70 (n/a)119.40 (n/a)41.02 (n/a)

transpose_2048_M_64_N_1_cols_2_channels_64_m_64_n_8_s

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:392.72 (-10.04%)2.30 (-13.01%)2.26 (-13.66%)1.88 (-21.25%)0.34 (+46.90%)278.20 (+26.97%)231.60 (+16.34%)232.20 (+15.81%)192.40 (+11.15%)34.64 (+109.20%)
dc66da7 — 2026-03-25 19:23:453.03 (n/a)2.65 (n/a)2.61 (n/a)2.39 (n/a)0.23 (n/a)219.10 (n/a)199.08 (n/a)200.50 (n/a)173.10 (n/a)16.56 (n/a)

transpose_2048_M_64_N_1_cols_2_channels_64_m_64_n_8_s0

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
897d04e — 2026-03-06 22:12:255.27 (-14.18%)4.35 (-2.49%)4.45 (+2.15%)2.78 (-22.18%)0.98 (-3.38%)188.60 (+28.47%)126.98 (+4.20%)117.90 (-2.08%)99.40 (+16.53%)36.02 (+50.15%)
84d3478 — 2026-02-17 23:25:136.14 (n/a)4.46 (n/a)4.35 (n/a)3.57 (n/a)1.02 (n/a)146.80 (n/a)121.86 (n/a)120.40 (n/a)85.30 (n/a)23.99 (n/a)

weighted_rms_norm_1_cols_2_channels_2048_weights_2048

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.09 (+6.38%)0.07 (-4.87%)0.07 (+1.68%)0.05 (-23.45%)0.01 (+101.20%)230.80 (+30.62%)177.68 (+7.40%)170.00 (-1.68%)138.30 (-5.98%)33.65 (+149.29%)
dc66da7 — 2026-03-25 19:23:450.08 (n/a)0.07 (n/a)0.07 (n/a)0.07 (n/a)0.01 (n/a)176.70 (n/a)165.44 (n/a)172.90 (n/a)147.10 (n/a)13.50 (n/a)

weighted_rms_norm_2_cols_2_channels_2048_weights_1024

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.07 (-25.98%)0.06 (-18.74%)0.06 (-7.04%)0.04 (-20.93%)0.01 (-34.18%)237.00 (+26.47%)189.06 (+22.07%)179.50 (+7.55%)151.40 (+35.06%)34.27 (+14.91%)
dc66da7 — 2026-03-25 19:23:450.09 (n/a)0.07 (n/a)0.06 (n/a)0.05 (n/a)0.01 (n/a)187.40 (n/a)154.88 (n/a)166.90 (n/a)112.10 (n/a)29.83 (n/a)

weighted_rms_norm_4_cols_2_channels_2048_weights_512

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.06 (-13.39%)0.05 (-21.43%)0.04 (-31.53%)0.04 (-12.12%)0.01 (-0.80%)237.90 (+13.77%)205.32 (+28.15%)225.30 (+46.01%)147.80 (+15.47%)38.15 (+27.36%)
dc66da7 — 2026-03-25 19:23:450.07 (n/a)0.06 (n/a)0.06 (n/a)0.04 (n/a)0.01 (n/a)209.10 (n/a)160.22 (n/a)154.30 (n/a)128.00 (n/a)29.95 (n/a)

weighted_rms_norm_8_cols_2_channels_2048_weights_256

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
8b6d0f4 — 2026-03-25 20:46:390.07 (+33.02%)0.04 (-11.48%)0.04 (-24.93%)0.02 (-32.99%)0.02 (+124.58%)371.50 (+49.26%)251.20 (+24.91%)247.70 (+33.17%)124.90 (-24.85%)88.76 (+136.07%)
dc66da7 — 2026-03-25 19:23:450.05 (n/a)0.04 (n/a)0.05 (n/a)0.03 (n/a)0.01 (n/a)248.90 (n/a)201.10 (n/a)186.00 (n/a)166.20 (n/a)37.60 (n/a)

@github-actions
Copy link
Contributor

📊 Test Results for Test Example Applications

8728b32 (2026_03_26_21_36_45)

IRONCLAD

Tested on 2026_03_26_21_36_45 at commit 8728b32.

Test Checks TTFT (mean)TPS (mean)
llama_3.2_1b_prompt_1024_tokens_1 ❌ 0/5 n/a n/a
llama_3.2_1b_prompt_1024_tokens_40 ❌ 0/5 n/a n/a
llama_3.2_1b_prompt_13_tokens_1 ❌ 0/5 n/a n/a
llama_3.2_1b_prompt_13_tokens_40 ❌ 0/5 n/a n/a
📈 Trends (vs main branch) for Test Example Applications

8728b32 (2026_03_26_21_36_45)

IRONCLAD Trends

llama_3.2_1b

Commit/Date Num Tokens (max)Num Tokens (mean)Num Tokens (median)Num Tokens (min)Num Tokens (stddev)TPS (max)TPS (mean)TPS (median)TPS (min)TPS (stddev)TTFT (max)TTFT (mean)TTFT (median)TTFT (min)TTFT (stddev)Total (max)Total (mean)Total (median)Total (min)Total (stddev)
130b6ea — 2025-12-05 21:33:1240.00 (+0.00%)40.00 (+0.00%)40.00 (+0.00%)40.00 (+0.00%)0.00 (n/a)4.71 (-0.42%)4.64 (-0.09%)4.64 (+0.65%)4.55 (-0.22%)0.05 (-17.66%)4.41 (-0.34%)4.39 (-0.19%)4.38 (-0.33%)4.37 (-0.15%)0.01 (-25.90%)12.96 (-0.00%)12.80 (+0.07%)12.80 (-0.23%)12.67 (+0.44%)0.09 (-21.12%)
0a6c11c — 2025-12-03 23:35:1540.00 (n/a)40.00 (n/a)40.00 (n/a)40.00 (n/a)0.00 (n/a)4.73 (n/a)4.64 (n/a)4.61 (n/a)4.56 (n/a)0.06 (n/a)4.42 (n/a)4.40 (n/a)4.40 (n/a)4.37 (n/a)0.02 (n/a)12.96 (n/a)12.79 (n/a)12.83 (n/a)12.62 (n/a)0.12 (n/a)

llama_3.2_1b_prompt_1024_tokens_1

Commit/Date TTFT (max)TTFT (mean)TTFT (median)TTFT (min)TTFT (stddev)
8728b32 — 2026-03-26 21:34:03n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)
8b6d0f4 — 2026-03-25 21:36:352.26 (n/a)2.25 (n/a)2.25 (n/a)2.23 (n/a)0.01 (n/a)

llama_3.2_1b_prompt_1024_tokens_40

Commit/Date TPS (max)TPS (mean)TPS (median)TPS (min)TPS (stddev)TTFT (max)TTFT (mean)TTFT (median)TTFT (min)TTFT (stddev)
8728b32 — 2026-03-26 21:34:03n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)
8b6d0f4 — 2026-03-25 21:36:354.30 (n/a)4.28 (n/a)4.28 (n/a)4.26 (n/a)0.02 (n/a)2.37 (n/a)2.27 (n/a)2.25 (n/a)2.23 (n/a)0.06 (n/a)

llama_3.2_1b_prompt_13_tokens_1

Commit/Date TTFT (max)TTFT (mean)TTFT (median)TTFT (min)TTFT (stddev)
8728b32 — 2026-03-26 21:34:03n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)
8b6d0f4 — 2026-03-25 21:36:352.19 (n/a)2.17 (n/a)2.17 (n/a)2.14 (n/a)0.02 (n/a)

llama_3.2_1b_prompt_13_tokens_40

Commit/Date TPS (max)TPS (mean)TPS (median)TPS (min)TPS (stddev)TTFT (max)TTFT (mean)TTFT (median)TTFT (min)TTFT (stddev)
8728b32 — 2026-03-26 21:34:03n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)n/a (n/a)
8b6d0f4 — 2026-03-25 21:36:354.30 (n/a)4.27 (n/a)4.27 (n/a)4.25 (n/a)0.02 (n/a)2.19 (n/a)2.17 (n/a)2.16 (n/a)2.15 (n/a)0.01 (n/a)

llama_3.2_1b_prompt_2048_tokens_1

Commit/Date Num_Tokens (max)Num_Tokens (mean)Num_Tokens (median)Num_Tokens (min)Num_Tokens (stddev)TPS (max)TPS (mean)TPS (median)TPS (min)TPS (stddev)TTFT (max)TTFT (mean)TTFT (median)TTFT (min)TTFT (stddev)
897d04e — 2026-03-06 22:56:071.00 (+0.00%)1.00 (+0.00%)1.00 (+0.00%)1.00 (+0.00%)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)2.68 (-1.06%)2.68 (-1.06%)2.68 (-1.06%)2.68 (-1.06%)0.00 (n/a)
84d3478 — 2026-02-17 23:16:231.00 (n/a)1.00 (n/a)1.00 (n/a)1.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)2.70 (n/a)2.70 (n/a)2.70 (n/a)2.70 (n/a)0.00 (n/a)

llama_3.2_1b_prompt_2048_tokens_40

Commit/Date Num_Tokens (max)Num_Tokens (mean)Num_Tokens (median)Num_Tokens (min)Num_Tokens (stddev)TPS (max)TPS (mean)TPS (median)TPS (min)TPS (stddev)TTFT (max)TTFT (mean)TTFT (median)TTFT (min)TTFT (stddev)
897d04e — 2026-03-06 22:56:0740.00 (+0.00%)40.00 (+0.00%)40.00 (+0.00%)40.00 (+0.00%)0.00 (n/a)4.00 (-1.72%)4.00 (-1.72%)4.00 (-1.72%)4.00 (-1.72%)0.00 (n/a)2.70 (-0.44%)2.70 (-0.44%)2.70 (-0.44%)2.70 (-0.44%)0.00 (n/a)
84d3478 — 2026-02-17 23:16:2340.00 (n/a)40.00 (n/a)40.00 (n/a)40.00 (n/a)0.00 (n/a)4.07 (n/a)4.07 (n/a)4.07 (n/a)4.07 (n/a)0.00 (n/a)2.71 (n/a)2.71 (n/a)2.71 (n/a)2.71 (n/a)0.00 (n/a)

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants