Skip to content

Update MLIR-AIE to pick up ObjectFIFO fix and re-enable MHA#133

Open
andrej wants to merge 1 commit into
amd:develfrom
andrej:reenable-mha
Open

Update MLIR-AIE to pick up ObjectFIFO fix and re-enable MHA#133
andrej wants to merge 1 commit into
amd:develfrom
andrej:reenable-mha

Conversation

@andrej

@andrej andrej commented Jul 9, 2026

Copy link
Copy Markdown
Collaborator

Pick up Xilinx/mlir-aie#3295 and re-enable MHA test that relies on this fix

closes #130

@andrej andrej requested review from hunhoffe and jgmelber as code owners July 9, 2026 22:46
@github-actions

github-actions Bot commented Jul 9, 2026

Copy link
Copy Markdown
Contributor

CI Test Results

1f35831 (2026_07_09_23_41_45)

IRON - CI Summary

Examples

iron/applications/llama_3.2_1b
Test Krackan Status Krackan Phoenix Status Phoenix
test_llama_3_2_1b[llama_3.2_1b_prompt_1024_tokens_1] - - -
test_llama_3_2_1b[llama_3.2_1b_prompt_1024_tokens_40] - - -
test_llama_3_2_1b[llama_3.2_1b_prompt_13_tokens_1] - - -
test_llama_3_2_1b[llama_3.2_1b_prompt_13_tokens_40] - - -

Small

iron/operators/axpy
Test Krackan Status Krackan Latency (mean) Phoenix Status Phoenix Latency (mean)
test_axpy[input_length_2048-num_aie_columns_1-tile_size_2048-scalar_factor_3.0] 158.82 387.68
test_axpy[input_length_2048-num_aie_columns_2-tile_size_1024-scalar_factor_3.0] 182.94 334.78
test_axpy[input_length_2048-num_aie_columns_4-tile_size_512-scalar_factor_3.0] 224.52 379.66
test_axpy[input_length_2048-num_aie_columns_8-tile_size_256-scalar_factor_3.0] 179.86 - -
iron/operators/dequant
Test Krackan Status Krackan Latency (mean) Phoenix Status Phoenix Latency (mean)
test_dequant[input_length_2048-num_aie_columns_1-num_channels_1-tile_size_2048-group_size_32] 175.18 386.68
test_dequant[input_length_2048-num_aie_columns_1-num_channels_2-tile_size_1024-group_size_32] 178.10 434.06
test_dequant[input_length_2048-num_aie_columns_2-num_channels_1-tile_size_1024-group_size_32] 153.58 623.08
test_dequant[input_length_2048-num_aie_columns_2-num_channels_2-tile_size_512-group_size_32] 175.48 335.58
test_dequant[input_length_2048-num_aie_columns_4-num_channels_1-tile_size_512-group_size_32] 166.24 461.32
test_dequant[input_length_2048-num_aie_columns_4-num_channels_2-tile_size_256-group_size_32] 172.16 437.74
test_dequant[input_length_2048-num_aie_columns_8-num_channels_1-tile_size_256-group_size_32] 172.18 - -
test_dequant[input_length_2048-num_aie_columns_8-num_channels_2-tile_size_128-group_size_32] 202.74 - -
iron/operators/elementwise_add
Test Krackan Status Krackan Latency (mean) Phoenix Status Phoenix Latency (mean)
test_elementwise_add[input_length_2048-num_aie_columns_1-tile_size_2048] 176.48 475.12
test_elementwise_add[input_length_2048-num_aie_columns_2-tile_size_1024] 186.72 454.70
test_elementwise_add[input_length_2048-num_aie_columns_4-tile_size_512] 206.48 471.90
test_elementwise_add[input_length_2048-num_aie_columns_8-tile_size_256] 187.02 - -
iron/operators/elementwise_mul
Test Krackan Status Krackan Latency (mean) Phoenix Status Phoenix Latency (mean)
test_elementwise_mul[input_length_2048-num_aie_columns_1-tile_size_2048] 189.02 302.90
test_elementwise_mul[input_length_2048-num_aie_columns_2-tile_size_1024] 188.64 688.78
test_elementwise_mul[input_length_2048-num_aie_columns_4-tile_size_512] 196.94 311.58
test_elementwise_mul[input_length_2048-num_aie_columns_8-tile_size_256] 190.38 - -
iron/operators/gelu
Test Krackan Status Krackan Latency (mean) Phoenix Status Phoenix Latency (mean)
test_gelu[input_length_2048-num_aie_columns_1-num_channels_1-tile_size_2048] 188.12 428.54
test_gelu[input_length_2048-num_aie_columns_1-num_channels_2-tile_size_1024] 176.92 334.84
test_gelu[input_length_2048-num_aie_columns_2-num_channels_1-tile_size_1024] 201.62 444.66
test_gelu[input_length_2048-num_aie_columns_2-num_channels_2-tile_size_512] 197.48 435.00
test_gelu[input_length_2048-num_aie_columns_4-num_channels_1-tile_size_512] 220.20 430.52
test_gelu[input_length_2048-num_aie_columns_4-num_channels_2-tile_size_256] 216.66 422.12
test_gelu[input_length_2048-num_aie_columns_8-num_channels_1-tile_size_256] 209.64 - -
test_gelu[input_length_2048-num_aie_columns_8-num_channels_2-tile_size_128] 224.22 - -
iron/operators/gemm
Test Krackan Status Krackan Latency (mean) Phoenix Status Phoenix Latency (mean)
test_gemm[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-partition_N_1] 2217.18 - -
test_gemm[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-partition_N_1] 231.10 467.74
test_gemm[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-partition_N_1] 262.46 482.34
test_gemm[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-partition_N_1] 48693.58 83029.96
test_gemm[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-partition_N_1] 28545.82 24836.62
test_gemm[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-partition_N_1] 7907.18 - -
test_gemm[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-partition_N_1] 2297.50 4901.04
test_gemm[M_64-K_512-N_256-num_aie_columns_4-b_col_maj_True-c_col_maj_False-m_16-k_64-n_64-trace_size_0-partition_N_4] 3888.50 5894.82
test_gemm[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-partition_N_1] 1444.06 - -
iron/operators/gemv
Test Krackan Status Krackan Latency (mean) Phoenix Status Phoenix Latency (mean)
test_gemv[M_128-K_128-num_aie_columns_1-tile_size_input_32-tile_size_output_128] 0.22 0.09
test_gemv[M_2048-K_8192-num_aie_columns_1-tile_size_input_1-tile_size_output_2048] 12.65 3.64
test_gemv[M_2048-K_8192-num_aie_columns_2-tile_size_input_1-tile_size_output_1024] 25.09 6.42
test_gemv[M_2048-K_8192-num_aie_columns_4-tile_size_input_1-tile_size_output_512] 41.63 10.68
test_gemv[M_2048-K_8192-num_aie_columns_8-tile_size_input_1-tile_size_output_256] 44.23 - -
test_gemv[M_8192-K_2048-num_aie_columns_1-tile_size_input_4-tile_size_output_1024] 12.78 3.66
test_gemv[M_8192-K_2048-num_aie_columns_2-tile_size_input_4-tile_size_output_1024] 24.68 5.93
test_gemv[M_8192-K_2048-num_aie_columns_4-tile_size_input_4-tile_size_output_1024] 41.58 9.62
test_gemv[M_8192-K_2048-num_aie_columns_8-tile_size_input_4-tile_size_output_1024] 43.97 - -
iron/operators/layer_norm
Test Krackan Status Krackan Latency (mean) Phoenix Status Phoenix Latency (mean)
test_layer_norm[input_length_2048-num_aie_columns_1-num_channels_1-tile_size_2048] 207.06 344.00
test_layer_norm[input_length_2048-num_aie_columns_1-num_channels_2-tile_size_1024] 170.74 348.34
test_layer_norm[input_length_2048-num_aie_columns_2-num_channels_1-tile_size_1024] 180.30 400.14
test_layer_norm[input_length_2048-num_aie_columns_2-num_channels_2-tile_size_512] 180.26 511.48
test_layer_norm[input_length_2048-num_aie_columns_4-num_channels_1-tile_size_512] 188.18 418.30
test_layer_norm[input_length_2048-num_aie_columns_4-num_channels_2-tile_size_256] 170.64 443.56
test_layer_norm[input_length_2048-num_aie_columns_8-num_channels_1-tile_size_256] 182.92 - -
test_layer_norm[input_length_2048-num_aie_columns_8-num_channels_2-tile_size_128] 219.06 - -
iron/operators/leaky_relu
Test Krackan Status Krackan Latency (mean) Phoenix Status Phoenix Latency (mean)
test_leaky_relu[input_length_2048-num_aie_columns_1-num_channels_1-tile_size_2048-alpha_0.01] 170.92 352.84
test_leaky_relu[input_length_2048-num_aie_columns_1-num_channels_1-tile_size_2048-alpha_0.1] 175.68 392.10
test_leaky_relu[input_length_2048-num_aie_columns_1-num_channels_1-tile_size_2048-alpha_0.25] 161.60 356.94
test_leaky_relu[input_length_2048-num_aie_columns_1-num_channels_2-tile_size_1024-alpha_0.01] 181.40 697.92
test_leaky_relu[input_length_2048-num_aie_columns_2-num_channels_1-tile_size_1024-alpha_0.01] 221.80 318.50
test_leaky_relu[input_length_2048-num_aie_columns_2-num_channels_2-tile_size_512-alpha_0.01] 184.60 369.16
test_leaky_relu[input_length_2048-num_aie_columns_4-num_channels_1-tile_size_512-alpha_0.01] 208.14 428.44
test_leaky_relu[input_length_2048-num_aie_columns_4-num_channels_2-tile_size_256-alpha_0.01] 213.88 475.20
test_leaky_relu[input_length_2048-num_aie_columns_8-num_channels_1-tile_size_256-alpha_0.01] 185.38 - -
test_leaky_relu[input_length_2048-num_aie_columns_8-num_channels_2-tile_size_128-alpha_0.01] 203.08 - -
iron/operators/mem_copy
Test Krackan Status Krackan Latency (mean) Phoenix Status Phoenix Latency (mean)
test_mem_copy[input_length_2048-num_cores_1-num_channels_1-bypass_False-tile_size_2048] 139.62 369.82
test_mem_copy[input_length_2048-num_cores_16-num_channels_2-bypass_False-tile_size_128] 255.54 - -
test_mem_copy[input_length_2048-num_cores_2-num_channels_1-bypass_False-tile_size_1024] 168.18 394.42
test_mem_copy[input_length_2048-num_cores_2-num_channels_2-bypass_False-tile_size_1024] 178.52 365.72
test_mem_copy[input_length_2048-num_cores_4-num_channels_1-bypass_False-tile_size_512] 192.90 541.88
test_mem_copy[input_length_2048-num_cores_4-num_channels_2-bypass_False-tile_size_512] 161.02 478.32
test_mem_copy[input_length_2048-num_cores_8-num_channels_1-bypass_False-tile_size_256] 201.30 - -
test_mem_copy[input_length_2048-num_cores_8-num_channels_2-bypass_False-tile_size_256] 197.18 489.60
iron/operators/mha
Test Krackan Status Krackan Latency (mean) Phoenix Status Phoenix Latency (mean)
test_mha[seq_len_16384-dim_64-num_heads_1-num_pipelines_8-num_kv_heads_0] 40865.30 - -
iron/operators/relu
Test Krackan Status Krackan Latency (mean) Phoenix Status Phoenix Latency (mean)
test_relu[input_length_2048-num_aie_columns_1-num_channels_1-tile_size_2048] 178.68 421.40
test_relu[input_length_2048-num_aie_columns_1-num_channels_2-tile_size_1024] 193.20 347.36
test_relu[input_length_2048-num_aie_columns_2-num_channels_1-tile_size_1024] 185.40 355.38
test_relu[input_length_2048-num_aie_columns_2-num_channels_2-tile_size_512] 207.20 333.38
test_relu[input_length_2048-num_aie_columns_4-num_channels_1-tile_size_512] 187.78 878.70
test_relu[input_length_2048-num_aie_columns_4-num_channels_2-tile_size_256] 175.70 546.12
test_relu[input_length_2048-num_aie_columns_8-num_channels_1-tile_size_256] 194.68 - -
test_relu[input_length_2048-num_aie_columns_8-num_channels_2-tile_size_128] 246.80 - -
iron/operators/rms_norm
Test Krackan Status Krackan Latency (mean) Phoenix Status Phoenix Latency (mean)
test_rms_norm[input_length_2048-num_aie_columns_1-num_channels_1-tile_size_2048-weighted_False] 160.60 358.74
test_rms_norm[input_length_2048-num_aie_columns_1-num_channels_1-tile_size_2048-weighted_True] 183.34 680.14
test_rms_norm[input_length_2048-num_aie_columns_1-num_channels_2-tile_size_1024-weighted_False] 195.38 441.22
test_rms_norm[input_length_2048-num_aie_columns_1-num_channels_2-tile_size_1024-weighted_True] 177.38 447.10
test_rms_norm[input_length_2048-num_aie_columns_2-num_channels_1-tile_size_1024-weighted_False] 189.20 661.16
test_rms_norm[input_length_2048-num_aie_columns_2-num_channels_1-tile_size_1024-weighted_True] 196.62 363.98
test_rms_norm[input_length_2048-num_aie_columns_2-num_channels_2-tile_size_512-weighted_False] 198.48 394.98
test_rms_norm[input_length_2048-num_aie_columns_2-num_channels_2-tile_size_512-weighted_True] 193.80 429.32
test_rms_norm[input_length_2048-num_aie_columns_4-num_channels_1-tile_size_512-weighted_False] 191.68 362.84
test_rms_norm[input_length_2048-num_aie_columns_4-num_channels_1-tile_size_512-weighted_True] 168.26 726.34
test_rms_norm[input_length_2048-num_aie_columns_4-num_channels_2-tile_size_256-weighted_False] 195.00 440.06
test_rms_norm[input_length_2048-num_aie_columns_4-num_channels_2-tile_size_256-weighted_True] 192.20 - -
test_rms_norm[input_length_2048-num_aie_columns_8-num_channels_1-tile_size_256-weighted_False] 204.02 - -
test_rms_norm[input_length_2048-num_aie_columns_8-num_channels_1-tile_size_256-weighted_True] 187.72 - -
test_rms_norm[input_length_2048-num_aie_columns_8-num_channels_2-tile_size_128-weighted_False] 226.92 - -
iron/operators/rope
Test Krackan Status Krackan Latency (mean) Phoenix Status Phoenix Latency (mean)
test_rope[rows_32-cols_512-angle_rows_32-aie_columns_1-method_type_0] 173.26 382.66
test_rope[rows_32-cols_512-angle_rows_32-aie_columns_2-method_type_0] 152.82 356.80
test_rope[rows_32-cols_512-angle_rows_32-aie_columns_4-method_type_0] 173.76 364.88
test_rope[rows_32-cols_512-angle_rows_32-aie_columns_8-method_type_0] 200.44 - -
test_rope[rows_32-cols_512-angle_rows_8-aie_columns_1-method_type_0] 175.82 378.18
test_rope[rows_32-cols_512-angle_rows_8-aie_columns_2-method_type_0] 161.62 337.50
test_rope[rows_32-cols_512-angle_rows_8-aie_columns_4-method_type_0] 161.66 380.24
test_rope[rows_32-cols_512-angle_rows_8-aie_columns_8-method_type_0] 183.52 - -
iron/operators/sigmoid
Test Krackan Status Krackan Latency (mean) Phoenix Status Phoenix Latency (mean)
test_sigmoid[input_length_2048-num_aie_columns_1-num_channels_1-tile_size_2048] 184.52 320.64
test_sigmoid[input_length_2048-num_aie_columns_1-num_channels_2-tile_size_1024] 174.22 368.66
test_sigmoid[input_length_2048-num_aie_columns_2-num_channels_1-tile_size_1024] 185.54 657.12
test_sigmoid[input_length_2048-num_aie_columns_2-num_channels_2-tile_size_512] 208.76 330.46
test_sigmoid[input_length_2048-num_aie_columns_4-num_channels_1-tile_size_512] 188.42 665.18
test_sigmoid[input_length_2048-num_aie_columns_4-num_channels_2-tile_size_256] 201.06 468.14
test_sigmoid[input_length_2048-num_aie_columns_8-num_channels_1-tile_size_256] 200.70 - -
test_sigmoid[input_length_2048-num_aie_columns_8-num_channels_2-tile_size_128] 257.98 - -
iron/operators/silu
Test Krackan Status Krackan Latency (mean) Phoenix Status Phoenix Latency (mean)
test_silu[input_length_2048-num_aie_columns_1-num_channels_1-tile_size_2048] 147.56 336.68
test_silu[input_length_2048-num_aie_columns_2-num_channels_1-tile_size_1024] 206.04 468.64
test_silu[input_length_2048-num_aie_columns_4-num_channels_1-tile_size_512] 150.40 489.28
test_silu[input_length_2048-num_aie_columns_8-num_channels_1-tile_size_256] 197.44 - -
iron/operators/softmax
Test Krackan Status Krackan Latency (mean) Phoenix Status Phoenix Latency (mean)
test_softmax[input_length_32768-num_aie_columns_2-num_channels_2-tile_size_1024] 164.04 433.68
test_softmax[input_length_32768-num_aie_columns_2-num_channels_2-tile_size_2048] 170.70 486.30
test_softmax[input_length_32768-num_aie_columns_2-num_channels_2-tile_size_512] 176.40 488.42
iron/operators/swiglu_decode
Test Krackan Status Krackan Latency (mean) Phoenix Status Phoenix Latency (mean)
test_swiglu_decode[embedding_dim_1024-hidden_dim_3584] 956.20 13415.13
test_swiglu_decode[embedding_dim_2048-hidden_dim_2048] 997.95 14558.66
iron/operators/swiglu_prefill
Test Krackan Status Krackan Latency (mean) Phoenix Status Phoenix Latency (mean)
test_swiglu_prefill[seq_len_256-embedding_dim_2048-hidden_dim_2048-prio_accuracy_False] 2210.54 20896.28
iron/operators/tanh
Test Krackan Status Krackan Latency (mean) Phoenix Status Phoenix Latency (mean)
test_tanh[input_length_2048-num_aie_columns_1-num_channels_1-tile_size_2048] 146.74 451.78
test_tanh[input_length_2048-num_aie_columns_1-num_channels_2-tile_size_1024] 162.82 790.58
test_tanh[input_length_2048-num_aie_columns_2-num_channels_1-tile_size_1024] 149.34 351.00
test_tanh[input_length_2048-num_aie_columns_2-num_channels_2-tile_size_512] 181.46 528.72
test_tanh[input_length_2048-num_aie_columns_4-num_channels_1-tile_size_512] 169.68 357.88
test_tanh[input_length_2048-num_aie_columns_4-num_channels_2-tile_size_256] 167.28 426.88
test_tanh[input_length_2048-num_aie_columns_8-num_channels_1-tile_size_256] 165.72 - -
test_tanh[input_length_2048-num_aie_columns_8-num_channels_2-tile_size_128] 270.52 - -
iron/operators/transpose
Test Krackan Status Krackan Latency (mean) Phoenix Status Phoenix Latency (mean)
test_transpose[M_2048-N_64-aie_columns_1-channels_1-m_64-n_64-s_8-num_batches_1] 186.62 448.78
test_transpose[M_2048-N_64-aie_columns_1-channels_1-m_64-n_64-s_8-num_batches_2] 214.58 965.74
test_transpose[M_2048-N_64-aie_columns_1-channels_2-m_64-n_64-s_8-num_batches_1] 185.18 383.00
Krackan - Small

IRON

Tested on 2026_07_09_23_41_45 at commit 1f35831.

iron/operators/axpy
TestChecksLatency (mean)Bandwidth (mean)Throughput (mean)
test_axpy[input_length_2048-num_aie_columns_1-tile_size_2048-scalar_factor_3.0]✅ 5/5158.820.08n/a
test_axpy[input_length_2048-num_aie_columns_2-tile_size_1024-scalar_factor_3.0]✅ 5/5182.940.07n/a
test_axpy[input_length_2048-num_aie_columns_4-tile_size_512-scalar_factor_3.0]✅ 5/5224.520.06n/a
test_axpy[input_length_2048-num_aie_columns_8-tile_size_256-scalar_factor_3.0]✅ 5/5179.860.07n/a
iron/operators/dequant
TestChecksLatency (mean)Bandwidth (mean)Throughput (mean)
test_dequant[input_length_2048-num_aie_columns_1-num_channels_1-tile_size_2048-group_size_32]✅ 5/5175.180.03n/a
test_dequant[input_length_2048-num_aie_columns_1-num_channels_2-tile_size_1024-group_size_32]✅ 5/5178.100.03n/a
test_dequant[input_length_2048-num_aie_columns_2-num_channels_1-tile_size_1024-group_size_32]✅ 5/5153.580.03n/a
test_dequant[input_length_2048-num_aie_columns_2-num_channels_2-tile_size_512-group_size_32]✅ 5/5175.480.03n/a
test_dequant[input_length_2048-num_aie_columns_4-num_channels_1-tile_size_512-group_size_32]✅ 5/5166.240.03n/a
test_dequant[input_length_2048-num_aie_columns_4-num_channels_2-tile_size_256-group_size_32]✅ 5/5172.160.03n/a
test_dequant[input_length_2048-num_aie_columns_8-num_channels_1-tile_size_256-group_size_32]✅ 5/5172.180.03n/a
test_dequant[input_length_2048-num_aie_columns_8-num_channels_2-tile_size_128-group_size_32]✅ 5/5202.740.03n/a
iron/operators/elementwise_add
TestChecksLatency (mean)Bandwidth (mean)Throughput (mean)
test_elementwise_add[input_length_2048-num_aie_columns_1-tile_size_2048]✅ 5/5176.480.07n/a
test_elementwise_add[input_length_2048-num_aie_columns_2-tile_size_1024]✅ 5/5186.720.07n/a
test_elementwise_add[input_length_2048-num_aie_columns_4-tile_size_512]✅ 5/5206.480.06n/a
test_elementwise_add[input_length_2048-num_aie_columns_8-tile_size_256]✅ 5/5187.020.07n/a
iron/operators/elementwise_mul
TestChecksLatency (mean)Bandwidth (mean)Throughput (mean)
test_elementwise_mul[input_length_2048-num_aie_columns_1-tile_size_2048]✅ 5/5189.020.07n/a
test_elementwise_mul[input_length_2048-num_aie_columns_2-tile_size_1024]✅ 5/5188.640.07n/a
test_elementwise_mul[input_length_2048-num_aie_columns_4-tile_size_512]✅ 5/5196.940.06n/a
test_elementwise_mul[input_length_2048-num_aie_columns_8-tile_size_256]✅ 5/5190.380.07n/a
iron/operators/gelu
TestChecksLatency (mean)Bandwidth (mean)Throughput (mean)
test_gelu[input_length_2048-num_aie_columns_1-num_channels_1-tile_size_2048]✅ 5/5188.120.04n/a
test_gelu[input_length_2048-num_aie_columns_1-num_channels_2-tile_size_1024]✅ 5/5176.920.05n/a
test_gelu[input_length_2048-num_aie_columns_2-num_channels_1-tile_size_1024]✅ 5/5201.620.04n/a
test_gelu[input_length_2048-num_aie_columns_2-num_channels_2-tile_size_512]✅ 5/5197.480.04n/a
test_gelu[input_length_2048-num_aie_columns_4-num_channels_1-tile_size_512]✅ 5/5220.200.04n/a
test_gelu[input_length_2048-num_aie_columns_4-num_channels_2-tile_size_256]✅ 5/5216.660.04n/a
test_gelu[input_length_2048-num_aie_columns_8-num_channels_1-tile_size_256]✅ 5/5209.640.04n/a
test_gelu[input_length_2048-num_aie_columns_8-num_channels_2-tile_size_128]✅ 5/5224.220.04n/a
iron/operators/gemm
TestChecksLatency (mean)Bandwidth (mean)Throughput (mean)
test_gemm[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-partition_N_1]✅ 5/52217.184.261677.57
test_gemm[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-partition_N_1]✅ 5/5231.100.9841.77
test_gemm[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-partition_N_1]✅ 5/5262.460.8737.04
test_gemm[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-partition_N_1]✅ 5/548693.580.52352.82
test_gemm[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-partition_N_1]✅ 5/528545.820.88601.85
test_gemm[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-partition_N_1]✅ 5/57907.183.182173.81
test_gemm[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-partition_N_1]✅ 5/52297.503.58939.88
test_gemm[M_64-K_512-N_256-num_aie_columns_4-b_col_maj_True-c_col_maj_False-m_16-k_64-n_64-trace_size_0-partition_N_4]✅ 5/53888.500.3317.53
test_gemm[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-partition_N_1]✅ 5/51444.064.661439.14
iron/operators/gemv
TestChecksLatency (mean)Bandwidth (mean)Throughput (mean)
test_gemv[M_128-K_128-num_aie_columns_1-tile_size_input_32-tile_size_output_128]✅ 5/5n/a0.220.21
test_gemv[M_2048-K_8192-num_aie_columns_1-tile_size_input_1-tile_size_output_2048]✅ 5/5n/a12.6512.64
test_gemv[M_2048-K_8192-num_aie_columns_2-tile_size_input_1-tile_size_output_1024]✅ 5/5n/a25.0925.08
test_gemv[M_2048-K_8192-num_aie_columns_4-tile_size_input_1-tile_size_output_512]✅ 5/5n/a41.6341.61
test_gemv[M_2048-K_8192-num_aie_columns_8-tile_size_input_1-tile_size_output_256]✅ 5/5n/a44.2344.21
test_gemv[M_8192-K_2048-num_aie_columns_1-tile_size_input_4-tile_size_output_1024]✅ 5/5n/a12.7812.77
test_gemv[M_8192-K_2048-num_aie_columns_2-tile_size_input_4-tile_size_output_1024]✅ 5/5n/a24.6824.67
test_gemv[M_8192-K_2048-num_aie_columns_4-tile_size_input_4-tile_size_output_1024]✅ 5/5n/a41.5841.55
test_gemv[M_8192-K_2048-num_aie_columns_8-tile_size_input_4-tile_size_output_1024]✅ 5/5n/a43.9743.95
iron/operators/layer_norm
TestChecksLatency (mean)Bandwidth (mean)Throughput (mean)
test_layer_norm[input_length_2048-num_aie_columns_1-num_channels_1-tile_size_2048]✅ 5/5207.060.04n/a
test_layer_norm[input_length_2048-num_aie_columns_1-num_channels_2-tile_size_1024]✅ 5/5170.740.05n/a
test_layer_norm[input_length_2048-num_aie_columns_2-num_channels_1-tile_size_1024]✅ 5/5180.300.05n/a
test_layer_norm[input_length_2048-num_aie_columns_2-num_channels_2-tile_size_512]✅ 5/5180.260.05n/a
test_layer_norm[input_length_2048-num_aie_columns_4-num_channels_1-tile_size_512]✅ 5/5188.180.04n/a
test_layer_norm[input_length_2048-num_aie_columns_4-num_channels_2-tile_size_256]✅ 5/5170.640.05n/a
test_layer_norm[input_length_2048-num_aie_columns_8-num_channels_1-tile_size_256]✅ 5/5182.920.05n/a
test_layer_norm[input_length_2048-num_aie_columns_8-num_channels_2-tile_size_128]✅ 5/5219.060.04n/a
iron/operators/leaky_relu
TestChecksLatency (mean)Bandwidth (mean)Throughput (mean)
test_leaky_relu[input_length_2048-num_aie_columns_1-num_channels_1-tile_size_2048-alpha_0.01]✅ 5/5170.920.05n/a
test_leaky_relu[input_length_2048-num_aie_columns_1-num_channels_1-tile_size_2048-alpha_0.1]✅ 5/5175.680.05n/a
test_leaky_relu[input_length_2048-num_aie_columns_1-num_channels_1-tile_size_2048-alpha_0.25]✅ 5/5161.600.05n/a
test_leaky_relu[input_length_2048-num_aie_columns_1-num_channels_2-tile_size_1024-alpha_0.01]✅ 5/5181.400.05n/a
test_leaky_relu[input_length_2048-num_aie_columns_2-num_channels_1-tile_size_1024-alpha_0.01]✅ 5/5221.800.04n/a
test_leaky_relu[input_length_2048-num_aie_columns_2-num_channels_2-tile_size_512-alpha_0.01]✅ 5/5184.600.04n/a
test_leaky_relu[input_length_2048-num_aie_columns_4-num_channels_1-tile_size_512-alpha_0.01]✅ 5/5208.140.04n/a
test_leaky_relu[input_length_2048-num_aie_columns_4-num_channels_2-tile_size_256-alpha_0.01]✅ 5/5213.880.04n/a
test_leaky_relu[input_length_2048-num_aie_columns_8-num_channels_1-tile_size_256-alpha_0.01]✅ 5/5185.380.05n/a
test_leaky_relu[input_length_2048-num_aie_columns_8-num_channels_2-tile_size_128-alpha_0.01]✅ 5/5203.080.04n/a
iron/operators/mem_copy
TestChecksLatency (mean)Bandwidth (mean)Throughput (mean)
test_mem_copy[input_length_2048-num_cores_1-num_channels_1-bypass_False-tile_size_2048]✅ 5/5139.620.06n/a
test_mem_copy[input_length_2048-num_cores_16-num_channels_2-bypass_False-tile_size_128]✅ 5/5255.540.03n/a
test_mem_copy[input_length_2048-num_cores_2-num_channels_1-bypass_False-tile_size_1024]✅ 5/5168.180.05n/a
test_mem_copy[input_length_2048-num_cores_2-num_channels_2-bypass_False-tile_size_1024]✅ 5/5178.520.05n/a
test_mem_copy[input_length_2048-num_cores_4-num_channels_1-bypass_False-tile_size_512]✅ 5/5192.900.04n/a
test_mem_copy[input_length_2048-num_cores_4-num_channels_2-bypass_False-tile_size_512]✅ 5/5161.020.05n/a
test_mem_copy[input_length_2048-num_cores_8-num_channels_1-bypass_False-tile_size_256]✅ 5/5201.300.04n/a
test_mem_copy[input_length_2048-num_cores_8-num_channels_2-bypass_False-tile_size_256]✅ 5/5197.180.04n/a
iron/operators/mha
TestChecksLatency (mean)Bandwidth (mean)Throughput (mean)
test_mha[seq_len_16384-dim_64-num_heads_1-num_pipelines_8-num_kv_heads_0]✅ 5/540865.300.21n/a
iron/operators/relu
TestChecksLatency (mean)Bandwidth (mean)Throughput (mean)
test_relu[input_length_2048-num_aie_columns_1-num_channels_1-tile_size_2048]✅ 5/5178.680.05n/a
test_relu[input_length_2048-num_aie_columns_1-num_channels_2-tile_size_1024]✅ 5/5193.200.04n/a
test_relu[input_length_2048-num_aie_columns_2-num_channels_1-tile_size_1024]✅ 5/5185.400.04n/a
test_relu[input_length_2048-num_aie_columns_2-num_channels_2-tile_size_512]✅ 5/5207.200.04n/a
test_relu[input_length_2048-num_aie_columns_4-num_channels_1-tile_size_512]✅ 5/5187.780.04n/a
test_relu[input_length_2048-num_aie_columns_4-num_channels_2-tile_size_256]✅ 5/5175.700.05n/a
test_relu[input_length_2048-num_aie_columns_8-num_channels_1-tile_size_256]✅ 5/5194.680.04n/a
test_relu[input_length_2048-num_aie_columns_8-num_channels_2-tile_size_128]✅ 5/5246.800.03n/a
iron/operators/rms_norm
TestChecksLatency (mean)Bandwidth (mean)Throughput (mean)
test_rms_norm[input_length_2048-num_aie_columns_1-num_channels_1-tile_size_2048-weighted_False]✅ 5/5160.600.05n/a
test_rms_norm[input_length_2048-num_aie_columns_1-num_channels_1-tile_size_2048-weighted_True]✅ 5/5183.340.07n/a
test_rms_norm[input_length_2048-num_aie_columns_1-num_channels_2-tile_size_1024-weighted_False]✅ 5/5195.380.04n/a
test_rms_norm[input_length_2048-num_aie_columns_1-num_channels_2-tile_size_1024-weighted_True]✅ 5/5177.380.06n/a
test_rms_norm[input_length_2048-num_aie_columns_2-num_channels_1-tile_size_1024-weighted_False]✅ 5/5189.200.04n/a
test_rms_norm[input_length_2048-num_aie_columns_2-num_channels_1-tile_size_1024-weighted_True]✅ 5/5196.620.05n/a
test_rms_norm[input_length_2048-num_aie_columns_2-num_channels_2-tile_size_512-weighted_False]✅ 5/5198.480.04n/a
test_rms_norm[input_length_2048-num_aie_columns_2-num_channels_2-tile_size_512-weighted_True]✅ 5/5193.800.05n/a
test_rms_norm[input_length_2048-num_aie_columns_4-num_channels_1-tile_size_512-weighted_False]✅ 5/5191.680.05n/a
test_rms_norm[input_length_2048-num_aie_columns_4-num_channels_1-tile_size_512-weighted_True]✅ 5/5168.260.06n/a
test_rms_norm[input_length_2048-num_aie_columns_4-num_channels_2-tile_size_256-weighted_False]✅ 5/5195.000.04n/a
test_rms_norm[input_length_2048-num_aie_columns_4-num_channels_2-tile_size_256-weighted_True]✅ 5/5192.200.05n/a
test_rms_norm[input_length_2048-num_aie_columns_8-num_channels_1-tile_size_256-weighted_False]✅ 5/5204.020.04n/a
test_rms_norm[input_length_2048-num_aie_columns_8-num_channels_1-tile_size_256-weighted_True]✅ 5/5187.720.05n/a
test_rms_norm[input_length_2048-num_aie_columns_8-num_channels_2-tile_size_128-weighted_False]✅ 5/5226.920.04n/a
iron/operators/rope
TestChecksLatency (mean)Bandwidth (mean)Throughput (mean)
test_rope[rows_32-cols_512-angle_rows_32-aie_columns_1-method_type_0]✅ 5/5173.260.57n/a
test_rope[rows_32-cols_512-angle_rows_32-aie_columns_2-method_type_0]✅ 5/5152.820.66n/a
test_rope[rows_32-cols_512-angle_rows_32-aie_columns_4-method_type_0]✅ 5/5173.760.58n/a
test_rope[rows_32-cols_512-angle_rows_32-aie_columns_8-method_type_0]✅ 5/5200.440.51n/a
test_rope[rows_32-cols_512-angle_rows_8-aie_columns_1-method_type_0]✅ 5/5175.820.45n/a
test_rope[rows_32-cols_512-angle_rows_8-aie_columns_2-method_type_0]✅ 5/5161.620.46n/a
test_rope[rows_32-cols_512-angle_rows_8-aie_columns_4-method_type_0]✅ 5/5161.660.47n/a
test_rope[rows_32-cols_512-angle_rows_8-aie_columns_8-method_type_0]✅ 5/5183.520.41n/a
iron/operators/sigmoid
TestChecksLatency (mean)Bandwidth (mean)Throughput (mean)
test_sigmoid[input_length_2048-num_aie_columns_1-num_channels_1-tile_size_2048]✅ 5/5184.520.04n/a
test_sigmoid[input_length_2048-num_aie_columns_1-num_channels_2-tile_size_1024]✅ 5/5174.220.05n/a
test_sigmoid[input_length_2048-num_aie_columns_2-num_channels_1-tile_size_1024]✅ 5/5185.540.05n/a
test_sigmoid[input_length_2048-num_aie_columns_2-num_channels_2-tile_size_512]✅ 5/5208.760.04n/a
test_sigmoid[input_length_2048-num_aie_columns_4-num_channels_1-tile_size_512]✅ 5/5188.420.04n/a
test_sigmoid[input_length_2048-num_aie_columns_4-num_channels_2-tile_size_256]✅ 5/5201.060.04n/a
test_sigmoid[input_length_2048-num_aie_columns_8-num_channels_1-tile_size_256]✅ 5/5200.700.04n/a
test_sigmoid[input_length_2048-num_aie_columns_8-num_channels_2-tile_size_128]✅ 5/5257.980.03n/a
iron/operators/silu
TestChecksLatency (mean)Bandwidth (mean)Throughput (mean)
test_silu[input_length_2048-num_aie_columns_1-num_channels_1-tile_size_2048]✅ 5/5147.560.06n/a
test_silu[input_length_2048-num_aie_columns_2-num_channels_1-tile_size_1024]✅ 5/5206.040.05n/a
test_silu[input_length_2048-num_aie_columns_4-num_channels_1-tile_size_512]✅ 5/5150.400.06n/a
test_silu[input_length_2048-num_aie_columns_8-num_channels_1-tile_size_256]✅ 5/5197.440.04n/a
iron/operators/softmax
TestChecksLatency (mean)Bandwidth (mean)Throughput (mean)
test_softmax[input_length_32768-num_aie_columns_2-num_channels_2-tile_size_1024]✅ 5/5164.040.83n/a
test_softmax[input_length_32768-num_aie_columns_2-num_channels_2-tile_size_2048]✅ 5/5170.700.77n/a
test_softmax[input_length_32768-num_aie_columns_2-num_channels_2-tile_size_512]✅ 5/5176.400.76n/a
iron/operators/swiglu_decode
TestChecksLatency (mean)Bandwidth (mean)Throughput (mean)
test_swiglu_decode[embedding_dim_1024-hidden_dim_3584]✅ 5/5956.200.00n/a
test_swiglu_decode[embedding_dim_2048-hidden_dim_2048]✅ 5/5997.950.01n/a
iron/operators/swiglu_prefill
TestChecksLatency (mean)Bandwidth (mean)Throughput (mean)
test_swiglu_prefill[seq_len_256-embedding_dim_2048-hidden_dim_2048-prio_accuracy_False]✅ 5/52210.540.95n/a
iron/operators/tanh
TestChecksLatency (mean)Bandwidth (mean)Throughput (mean)
test_tanh[input_length_2048-num_aie_columns_1-num_channels_1-tile_size_2048]✅ 5/5146.740.06n/a
test_tanh[input_length_2048-num_aie_columns_1-num_channels_2-tile_size_1024]✅ 5/5162.820.05n/a
test_tanh[input_length_2048-num_aie_columns_2-num_channels_1-tile_size_1024]✅ 5/5149.340.06n/a
test_tanh[input_length_2048-num_aie_columns_2-num_channels_2-tile_size_512]✅ 5/5181.460.05n/a
test_tanh[input_length_2048-num_aie_columns_4-num_channels_1-tile_size_512]✅ 5/5169.680.05n/a
test_tanh[input_length_2048-num_aie_columns_4-num_channels_2-tile_size_256]✅ 5/5167.280.05n/a
test_tanh[input_length_2048-num_aie_columns_8-num_channels_1-tile_size_256]✅ 5/5165.720.05n/a
test_tanh[input_length_2048-num_aie_columns_8-num_channels_2-tile_size_128]✅ 5/5270.520.03n/a
iron/operators/transpose
TestChecksLatency (mean)Bandwidth (mean)Throughput (mean)
test_transpose[M_2048-N_64-aie_columns_1-channels_1-m_64-n_64-s_8-num_batches_1]✅ 5/5186.622.87n/a
test_transpose[M_2048-N_64-aie_columns_1-channels_1-m_64-n_64-s_8-num_batches_2]✅ 5/5214.584.93n/a
test_transpose[M_2048-N_64-aie_columns_1-channels_2-m_64-n_64-s_8-num_batches_1]✅ 5/5185.182.90n/a

Trends:

IRON Trends

iron/operators/axpy

test_axpy[input_length_2048-num_aie_columns_1-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)
1f35831 — 2026-07-09 23:36:090.09 (+4.46%)0.08 (+0.81%)0.07 (-2.98%)0.07 (-2.16%)0.01 (+58.60%)184.40 (+2.16%)158.82 (+0.61%)165.20 (+3.06%)129.90 (-4.27%)27.02 (+54.31%)
17769b8 — 2026-07-09 23:01:350.09 (n/a)0.08 (n/a)0.08 (n/a)0.07 (n/a)0.01 (n/a)180.50 (n/a)157.86 (n/a)160.30 (n/a)135.70 (n/a)17.51 (n/a)

test_axpy[input_length_2048-num_aie_columns_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)
1f35831 — 2026-07-09 23:36:090.10 (+32.68%)0.07 (+7.65%)0.06 (-5.66%)0.06 (+8.86%)0.02 (+88.97%)208.30 (-8.16%)182.94 (-5.40%)196.90 (+6.03%)128.60 (-24.66%)32.34 (+29.61%)
17769b8 — 2026-07-09 23:01:350.07 (n/a)0.06 (n/a)0.07 (n/a)0.05 (n/a)0.01 (n/a)226.80 (n/a)193.38 (n/a)185.70 (n/a)170.70 (n/a)24.95 (n/a)

test_axpy[input_length_2048-num_aie_columns_4-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)
1f35831 — 2026-07-09 23:36:090.07 (-31.19%)0.06 (-24.69%)0.06 (-25.23%)0.05 (-7.35%)0.01 (-50.16%)263.70 (+7.94%)224.52 (+28.59%)222.30 (+33.75%)173.20 (+45.42%)35.65 (-22.25%)
17769b8 — 2026-07-09 23:01:350.10 (n/a)0.07 (n/a)0.07 (n/a)0.05 (n/a)0.02 (n/a)244.30 (n/a)174.60 (n/a)166.20 (n/a)119.10 (n/a)45.86 (n/a)

test_axpy[input_length_2048-num_aie_columns_8-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)
1f35831 — 2026-07-09 23:36:090.08 (-9.73%)0.07 (+9.43%)0.07 (+6.34%)0.06 (+33.87%)0.01 (-52.59%)198.80 (-25.29%)179.86 (-11.70%)185.00 (-5.95%)155.80 (+10.73%)18.21 (-59.83%)
17769b8 — 2026-07-09 23:01:350.09 (n/a)0.06 (n/a)0.06 (n/a)0.05 (n/a)0.02 (n/a)266.10 (n/a)203.70 (n/a)196.70 (n/a)140.70 (n/a)45.32 (n/a)
iron/operators/dequant

test_dequant[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)
1f35831 — 2026-07-09 23:36:090.04 (-12.31%)0.03 (-17.73%)0.03 (-22.39%)0.03 (-16.50%)0.01 (-8.27%)206.00 (+19.77%)175.18 (+21.79%)176.00 (+28.84%)129.90 (+14.05%)28.34 (+19.39%)
17769b8 — 2026-07-09 23:01:350.05 (n/a)0.04 (n/a)0.04 (n/a)0.03 (n/a)0.01 (n/a)172.00 (n/a)143.84 (n/a)136.60 (n/a)113.90 (n/a)23.74 (n/a)

test_dequant[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)
1f35831 — 2026-07-09 23:36:090.03 (-37.65%)0.03 (-22.39%)0.03 (-19.17%)0.03 (-19.39%)0.00 (-58.92%)201.50 (+24.08%)178.10 (+25.40%)190.30 (+23.73%)150.90 (+60.36%)23.18 (-17.89%)
17769b8 — 2026-07-09 23:01:350.06 (n/a)0.04 (n/a)0.03 (n/a)0.03 (n/a)0.01 (n/a)162.40 (n/a)142.02 (n/a)153.80 (n/a)94.10 (n/a)28.23 (n/a)

test_dequant[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)
1f35831 — 2026-07-09 23:36:090.04 (-9.99%)0.03 (-7.68%)0.03 (-8.59%)0.03 (-4.76%)0.00 (-26.37%)169.60 (+5.02%)153.58 (+7.61%)164.80 (+9.43%)130.70 (+11.14%)18.04 (-14.49%)
17769b8 — 2026-07-09 23:01:350.04 (n/a)0.04 (n/a)0.03 (n/a)0.03 (n/a)0.01 (n/a)161.50 (n/a)142.72 (n/a)150.60 (n/a)117.60 (n/a)21.10 (n/a)

test_dequant[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)
1f35831 — 2026-07-09 23:36:090.04 (-11.75%)0.03 (-1.84%)0.03 (+1.78%)0.03 (+5.95%)0.00 (-39.24%)192.80 (-5.63%)175.48 (+0.35%)176.90 (-1.78%)145.40 (+13.33%)19.27 (-32.82%)
17769b8 — 2026-07-09 23:01:350.04 (n/a)0.03 (n/a)0.03 (n/a)0.03 (n/a)0.01 (n/a)204.30 (n/a)174.86 (n/a)180.10 (n/a)128.30 (n/a)28.68 (n/a)

test_dequant[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)
1f35831 — 2026-07-09 23:36:090.04 (+17.38%)0.03 (+11.69%)0.03 (+4.10%)0.03 (+23.57%)0.01 (+16.54%)192.30 (-19.10%)166.24 (-10.67%)169.90 (-3.96%)120.20 (-14.81%)28.86 (-21.01%)
17769b8 — 2026-07-09 23:01:350.04 (n/a)0.03 (n/a)0.03 (n/a)0.02 (n/a)0.01 (n/a)237.70 (n/a)186.10 (n/a)176.90 (n/a)141.10 (n/a)36.54 (n/a)

test_dequant[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)
1f35831 — 2026-07-09 23:36:090.04 (+22.07%)0.03 (+12.01%)0.03 (+1.95%)0.03 (+5.12%)0.00 (+146.47%)192.90 (-4.88%)172.16 (-9.81%)184.30 (-1.92%)146.00 (-18.07%)21.34 (+88.52%)
17769b8 — 2026-07-09 23:01:350.03 (n/a)0.03 (n/a)0.03 (n/a)0.03 (n/a)0.00 (n/a)202.80 (n/a)190.88 (n/a)187.90 (n/a)178.20 (n/a)11.32 (n/a)

test_dequant[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)
1f35831 — 2026-07-09 23:36:090.04 (-23.54%)0.03 (-1.19%)0.03 (+12.90%)0.03 (-2.91%)0.00 (-54.36%)208.20 (+2.97%)172.18 (-1.96%)161.90 (-11.43%)149.30 (+30.74%)22.94 (-36.48%)
17769b8 — 2026-07-09 23:01:350.05 (n/a)0.03 (n/a)0.03 (n/a)0.03 (n/a)0.01 (n/a)202.20 (n/a)175.62 (n/a)182.80 (n/a)114.20 (n/a)36.12 (n/a)

test_dequant[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)
1f35831 — 2026-07-09 23:36:090.03 (+11.06%)0.03 (+12.28%)0.03 (+13.26%)0.02 (+13.84%)0.00 (+9.12%)213.50 (-12.18%)202.74 (-10.95%)197.60 (-11.71%)195.50 (-9.95%)8.80 (-14.24%)
17769b8 — 2026-07-09 23:01:350.02 (n/a)0.02 (n/a)0.02 (n/a)0.02 (n/a)0.00 (n/a)243.10 (n/a)227.68 (n/a)223.80 (n/a)217.10 (n/a)10.26 (n/a)
iron/operators/elementwise_add

test_elementwise_add[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)
1f35831 — 2026-07-09 23:36:090.08 (n/a)0.07 (n/a)0.07 (n/a)0.06 (n/a)0.01 (n/a)200.30 (n/a)176.48 (n/a)169.30 (n/a)160.50 (n/a)17.43 (n/a)

test_elementwise_add[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)
1f35831 — 2026-07-09 23:36:090.09 (n/a)0.07 (n/a)0.06 (n/a)0.06 (n/a)0.01 (n/a)212.30 (n/a)186.72 (n/a)193.70 (n/a)143.60 (n/a)29.16 (n/a)

test_elementwise_add[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)
1f35831 — 2026-07-09 23:36:090.07 (n/a)0.06 (n/a)0.06 (n/a)0.05 (n/a)0.01 (n/a)250.10 (n/a)206.48 (n/a)208.20 (n/a)167.70 (n/a)29.69 (n/a)

test_elementwise_add[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)
1f35831 — 2026-07-09 23:36:090.08 (n/a)0.07 (n/a)0.07 (n/a)0.06 (n/a)0.01 (n/a)215.20 (n/a)187.02 (n/a)187.50 (n/a)149.90 (n/a)26.09 (n/a)
iron/operators/elementwise_mul

test_elementwise_mul[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)
1f35831 — 2026-07-09 23:36:090.07 (n/a)0.07 (n/a)0.07 (n/a)0.06 (n/a)0.01 (n/a)218.90 (n/a)189.02 (n/a)184.50 (n/a)172.30 (n/a)18.58 (n/a)

test_elementwise_mul[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)
1f35831 — 2026-07-09 23:36:090.08 (n/a)0.07 (n/a)0.07 (n/a)0.05 (n/a)0.01 (n/a)232.60 (n/a)188.64 (n/a)179.00 (n/a)161.70 (n/a)30.62 (n/a)

test_elementwise_mul[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)
1f35831 — 2026-07-09 23:36:090.07 (n/a)0.06 (n/a)0.06 (n/a)0.06 (n/a)0.01 (n/a)218.70 (n/a)196.94 (n/a)198.10 (n/a)176.10 (n/a)18.82 (n/a)

test_elementwise_mul[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)
1f35831 — 2026-07-09 23:36:090.09 (n/a)0.07 (n/a)0.06 (n/a)0.06 (n/a)0.01 (n/a)215.60 (n/a)190.38 (n/a)195.80 (n/a)143.30 (n/a)27.69 (n/a)
iron/operators/gelu

test_gelu[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)
1f35831 — 2026-07-09 23:36:090.05 (n/a)0.04 (n/a)0.05 (n/a)0.04 (n/a)0.01 (n/a)225.00 (n/a)188.12 (n/a)181.50 (n/a)160.10 (n/a)24.43 (n/a)

test_gelu[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)
1f35831 — 2026-07-09 23:36:090.05 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.00 (n/a)194.50 (n/a)176.92 (n/a)176.00 (n/a)160.10 (n/a)13.37 (n/a)

test_gelu[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)
1f35831 — 2026-07-09 23:36:090.04 (n/a)0.04 (n/a)0.04 (n/a)0.04 (n/a)0.00 (n/a)209.70 (n/a)201.62 (n/a)204.30 (n/a)189.50 (n/a)7.94 (n/a)

test_gelu[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)
1f35831 — 2026-07-09 23:36:090.06 (n/a)0.04 (n/a)0.05 (n/a)0.03 (n/a)0.01 (n/a)250.70 (n/a)197.48 (n/a)180.30 (n/a)135.50 (n/a)49.23 (n/a)

test_gelu[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)
1f35831 — 2026-07-09 23:36:090.05 (n/a)0.04 (n/a)0.04 (n/a)0.02 (n/a)0.01 (n/a)353.10 (n/a)220.20 (n/a)203.60 (n/a)156.40 (n/a)77.18 (n/a)

test_gelu[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)
1f35831 — 2026-07-09 23:36:090.04 (n/a)0.04 (n/a)0.04 (n/a)0.03 (n/a)0.00 (n/a)246.10 (n/a)216.66 (n/a)208.80 (n/a)203.20 (n/a)18.08 (n/a)

test_gelu[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)
1f35831 — 2026-07-09 23:36:090.05 (n/a)0.04 (n/a)0.04 (n/a)0.03 (n/a)0.01 (n/a)243.40 (n/a)209.64 (n/a)212.70 (n/a)161.20 (n/a)31.69 (n/a)

test_gelu[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)
1f35831 — 2026-07-09 23:36:090.04 (n/a)0.04 (n/a)0.04 (n/a)0.03 (n/a)0.00 (n/a)244.50 (n/a)224.22 (n/a)233.90 (n/a)188.50 (n/a)22.30 (n/a)
iron/operators/gemm

test_gemm[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-partition_N_1]

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)
1f35831 — 2026-07-09 23:36:094.91 (+19.07%)4.26 (+4.74%)4.16 (+1.87%)4.03 (+2.01%)0.37 (+436.45%)2330.80 (-1.97%)2217.18 (-4.02%)2260.40 (-1.83%)1913.80 (-16.01%)173.24 (+336.87%)1933.04 (+19.07%)1677.57 (+4.74%)1636.59 (+1.87%)1587.18 (+2.01%)144.97 (+436.45%)
17769b8 — 2026-07-09 23:01:354.13 (n/a)4.07 (n/a)4.08 (n/a)3.96 (n/a)0.07 (n/a)2377.60 (n/a)2310.12 (n/a)2302.60 (n/a)2278.70 (n/a)39.66 (n/a)1623.43 (n/a)1601.73 (n/a)1606.61 (n/a)1555.90 (n/a)27.02 (n/a)

test_gemm[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-partition_N_1]

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)
1f35831 — 2026-07-09 23:36:091.12 (-16.73%)0.98 (+3.26%)1.03 (+3.53%)0.73 (+6.46%)0.15 (-43.69%)305.10 (-6.07%)231.10 (-6.80%)215.00 (-3.41%)197.50 (+20.06%)42.53 (-36.09%)47.78 (-16.73%)41.77 (+3.26%)43.89 (+3.53%)30.94 (+6.46%)6.42 (-43.69%)
17769b8 — 2026-07-09 23:01:351.34 (n/a)0.95 (n/a)0.99 (n/a)0.68 (n/a)0.27 (n/a)324.80 (n/a)247.96 (n/a)222.60 (n/a)164.50 (n/a)66.55 (n/a)57.38 (n/a)40.46 (n/a)42.39 (n/a)29.06 (n/a)11.39 (n/a)

test_gemm[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-partition_N_1]

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)
1f35831 — 2026-07-09 23:36:091.05 (-23.19%)0.87 (-18.96%)0.94 (-21.18%)0.67 (+11.15%)0.16 (-44.97%)328.30 (-10.05%)262.46 (+17.08%)234.90 (+26.84%)210.90 (+30.19%)51.68 (-37.08%)44.74 (-23.19%)37.04 (-18.96%)40.17 (-21.18%)28.74 (+11.15%)6.92 (-44.97%)
17769b8 — 2026-07-09 23:01:351.37 (n/a)1.07 (n/a)1.19 (n/a)0.61 (n/a)0.29 (n/a)365.00 (n/a)224.18 (n/a)185.20 (n/a)162.00 (n/a)82.13 (n/a)58.25 (n/a)45.71 (n/a)50.96 (n/a)25.86 (n/a)12.57 (n/a)

test_gemm[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-partition_N_1]

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)
1f35831 — 2026-07-09 23:36:090.52 (-0.02%)0.52 (-0.13%)0.52 (-0.03%)0.51 (-0.47%)0.00 (+539.14%)48876.90 (+0.47%)48693.58 (+0.13%)48632.90 (+0.03%)48621.90 (+0.02%)108.15 (+543.12%)353.34 (-0.02%)352.82 (-0.13%)353.26 (-0.03%)351.49 (-0.47%)0.78 (+539.15%)
17769b8 — 2026-07-09 23:01:350.52 (n/a)0.52 (n/a)0.52 (n/a)0.52 (n/a)0.00 (n/a)48648.30 (n/a)48629.88 (n/a)48619.90 (n/a)48613.70 (n/a)16.82 (n/a)353.40 (n/a)353.28 (n/a)353.35 (n/a)353.14 (n/a)0.12 (n/a)

test_gemm[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-partition_N_1]

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)
1f35831 — 2026-07-09 23:36:090.89 (-0.43%)0.88 (-0.02%)0.88 (+0.15%)0.87 (-0.10%)0.01 (-13.94%)28832.50 (+0.10%)28545.82 (+0.02%)28554.10 (-0.15%)28382.10 (+0.43%)182.29 (-13.45%)605.31 (-0.43%)601.85 (-0.02%)601.66 (+0.15%)595.85 (-0.10%)3.83 (-13.94%)
17769b8 — 2026-07-09 23:01:350.89 (n/a)0.88 (n/a)0.88 (n/a)0.87 (n/a)0.01 (n/a)28803.80 (n/a)28539.90 (n/a)28597.40 (n/a)28260.10 (n/a)210.60 (n/a)607.92 (n/a)601.99 (n/a)600.75 (n/a)596.44 (n/a)4.45 (n/a)

test_gemm[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-partition_N_1]

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)
1f35831 — 2026-07-09 23:36:093.28 (-1.29%)3.18 (-1.20%)3.18 (-0.53%)3.07 (-1.30%)0.08 (-13.33%)8201.10 (+1.32%)7907.18 (+1.20%)7926.00 (+0.54%)7671.10 (+1.31%)201.65 (-10.48%)2239.55 (-1.29%)2173.81 (-1.20%)2167.53 (-0.53%)2094.83 (-1.30%)55.04 (-13.33%)
17769b8 — 2026-07-09 23:01:353.32 (n/a)3.22 (n/a)3.19 (n/a)3.11 (n/a)0.09 (n/a)8094.50 (n/a)7813.80 (n/a)7883.70 (n/a)7572.10 (n/a)225.26 (n/a)2268.84 (n/a)2200.13 (n/a)2179.17 (n/a)2122.40 (n/a)63.50 (n/a)

test_gemm[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-partition_N_1]

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)
1f35831 — 2026-07-09 23:36:094.23 (+7.62%)3.58 (+4.13%)3.55 (+0.01%)2.94 (-3.04%)0.58 (+50.92%)2746.40 (+3.13%)2297.50 (-2.88%)2270.90 (-0.00%)1907.10 (-7.08%)374.42 (+40.62%)1108.42 (+7.62%)939.88 (+4.13%)930.87 (+0.01%)769.71 (-3.04%)152.31 (+50.92%)
17769b8 — 2026-07-09 23:01:353.93 (n/a)3.44 (n/a)3.55 (n/a)3.03 (n/a)0.38 (n/a)2663.00 (n/a)2365.70 (n/a)2271.00 (n/a)2052.50 (n/a)266.26 (n/a)1029.92 (n/a)902.64 (n/a)930.82 (n/a)793.82 (n/a)100.92 (n/a)

test_gemm[M_64-K_512-N_256-num_aie_columns_4-b_col_maj_True-c_col_maj_False-m_16-k_64-n_64-trace_size_0-partition_N_4]

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)
1f35831 — 2026-07-09 23:36:090.37 (-27.79%)0.33 (-20.17%)0.32 (-12.86%)0.27 (-16.38%)0.05 (-51.68%)4608.90 (+19.58%)3888.50 (+22.21%)3897.40 (+14.75%)3335.80 (+38.50%)545.61 (-20.64%)20.12 (-27.79%)17.53 (-20.17%)17.22 (-12.86%)14.56 (-16.38%)2.43 (-51.68%)
17769b8 — 2026-07-09 23:01:350.52 (n/a)0.41 (n/a)0.37 (n/a)0.32 (n/a)0.09 (n/a)3854.10 (n/a)3181.74 (n/a)3396.30 (n/a)2408.60 (n/a)687.47 (n/a)27.86 (n/a)21.96 (n/a)19.76 (n/a)17.41 (n/a)5.03 (n/a)

test_gemm[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-partition_N_1]

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)
1f35831 — 2026-07-09 23:36:095.21 (-14.53%)4.66 (+1.78%)4.75 (-2.04%)3.97 (+19.59%)0.54 (-51.25%)1675.10 (-16.38%)1444.06 (-5.28%)1401.60 (+2.08%)1275.90 (+17.00%)172.69 (-53.60%)1610.73 (-14.53%)1439.14 (+1.78%)1466.30 (-2.04%)1226.91 (+19.59%)166.99 (-51.25%)
17769b8 — 2026-07-09 23:01:356.10 (n/a)4.58 (n/a)4.84 (n/a)3.32 (n/a)1.11 (n/a)2003.30 (n/a)1524.52 (n/a)1373.10 (n/a)1090.50 (n/a)372.16 (n/a)1884.65 (n/a)1414.00 (n/a)1496.77 (n/a)1025.91 (n/a)342.56 (n/a)
iron/operators/gemv

test_gemv[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)
1f35831 — 2026-07-09 23:36:090.25 (+8.87%)0.22 (+5.57%)0.22 (+10.33%)0.17 (-14.88%)0.03 (+107.74%)0.25 (+8.87%)0.21 (+5.57%)0.21 (+10.33%)0.16 (-14.88%)0.03 (+107.74%)
17769b8 — 2026-07-09 23:01:350.23 (n/a)0.21 (n/a)0.20 (n/a)0.20 (n/a)0.02 (n/a)0.23 (n/a)0.20 (n/a)0.19 (n/a)0.19 (n/a)0.02 (n/a)

test_gemv[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)
1f35831 — 2026-07-09 23:36:0913.46 (+1.09%)12.65 (-1.04%)12.29 (-7.26%)12.11 (+11.48%)0.62 (-42.42%)13.45 (+1.09%)12.64 (-1.04%)12.29 (-7.26%)12.10 (+11.48%)0.62 (-42.42%)
17769b8 — 2026-07-09 23:01:3513.32 (n/a)12.79 (n/a)13.26 (n/a)10.86 (n/a)1.08 (n/a)13.31 (n/a)12.78 (n/a)13.25 (n/a)10.86 (n/a)1.07 (n/a)

test_gemv[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)
1f35831 — 2026-07-09 23:36:0925.33 (-0.09%)25.09 (+1.68%)25.22 (+2.78%)24.66 (+1.82%)0.28 (-42.72%)25.31 (-0.09%)25.08 (+1.68%)25.20 (+2.78%)24.65 (+1.82%)0.28 (-42.72%)
17769b8 — 2026-07-09 23:01:3525.35 (n/a)24.68 (n/a)24.54 (n/a)24.22 (n/a)0.49 (n/a)25.33 (n/a)24.66 (n/a)24.52 (n/a)24.20 (n/a)0.49 (n/a)

test_gemv[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)
1f35831 — 2026-07-09 23:36:0942.67 (-3.99%)41.63 (+1.54%)41.52 (+0.62%)40.27 (+8.89%)0.93 (-66.25%)42.65 (-3.99%)41.61 (+1.54%)41.50 (+0.62%)40.24 (+8.89%)0.93 (-66.25%)
17769b8 — 2026-07-09 23:01:3544.44 (n/a)41.00 (n/a)41.27 (n/a)36.98 (n/a)2.77 (n/a)44.42 (n/a)40.98 (n/a)41.24 (n/a)36.96 (n/a)2.77 (n/a)

test_gemv[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)
1f35831 — 2026-07-09 23:36:0945.16 (-0.89%)44.23 (+1.71%)44.77 (+4.43%)43.03 (+0.84%)1.07 (-12.77%)45.13 (-0.89%)44.21 (+1.71%)44.74 (+4.43%)43.00 (+0.84%)1.07 (-12.77%)
17769b8 — 2026-07-09 23:01:3545.57 (n/a)43.49 (n/a)42.87 (n/a)42.67 (n/a)1.23 (n/a)45.54 (n/a)43.46 (n/a)42.85 (n/a)42.64 (n/a)1.23 (n/a)

test_gemv[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)
1f35831 — 2026-07-09 23:36:0913.52 (+1.69%)12.78 (+0.40%)13.27 (+0.35%)11.61 (-2.27%)0.84 (+16.44%)13.51 (+1.69%)12.77 (+0.40%)13.26 (+0.35%)11.60 (-2.27%)0.84 (+16.44%)
17769b8 — 2026-07-09 23:01:3513.30 (n/a)12.73 (n/a)13.22 (n/a)11.88 (n/a)0.72 (n/a)13.29 (n/a)12.72 (n/a)13.21 (n/a)11.87 (n/a)0.72 (n/a)

test_gemv[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)
1f35831 — 2026-07-09 23:36:0925.26 (+0.35%)24.68 (-0.35%)24.73 (-0.18%)24.14 (-0.29%)0.50 (+32.72%)25.25 (+0.35%)24.67 (-0.35%)24.71 (-0.18%)24.13 (-0.29%)0.50 (+32.72%)
17769b8 — 2026-07-09 23:01:3525.17 (n/a)24.77 (n/a)24.77 (n/a)24.21 (n/a)0.38 (n/a)25.16 (n/a)24.75 (n/a)24.76 (n/a)24.20 (n/a)0.38 (n/a)

test_gemv[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)
1f35831 — 2026-07-09 23:36:0942.86 (+3.55%)41.58 (+4.26%)41.90 (+5.22%)39.37 (+3.01%)1.35 (-2.45%)42.83 (+3.55%)41.55 (+4.26%)41.87 (+5.22%)39.35 (+3.01%)1.34 (-2.45%)
17769b8 — 2026-07-09 23:01:3541.39 (n/a)39.88 (n/a)39.82 (n/a)38.22 (n/a)1.38 (n/a)41.36 (n/a)39.85 (n/a)39.79 (n/a)38.20 (n/a)1.38 (n/a)

test_gemv[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)
1f35831 — 2026-07-09 23:36:0945.56 (+6.84%)43.97 (+4.15%)44.42 (+4.79%)40.79 (-1.54%)1.92 (+308.99%)45.53 (+6.84%)43.95 (+4.15%)44.40 (+4.79%)40.77 (-1.54%)1.92 (+308.99%)
17769b8 — 2026-07-09 23:01:3542.64 (n/a)42.22 (n/a)42.39 (n/a)41.43 (n/a)0.47 (n/a)42.62 (n/a)42.19 (n/a)42.37 (n/a)41.41 (n/a)0.47 (n/a)
iron/operators/layer_norm

test_layer_norm[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)
1f35831 — 2026-07-09 23:36:090.06 (n/a)0.04 (n/a)0.05 (n/a)0.03 (n/a)0.01 (n/a)292.10 (n/a)207.06 (n/a)176.30 (n/a)144.00 (n/a)62.54 (n/a)

test_layer_norm[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)
1f35831 — 2026-07-09 23:36:090.06 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.00 (n/a)185.90 (n/a)170.74 (n/a)175.70 (n/a)146.60 (n/a)14.73 (n/a)

test_layer_norm[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)
1f35831 — 2026-07-09 23:36:090.05 (n/a)0.05 (n/a)0.04 (n/a)0.04 (n/a)0.01 (n/a)204.20 (n/a)180.30 (n/a)184.70 (n/a)150.90 (n/a)19.39 (n/a)

test_layer_norm[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)
1f35831 — 2026-07-09 23:36:090.05 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.00 (n/a)203.10 (n/a)180.26 (n/a)172.20 (n/a)169.70 (n/a)14.03 (n/a)

test_layer_norm[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)
1f35831 — 2026-07-09 23:36:090.05 (n/a)0.04 (n/a)0.05 (n/a)0.03 (n/a)0.01 (n/a)243.90 (n/a)188.18 (n/a)178.30 (n/a)153.10 (n/a)33.87 (n/a)

test_layer_norm[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)
1f35831 — 2026-07-09 23:36:090.06 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.01 (n/a)219.40 (n/a)170.64 (n/a)161.50 (n/a)145.40 (n/a)28.82 (n/a)

test_layer_norm[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)
1f35831 — 2026-07-09 23:36:090.05 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.00 (n/a)216.40 (n/a)182.92 (n/a)174.30 (n/a)168.60 (n/a)19.30 (n/a)

test_layer_norm[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)
1f35831 — 2026-07-09 23:36:090.04 (n/a)0.04 (n/a)0.04 (n/a)0.04 (n/a)0.00 (n/a)230.40 (n/a)219.06 (n/a)220.70 (n/a)194.70 (n/a)14.41 (n/a)
iron/operators/leaky_relu

test_leaky_relu[input_length_2048-num_aie_columns_1-num_channels_1-tile_size_2048-alpha_0.01]

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
1f35831 — 2026-07-09 23:36:090.05 (-7.09%)0.05 (+13.77%)0.05 (+11.29%)0.04 (+108.22%)0.00 (-70.68%)184.70 (-51.98%)170.92 (-21.07%)169.70 (-10.16%)154.50 (+7.67%)13.60 (-85.84%)
17769b8 — 2026-07-09 23:01:350.06 (n/a)0.04 (n/a)0.04 (n/a)0.02 (n/a)0.01 (n/a)384.60 (n/a)216.56 (n/a)188.90 (n/a)143.50 (n/a)96.03 (n/a)

test_leaky_relu[input_length_2048-num_aie_columns_1-num_channels_1-tile_size_2048-alpha_0.1]

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
1f35831 — 2026-07-09 23:36:090.06 (+7.32%)0.05 (-2.55%)0.04 (-7.46%)0.04 (+3.31%)0.01 (+24.66%)198.50 (-3.17%)175.68 (+3.26%)185.50 (+8.04%)133.70 (-6.83%)26.68 (+12.33%)
17769b8 — 2026-07-09 23:01:350.06 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.01 (n/a)205.00 (n/a)170.14 (n/a)171.70 (n/a)143.50 (n/a)23.75 (n/a)

test_leaky_relu[input_length_2048-num_aie_columns_1-num_channels_1-tile_size_2048-alpha_0.25]

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
1f35831 — 2026-07-09 23:36:090.05 (-20.68%)0.05 (+7.89%)0.05 (+19.82%)0.05 (+28.49%)0.00 (-71.22%)175.70 (-22.19%)161.60 (-11.74%)153.70 (-16.51%)151.50 (+26.04%)12.22 (-72.09%)
17769b8 — 2026-07-09 23:01:350.07 (n/a)0.05 (n/a)0.04 (n/a)0.04 (n/a)0.01 (n/a)225.80 (n/a)183.10 (n/a)184.10 (n/a)120.20 (n/a)43.79 (n/a)

test_leaky_relu[input_length_2048-num_aie_columns_1-num_channels_2-tile_size_1024-alpha_0.01]

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
1f35831 — 2026-07-09 23:36:090.05 (-17.29%)0.05 (-6.13%)0.05 (-3.17%)0.04 (+16.65%)0.01 (-52.48%)210.20 (-14.27%)181.40 (+3.12%)173.70 (+3.27%)157.00 (+20.86%)20.66 (-51.67%)
17769b8 — 2026-07-09 23:01:350.06 (n/a)0.05 (n/a)0.05 (n/a)0.03 (n/a)0.01 (n/a)245.20 (n/a)175.92 (n/a)168.20 (n/a)129.90 (n/a)42.76 (n/a)

test_leaky_relu[input_length_2048-num_aie_columns_2-num_channels_1-tile_size_1024-alpha_0.01]

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
1f35831 — 2026-07-09 23:36:090.05 (-17.38%)0.04 (-22.01%)0.04 (-14.74%)0.02 (-51.41%)0.01 (+79.89%)365.50 (+105.80%)221.80 (+38.83%)188.60 (+17.29%)158.70 (+21.05%)84.93 (+355.52%)
17769b8 — 2026-07-09 23:01:350.06 (n/a)0.05 (n/a)0.05 (n/a)0.05 (n/a)0.01 (n/a)177.60 (n/a)159.76 (n/a)160.80 (n/a)131.10 (n/a)18.65 (n/a)

test_leaky_relu[input_length_2048-num_aie_columns_2-num_channels_2-tile_size_512-alpha_0.01]

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
1f35831 — 2026-07-09 23:36:090.05 (-7.29%)0.04 (-7.17%)0.04 (-6.76%)0.04 (-2.65%)0.01 (-23.58%)219.60 (+2.76%)184.60 (+7.09%)184.60 (+7.26%)159.50 (+7.92%)22.90 (-13.99%)
17769b8 — 2026-07-09 23:01:350.06 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.01 (n/a)213.70 (n/a)172.38 (n/a)172.10 (n/a)147.80 (n/a)26.63 (n/a)

test_leaky_relu[input_length_2048-num_aie_columns_4-num_channels_1-tile_size_512-alpha_0.01]

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
1f35831 — 2026-07-09 23:36:090.05 (-5.94%)0.04 (-12.24%)0.04 (-17.15%)0.03 (-17.22%)0.01 (+29.77%)255.50 (+20.80%)208.14 (+15.81%)220.80 (+20.72%)156.20 (+6.33%)39.31 (+66.05%)
17769b8 — 2026-07-09 23:01:350.06 (n/a)0.05 (n/a)0.04 (n/a)0.04 (n/a)0.01 (n/a)211.50 (n/a)179.72 (n/a)182.90 (n/a)146.90 (n/a)23.68 (n/a)

test_leaky_relu[input_length_2048-num_aie_columns_4-num_channels_2-tile_size_256-alpha_0.01]

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
1f35831 — 2026-07-09 23:36:090.04 (-14.83%)0.04 (-12.30%)0.04 (-8.87%)0.03 (-16.67%)0.00 (-5.88%)237.90 (+20.03%)213.88 (+14.15%)209.00 (+9.77%)195.00 (+17.40%)17.63 (+33.00%)
17769b8 — 2026-07-09 23:01:350.05 (n/a)0.04 (n/a)0.04 (n/a)0.04 (n/a)0.00 (n/a)198.20 (n/a)187.36 (n/a)190.40 (n/a)166.10 (n/a)13.25 (n/a)

test_leaky_relu[input_length_2048-num_aie_columns_8-num_channels_1-tile_size_256-alpha_0.01]

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
1f35831 — 2026-07-09 23:36:090.06 (-15.45%)0.05 (+0.66%)0.04 (+6.01%)0.04 (+10.83%)0.01 (-35.05%)224.20 (-9.78%)185.38 (-3.94%)182.80 (-5.68%)136.30 (+18.32%)36.74 (-25.10%)
17769b8 — 2026-07-09 23:01:350.07 (n/a)0.05 (n/a)0.04 (n/a)0.03 (n/a)0.01 (n/a)248.50 (n/a)192.98 (n/a)193.80 (n/a)115.20 (n/a)49.05 (n/a)

test_leaky_relu[input_length_2048-num_aie_columns_8-num_channels_2-tile_size_128-alpha_0.01]

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
1f35831 — 2026-07-09 23:36:090.05 (-2.53%)0.04 (+0.76%)0.04 (+4.83%)0.03 (+5.95%)0.00 (-19.77%)238.60 (-5.62%)203.08 (-1.41%)196.80 (-4.60%)173.70 (+2.60%)24.34 (-22.16%)
17769b8 — 2026-07-09 23:01:350.05 (n/a)0.04 (n/a)0.04 (n/a)0.03 (n/a)0.01 (n/a)252.80 (n/a)205.98 (n/a)206.30 (n/a)169.30 (n/a)31.27 (n/a)
iron/operators/mem_copy

test_mem_copy[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)
1f35831 — 2026-07-09 23:36:090.07 (+14.18%)0.06 (+17.44%)0.06 (+18.83%)0.05 (+22.16%)0.01 (+7.88%)171.10 (-18.13%)139.62 (-15.22%)134.60 (-15.82%)110.80 (-12.41%)24.36 (-22.23%)
17769b8 — 2026-07-09 23:01:350.06 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.01 (n/a)209.00 (n/a)164.68 (n/a)159.90 (n/a)126.50 (n/a)31.33 (n/a)

test_mem_copy[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)
1f35831 — 2026-07-09 23:36:090.04 (-7.70%)0.03 (-21.34%)0.04 (-15.37%)0.02 (-39.16%)0.01 (+184.05%)336.10 (+64.35%)255.54 (+32.65%)230.50 (+18.14%)187.80 (+8.30%)61.81 (+422.69%)
17769b8 — 2026-07-09 23:01:350.05 (n/a)0.04 (n/a)0.04 (n/a)0.04 (n/a)0.00 (n/a)204.50 (n/a)192.64 (n/a)195.10 (n/a)173.40 (n/a)11.83 (n/a)

test_mem_copy[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)
1f35831 — 2026-07-09 23:36:090.06 (-2.14%)0.05 (-0.33%)0.05 (-8.55%)0.04 (+5.10%)0.01 (-0.90%)188.70 (-4.84%)168.18 (+0.27%)179.20 (+9.33%)141.00 (+2.17%)22.60 (-4.14%)
17769b8 — 2026-07-09 23:01:350.06 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.01 (n/a)198.30 (n/a)167.72 (n/a)163.90 (n/a)138.00 (n/a)23.58 (n/a)

test_mem_copy[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)
1f35831 — 2026-07-09 23:36:090.07 (-5.31%)0.05 (-9.44%)0.05 (-4.67%)0.04 (-14.84%)0.01 (+12.38%)209.30 (+17.45%)178.52 (+11.90%)176.20 (+4.88%)124.10 (+5.62%)34.46 (+43.09%)
17769b8 — 2026-07-09 23:01:350.07 (n/a)0.05 (n/a)0.05 (n/a)0.05 (n/a)0.01 (n/a)178.20 (n/a)159.54 (n/a)168.00 (n/a)117.50 (n/a)24.08 (n/a)

test_mem_copy[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)
1f35831 — 2026-07-09 23:36:090.06 (+0.16%)0.04 (-7.34%)0.04 (-21.60%)0.03 (+63.72%)0.01 (-37.65%)236.70 (-38.93%)192.90 (-3.32%)206.40 (+27.56%)140.40 (-0.14%)37.83 (-64.20%)
17769b8 — 2026-07-09 23:01:350.06 (n/a)0.05 (n/a)0.05 (n/a)0.02 (n/a)0.02 (n/a)387.60 (n/a)199.52 (n/a)161.80 (n/a)140.60 (n/a)105.69 (n/a)

test_mem_copy[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)
1f35831 — 2026-07-09 23:36:090.06 (+2.88%)0.05 (-3.14%)0.05 (+5.79%)0.04 (-21.35%)0.01 (+79.08%)207.90 (+27.16%)161.02 (+5.34%)151.40 (-5.43%)127.20 (-2.75%)30.34 (+125.73%)
17769b8 — 2026-07-09 23:01:350.06 (n/a)0.05 (n/a)0.05 (n/a)0.05 (n/a)0.01 (n/a)163.50 (n/a)152.86 (n/a)160.10 (n/a)130.80 (n/a)13.44 (n/a)

test_mem_copy[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)
1f35831 — 2026-07-09 23:36:090.05 (-16.72%)0.04 (-12.31%)0.05 (+1.07%)0.03 (-14.20%)0.01 (-11.15%)252.40 (+16.53%)201.30 (+14.51%)175.30 (-1.07%)162.40 (+20.12%)43.47 (+27.12%)
17769b8 — 2026-07-09 23:01:350.06 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.01 (n/a)216.60 (n/a)175.80 (n/a)177.20 (n/a)135.20 (n/a)34.20 (n/a)

test_mem_copy[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)
1f35831 — 2026-07-09 23:36:090.06 (-8.15%)0.04 (-2.09%)0.05 (+9.85%)0.03 (-4.09%)0.01 (-4.48%)290.70 (+4.27%)197.18 (+2.37%)170.90 (-9.00%)146.70 (+8.83%)60.78 (+7.57%)
17769b8 — 2026-07-09 23:01:350.06 (n/a)0.05 (n/a)0.04 (n/a)0.03 (n/a)0.01 (n/a)278.80 (n/a)192.62 (n/a)187.80 (n/a)134.80 (n/a)56.50 (n/a)
iron/operators/mha

test_mha[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)
1f35831 — 2026-07-09 23:36:090.21 (+0.09%)0.21 (+0.15%)0.21 (+0.21%)0.21 (+0.11%)0.00 (-15.81%)40914.60 (-0.11%)40865.30 (-0.15%)40855.90 (-0.21%)40821.20 (-0.09%)35.00 (-16.00%)
573678d — 2026-06-29 18:46:460.21 (n/a)0.20 (n/a)0.20 (n/a)0.20 (n/a)0.00 (n/a)40961.20 (n/a)40924.92 (n/a)40941.50 (n/a)40859.60 (n/a)41.66 (n/a)
iron/operators/rms_norm

test_rms_norm[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)
1f35831 — 2026-07-09 23:36:090.06 (-3.02%)0.05 (+11.77%)0.05 (+14.58%)0.04 (+20.05%)0.01 (-25.19%)184.10 (-16.66%)160.60 (-12.21%)167.10 (-12.74%)129.70 (+3.10%)24.42 (-33.38%)
17769b8 — 2026-07-09 23:01:350.07 (n/a)0.05 (n/a)0.04 (n/a)0.04 (n/a)0.01 (n/a)220.90 (n/a)182.94 (n/a)191.50 (n/a)125.80 (n/a)36.66 (n/a)

test_rms_norm[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)
1f35831 — 2026-07-09 23:36:090.08 (-5.63%)0.07 (-5.24%)0.07 (-0.16%)0.06 (+0.60%)0.01 (-9.39%)211.30 (-0.61%)183.34 (+5.30%)177.00 (+0.11%)148.60 (+5.92%)25.50 (-3.19%)
17769b8 — 2026-07-09 23:01:350.09 (n/a)0.07 (n/a)0.07 (n/a)0.06 (n/a)0.01 (n/a)212.60 (n/a)174.12 (n/a)176.80 (n/a)140.30 (n/a)26.34 (n/a)

test_rms_norm[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)
1f35831 — 2026-07-09 23:36:090.04 (-31.45%)0.04 (-15.27%)0.04 (-4.22%)0.04 (+9.49%)0.00 (-83.05%)205.40 (-8.67%)195.38 (+12.33%)196.60 (+4.41%)183.20 (+45.86%)9.96 (-76.64%)
17769b8 — 2026-07-09 23:01:350.07 (n/a)0.05 (n/a)0.04 (n/a)0.04 (n/a)0.01 (n/a)224.90 (n/a)173.94 (n/a)188.30 (n/a)125.60 (n/a)42.65 (n/a)

test_rms_norm[input_length_2048-num_aie_columns_1-num_channels_2-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)
1f35831 — 2026-07-09 23:36:090.06 (+12.48%)0.06 (+16.93%)0.06 (+22.53%)0.05 (+6.71%)0.01 (+28.99%)227.40 (-6.30%)177.38 (-14.06%)166.40 (-18.39%)160.80 (-11.11%)28.09 (+10.52%)
17769b8 — 2026-07-09 23:01:350.06 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.01 (n/a)242.70 (n/a)206.40 (n/a)203.90 (n/a)180.90 (n/a)25.41 (n/a)

test_rms_norm[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)
1f35831 — 2026-07-09 23:36:090.06 (-18.97%)0.04 (-18.08%)0.04 (-16.65%)0.04 (-12.72%)0.01 (-25.83%)213.60 (+14.53%)189.20 (+21.30%)193.50 (+19.96%)140.40 (+23.37%)28.83 (+4.56%)
17769b8 — 2026-07-09 23:01:350.07 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.01 (n/a)186.50 (n/a)155.98 (n/a)161.30 (n/a)113.80 (n/a)27.57 (n/a)

test_rms_norm[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)
1f35831 — 2026-07-09 23:36:090.07 (+0.64%)0.05 (-7.95%)0.05 (-20.77%)0.04 (+45.99%)0.01 (-36.04%)244.10 (-31.51%)196.62 (+0.34%)198.50 (+26.27%)149.60 (-0.60%)37.80 (-57.90%)
17769b8 — 2026-07-09 23:01:350.07 (n/a)0.06 (n/a)0.07 (n/a)0.03 (n/a)0.02 (n/a)356.40 (n/a)195.96 (n/a)157.20 (n/a)150.50 (n/a)89.79 (n/a)

test_rms_norm[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)
1f35831 — 2026-07-09 23:36:090.05 (-13.06%)0.04 (-11.15%)0.04 (-10.46%)0.04 (-6.00%)0.01 (-27.39%)233.30 (+6.38%)198.48 (+11.47%)198.60 (+11.70%)155.80 (+15.07%)27.80 (-13.01%)
17769b8 — 2026-07-09 23:01:350.06 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.01 (n/a)219.30 (n/a)178.06 (n/a)177.80 (n/a)135.40 (n/a)31.96 (n/a)

test_rms_norm[input_length_2048-num_aie_columns_2-num_channels_2-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)
1f35831 — 2026-07-09 23:36:090.05 (-24.94%)0.05 (-7.69%)0.05 (-2.57%)0.04 (-7.02%)0.01 (-42.13%)246.60 (+7.54%)193.80 (+6.30%)191.50 (+2.68%)168.00 (+33.23%)32.03 (-14.30%)
17769b8 — 2026-07-09 23:01:350.07 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.01 (n/a)229.30 (n/a)182.32 (n/a)186.50 (n/a)126.10 (n/a)37.38 (n/a)

test_rms_norm[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)
1f35831 — 2026-07-09 23:36:090.06 (+3.58%)0.05 (+1.65%)0.04 (+6.35%)0.03 (-12.01%)0.01 (+18.76%)255.20 (+13.67%)191.68 (+0.29%)191.30 (-6.00%)127.10 (-3.42%)49.02 (+30.69%)
17769b8 — 2026-07-09 23:01:350.06 (n/a)0.04 (n/a)0.04 (n/a)0.04 (n/a)0.01 (n/a)224.50 (n/a)191.12 (n/a)203.50 (n/a)131.60 (n/a)37.51 (n/a)

test_rms_norm[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)
1f35831 — 2026-07-09 23:36:090.08 (+18.72%)0.06 (+9.92%)0.06 (+16.82%)0.04 (-5.77%)0.01 (+83.68%)220.80 (+6.15%)168.26 (-6.10%)153.00 (-14.43%)122.80 (-15.78%)41.21 (+68.18%)
17769b8 — 2026-07-09 23:01:350.06 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.01 (n/a)208.00 (n/a)179.20 (n/a)178.80 (n/a)145.80 (n/a)24.50 (n/a)

test_rms_norm[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)
1f35831 — 2026-07-09 23:36:090.05 (-5.36%)0.04 (-2.14%)0.04 (-3.41%)0.03 (-3.37%)0.01 (-10.45%)241.30 (+3.47%)195.00 (+1.93%)196.30 (+3.53%)164.80 (+5.71%)30.12 (-2.26%)
17769b8 — 2026-07-09 23:01:350.05 (n/a)0.04 (n/a)0.04 (n/a)0.04 (n/a)0.01 (n/a)233.20 (n/a)191.30 (n/a)189.60 (n/a)155.90 (n/a)30.82 (n/a)

test_rms_norm[input_length_2048-num_aie_columns_4-num_channels_2-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)
1f35831 — 2026-07-09 23:36:090.05 (+15.32%)0.05 (+6.83%)0.04 (+2.67%)0.04 (-0.17%)0.01 (+96.85%)222.00 (+0.18%)192.20 (-5.73%)197.70 (-2.61%)164.60 (-13.32%)21.73 (+70.82%)
17769b8 — 2026-07-09 23:01:350.05 (n/a)0.04 (n/a)0.04 (n/a)0.04 (n/a)0.00 (n/a)221.60 (n/a)203.88 (n/a)203.00 (n/a)189.90 (n/a)12.72 (n/a)

test_rms_norm[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)
1f35831 — 2026-07-09 23:36:090.05 (+9.94%)0.04 (-4.06%)0.04 (+0.36%)0.03 (-31.53%)0.01 (+263.59%)290.40 (+46.00%)204.02 (+8.67%)187.60 (-0.37%)156.60 (-9.01%)51.98 (+398.20%)
17769b8 — 2026-07-09 23:01:350.05 (n/a)0.04 (n/a)0.04 (n/a)0.04 (n/a)0.00 (n/a)198.90 (n/a)187.74 (n/a)188.30 (n/a)172.10 (n/a)10.43 (n/a)

test_rms_norm[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)
1f35831 — 2026-07-09 23:36:090.06 (+4.61%)0.05 (-0.35%)0.05 (-7.38%)0.04 (+9.99%)0.01 (-14.60%)203.50 (-9.07%)187.72 (-0.16%)189.20 (+7.93%)157.40 (-4.43%)18.70 (-26.00%)
17769b8 — 2026-07-09 23:01:350.05 (n/a)0.05 (n/a)0.05 (n/a)0.04 (n/a)0.01 (n/a)223.80 (n/a)188.02 (n/a)175.30 (n/a)164.70 (n/a)25.27 (n/a)

test_rms_norm[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)
1f35831 — 2026-07-09 23:36:090.05 (+1.88%)0.04 (-10.22%)0.04 (-10.24%)0.03 (-11.94%)0.01 (+25.87%)281.60 (+13.55%)226.92 (+12.67%)226.80 (+11.39%)168.70 (-1.86%)40.96 (+37.54%)
17769b8 — 2026-07-09 23:01:350.05 (n/a)0.04 (n/a)0.04 (n/a)0.03 (n/a)0.01 (n/a)248.00 (n/a)201.40 (n/a)203.60 (n/a)171.90 (n/a)29.78 (n/a)
iron/operators/rope

test_rope[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)
1f35831 — 2026-07-09 23:36:090.64 (-12.00%)0.57 (-10.64%)0.58 (-12.94%)0.50 (-9.86%)0.06 (-32.86%)198.30 (+10.97%)173.26 (+11.21%)168.30 (+14.80%)153.40 (+13.63%)17.44 (-16.72%)
17769b8 — 2026-07-09 23:01:350.73 (n/a)0.64 (n/a)0.67 (n/a)0.55 (n/a)0.08 (n/a)178.70 (n/a)155.80 (n/a)146.60 (n/a)135.00 (n/a)20.95 (n/a)

test_rope[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)
1f35831 — 2026-07-09 23:36:090.90 (+37.96%)0.66 (+12.32%)0.59 (+2.32%)0.57 (+12.72%)0.14 (+113.99%)173.80 (-11.28%)152.82 (-9.19%)165.30 (-2.30%)108.70 (-27.53%)26.81 (+38.87%)
17769b8 — 2026-07-09 23:01:350.66 (n/a)0.59 (n/a)0.58 (n/a)0.50 (n/a)0.07 (n/a)195.90 (n/a)168.28 (n/a)169.20 (n/a)150.00 (n/a)19.31 (n/a)

test_rope[rows_32-cols_512-angle_rows_32-aie_columns_4-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)
1f35831 — 2026-07-09 23:36:090.73 (-3.84%)0.58 (-3.40%)0.55 (-0.55%)0.46 (-9.59%)0.10 (+2.91%)214.10 (+10.59%)173.76 (+3.95%)177.50 (+0.57%)135.40 (+3.99%)29.33 (+19.14%)
17769b8 — 2026-07-09 23:01:350.76 (n/a)0.60 (n/a)0.56 (n/a)0.51 (n/a)0.10 (n/a)193.60 (n/a)167.16 (n/a)176.50 (n/a)130.20 (n/a)24.62 (n/a)

test_rope[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)
1f35831 — 2026-07-09 23:36:090.70 (+13.15%)0.51 (-2.90%)0.52 (+0.59%)0.39 (-18.85%)0.13 (+135.67%)251.60 (+23.21%)200.44 (+7.23%)188.40 (-0.58%)141.20 (-11.64%)48.30 (+168.11%)
17769b8 — 2026-07-09 23:01:350.62 (n/a)0.53 (n/a)0.52 (n/a)0.48 (n/a)0.05 (n/a)204.20 (n/a)186.92 (n/a)189.50 (n/a)159.80 (n/a)18.02 (n/a)

test_rope[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)
1f35831 — 2026-07-09 23:36:090.52 (+8.63%)0.45 (+8.00%)0.50 (+22.30%)0.27 (-27.51%)0.11 (+157.07%)277.00 (+37.95%)175.82 (-2.12%)146.20 (-18.23%)142.10 (-7.97%)57.44 (+236.14%)
17769b8 — 2026-07-09 23:01:350.48 (n/a)0.41 (n/a)0.41 (n/a)0.37 (n/a)0.04 (n/a)200.80 (n/a)179.62 (n/a)178.80 (n/a)154.40 (n/a)17.09 (n/a)

test_rope[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)
1f35831 — 2026-07-09 23:36:090.50 (-3.23%)0.46 (-0.40%)0.47 (+4.16%)0.41 (-5.77%)0.04 (+32.18%)178.30 (+6.13%)161.62 (+0.72%)157.40 (-4.02%)147.80 (+3.36%)14.97 (+45.92%)
17769b8 — 2026-07-09 23:01:350.52 (n/a)0.46 (n/a)0.45 (n/a)0.44 (n/a)0.03 (n/a)168.00 (n/a)160.46 (n/a)164.00 (n/a)143.00 (n/a)10.26 (n/a)

test_rope[rows_32-cols_512-angle_rows_8-aie_columns_4-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)
1f35831 — 2026-07-09 23:36:090.58 (+5.08%)0.47 (+5.26%)0.49 (+10.27%)0.37 (+7.26%)0.09 (+19.16%)198.30 (-6.77%)161.66 (-4.43%)151.70 (-9.32%)127.90 (-4.84%)30.27 (+7.03%)
17769b8 — 2026-07-09 23:01:350.55 (n/a)0.45 (n/a)0.44 (n/a)0.35 (n/a)0.07 (n/a)212.70 (n/a)169.16 (n/a)167.30 (n/a)134.40 (n/a)28.28 (n/a)

test_rope[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)
1f35831 — 2026-07-09 23:36:090.47 (-11.54%)0.41 (-0.90%)0.39 (-3.14%)0.36 (+9.13%)0.04 (-42.28%)202.10 (-8.34%)183.52 (-0.57%)186.90 (+3.26%)158.50 (+13.05%)18.40 (-38.98%)
17769b8 — 2026-07-09 23:01:350.53 (n/a)0.41 (n/a)0.41 (n/a)0.33 (n/a)0.07 (n/a)220.50 (n/a)184.58 (n/a)181.00 (n/a)140.20 (n/a)30.16 (n/a)
iron/operators/softmax

test_softmax[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)
1f35831 — 2026-07-09 23:36:091.00 (+18.96%)0.83 (+11.48%)0.92 (+23.18%)0.62 (-1.83%)0.18 (+117.14%)212.60 (+1.87%)164.04 (-7.58%)142.20 (-18.84%)131.30 (-15.94%)37.73 (+85.60%)
17769b8 — 2026-07-09 23:01:350.84 (n/a)0.75 (n/a)0.75 (n/a)0.63 (n/a)0.08 (n/a)208.70 (n/a)177.50 (n/a)175.20 (n/a)156.20 (n/a)20.33 (n/a)

test_softmax[input_length_32768-num_aie_columns_2-num_channels_2-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)
1f35831 — 2026-07-09 23:36:090.90 (-5.20%)0.77 (+5.17%)0.76 (+7.26%)0.69 (+26.26%)0.08 (-51.36%)189.40 (-20.82%)170.70 (-8.09%)173.40 (-6.77%)146.20 (+5.48%)17.43 (-58.92%)
17769b8 — 2026-07-09 23:01:350.95 (n/a)0.74 (n/a)0.70 (n/a)0.55 (n/a)0.17 (n/a)239.20 (n/a)185.72 (n/a)186.00 (n/a)138.60 (n/a)42.43 (n/a)

test_softmax[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)
1f35831 — 2026-07-09 23:36:090.90 (+21.48%)0.76 (+20.18%)0.78 (+27.54%)0.56 (+12.84%)0.13 (+35.59%)232.20 (-11.37%)176.40 (-16.21%)168.00 (-21.61%)145.70 (-17.68%)33.84 (+1.22%)
17769b8 — 2026-07-09 23:01:350.74 (n/a)0.63 (n/a)0.61 (n/a)0.50 (n/a)0.09 (n/a)262.00 (n/a)210.52 (n/a)214.30 (n/a)177.00 (n/a)33.43 (n/a)
iron/operators/swiglu_decode

test_swiglu_decode[embedding_dim_1024-hidden_dim_3584]

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
1f35831 — 2026-07-09 23:36:090.00 (+2.33%)0.00 (+1.90%)0.00 (+0.00%)0.00 (+7.89%)0.00 (-51.01%)1004.46 (-6.91%)956.20 (-2.30%)949.55 (-0.33%)926.41 (-2.53%)28.87 (-48.66%)
17769b8 — 2026-07-09 23:01:350.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)1079.03 (n/a)978.73 (n/a)952.71 (n/a)950.43 (n/a)56.23 (n/a)

test_swiglu_decode[embedding_dim_2048-hidden_dim_2048]

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
1f35831 — 2026-07-09 23:36:090.01 (+1.22%)0.01 (+2.76%)0.01 (+1.23%)0.01 (+5.19%)0.00 (-53.87%)1006.40 (-5.37%)997.95 (-2.69%)1003.06 (-0.99%)987.44 (-0.99%)9.64 (-64.58%)
17769b8 — 2026-07-09 23:01:350.01 (n/a)0.01 (n/a)0.01 (n/a)0.01 (n/a)0.00 (n/a)1063.48 (n/a)1025.53 (n/a)1013.10 (n/a)997.32 (n/a)27.23 (n/a)
iron/operators/swiglu_prefill

test_swiglu_prefill[seq_len_256-embedding_dim_2048-hidden_dim_2048-prio_accuracy_False]

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
1f35831 — 2026-07-09 23:36:090.97 (+2.36%)0.95 (-0.12%)0.94 (-0.72%)0.94 (-0.99%)0.01 (+510.57%)2238.90 (+1.01%)2210.54 (+0.14%)2221.74 (+0.72%)2152.08 (-2.30%)33.91 (+501.16%)
17769b8 — 2026-07-09 23:01:350.95 (n/a)0.95 (n/a)0.95 (n/a)0.95 (n/a)0.00 (n/a)2216.53 (n/a)2207.36 (n/a)2205.81 (n/a)2202.82 (n/a)5.64 (n/a)
iron/operators/transpose

test_transpose[M_2048-N_64-aie_columns_1-channels_1-m_64-n_64-s_8-num_batches_1]

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
1f35831 — 2026-07-09 23:36:093.73 (+21.89%)2.87 (+3.50%)2.74 (+3.16%)2.50 (-3.23%)0.51 (+131.72%)210.00 (+3.35%)186.62 (-1.77%)191.60 (-3.09%)140.50 (-17.98%)28.19 (+94.65%)
17769b8 — 2026-07-09 23:01:353.06 (n/a)2.77 (n/a)2.65 (n/a)2.58 (n/a)0.22 (n/a)203.20 (n/a)189.98 (n/a)197.70 (n/a)171.30 (n/a)14.48 (n/a)

test_transpose[M_2048-N_64-aie_columns_1-channels_1-m_64-n_64-s_8-num_batches_2]

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
1f35831 — 2026-07-09 23:36:095.36 (-14.13%)4.93 (+9.37%)5.17 (+20.27%)4.22 (+16.28%)0.50 (-50.91%)248.40 (-13.99%)214.58 (-10.83%)202.70 (-16.82%)195.80 (+16.48%)22.87 (-49.18%)
17769b8 — 2026-07-09 23:01:356.24 (n/a)4.51 (n/a)4.30 (n/a)3.63 (n/a)1.01 (n/a)288.80 (n/a)240.64 (n/a)243.70 (n/a)168.10 (n/a)45.00 (n/a)

test_transpose[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)
4bb8427 — 2026-06-23 23:08:223.72 (+20.32%)3.12 (+16.02%)3.04 (+6.32%)2.73 (+32.06%)0.43 (+8.15%)192.20 (-24.30%)170.32 (-14.26%)172.50 (-5.94%)141.00 (-16.86%)22.57 (-32.31%)
4d4b803 — 2026-06-22 17:54:573.09 (n/a)2.69 (n/a)2.86 (n/a)2.07 (n/a)0.40 (n/a)253.90 (n/a)198.64 (n/a)183.40 (n/a)169.60 (n/a)33.34 (n/a)

test_transpose[M_2048-N_64-aie_columns_1-channels_2-m_64-n_64-s_8-num_batches_1]

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
1f35831 — 2026-07-09 23:36:093.43 (+13.75%)2.90 (+14.04%)2.82 (+13.99%)2.26 (+0.82%)0.50 (+71.10%)231.90 (-0.81%)185.18 (-10.99%)186.00 (-12.31%)152.90 (-12.08%)33.15 (+46.67%)
17769b8 — 2026-07-09 23:01:353.01 (n/a)2.55 (n/a)2.47 (n/a)2.24 (n/a)0.29 (n/a)233.80 (n/a)208.04 (n/a)212.10 (n/a)173.90 (n/a)22.60 (n/a)

test_transpose[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)
4bb8427 — 2026-06-23 23:08:224.36 (+7.95%)3.26 (+13.04%)3.33 (+22.37%)2.38 (+18.97%)0.79 (-14.92%)220.70 (-15.96%)168.40 (-14.68%)157.40 (-18.28%)120.30 (-7.39%)40.78 (-34.38%)
4d4b803 — 2026-06-22 17:54:574.04 (n/a)2.89 (n/a)2.72 (n/a)2.00 (n/a)0.93 (n/a)262.60 (n/a)197.38 (n/a)192.60 (n/a)129.90 (n/a)62.14 (n/a)
Krackan - Examples

IRON

Tested on 2026_07_09_22_59_02 at commit 1f35831.

iron/applications/llama_3.2_1b
TestChecksTTFT (mean)TPS (mean)
test_llama_3_2_1b[llama_3.2_1b_prompt_1024_tokens_1]✅ 5/52.13n/a
test_llama_3_2_1b[llama_3.2_1b_prompt_1024_tokens_40]✅ 5/52.177.41
test_llama_3_2_1b[llama_3.2_1b_prompt_13_tokens_1]✅ 5/52.10n/a
test_llama_3_2_1b[llama_3.2_1b_prompt_13_tokens_40]✅ 5/52.097.22

Trends:

IRON Trends

iron/applications/llama_3.2_1b

test_llama_3_2_1b[llama_3.2_1b_prompt_1024_tokens_1]

Commit/Date TTFT (max)TTFT (mean)TTFT (median)TTFT (min)TTFT (stddev)
1f35831 — 2026-07-09 22:53:532.15 (+0.51%)2.13 (+0.47%)2.13 (+0.38%)2.12 (+0.52%)0.01 (+18.16%)
17769b8 — 2026-07-09 22:38:402.14 (n/a)2.12 (n/a)2.12 (n/a)2.11 (n/a)0.01 (n/a)

test_llama_3_2_1b[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)
1f35831 — 2026-07-09 22:53:537.43 (+0.00%)7.41 (+0.25%)7.42 (+0.52%)7.38 (+0.09%)0.02 (-26.08%)2.28 (+0.31%)2.17 (+0.68%)2.15 (+1.08%)2.12 (+0.28%)0.06 (-5.09%)
17769b8 — 2026-07-09 22:38:407.43 (n/a)7.39 (n/a)7.38 (n/a)7.37 (n/a)0.03 (n/a)2.28 (n/a)2.16 (n/a)2.13 (n/a)2.12 (n/a)0.07 (n/a)

test_llama_3_2_1b[llama_3.2_1b_prompt_13_tokens_1]

Commit/Date TTFT (max)TTFT (mean)TTFT (median)TTFT (min)TTFT (stddev)
1f35831 — 2026-07-09 22:53:532.10 (-0.19%)2.10 (+0.74%)2.10 (+0.53%)2.09 (+1.60%)0.00 (-77.95%)
17769b8 — 2026-07-09 22:38:402.11 (n/a)2.08 (n/a)2.09 (n/a)2.06 (n/a)0.02 (n/a)

test_llama_3_2_1b[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)
1f35831 — 2026-07-09 22:53:537.27 (+0.26%)7.22 (-0.20%)7.21 (-0.39%)7.20 (-0.33%)0.03 (+212.87%)2.10 (-0.14%)2.09 (-0.10%)2.08 (-0.71%)2.08 (+0.44%)0.01 (-32.30%)
17769b8 — 2026-07-09 22:38:407.25 (n/a)7.24 (n/a)7.24 (n/a)7.22 (n/a)0.01 (n/a)2.10 (n/a)2.09 (n/a)2.10 (n/a)2.07 (n/a)0.02 (n/a)
Phoenix - Small

IRON

Tested on 2026_07_09_23_09_31 at commit 1f35831.

iron/operators/axpy
TestChecksLatency (mean)Bandwidth (mean)Throughput (mean)
test_axpy[input_length_2048-num_aie_columns_1-tile_size_2048-scalar_factor_3.0]✅ 5/5387.680.04n/a
test_axpy[input_length_2048-num_aie_columns_2-tile_size_1024-scalar_factor_3.0]✅ 5/5334.780.04n/a
test_axpy[input_length_2048-num_aie_columns_4-tile_size_512-scalar_factor_3.0]✅ 5/5379.660.04n/a
iron/operators/dequant
TestChecksLatency (mean)Bandwidth (mean)Throughput (mean)
test_dequant[input_length_2048-num_aie_columns_1-num_channels_1-tile_size_2048-group_size_32]✅ 5/5386.680.01n/a
test_dequant[input_length_2048-num_aie_columns_1-num_channels_2-tile_size_1024-group_size_32]✅ 5/5434.060.01n/a
test_dequant[input_length_2048-num_aie_columns_2-num_channels_1-tile_size_1024-group_size_32]✅ 5/5623.080.02n/a
test_dequant[input_length_2048-num_aie_columns_2-num_channels_2-tile_size_512-group_size_32]✅ 5/5335.580.02n/a
test_dequant[input_length_2048-num_aie_columns_4-num_channels_1-tile_size_512-group_size_32]✅ 5/5461.320.01n/a
test_dequant[input_length_2048-num_aie_columns_4-num_channels_2-tile_size_256-group_size_32]✅ 5/5437.740.01n/a
iron/operators/elementwise_add
TestChecksLatency (mean)Bandwidth (mean)Throughput (mean)
test_elementwise_add[input_length_2048-num_aie_columns_1-tile_size_2048]✅ 5/5475.120.03n/a
test_elementwise_add[input_length_2048-num_aie_columns_2-tile_size_1024]✅ 5/5454.700.03n/a
test_elementwise_add[input_length_2048-num_aie_columns_4-tile_size_512]✅ 5/5471.900.03n/a
iron/operators/elementwise_mul
TestChecksLatency (mean)Bandwidth (mean)Throughput (mean)
test_elementwise_mul[input_length_2048-num_aie_columns_1-tile_size_2048]✅ 5/5302.900.04n/a
test_elementwise_mul[input_length_2048-num_aie_columns_2-tile_size_1024]✅ 5/5688.780.03n/a
test_elementwise_mul[input_length_2048-num_aie_columns_4-tile_size_512]✅ 5/5311.580.04n/a
iron/operators/gelu
TestChecksLatency (mean)Bandwidth (mean)Throughput (mean)
test_gelu[input_length_2048-num_aie_columns_1-num_channels_1-tile_size_2048]✅ 5/5428.540.02n/a
test_gelu[input_length_2048-num_aie_columns_1-num_channels_2-tile_size_1024]✅ 5/5334.840.03n/a
test_gelu[input_length_2048-num_aie_columns_2-num_channels_1-tile_size_1024]✅ 5/5444.660.02n/a
test_gelu[input_length_2048-num_aie_columns_2-num_channels_2-tile_size_512]✅ 5/5435.000.02n/a
test_gelu[input_length_2048-num_aie_columns_4-num_channels_1-tile_size_512]✅ 5/5430.520.02n/a
test_gelu[input_length_2048-num_aie_columns_4-num_channels_2-tile_size_256]✅ 5/5422.120.02n/a
iron/operators/gemm
TestChecksLatency (mean)Bandwidth (mean)Throughput (mean)
test_gemm[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-partition_N_1]✅ 5/5467.740.4820.55
test_gemm[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-partition_N_1]✅ 5/5482.340.4920.77
test_gemm[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-partition_N_1]✅ 5/583029.960.30206.99
test_gemm[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-partition_N_1]✅ 5/524836.621.01691.89
test_gemm[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-partition_N_1]✅ 5/54901.041.99521.11
test_gemm[M_64-K_512-N_256-num_aie_columns_4-b_col_maj_True-c_col_maj_False-m_16-k_64-n_64-trace_size_0-partition_N_4]✅ 5/55894.820.2312.32
iron/operators/gemv
TestChecksLatency (mean)Bandwidth (mean)Throughput (mean)
test_gemv[M_128-K_128-num_aie_columns_1-tile_size_input_32-tile_size_output_128]✅ 5/5n/a0.090.08
test_gemv[M_2048-K_8192-num_aie_columns_1-tile_size_input_1-tile_size_output_2048]✅ 5/5n/a3.643.64
test_gemv[M_2048-K_8192-num_aie_columns_2-tile_size_input_1-tile_size_output_1024]✅ 5/5n/a6.426.41
test_gemv[M_2048-K_8192-num_aie_columns_4-tile_size_input_1-tile_size_output_512]✅ 5/5n/a10.6810.68
test_gemv[M_8192-K_2048-num_aie_columns_1-tile_size_input_4-tile_size_output_1024]✅ 5/5n/a3.663.65
test_gemv[M_8192-K_2048-num_aie_columns_2-tile_size_input_4-tile_size_output_1024]✅ 5/5n/a5.935.93
test_gemv[M_8192-K_2048-num_aie_columns_4-tile_size_input_4-tile_size_output_1024]✅ 5/5n/a9.629.61
iron/operators/layer_norm
TestChecksLatency (mean)Bandwidth (mean)Throughput (mean)
test_layer_norm[input_length_2048-num_aie_columns_1-num_channels_1-tile_size_2048]✅ 5/5344.000.03n/a
test_layer_norm[input_length_2048-num_aie_columns_1-num_channels_2-tile_size_1024]✅ 5/5348.340.03n/a
test_layer_norm[input_length_2048-num_aie_columns_2-num_channels_1-tile_size_1024]✅ 5/5400.140.02n/a
test_layer_norm[input_length_2048-num_aie_columns_2-num_channels_2-tile_size_512]✅ 5/5511.480.02n/a
test_layer_norm[input_length_2048-num_aie_columns_4-num_channels_1-tile_size_512]✅ 5/5418.300.02n/a
test_layer_norm[input_length_2048-num_aie_columns_4-num_channels_2-tile_size_256]✅ 5/5443.560.02n/a
iron/operators/leaky_relu
TestChecksLatency (mean)Bandwidth (mean)Throughput (mean)
test_leaky_relu[input_length_2048-num_aie_columns_1-num_channels_1-tile_size_2048-alpha_0.01]✅ 5/5352.840.03n/a
test_leaky_relu[input_length_2048-num_aie_columns_1-num_channels_1-tile_size_2048-alpha_0.1]✅ 5/5392.100.02n/a
test_leaky_relu[input_length_2048-num_aie_columns_1-num_channels_1-tile_size_2048-alpha_0.25]✅ 5/5356.940.03n/a
test_leaky_relu[input_length_2048-num_aie_columns_1-num_channels_2-tile_size_1024-alpha_0.01]✅ 5/5697.920.02n/a
test_leaky_relu[input_length_2048-num_aie_columns_2-num_channels_1-tile_size_1024-alpha_0.01]✅ 5/5318.500.03n/a
test_leaky_relu[input_length_2048-num_aie_columns_2-num_channels_2-tile_size_512-alpha_0.01]✅ 5/5369.160.03n/a
test_leaky_relu[input_length_2048-num_aie_columns_4-num_channels_1-tile_size_512-alpha_0.01]✅ 5/5428.440.02n/a
test_leaky_relu[input_length_2048-num_aie_columns_4-num_channels_2-tile_size_256-alpha_0.01]✅ 5/5475.200.02n/a
iron/operators/mem_copy
TestChecksLatency (mean)Bandwidth (mean)Throughput (mean)
test_mem_copy[input_length_2048-num_cores_1-num_channels_1-bypass_False-tile_size_2048]✅ 5/5369.820.02n/a
test_mem_copy[input_length_2048-num_cores_2-num_channels_1-bypass_False-tile_size_1024]✅ 5/5394.420.02n/a
test_mem_copy[input_length_2048-num_cores_2-num_channels_2-bypass_False-tile_size_1024]✅ 5/5365.720.02n/a
test_mem_copy[input_length_2048-num_cores_4-num_channels_1-bypass_False-tile_size_512]✅ 5/5541.880.02n/a
test_mem_copy[input_length_2048-num_cores_4-num_channels_2-bypass_False-tile_size_512]✅ 5/5478.320.02n/a
test_mem_copy[input_length_2048-num_cores_8-num_channels_2-bypass_False-tile_size_256]✅ 5/5489.600.02n/a
iron/operators/relu
TestChecksLatency (mean)Bandwidth (mean)Throughput (mean)
test_relu[input_length_2048-num_aie_columns_1-num_channels_1-tile_size_2048]✅ 5/5421.400.02n/a
test_relu[input_length_2048-num_aie_columns_1-num_channels_2-tile_size_1024]✅ 5/5347.360.03n/a
test_relu[input_length_2048-num_aie_columns_2-num_channels_1-tile_size_1024]✅ 5/5355.380.03n/a
test_relu[input_length_2048-num_aie_columns_2-num_channels_2-tile_size_512]✅ 5/5333.380.03n/a
test_relu[input_length_2048-num_aie_columns_4-num_channels_1-tile_size_512]✅ 5/5878.700.01n/a
test_relu[input_length_2048-num_aie_columns_4-num_channels_2-tile_size_256]✅ 5/5546.120.02n/a
iron/operators/rms_norm
TestChecksLatency (mean)Bandwidth (mean)Throughput (mean)
test_rms_norm[input_length_2048-num_aie_columns_1-num_channels_1-tile_size_2048-weighted_False]✅ 5/5358.740.03n/a
test_rms_norm[input_length_2048-num_aie_columns_1-num_channels_1-tile_size_2048-weighted_True]✅ 5/5680.140.03n/a
test_rms_norm[input_length_2048-num_aie_columns_1-num_channels_2-tile_size_1024-weighted_False]✅ 5/5441.220.02n/a
test_rms_norm[input_length_2048-num_aie_columns_1-num_channels_2-tile_size_1024-weighted_True]✅ 5/5447.100.03n/a
test_rms_norm[input_length_2048-num_aie_columns_2-num_channels_1-tile_size_1024-weighted_False]✅ 5/5661.160.02n/a
test_rms_norm[input_length_2048-num_aie_columns_2-num_channels_1-tile_size_1024-weighted_True]✅ 5/5363.980.03n/a
test_rms_norm[input_length_2048-num_aie_columns_2-num_channels_2-tile_size_512-weighted_False]✅ 5/5394.980.02n/a
test_rms_norm[input_length_2048-num_aie_columns_2-num_channels_2-tile_size_512-weighted_True]✅ 5/5429.320.02n/a
test_rms_norm[input_length_2048-num_aie_columns_4-num_channels_1-tile_size_512-weighted_False]✅ 5/5362.840.02n/a
test_rms_norm[input_length_2048-num_aie_columns_4-num_channels_1-tile_size_512-weighted_True]✅ 5/5726.340.02n/a
test_rms_norm[input_length_2048-num_aie_columns_4-num_channels_2-tile_size_256-weighted_False]✅ 5/5440.060.02n/a
iron/operators/rope
TestChecksLatency (mean)Bandwidth (mean)Throughput (mean)
test_rope[rows_32-cols_512-angle_rows_32-aie_columns_1-method_type_0]✅ 5/5382.660.30n/a
test_rope[rows_32-cols_512-angle_rows_32-aie_columns_2-method_type_0]✅ 5/5356.800.31n/a
test_rope[rows_32-cols_512-angle_rows_32-aie_columns_4-method_type_0]✅ 5/5364.880.29n/a
test_rope[rows_32-cols_512-angle_rows_8-aie_columns_1-method_type_0]✅ 5/5378.180.21n/a
test_rope[rows_32-cols_512-angle_rows_8-aie_columns_2-method_type_0]✅ 5/5337.500.24n/a
test_rope[rows_32-cols_512-angle_rows_8-aie_columns_4-method_type_0]✅ 5/5380.240.23n/a
iron/operators/sigmoid
TestChecksLatency (mean)Bandwidth (mean)Throughput (mean)
test_sigmoid[input_length_2048-num_aie_columns_1-num_channels_1-tile_size_2048]✅ 5/5320.640.03n/a
test_sigmoid[input_length_2048-num_aie_columns_1-num_channels_2-tile_size_1024]✅ 5/5368.660.03n/a
test_sigmoid[input_length_2048-num_aie_columns_2-num_channels_1-tile_size_1024]✅ 5/5657.120.02n/a
test_sigmoid[input_length_2048-num_aie_columns_2-num_channels_2-tile_size_512]✅ 5/5330.460.03n/a
test_sigmoid[input_length_2048-num_aie_columns_4-num_channels_1-tile_size_512]✅ 5/5665.180.02n/a
test_sigmoid[input_length_2048-num_aie_columns_4-num_channels_2-tile_size_256]✅ 5/5468.140.02n/a
iron/operators/silu
TestChecksLatency (mean)Bandwidth (mean)Throughput (mean)
test_silu[input_length_2048-num_aie_columns_1-num_channels_1-tile_size_2048]✅ 5/5336.680.03n/a
test_silu[input_length_2048-num_aie_columns_2-num_channels_1-tile_size_1024]✅ 5/5468.640.02n/a
test_silu[input_length_2048-num_aie_columns_4-num_channels_1-tile_size_512]✅ 5/5489.280.02n/a
iron/operators/softmax
TestChecksLatency (mean)Bandwidth (mean)Throughput (mean)
test_softmax[input_length_32768-num_aie_columns_2-num_channels_2-tile_size_1024]✅ 5/5433.680.33n/a
test_softmax[input_length_32768-num_aie_columns_2-num_channels_2-tile_size_2048]✅ 5/5486.300.28n/a
test_softmax[input_length_32768-num_aie_columns_2-num_channels_2-tile_size_512]✅ 5/5488.420.27n/a
iron/operators/swiglu_decode
TestChecksLatency (mean)Bandwidth (mean)Throughput (mean)
test_swiglu_decode[embedding_dim_1024-hidden_dim_3584]✅ 5/513415.130.00n/a
test_swiglu_decode[embedding_dim_2048-hidden_dim_2048]✅ 5/514558.660.00n/a
iron/operators/swiglu_prefill
TestChecksLatency (mean)Bandwidth (mean)Throughput (mean)
test_swiglu_prefill[seq_len_256-embedding_dim_2048-hidden_dim_2048-prio_accuracy_False]✅ 5/520896.280.11n/a
iron/operators/tanh
TestChecksLatency (mean)Bandwidth (mean)Throughput (mean)
test_tanh[input_length_2048-num_aie_columns_1-num_channels_1-tile_size_2048]✅ 5/5451.780.02n/a
test_tanh[input_length_2048-num_aie_columns_1-num_channels_2-tile_size_1024]✅ 5/5790.580.02n/a
test_tanh[input_length_2048-num_aie_columns_2-num_channels_1-tile_size_1024]✅ 5/5351.000.02n/a
test_tanh[input_length_2048-num_aie_columns_2-num_channels_2-tile_size_512]✅ 5/5528.720.02n/a
test_tanh[input_length_2048-num_aie_columns_4-num_channels_1-tile_size_512]✅ 5/5357.880.03n/a
test_tanh[input_length_2048-num_aie_columns_4-num_channels_2-tile_size_256]✅ 5/5426.880.02n/a
iron/operators/transpose
TestChecksLatency (mean)Bandwidth (mean)Throughput (mean)
test_transpose[M_2048-N_64-aie_columns_1-channels_1-m_64-n_64-s_8-num_batches_1]✅ 5/5448.781.27n/a
test_transpose[M_2048-N_64-aie_columns_1-channels_1-m_64-n_64-s_8-num_batches_2]✅ 5/5965.741.66n/a
test_transpose[M_2048-N_64-aie_columns_1-channels_2-m_64-n_64-s_8-num_batches_1]✅ 5/5383.001.45n/a

Trends:

IRON Trends

iron/operators/axpy

test_axpy[input_length_2048-num_aie_columns_1-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)
1f35831 — 2026-07-09 23:05:560.05 (-1.01%)0.04 (+6.45%)0.04 (+18.04%)0.02 (+13.06%)0.01 (+2.15%)559.20 (-11.55%)387.68 (-6.44%)303.20 (-15.28%)268.40 (+1.02%)144.53 (-7.67%)
17769b8 — 2026-07-09 22:38:300.05 (n/a)0.03 (n/a)0.03 (n/a)0.02 (n/a)0.01 (n/a)632.20 (n/a)414.38 (n/a)357.90 (n/a)265.70 (n/a)156.54 (n/a)

test_axpy[input_length_2048-num_aie_columns_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)
1f35831 — 2026-07-09 23:05:560.05 (-4.20%)0.04 (+1.82%)0.05 (+7.19%)0.03 (+28.67%)0.01 (-18.24%)487.60 (-22.27%)334.78 (-7.44%)260.80 (-6.72%)244.10 (+4.36%)111.82 (-32.99%)
17769b8 — 2026-07-09 22:38:300.05 (n/a)0.04 (n/a)0.04 (n/a)0.02 (n/a)0.01 (n/a)627.30 (n/a)361.70 (n/a)279.60 (n/a)233.90 (n/a)166.87 (n/a)

test_axpy[input_length_2048-num_aie_columns_4-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)
1f35831 — 2026-07-09 23:05:560.05 (+121.97%)0.04 (+78.10%)0.04 (+86.99%)0.02 (+4.72%)0.01 (+535.11%)663.40 (-4.51%)379.66 (-36.89%)321.40 (-46.52%)242.40 (-54.96%)169.14 (+180.21%)
17769b8 — 2026-07-09 22:38:300.02 (n/a)0.02 (n/a)0.02 (n/a)0.02 (n/a)0.00 (n/a)694.70 (n/a)601.60 (n/a)601.00 (n/a)538.20 (n/a)60.36 (n/a)
iron/operators/dequant

test_dequant[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)
1f35831 — 2026-07-09 23:05:560.02 (-26.60%)0.01 (-19.49%)0.01 (-34.16%)0.01 (+4.38%)0.00 (-50.77%)474.80 (-4.20%)386.68 (+14.15%)380.50 (+51.84%)299.40 (+36.21%)84.61 (-38.43%)
17769b8 — 2026-07-09 22:38:300.02 (n/a)0.02 (n/a)0.02 (n/a)0.01 (n/a)0.01 (n/a)495.60 (n/a)338.76 (n/a)250.60 (n/a)219.80 (n/a)137.42 (n/a)

test_dequant[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)
1f35831 — 2026-07-09 23:05:560.02 (-9.96%)0.01 (-15.94%)0.01 (-19.87%)0.01 (-5.36%)0.00 (-7.36%)555.40 (+5.67%)434.06 (+18.76%)433.70 (+24.81%)275.70 (+11.04%)104.19 (+3.10%)
17769b8 — 2026-07-09 22:38:300.02 (n/a)0.02 (n/a)0.02 (n/a)0.01 (n/a)0.00 (n/a)525.60 (n/a)365.48 (n/a)347.50 (n/a)248.30 (n/a)101.06 (n/a)

test_dequant[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)
1f35831 — 2026-07-09 23:05:560.02 (-14.90%)0.02 (-26.85%)0.02 (-14.68%)0.00 (-74.98%)0.01 (+18.45%)1915.40 (+299.62%)623.08 (+122.93%)287.30 (+17.22%)212.90 (+17.49%)726.56 (+509.57%)
17769b8 — 2026-07-09 22:38:300.03 (n/a)0.02 (n/a)0.02 (n/a)0.01 (n/a)0.01 (n/a)479.30 (n/a)279.50 (n/a)245.10 (n/a)181.20 (n/a)119.19 (n/a)

test_dequant[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)
1f35831 — 2026-07-09 23:05:560.02 (+78.70%)0.02 (+50.33%)0.02 (+59.10%)0.01 (+11.57%)0.00 (+323.57%)474.20 (-10.38%)335.58 (-30.52%)313.10 (-37.14%)240.00 (-44.04%)86.80 (+118.63%)
17769b8 — 2026-07-09 22:38:300.01 (n/a)0.01 (n/a)0.01 (n/a)0.01 (n/a)0.00 (n/a)529.10 (n/a)483.00 (n/a)498.10 (n/a)428.90 (n/a)39.70 (n/a)

test_dequant[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)
1f35831 — 2026-07-09 23:05:560.02 (-20.88%)0.01 (-29.30%)0.01 (-47.64%)0.01 (-20.43%)0.01 (-19.11%)592.60 (+25.68%)461.32 (+41.21%)521.30 (+91.02%)225.70 (+26.37%)157.16 (+21.03%)
17769b8 — 2026-07-09 22:38:300.03 (n/a)0.02 (n/a)0.02 (n/a)0.01 (n/a)0.01 (n/a)471.50 (n/a)326.70 (n/a)272.90 (n/a)178.60 (n/a)129.85 (n/a)

test_dequant[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)
1f35831 — 2026-07-09 23:05:560.02 (+17.26%)0.01 (+15.19%)0.01 (+0.95%)0.01 (+32.63%)0.01 (+17.05%)522.70 (-24.61%)437.74 (-13.77%)493.40 (-0.94%)217.90 (-14.72%)128.02 (-22.04%)
17769b8 — 2026-07-09 22:38:300.02 (n/a)0.01 (n/a)0.01 (n/a)0.01 (n/a)0.01 (n/a)693.30 (n/a)507.64 (n/a)498.10 (n/a)255.50 (n/a)164.23 (n/a)
iron/operators/elementwise_add

test_elementwise_add[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)
1f35831 — 2026-07-09 23:05:560.04 (n/a)0.03 (n/a)0.02 (n/a)0.02 (n/a)0.01 (n/a)543.00 (n/a)475.12 (n/a)514.10 (n/a)288.90 (n/a)105.41 (n/a)

test_elementwise_add[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)
1f35831 — 2026-07-09 23:05:560.03 (n/a)0.03 (n/a)0.03 (n/a)0.02 (n/a)0.00 (n/a)526.60 (n/a)454.70 (n/a)450.80 (n/a)367.70 (n/a)67.53 (n/a)

test_elementwise_add[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)
1f35831 — 2026-07-09 23:05:560.05 (n/a)0.03 (n/a)0.02 (n/a)0.02 (n/a)0.01 (n/a)573.00 (n/a)471.90 (n/a)497.80 (n/a)245.10 (n/a)132.83 (n/a)
iron/operators/elementwise_mul

test_elementwise_mul[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)
1f35831 — 2026-07-09 23:05:560.05 (n/a)0.04 (n/a)0.05 (n/a)0.03 (n/a)0.01 (n/a)475.30 (n/a)302.90 (n/a)263.80 (n/a)226.10 (n/a)99.47 (n/a)

test_elementwise_mul[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)
1f35831 — 2026-07-09 23:05:560.06 (n/a)0.03 (n/a)0.03 (n/a)0.01 (n/a)0.02 (n/a)1881.90 (n/a)688.78 (n/a)450.20 (n/a)191.20 (n/a)679.96 (n/a)

test_elementwise_mul[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)
1f35831 — 2026-07-09 23:05:560.05 (n/a)0.04 (n/a)0.04 (n/a)0.03 (n/a)0.01 (n/a)363.00 (n/a)311.58 (n/a)307.70 (n/a)234.60 (n/a)51.86 (n/a)
iron/operators/gelu

test_gelu[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)
1f35831 — 2026-07-09 23:05:560.03 (n/a)0.02 (n/a)0.02 (n/a)0.02 (n/a)0.01 (n/a)528.10 (n/a)428.54 (n/a)490.20 (n/a)290.60 (n/a)116.08 (n/a)

test_gelu[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)
1f35831 — 2026-07-09 23:05:560.03 (n/a)0.03 (n/a)0.03 (n/a)0.02 (n/a)0.01 (n/a)528.70 (n/a)334.84 (n/a)291.80 (n/a)264.20 (n/a)109.22 (n/a)

test_gelu[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)
1f35831 — 2026-07-09 23:05:560.04 (n/a)0.02 (n/a)0.02 (n/a)0.01 (n/a)0.01 (n/a)710.00 (n/a)444.66 (n/a)419.80 (n/a)233.50 (n/a)171.20 (n/a)

test_gelu[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)
1f35831 — 2026-07-09 23:05:560.04 (n/a)0.02 (n/a)0.02 (n/a)0.01 (n/a)0.01 (n/a)601.90 (n/a)435.00 (n/a)426.20 (n/a)197.70 (n/a)153.29 (n/a)

test_gelu[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)
1f35831 — 2026-07-09 23:05:560.03 (n/a)0.02 (n/a)0.02 (n/a)0.01 (n/a)0.01 (n/a)568.30 (n/a)430.52 (n/a)413.10 (n/a)315.90 (n/a)113.96 (n/a)

test_gelu[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)
1f35831 — 2026-07-09 23:05:560.03 (n/a)0.02 (n/a)0.03 (n/a)0.01 (n/a)0.01 (n/a)644.00 (n/a)422.12 (n/a)313.30 (n/a)292.30 (n/a)163.31 (n/a)
iron/operators/gemm

test_gemm[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-partition_N_1]

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)
1f35831 — 2026-07-09 23:05:560.60 (+10.53%)0.48 (+17.21%)0.47 (+20.84%)0.41 (+22.81%)0.08 (-9.12%)537.40 (-18.56%)467.74 (-15.65%)474.10 (-17.25%)368.70 (-9.52%)68.36 (-32.64%)25.59 (+10.53%)20.55 (+17.21%)19.90 (+20.84%)17.56 (+22.81%)3.25 (-9.12%)
17769b8 — 2026-07-09 22:38:300.54 (n/a)0.41 (n/a)0.39 (n/a)0.34 (n/a)0.08 (n/a)659.90 (n/a)554.54 (n/a)572.90 (n/a)407.50 (n/a)101.48 (n/a)23.16 (n/a)17.54 (n/a)16.47 (n/a)14.30 (n/a)3.57 (n/a)

test_gemm[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-partition_N_1]

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)
1f35831 — 2026-07-09 23:05:560.66 (+4.57%)0.49 (+11.48%)0.48 (+15.53%)0.31 (-2.55%)0.13 (-2.25%)706.70 (+2.61%)482.34 (-10.81%)460.40 (-13.44%)335.80 (-4.38%)139.27 (-5.57%)28.10 (+4.57%)20.77 (+11.48%)20.50 (+15.53%)13.35 (-2.55%)5.40 (-2.25%)
17769b8 — 2026-07-09 22:38:300.63 (n/a)0.44 (n/a)0.42 (n/a)0.32 (n/a)0.13 (n/a)688.70 (n/a)540.78 (n/a)531.90 (n/a)351.20 (n/a)147.48 (n/a)26.87 (n/a)18.63 (n/a)17.74 (n/a)13.70 (n/a)5.53 (n/a)

test_gemm[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-partition_N_1]

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)
1f35831 — 2026-07-09 23:05:560.31 (-0.77%)0.30 (-1.08%)0.31 (-0.96%)0.29 (-1.32%)0.01 (-9.08%)85743.80 (+1.34%)83029.96 (+1.09%)82354.60 (+0.97%)80891.20 (+0.78%)1865.43 (-6.87%)212.38 (-0.77%)206.99 (-1.08%)208.61 (-0.96%)200.36 (-1.32%)4.61 (-9.08%)
17769b8 — 2026-07-09 22:38:300.31 (n/a)0.31 (n/a)0.31 (n/a)0.30 (n/a)0.01 (n/a)84611.30 (n/a)82135.30 (n/a)81566.10 (n/a)80266.00 (n/a)2002.97 (n/a)214.04 (n/a)209.26 (n/a)210.62 (n/a)203.04 (n/a)5.07 (n/a)

test_gemm[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-partition_N_1]

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)
1f35831 — 2026-07-09 23:05:561.03 (+0.28%)1.01 (-0.38%)1.01 (-0.51%)0.99 (-0.35%)0.02 (+54.07%)25293.00 (+0.36%)24836.62 (+0.40%)24810.30 (+0.52%)24351.00 (-0.28%)445.15 (+54.11%)705.51 (+0.28%)691.89 (-0.38%)692.45 (-0.51%)679.23 (-0.35%)12.40 (+54.07%)
17769b8 — 2026-07-09 22:38:301.03 (n/a)1.02 (n/a)1.02 (n/a)1.00 (n/a)0.01 (n/a)25203.40 (n/a)24737.36 (n/a)24682.70 (n/a)24419.10 (n/a)288.84 (n/a)703.54 (n/a)694.57 (n/a)696.03 (n/a)681.65 (n/a)8.05 (n/a)

test_gemm[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-partition_N_1]

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)
1f35831 — 2026-07-09 23:05:564.17 (+10.13%)1.99 (-19.09%)1.52 (-16.06%)1.27 (-18.95%)1.23 (+18.59%)6335.40 (+23.38%)4901.04 (+30.93%)5293.10 (+19.14%)1933.10 (-9.19%)1743.20 (+26.50%)1093.55 (+10.13%)521.11 (-19.09%)399.37 (-16.06%)333.67 (-18.95%)322.01 (+18.59%)
17769b8 — 2026-07-09 22:38:303.79 (n/a)2.46 (n/a)1.81 (n/a)1.57 (n/a)1.04 (n/a)5134.70 (n/a)3743.24 (n/a)4442.90 (n/a)2128.80 (n/a)1377.98 (n/a)992.99 (n/a)644.07 (n/a)475.80 (n/a)411.70 (n/a)271.52 (n/a)

test_gemm[M_64-K_512-N_256-num_aie_columns_4-b_col_maj_True-c_col_maj_False-m_16-k_64-n_64-trace_size_0-partition_N_4]

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)
1f35831 — 2026-07-09 23:05:560.34 (+58.14%)0.23 (+13.15%)0.19 (-3.56%)0.15 (-16.48%)0.07 (+531.61%)8088.20 (+19.73%)5894.82 (-4.64%)6402.30 (+3.69%)3679.60 (-36.76%)1750.42 (+369.39%)18.24 (+58.14%)12.32 (+13.15%)10.48 (-3.56%)8.30 (-16.48%)4.03 (+531.61%)
17769b8 — 2026-07-09 22:38:300.21 (n/a)0.20 (n/a)0.20 (n/a)0.18 (n/a)0.01 (n/a)6755.60 (n/a)6181.38 (n/a)6174.70 (n/a)5818.90 (n/a)372.92 (n/a)11.53 (n/a)10.89 (n/a)10.87 (n/a)9.93 (n/a)0.64 (n/a)
iron/operators/gemv

test_gemv[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)
1f35831 — 2026-07-09 23:05:560.12 (+3.87%)0.09 (+13.63%)0.08 (+26.55%)0.04 (-30.10%)0.03 (+35.17%)0.12 (+3.87%)0.08 (+13.63%)0.08 (+26.55%)0.04 (-30.10%)0.03 (+35.18%)
17769b8 — 2026-07-09 22:38:300.12 (n/a)0.07 (n/a)0.06 (n/a)0.06 (n/a)0.03 (n/a)0.12 (n/a)0.07 (n/a)0.06 (n/a)0.06 (n/a)0.03 (n/a)

test_gemv[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)
1f35831 — 2026-07-09 23:05:563.92 (-0.40%)3.64 (-3.76%)3.69 (-5.16%)3.37 (+1.48%)0.25 (-2.00%)3.92 (-0.40%)3.64 (-3.76%)3.69 (-5.16%)3.37 (+1.48%)0.25 (-2.00%)
17769b8 — 2026-07-09 22:38:303.94 (n/a)3.78 (n/a)3.89 (n/a)3.32 (n/a)0.26 (n/a)3.94 (n/a)3.78 (n/a)3.89 (n/a)3.32 (n/a)0.26 (n/a)

test_gemv[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)
1f35831 — 2026-07-09 23:05:567.04 (-6.83%)6.42 (-5.92%)6.05 (-14.28%)5.98 (+5.80%)0.55 (-23.66%)7.03 (-6.83%)6.41 (-5.92%)6.05 (-14.28%)5.97 (+5.80%)0.55 (-23.66%)
17769b8 — 2026-07-09 22:38:307.55 (n/a)6.82 (n/a)7.06 (n/a)5.65 (n/a)0.73 (n/a)7.55 (n/a)6.82 (n/a)7.06 (n/a)5.65 (n/a)0.73 (n/a)

test_gemv[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)
1f35831 — 2026-07-09 23:05:5613.51 (-2.15%)10.68 (+11.48%)11.58 (+33.52%)7.68 (-1.70%)2.55 (+3.34%)13.50 (-2.15%)10.68 (+11.48%)11.57 (+33.52%)7.68 (-1.70%)2.55 (+3.34%)
17769b8 — 2026-07-09 22:38:3013.81 (n/a)9.58 (n/a)8.67 (n/a)7.82 (n/a)2.47 (n/a)13.80 (n/a)9.58 (n/a)8.67 (n/a)7.81 (n/a)2.47 (n/a)

test_gemv[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)
1f35831 — 2026-07-09 23:05:563.84 (-3.33%)3.66 (-3.19%)3.73 (-2.60%)3.41 (-1.03%)0.17 (-16.29%)3.83 (-3.33%)3.65 (-3.19%)3.72 (-2.60%)3.41 (-1.03%)0.17 (-16.29%)
17769b8 — 2026-07-09 22:38:303.97 (n/a)3.78 (n/a)3.83 (n/a)3.45 (n/a)0.21 (n/a)3.97 (n/a)3.77 (n/a)3.82 (n/a)3.45 (n/a)0.21 (n/a)

test_gemv[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)
1f35831 — 2026-07-09 23:05:567.06 (-0.95%)5.93 (-2.21%)5.67 (-6.97%)4.55 (+1.07%)1.06 (-2.45%)7.06 (-0.95%)5.93 (-2.21%)5.67 (-6.97%)4.55 (+1.07%)1.06 (-2.45%)
17769b8 — 2026-07-09 22:38:307.13 (n/a)6.06 (n/a)6.10 (n/a)4.50 (n/a)1.09 (n/a)7.13 (n/a)6.06 (n/a)6.09 (n/a)4.50 (n/a)1.09 (n/a)

test_gemv[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)
1f35831 — 2026-07-09 23:05:5611.57 (-16.66%)9.62 (-11.48%)9.78 (+4.14%)8.17 (-0.18%)1.29 (-51.82%)11.56 (-16.66%)9.61 (-11.48%)9.77 (+4.14%)8.16 (-0.18%)1.29 (-51.82%)
17769b8 — 2026-07-09 22:38:3013.88 (n/a)10.86 (n/a)9.39 (n/a)8.18 (n/a)2.68 (n/a)13.88 (n/a)10.86 (n/a)9.38 (n/a)8.18 (n/a)2.68 (n/a)
iron/operators/layer_norm

test_layer_norm[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)
1f35831 — 2026-07-09 23:05:560.03 (n/a)0.03 (n/a)0.03 (n/a)0.02 (n/a)0.01 (n/a)484.40 (n/a)344.00 (n/a)299.00 (n/a)240.90 (n/a)116.65 (n/a)

test_layer_norm[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)
1f35831 — 2026-07-09 23:05:560.03 (n/a)0.03 (n/a)0.03 (n/a)0.02 (n/a)0.01 (n/a)484.00 (n/a)348.34 (n/a)270.00 (n/a)245.00 (n/a)122.41 (n/a)

test_layer_norm[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)
1f35831 — 2026-07-09 23:05:560.04 (n/a)0.02 (n/a)0.02 (n/a)0.02 (n/a)0.01 (n/a)541.60 (n/a)400.14 (n/a)482.70 (n/a)228.70 (n/a)155.95 (n/a)

test_layer_norm[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)
1f35831 — 2026-07-09 23:05:560.02 (n/a)0.02 (n/a)0.02 (n/a)0.01 (n/a)0.00 (n/a)679.90 (n/a)511.48 (n/a)524.50 (n/a)349.50 (n/a)120.95 (n/a)

test_layer_norm[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)
1f35831 — 2026-07-09 23:05:560.03 (n/a)0.02 (n/a)0.02 (n/a)0.02 (n/a)0.00 (n/a)489.30 (n/a)418.30 (n/a)417.20 (n/a)293.40 (n/a)77.43 (n/a)

test_layer_norm[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)
1f35831 — 2026-07-09 23:05:560.03 (n/a)0.02 (n/a)0.02 (n/a)0.01 (n/a)0.01 (n/a)668.40 (n/a)443.56 (n/a)404.30 (n/a)253.60 (n/a)164.11 (n/a)
iron/operators/leaky_relu

test_leaky_relu[input_length_2048-num_aie_columns_1-num_channels_1-tile_size_2048-alpha_0.01]

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
1f35831 — 2026-07-09 23:05:560.04 (+20.02%)0.03 (-2.18%)0.02 (-25.38%)0.02 (-2.75%)0.01 (+90.32%)468.40 (+2.83%)352.84 (+10.10%)396.30 (+34.02%)211.50 (-16.70%)122.63 (+55.89%)
17769b8 — 2026-07-09 22:38:300.03 (n/a)0.03 (n/a)0.03 (n/a)0.02 (n/a)0.01 (n/a)455.50 (n/a)320.46 (n/a)295.70 (n/a)253.90 (n/a)78.67 (n/a)

test_leaky_relu[input_length_2048-num_aie_columns_1-num_channels_1-tile_size_2048-alpha_0.1]

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
1f35831 — 2026-07-09 23:05:560.03 (+1.37%)0.02 (-12.35%)0.02 (-24.40%)0.01 (-19.69%)0.01 (+34.48%)607.10 (+24.51%)392.10 (+21.18%)389.40 (+32.27%)241.50 (-1.35%)151.87 (+54.89%)
17769b8 — 2026-07-09 22:38:300.03 (n/a)0.03 (n/a)0.03 (n/a)0.02 (n/a)0.01 (n/a)487.60 (n/a)323.56 (n/a)294.40 (n/a)244.80 (n/a)98.05 (n/a)

test_leaky_relu[input_length_2048-num_aie_columns_1-num_channels_1-tile_size_2048-alpha_0.25]

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
1f35831 — 2026-07-09 23:05:560.04 (-0.57%)0.03 (+12.91%)0.03 (+63.64%)0.02 (+16.88%)0.01 (-19.96%)545.50 (-14.44%)356.94 (-18.11%)294.10 (-38.88%)231.70 (+0.56%)142.58 (-27.32%)
17769b8 — 2026-07-09 22:38:300.04 (n/a)0.02 (n/a)0.02 (n/a)0.01 (n/a)0.01 (n/a)637.60 (n/a)435.86 (n/a)481.20 (n/a)230.40 (n/a)196.18 (n/a)

test_leaky_relu[input_length_2048-num_aie_columns_1-num_channels_2-tile_size_1024-alpha_0.01]

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
1f35831 — 2026-07-09 23:05:560.04 (+52.43%)0.02 (+16.30%)0.01 (-4.00%)0.00 (-67.97%)0.01 (+171.76%)1850.50 (+212.27%)697.92 (+40.85%)569.50 (+4.17%)205.70 (-34.39%)665.76 (+473.06%)
17769b8 — 2026-07-09 22:38:300.03 (n/a)0.02 (n/a)0.01 (n/a)0.01 (n/a)0.01 (n/a)592.60 (n/a)495.52 (n/a)546.70 (n/a)313.50 (n/a)116.18 (n/a)

test_leaky_relu[input_length_2048-num_aie_columns_2-num_channels_1-tile_size_1024-alpha_0.01]

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
1f35831 — 2026-07-09 23:05:560.05 (+53.36%)0.03 (+57.95%)0.03 (+107.27%)0.02 (+40.29%)0.01 (+60.63%)467.80 (-28.72%)318.50 (-34.99%)265.40 (-51.75%)176.10 (-34.78%)126.14 (-20.08%)
17769b8 — 2026-07-09 22:38:300.03 (n/a)0.02 (n/a)0.01 (n/a)0.01 (n/a)0.01 (n/a)656.30 (n/a)489.94 (n/a)550.10 (n/a)270.00 (n/a)157.82 (n/a)

test_leaky_relu[input_length_2048-num_aie_columns_2-num_channels_2-tile_size_512-alpha_0.01]

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
1f35831 — 2026-07-09 23:05:560.04 (+13.74%)0.03 (-0.26%)0.03 (-7.89%)0.01 (-0.19%)0.01 (-1.61%)601.70 (+0.20%)369.16 (-1.85%)300.30 (+8.57%)207.20 (-12.09%)159.27 (-9.46%)
17769b8 — 2026-07-09 22:38:300.03 (n/a)0.03 (n/a)0.03 (n/a)0.01 (n/a)0.01 (n/a)600.50 (n/a)376.12 (n/a)276.60 (n/a)235.70 (n/a)175.91 (n/a)

test_leaky_relu[input_length_2048-num_aie_columns_4-num_channels_1-tile_size_512-alpha_0.01]

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
1f35831 — 2026-07-09 23:05:560.03 (+6.97%)0.02 (-9.81%)0.02 (-17.50%)0.02 (+26.51%)0.01 (-5.91%)496.80 (-20.94%)428.44 (+7.48%)468.80 (+21.23%)247.60 (-6.53%)103.06 (-30.19%)
17769b8 — 2026-07-09 22:38:300.03 (n/a)0.02 (n/a)0.02 (n/a)0.01 (n/a)0.01 (n/a)628.40 (n/a)398.64 (n/a)386.70 (n/a)264.90 (n/a)147.64 (n/a)

test_leaky_relu[input_length_2048-num_aie_columns_4-num_channels_2-tile_size_256-alpha_0.01]

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
1f35831 — 2026-07-09 23:05:560.02 (-29.64%)0.02 (-17.51%)0.02 (-20.05%)0.01 (+15.24%)0.00 (-65.79%)575.10 (-13.22%)475.20 (+7.76%)479.20 (+25.08%)368.90 (+42.10%)73.79 (-59.62%)
17769b8 — 2026-07-09 22:38:300.03 (n/a)0.02 (n/a)0.02 (n/a)0.01 (n/a)0.01 (n/a)662.70 (n/a)440.96 (n/a)383.10 (n/a)259.60 (n/a)182.72 (n/a)
iron/operators/mem_copy

test_mem_copy[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)
1f35831 — 2026-07-09 23:05:560.04 (-2.92%)0.02 (-2.55%)0.03 (+5.28%)0.02 (+1.64%)0.01 (-3.90%)521.10 (-1.62%)369.82 (+2.72%)319.70 (-5.02%)233.30 (+3.00%)139.96 (+4.51%)
17769b8 — 2026-07-09 22:38:300.04 (n/a)0.03 (n/a)0.02 (n/a)0.02 (n/a)0.01 (n/a)529.70 (n/a)360.02 (n/a)336.60 (n/a)226.50 (n/a)133.92 (n/a)

test_mem_copy[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)
1f35831 — 2026-07-09 23:05:560.03 (-10.22%)0.02 (-3.41%)0.02 (-6.99%)0.02 (+15.21%)0.01 (-20.12%)511.90 (-13.19%)394.42 (+0.08%)429.30 (+7.51%)273.40 (+11.36%)103.30 (-23.37%)
17769b8 — 2026-07-09 22:38:300.03 (n/a)0.02 (n/a)0.02 (n/a)0.01 (n/a)0.01 (n/a)589.70 (n/a)394.12 (n/a)399.30 (n/a)245.50 (n/a)134.81 (n/a)

test_mem_copy[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)
1f35831 — 2026-07-09 23:05:560.03 (+23.51%)0.02 (+13.58%)0.03 (+20.41%)0.01 (-12.96%)0.01 (+117.19%)564.10 (+14.89%)365.72 (-6.14%)315.50 (-16.93%)255.10 (-19.04%)127.11 (+98.86%)
17769b8 — 2026-07-09 22:38:300.03 (n/a)0.02 (n/a)0.02 (n/a)0.02 (n/a)0.00 (n/a)491.00 (n/a)389.66 (n/a)379.80 (n/a)315.10 (n/a)63.92 (n/a)

test_mem_copy[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)
1f35831 — 2026-07-09 23:05:560.02 (-32.34%)0.02 (-11.88%)0.01 (-10.45%)0.01 (+70.57%)0.00 (-55.70%)654.80 (-41.37%)541.88 (-5.65%)585.30 (+11.68%)360.00 (+47.78%)123.74 (-62.02%)
17769b8 — 2026-07-09 22:38:300.03 (n/a)0.02 (n/a)0.02 (n/a)0.01 (n/a)0.01 (n/a)1116.90 (n/a)574.36 (n/a)524.10 (n/a)243.60 (n/a)325.84 (n/a)

test_mem_copy[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)
1f35831 — 2026-07-09 23:05:560.03 (+21.07%)0.02 (-12.54%)0.02 (-29.99%)0.01 (-6.73%)0.01 (+46.78%)642.70 (+7.22%)478.32 (+19.72%)494.20 (+42.83%)237.30 (-17.43%)148.26 (+17.40%)
17769b8 — 2026-07-09 22:38:300.03 (n/a)0.02 (n/a)0.02 (n/a)0.01 (n/a)0.01 (n/a)599.40 (n/a)399.54 (n/a)346.00 (n/a)287.40 (n/a)126.29 (n/a)

test_mem_copy[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)
1f35831 — 2026-07-09 23:05:560.03 (+2.99%)0.02 (+9.32%)0.02 (-12.02%)0.01 (+168.51%)0.01 (-16.18%)743.00 (-62.76%)489.60 (-33.68%)532.00 (+13.68%)263.00 (-2.92%)183.26 (-74.18%)
17769b8 — 2026-07-09 22:38:300.03 (n/a)0.02 (n/a)0.02 (n/a)0.00 (n/a)0.01 (n/a)1995.00 (n/a)738.24 (n/a)468.00 (n/a)270.90 (n/a)709.78 (n/a)
iron/operators/rms_norm

test_rms_norm[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)
1f35831 — 2026-07-09 23:05:560.04 (+4.70%)0.03 (+2.72%)0.03 (+2.63%)0.02 (-6.40%)0.01 (+22.40%)525.30 (+6.83%)358.74 (+1.63%)284.00 (-2.57%)231.10 (-4.46%)150.09 (+23.33%)
17769b8 — 2026-07-09 22:38:300.03 (n/a)0.03 (n/a)0.03 (n/a)0.02 (n/a)0.01 (n/a)491.70 (n/a)353.00 (n/a)291.50 (n/a)241.90 (n/a)121.70 (n/a)

test_rms_norm[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)
1f35831 — 2026-07-09 23:05:560.06 (+15.91%)0.03 (-14.89%)0.03 (-28.50%)0.01 (-71.36%)0.02 (+53.57%)1962.00 (+249.11%)680.14 (+89.23%)377.50 (+39.87%)208.30 (-13.71%)729.83 (+404.10%)
17769b8 — 2026-07-09 22:38:300.05 (n/a)0.04 (n/a)0.05 (n/a)0.02 (n/a)0.01 (n/a)562.00 (n/a)359.42 (n/a)269.90 (n/a)241.40 (n/a)144.78 (n/a)

test_rms_norm[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)
1f35831 — 2026-07-09 23:05:560.03 (-15.93%)0.02 (-20.14%)0.02 (-34.41%)0.01 (+5.43%)0.01 (-39.46%)549.40 (-5.15%)441.22 (+14.80%)448.90 (+52.43%)277.90 (+18.96%)104.73 (-37.19%)
17769b8 — 2026-07-09 22:38:300.04 (n/a)0.02 (n/a)0.03 (n/a)0.01 (n/a)0.01 (n/a)579.20 (n/a)384.34 (n/a)294.50 (n/a)233.60 (n/a)166.75 (n/a)

test_rms_norm[input_length_2048-num_aie_columns_1-num_channels_2-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)
1f35831 — 2026-07-09 23:05:560.05 (+30.88%)0.03 (-1.63%)0.02 (-14.30%)0.02 (-0.64%)0.01 (+44.20%)566.60 (+0.64%)447.10 (+5.80%)488.50 (+16.70%)217.40 (-23.61%)140.19 (+6.49%)
17769b8 — 2026-07-09 22:38:300.04 (n/a)0.03 (n/a)0.02 (n/a)0.02 (n/a)0.01 (n/a)563.00 (n/a)422.60 (n/a)418.60 (n/a)284.60 (n/a)131.65 (n/a)

test_rms_norm[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)
1f35831 — 2026-07-09 23:05:560.04 (+37.70%)0.02 (+25.69%)0.02 (+12.96%)0.00 (-45.84%)0.01 (+74.17%)1937.00 (+84.63%)661.16 (+20.37%)428.90 (-11.48%)191.70 (-27.39%)723.13 (+143.46%)
17769b8 — 2026-07-09 22:38:300.03 (n/a)0.02 (n/a)0.02 (n/a)0.01 (n/a)0.01 (n/a)1049.10 (n/a)549.26 (n/a)484.50 (n/a)264.00 (n/a)297.02 (n/a)

test_rms_norm[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)
1f35831 — 2026-07-09 23:05:560.05 (+40.80%)0.03 (+24.53%)0.02 (-23.53%)0.02 (+314.58%)0.01 (+4.38%)486.00 (-75.88%)363.98 (-46.76%)431.60 (+30.75%)193.60 (-28.98%)125.04 (-83.30%)
17769b8 — 2026-07-09 22:38:300.04 (n/a)0.03 (n/a)0.03 (n/a)0.01 (n/a)0.01 (n/a)2014.80 (n/a)683.68 (n/a)330.10 (n/a)272.60 (n/a)748.79 (n/a)

test_rms_norm[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)
1f35831 — 2026-07-09 23:05:560.03 (-25.92%)0.02 (-7.78%)0.02 (+10.21%)0.01 (+4.05%)0.01 (-36.42%)551.50 (-3.89%)394.98 (+1.99%)385.30 (-9.28%)272.00 (+34.99%)121.08 (-17.69%)
17769b8 — 2026-07-09 22:38:300.04 (n/a)0.02 (n/a)0.02 (n/a)0.01 (n/a)0.01 (n/a)573.80 (n/a)387.28 (n/a)424.70 (n/a)201.50 (n/a)147.11 (n/a)

test_rms_norm[input_length_2048-num_aie_columns_2-num_channels_2-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)
1f35831 — 2026-07-09 23:05:560.04 (+59.17%)0.02 (+10.58%)0.02 (-8.31%)0.02 (-6.27%)0.01 (+312.31%)528.80 (+6.68%)429.32 (-2.40%)473.60 (+9.07%)237.00 (-37.17%)115.89 (+167.11%)
17769b8 — 2026-07-09 22:38:300.02 (n/a)0.02 (n/a)0.02 (n/a)0.02 (n/a)0.00 (n/a)495.70 (n/a)439.86 (n/a)434.20 (n/a)377.20 (n/a)43.39 (n/a)

test_rms_norm[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)
1f35831 — 2026-07-09 23:05:560.03 (+64.54%)0.02 (+68.94%)0.03 (+79.46%)0.02 (+89.11%)0.01 (+78.17%)542.80 (-47.13%)362.84 (-40.92%)295.00 (-44.28%)250.00 (-39.22%)127.92 (-46.45%)
17769b8 — 2026-07-09 22:38:300.02 (n/a)0.01 (n/a)0.02 (n/a)0.01 (n/a)0.00 (n/a)1026.60 (n/a)614.14 (n/a)529.40 (n/a)411.30 (n/a)238.90 (n/a)

test_rms_norm[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)
1f35831 — 2026-07-09 23:05:560.03 (-7.70%)0.02 (-30.71%)0.02 (-38.19%)0.00 (-69.76%)0.01 (+17.16%)1904.00 (+230.67%)726.34 (+96.82%)492.70 (+61.81%)285.50 (+8.35%)663.92 (+396.38%)
17769b8 — 2026-07-09 22:38:300.03 (n/a)0.03 (n/a)0.03 (n/a)0.02 (n/a)0.01 (n/a)575.80 (n/a)369.04 (n/a)304.50 (n/a)263.50 (n/a)133.75 (n/a)

test_rms_norm[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)
1f35831 — 2026-07-09 23:05:560.03 (+30.06%)0.02 (+26.51%)0.02 (+18.34%)0.02 (+101.00%)0.01 (-7.58%)541.40 (-50.24%)440.06 (-30.37%)479.80 (-15.50%)256.70 (-23.10%)110.37 (-65.05%)
17769b8 — 2026-07-09 22:38:300.02 (n/a)0.02 (n/a)0.01 (n/a)0.01 (n/a)0.01 (n/a)1088.10 (n/a)632.02 (n/a)567.80 (n/a)333.80 (n/a)315.84 (n/a)
iron/operators/rope

test_rope[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)
1f35831 — 2026-07-09 23:05:560.47 (+25.09%)0.30 (+25.50%)0.34 (+65.15%)0.17 (+36.13%)0.13 (+10.53%)573.50 (-26.54%)382.66 (-22.76%)291.90 (-39.44%)209.20 (-20.06%)171.65 (-25.62%)
17769b8 — 2026-07-09 22:38:300.38 (n/a)0.24 (n/a)0.20 (n/a)0.13 (n/a)0.12 (n/a)780.70 (n/a)495.44 (n/a)482.00 (n/a)261.70 (n/a)230.78 (n/a)

test_rope[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)
1f35831 — 2026-07-09 23:05:560.43 (+49.55%)0.31 (+32.87%)0.32 (+36.62%)0.16 (-17.12%)0.10 (+165.18%)617.60 (+20.65%)356.80 (-17.26%)303.00 (-26.81%)226.90 (-33.13%)153.96 (+122.47%)
17769b8 — 2026-07-09 22:38:300.29 (n/a)0.23 (n/a)0.24 (n/a)0.19 (n/a)0.04 (n/a)511.90 (n/a)431.24 (n/a)414.00 (n/a)339.30 (n/a)69.21 (n/a)

test_rope[rows_32-cols_512-angle_rows_32-aie_columns_4-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)
1f35831 — 2026-07-09 23:05:560.39 (+2.70%)0.29 (+2.14%)0.33 (-1.35%)0.19 (+18.60%)0.09 (-21.33%)508.50 (-15.69%)364.88 (-8.41%)295.50 (+1.37%)252.60 (-2.62%)116.96 (-33.39%)
17769b8 — 2026-07-09 22:38:300.38 (n/a)0.29 (n/a)0.34 (n/a)0.16 (n/a)0.11 (n/a)603.10 (n/a)398.38 (n/a)291.50 (n/a)259.40 (n/a)175.59 (n/a)

test_rope[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)
1f35831 — 2026-07-09 23:05:560.29 (+6.03%)0.21 (+6.11%)0.21 (+31.12%)0.13 (-5.05%)0.06 (+3.20%)546.20 (+5.30%)378.18 (-5.26%)345.60 (-23.74%)254.40 (-5.71%)118.55 (+6.25%)
17769b8 — 2026-07-09 22:38:300.27 (n/a)0.20 (n/a)0.16 (n/a)0.14 (n/a)0.06 (n/a)518.70 (n/a)399.16 (n/a)453.20 (n/a)269.80 (n/a)111.58 (n/a)

test_rope[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)
1f35831 — 2026-07-09 23:05:560.30 (+23.76%)0.24 (+37.60%)0.27 (+75.50%)0.14 (+23.28%)0.07 (+37.67%)526.30 (-18.88%)337.50 (-26.05%)277.40 (-43.02%)242.00 (-19.20%)123.68 (-10.66%)
17769b8 — 2026-07-09 22:38:300.25 (n/a)0.17 (n/a)0.15 (n/a)0.11 (n/a)0.05 (n/a)648.80 (n/a)456.40 (n/a)486.80 (n/a)299.50 (n/a)138.44 (n/a)

test_rope[rows_32-cols_512-angle_rows_8-aie_columns_4-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)
1f35831 — 2026-07-09 23:05:560.38 (+30.89%)0.23 (+2.23%)0.20 (-30.51%)0.12 (+26.71%)0.10 (+16.45%)612.50 (-21.08%)380.24 (-5.02%)377.10 (+43.93%)194.30 (-23.59%)165.47 (-26.90%)
17769b8 — 2026-07-09 22:38:300.29 (n/a)0.22 (n/a)0.28 (n/a)0.10 (n/a)0.09 (n/a)776.10 (n/a)400.32 (n/a)262.00 (n/a)254.30 (n/a)226.37 (n/a)
iron/operators/softmax

test_softmax[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)
1f35831 — 2026-07-09 23:05:560.49 (+88.93%)0.33 (+43.52%)0.28 (+23.50%)0.21 (+3.45%)0.11 (+367.68%)614.30 (-3.34%)433.68 (-24.52%)475.80 (-19.03%)265.40 (-47.07%)137.63 (+133.45%)
17769b8 — 2026-07-09 22:38:300.26 (n/a)0.23 (n/a)0.22 (n/a)0.21 (n/a)0.02 (n/a)635.50 (n/a)574.58 (n/a)587.60 (n/a)501.40 (n/a)58.95 (n/a)

test_softmax[input_length_32768-num_aie_columns_2-num_channels_2-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)
1f35831 — 2026-07-09 23:05:560.44 (+8.94%)0.28 (+7.05%)0.25 (-7.03%)0.24 (+50.83%)0.09 (-3.75%)548.50 (-33.70%)486.30 (-9.99%)528.60 (+7.57%)300.50 (-8.19%)105.23 (-43.00%)
17769b8 — 2026-07-09 22:38:300.40 (n/a)0.27 (n/a)0.27 (n/a)0.16 (n/a)0.09 (n/a)827.30 (n/a)540.28 (n/a)491.40 (n/a)327.30 (n/a)184.62 (n/a)

test_softmax[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)
1f35831 — 2026-07-09 23:05:560.33 (-23.67%)0.27 (-17.74%)0.26 (-33.98%)0.25 (+103.68%)0.03 (-72.73%)524.90 (-50.90%)488.42 (-1.11%)510.80 (+51.48%)394.30 (+31.00%)53.42 (-83.60%)
17769b8 — 2026-07-09 22:38:300.44 (n/a)0.33 (n/a)0.39 (n/a)0.12 (n/a)0.13 (n/a)1069.10 (n/a)493.90 (n/a)337.20 (n/a)301.00 (n/a)325.66 (n/a)
iron/operators/swiglu_decode

test_swiglu_decode[embedding_dim_1024-hidden_dim_3584]

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
1f35831 — 2026-07-09 23:05:560.00 (+16.67%)0.00 (+42.86%)0.00 (+50.00%)0.00 (+0.00%)0.00 (+31.10%)20094.38 (-6.97%)13415.13 (-15.17%)14694.31 (-14.05%)6247.94 (-6.69%)6524.25 (+18.81%)
17769b8 — 2026-07-09 22:38:300.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)21598.80 (n/a)15813.49 (n/a)17097.23 (n/a)6695.96 (n/a)5491.52 (n/a)

test_swiglu_decode[embedding_dim_2048-hidden_dim_2048]

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
1f35831 — 2026-07-09 23:05:560.00 (-23.08%)0.00 (-24.39%)0.00 (-16.67%)0.00 (+25.00%)0.00 (-46.14%)17327.64 (-19.26%)14558.66 (+18.72%)16425.11 (+23.56%)8286.11 (+31.96%)3683.60 (-40.65%)
17769b8 — 2026-07-09 22:38:300.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)0.00 (n/a)21462.12 (n/a)12263.19 (n/a)13292.95 (n/a)6279.46 (n/a)6206.94 (n/a)
iron/operators/swiglu_prefill

test_swiglu_prefill[seq_len_256-embedding_dim_2048-hidden_dim_2048-prio_accuracy_False]

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
1f35831 — 2026-07-09 23:05:560.14 (+4.98%)0.11 (+16.28%)0.10 (+14.45%)0.07 (-10.80%)0.03 (+28.86%)30973.61 (+12.07%)20896.28 (-11.55%)21364.96 (-12.62%)15088.79 (-4.77%)6449.33 (+34.23%)
17769b8 — 2026-07-09 22:38:300.13 (n/a)0.09 (n/a)0.09 (n/a)0.08 (n/a)0.02 (n/a)27636.97 (n/a)23625.81 (n/a)24451.70 (n/a)15844.81 (n/a)4804.73 (n/a)
iron/operators/transpose

test_transpose[M_2048-N_64-aie_columns_1-channels_1-m_64-n_64-s_8-num_batches_1]

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
1f35831 — 2026-07-09 23:05:561.60 (-10.31%)1.27 (-13.51%)1.45 (-4.75%)0.79 (-20.12%)0.37 (+27.22%)660.80 (+25.18%)448.78 (+20.69%)360.80 (+4.98%)328.60 (+11.50%)151.65 (+66.31%)
17769b8 — 2026-07-09 22:38:301.78 (n/a)1.47 (n/a)1.53 (n/a)0.99 (n/a)0.29 (n/a)527.90 (n/a)371.84 (n/a)343.70 (n/a)294.70 (n/a)91.18 (n/a)

test_transpose[M_2048-N_64-aie_columns_1-channels_1-m_64-n_64-s_8-num_batches_2]

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
1f35831 — 2026-07-09 23:05:562.56 (-1.08%)1.66 (-21.62%)1.52 (-22.47%)0.42 (-76.32%)0.90 (+162.37%)2497.40 (+322.36%)965.74 (+91.37%)691.40 (+28.99%)410.00 (+1.11%)874.13 (+1021.75%)
17769b8 — 2026-07-09 22:38:302.59 (n/a)2.12 (n/a)1.96 (n/a)1.77 (n/a)0.34 (n/a)591.30 (n/a)504.64 (n/a)536.00 (n/a)405.50 (n/a)77.93 (n/a)

test_transpose[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)
4bb8427 — 2026-06-25 20:01:371.58 (-2.67%)1.21 (+5.65%)1.29 (+41.89%)0.66 (-20.05%)0.38 (-2.53%)792.90 (+25.08%)479.12 (-3.80%)406.80 (-29.52%)331.10 (+2.76%)189.18 (+26.62%)
4bb8427 — 2026-06-23 22:46:491.63 (n/a)1.14 (n/a)0.91 (n/a)0.83 (n/a)0.38 (n/a)633.90 (n/a)498.02 (n/a)577.20 (n/a)322.20 (n/a)149.40 (n/a)

test_transpose[M_2048-N_64-aie_columns_1-channels_2-m_64-n_64-s_8-num_batches_1]

Commit/Date Bandwidth (max)Bandwidth (mean)Bandwidth (median)Bandwidth (min)Bandwidth (stddev)Latency (max)Latency (mean)Latency (median)Latency (min)Latency (stddev)
1f35831 — 2026-07-09 23:05:561.82 (-8.73%)1.45 (+8.48%)1.49 (+19.18%)0.89 (-3.39%)0.34 (-21.34%)586.20 (+3.51%)383.00 (-9.52%)351.20 (-16.10%)288.30 (+9.58%)116.55 (-5.70%)
17769b8 — 2026-07-09 22:38:301.99 (n/a)1.34 (n/a)1.25 (n/a)0.93 (n/a)0.43 (n/a)566.30 (n/a)423.32 (n/a)418.60 (n/a)263.10 (n/a)123.59 (n/a)

test_transpose[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)
4bb8427 — 2026-06-25 20:01:371.75 (-15.44%)1.32 (-0.50%)1.36 (+22.47%)0.77 (-19.20%)0.36 (-21.75%)678.10 (+23.76%)430.02 (-0.09%)385.10 (-18.34%)300.10 (+18.29%)146.96 (+19.46%)
4bb8427 — 2026-06-23 22:46:492.07 (n/a)1.32 (n/a)1.11 (n/a)0.96 (n/a)0.46 (n/a)547.90 (n/a)430.40 (n/a)471.60 (n/a)253.70 (n/a)123.02 (n/a)
Phoenix - Examples

IRON

Tested on 2026_07_09_23_15_22 at commit 1f35831.

Trends:

IRON Trends

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.

Multi-head Attention Temporarily Disabled -- ObjectFIFO can't handle conditionals with newer MLIR-AIE

2 participants