Skip to content

Instantly share code, notes, and snippets.

Show Gist options
  • Save iree-github-actions-bot/8f05656196a61be8b2e09e606f5862cb to your computer and use it in GitHub Desktop.
Save iree-github-actions-bot/8f05656196a61be8b2e09e606f5862cb to your computer and use it in GitHub Desktop.

Full Benchmark Summary

Data-Tiling Comparison Table

Name No-DT (baseline) DT-Only DT-UK
BertForMaskedLMTF(stablehlo) [x86_64-cascadelake-linux_gnu-llvm_cpu] local_task(embedded_elf)[30-thread,full-inference,default-flags] with default @ c2-standard-60[cpu] 218.814 (1.0X) N/A 106.242 (2.1X)
BertLargeTF(stablehlo) [x86_64-cascadelake-linux_gnu-llvm_cpu] local_task(embedded_elf)[30-thread,full-inference,default-flags] with default @ c2-standard-60[cpu] 780.636 (1.0X) N/A 222.751 (3.5X)
DeepLabV3_fp32(tflite) [x86_64-cascadelake-linux_gnu-llvm_cpu] local_task(embedded_elf)[8-thread,full-inference,default-flags] with default @ c2-standard-60[cpu] 6.905 (1.0X) N/A 8.525 (0.8X)
DeepLabV3_fp32(tflite) [x86_64-cascadelake-linux_gnu-llvm_cpu] local_task(embedded_elf)[1-thread,full-inference,default-flags] with default @ c2-standard-60[cpu] 32.027 (1.0X) N/A 30.118 (1.1X)
EfficientNetV2STF(stablehlo) [x86_64-cascadelake-linux_gnu-llvm_cpu] local_task(embedded_elf)[15-thread,full-inference,default-flags] with default @ c2-standard-60[cpu] 35.762 (1.0X) N/A 34.370 (1.0X)
EfficientNetV2STF(stablehlo) [x86_64-cascadelake-linux_gnu-llvm_cpu] local_task(embedded_elf)[1-thread,full-inference,default-flags] with default @ c2-standard-60[cpu] 276.144 (1.0X) N/A 236.681 (1.2X)
EfficientNet_int8(tflite) [x86_64-cascadelake-linux_gnu-llvm_cpu] local_task(embedded_elf)[8-thread,full-inference,default-flags] with default @ c2-standard-60[cpu] 5.804 (1.0X) N/A 5.033 (1.2X)
EfficientNet_int8(tflite) [x86_64-cascadelake-linux_gnu-llvm_cpu] local_task(embedded_elf)[1-thread,full-inference,default-flags] with default @ c2-standard-60[cpu] 26.879 (1.0X) N/A 13.231 (2.0X)
GPT2_117M_TF_1X1XI32(stablehlo) [x86_64-cascadelake-linux_gnu-llvm_cpu] local_task(embedded_elf)[15-thread,full-inference,default-flags] with default @ c2-standard-60[cpu] 9.085 (1.0X) N/A 8.749 (1.0X)
GPT2_117M_TF_1X1XI32(stablehlo) [x86_64-cascadelake-linux_gnu-llvm_cpu] local_task(embedded_elf)[1-thread,full-inference,default-flags] with default @ c2-standard-60[cpu] 70.114 (1.0X) N/A 37.586 (1.9X)
GPT2_117M_TF_1X4XI32(stablehlo) [x86_64-cascadelake-linux_gnu-llvm_cpu] local_task(embedded_elf)[15-thread,full-inference,default-flags] with default @ c2-standard-60[cpu] 10.986 (1.0X) N/A 8.569 (1.3X)
GPT2_117M_TF_1X4XI32(stablehlo) [x86_64-cascadelake-linux_gnu-llvm_cpu] local_task(embedded_elf)[1-thread,full-inference,default-flags] with default @ c2-standard-60[cpu] 87.114 (1.0X) N/A 39.165 (2.2X)
MiniLML12H384Uncased(stablehlo) [x86_64-cascadelake-linux_gnu-llvm_cpu] local_task(embedded_elf)[15-thread,full-inference,default-flags] with default @ c2-standard-60[cpu] 12.215 (1.0X) N/A 13.035 (0.9X)
MiniLML12H384Uncased(stablehlo) [x86_64-cascadelake-linux_gnu-llvm_cpu] local_task(embedded_elf)[1-thread,full-inference,default-flags] with default @ c2-standard-60[cpu] 80.271 (1.0X) N/A 56.198 (1.4X)
MobileBertSquad_fp16(tflite) [x86_64-cascadelake-linux_gnu-llvm_cpu] local_task(embedded_elf)[15-thread,full-inference,default-flags] with default @ c2-standard-60[cpu] 34.045 (1.0X) N/A 61.171 (0.6X)
MobileBertSquad_fp16(tflite) [x86_64-cascadelake-linux_gnu-llvm_cpu] local_task(embedded_elf)[1-thread,full-inference,default-flags] with default @ c2-standard-60[cpu] 181.271 (1.0X) N/A 185.715 (1.0X)
MobileBertSquad_fp32(tflite) [x86_64-cascadelake-linux_gnu-llvm_cpu] local_task(embedded_elf)[15-thread,full-inference,default-flags] with default @ c2-standard-60[cpu] 34.229 (1.0X) N/A 61.197 (0.6X)
MobileBertSquad_fp32(tflite) [x86_64-cascadelake-linux_gnu-llvm_cpu] local_task(embedded_elf)[1-thread,full-inference,default-flags] with default @ c2-standard-60[cpu] 181.289 (1.0X) N/A 189.493 (1.0X)
MobileBertSquad_int8(tflite) [x86_64-cascadelake-linux_gnu-llvm_cpu] local_task(embedded_elf)[15-thread,full-inference,default-flags] with default @ c2-standard-60[cpu] 68.484 (1.0X) N/A 63.960 (1.1X)
MobileBertSquad_int8(tflite) [x86_64-cascadelake-linux_gnu-llvm_cpu] local_task(embedded_elf)[1-thread,full-inference,default-flags] with default @ c2-standard-60[cpu] 517.199 (1.0X) N/A 240.747 (2.1X)
MobileNetV1_fp32(tflite) [x86_64-cascadelake-linux_gnu-llvm_cpu] local_task(embedded_elf)[8-thread,full-inference,default-flags] with default @ c2-standard-60[cpu] 4.830 (1.0X) N/A 4.569 (1.1X)
MobileNetV1_fp32(tflite) [x86_64-cascadelake-linux_gnu-llvm_cpu] local_task(embedded_elf)[1-thread,full-inference,default-flags] with default @ c2-standard-60[cpu] 24.879 (1.0X) N/A 17.823 (1.4X)
MobileNetV2_fp32(tflite) [x86_64-cascadelake-linux_gnu-llvm_cpu] local_task(embedded_elf)[8-thread,full-inference,default-flags] with default @ c2-standard-60[cpu] 3.759 (1.0X) N/A 4.880 (0.8X)
MobileNetV2_fp32(tflite) [x86_64-cascadelake-linux_gnu-llvm_cpu] local_task(embedded_elf)[1-thread,full-inference,default-flags] with default @ c2-standard-60[cpu] 11.550 (1.0X) N/A 11.545 (1.0X)
MobileNetV2_int8(tflite) [x86_64-cascadelake-linux_gnu-llvm_cpu] local_task(embedded_elf)[8-thread,full-inference,default-flags] with default @ c2-standard-60[cpu] 5.832 (1.0X) N/A 5.419 (1.1X)
MobileNetV2_int8(tflite) [x86_64-cascadelake-linux_gnu-llvm_cpu] local_task(embedded_elf)[1-thread,full-inference,default-flags] with default @ c2-standard-60[cpu] 21.571 (1.0X) N/A 11.802 (1.8X)
MobileNetV3Small_fp32(tflite) [x86_64-cascadelake-linux_gnu-llvm_cpu] local_task(embedded_elf)[1-thread,full-inference,default-flags] with default @ c2-standard-60[cpu] 2.848 (1.0X) N/A 2.815 (1.0X)
MobileNetV3Small_fp32(tflite) [x86_64-cascadelake-linux_gnu-llvm_cpu] local_sync(embedded_elf)[full-inference,default-flags] with default @ c2-standard-60[cpu] 2.843 (1.0X) N/A 2.724 (1.0X)
MobileSSD_fp32(tflite) [x86_64-cascadelake-linux_gnu-llvm_cpu] local_task(embedded_elf)[8-thread,full-inference,default-flags] with default @ c2-standard-60[cpu] 8.462 (1.0X) N/A 9.823 (0.9X)
MobileSSD_fp32(tflite) [x86_64-cascadelake-linux_gnu-llvm_cpu] local_task(embedded_elf)[1-thread,full-inference,default-flags] with default @ c2-standard-60[cpu] 34.376 (1.0X) N/A 31.267 (1.1X)
PersonDetect_int8(tflite) [x86_64-cascadelake-linux_gnu-llvm_cpu] local_task(embedded_elf)[1-thread,full-inference,default-flags] with default @ c2-standard-60[cpu] 0.767 (1.0X) N/A 0.631 (1.2X)
PersonDetect_int8(tflite) [x86_64-cascadelake-linux_gnu-llvm_cpu] local_sync(embedded_elf)[full-inference,default-flags] with default @ c2-standard-60[cpu] 0.699 (1.0X) N/A 0.568 (1.2X)
PoseNet_fp32(tflite) [x86_64-cascadelake-linux_gnu-llvm_cpu] local_task(embedded_elf)[8-thread,full-inference,default-flags] with default @ c2-standard-60[cpu] 4.128 (1.0X) N/A 5.156 (0.8X)
PoseNet_fp32(tflite) [x86_64-cascadelake-linux_gnu-llvm_cpu] local_task(embedded_elf)[1-thread,full-inference,default-flags] with default @ c2-standard-60[cpu] 17.644 (1.0X) N/A 19.645 (0.9X)
matmul_256x256x2048_i8_i4_i32_tile_config_default(linalg) [x86_64-cascadelake-linux_gnu-llvm_cpu] local_sync(embedded_elf)[full-inference,default-flags] with default @ c2-standard-60[cpu] 7.556 (1.0X) N/A 7.571 (1.0X)
DeepLabV3_fp32(tflite) [armv8.2-a-generic-linux_android29-llvm_cpu] local_sync(embedded_elf)[full-inference,default-flags] with default @ pixel-6-pro[big-cores] 49.011 (1.0X) N/A 44.412 (1.1X)
DeepLabV3_fp32(tflite) [armv8.2-a-generic-linux_android29-llvm_cpu] local_task(embedded_elf)[1-thread,full-inference,system-scheduling] with default @ pixel-6-pro[big-cores] 50.220 (1.0X) N/A 44.950 (1.1X)
DeepLabV3_fp32(tflite) [armv8.2-a-generic-linux_android29-llvm_cpu] local_task(embedded_elf)[2-thread,full-inference,system-scheduling] with default @ pixel-6-pro[big-cores] 30.156 (1.0X) N/A 27.873 (1.1X)
GPT2_117M_TF_1X1XI32(stablehlo) [armv8.2-a-generic-linux_android29-llvm_cpu] local_sync(embedded_elf)[full-inference,default-flags] with default @ pixel-6-pro[big-cores] 92.587 (1.0X) N/A 21.878 (4.2X)
GPT2_117M_TF_1X1XI32(stablehlo) [armv8.2-a-generic-linux_android29-llvm_cpu] local_task(embedded_elf)[1-thread,full-inference,system-scheduling] with default @ pixel-6-pro[big-cores] 91.902 (1.0X) N/A 22.881 (4.0X)
GPT2_117M_TF_1X1XI32(stablehlo) [armv8.2-a-generic-linux_android29-llvm_cpu] local_task(embedded_elf)[2-thread,full-inference,system-scheduling] with default @ pixel-6-pro[big-cores] 52.074 (1.0X) N/A 22.052 (2.4X)
GPT2_117M_TF_1X4XI32(stablehlo) [armv8.2-a-generic-linux_android29-llvm_cpu] local_sync(embedded_elf)[full-inference,default-flags] with default @ pixel-6-pro[big-cores] 138.227 (1.0X) N/A 27.606 (5.0X)
GPT2_117M_TF_1X4XI32(stablehlo) [armv8.2-a-generic-linux_android29-llvm_cpu] local_task(embedded_elf)[1-thread,full-inference,system-scheduling] with default @ pixel-6-pro[big-cores] 124.005 (1.0X) N/A 29.448 (4.2X)
GPT2_117M_TF_1X4XI32(stablehlo) [armv8.2-a-generic-linux_android29-llvm_cpu] local_task(embedded_elf)[2-thread,full-inference,system-scheduling] with default @ pixel-6-pro[big-cores] 76.970 (1.0X) N/A 26.520 (2.9X)
MobileBertSquad_fp32(tflite) [armv8.2-a-generic-linux_android29-llvm_cpu] local_sync(embedded_elf)[full-inference,default-flags] with default @ pixel-6-pro[big-cores] 708.956 (1.0X) N/A 356.168 (2.0X)
MobileBertSquad_fp32(tflite) [armv8.2-a-generic-linux_android29-llvm_cpu] local_task(embedded_elf)[1-thread,full-inference,system-scheduling] with default @ pixel-6-pro[big-cores] 704.307 (1.0X) N/A 361.714 (1.9X)
MobileBertSquad_fp32(tflite) [armv8.2-a-generic-linux_android29-llvm_cpu] local_task(embedded_elf)[2-thread,full-inference,system-scheduling] with default @ pixel-6-pro[big-cores] 397.287 (1.0X) N/A 216.870 (1.8X)
MobileBertSquad_int8(tflite) [armv8.2-a-generic-linux_android29-llvm_cpu] local_sync(embedded_elf)[full-inference,default-flags] with default @ pixel-6-pro[big-cores] 1059.008 (1.0X) N/A 266.730 (4.0X)
MobileBertSquad_int8(tflite) [armv8.2-a-generic-linux_android29-llvm_cpu] local_task(embedded_elf)[1-thread,full-inference,system-scheduling] with default @ pixel-6-pro[big-cores] 1058.553 (1.0X) N/A 265.371 (4.0X)
MobileBertSquad_int8(tflite) [armv8.2-a-generic-linux_android29-llvm_cpu] local_task(embedded_elf)[2-thread,full-inference,system-scheduling] with default @ pixel-6-pro[big-cores] 552.999 (1.0X) N/A 157.294 (3.5X)
Vit_int8(tflite) [armv8.2-a-generic-linux_android29-llvm_cpu] local_sync(embedded_elf)[full-inference,default-flags] with default @ pixel-6-pro[big-cores] 2104.567 (1.0X) N/A 305.475 (6.9X)
Vit_int8(tflite) [armv8.2-a-generic-linux_android29-llvm_cpu] local_task(embedded_elf)[1-thread,full-inference,system-scheduling] with default @ pixel-6-pro[big-cores] 2105.228 (1.0X) N/A 308.456 (6.8X)
Vit_int8(tflite) [armv8.2-a-generic-linux_android29-llvm_cpu] local_task(embedded_elf)[2-thread,full-inference,system-scheduling] with default @ pixel-6-pro[big-cores] 1134.590 (1.0X) N/A 185.222 (6.1X)
matmul_256x256x2048_i8_i4_i32_tile_config_default(linalg) [armv8.2-a-generic-linux_android29-llvm_cpu] local_sync(embedded_elf)[full-inference,default-flags] with default @ pixel-6-pro[big-cores] 12.274 (1.0X) N/A 1.323 (9.3X)

Regressed Latencies 🚩

Benchmark Name Average Latency (ms) Median Latency (ms) Latency Standard Deviation (ms)
MobileBertSquad\_int8(tflite) [arm-valhall-vulkan\_android31-vulkan\_spirv][experimental-flags,fuse-padding,max-concurrency] vulkan(none)[full-inference,default-flags] with default @ pixel-6-pro[gpu] 103.830 (vs. 75.807, 36.97%↑) 104.044 0.845
MobileBertSquad\_int8(tflite) [arm-valhall-vulkan\_android31-vulkan\_spirv][default-flags] vulkan(none)[full-inference,default-flags] with default @ pixel-6-pro[gpu] 113.657 (vs. 87.071, 30.53%↑) 113.557 0.731
MobileBertSquad\_fp16(tflite) [arm-valhall-vulkan\_android31-vulkan\_spirv][experimental-flags,fuse-padding,max-concurrency,demote-f32-to-f16] vulkan(none)[full-inference,default-flags] with default @ pixel-6-pro[gpu] 108.362 (vs. 93.941, 15.35%↑) 110.493 5.122
MobileBertSquad\_int8(tflite) [arm-valhall-vulkan\_android31-vulkan\_spirv][experimental-flags,fuse-padding,max-concurrency,repeated-kernel] vulkan(none)[full-inference,experimental-flags] with default @ pixel-6-pro[gpu] 85.032 (vs. 74.736, 13.78%↑) 85.031 0.673
MobileBertSquad\_int8(tflite) [x86\_64-cascadelake-linux\_gnu-llvm\_cpu][default-flags,dt-uk] local\_task(embedded\_elf)[1-thread,full-inference,default-flags] with default @ c2-standard-60[cpu] 240.747 (vs. 214.036, 12.48%↑) 240.229 1.431
GPT2\_117M\_TF\_1X4XI32(stablehlo) [armv8.2-a-generic-linux\_android29-llvm\_cpu][experimental-flags,no-dt] local\_task(embedded\_elf)[2-thread,full-inference,system-scheduling] with default @ pixel-6-pro[big-cores] 76.970 (vs. 71.196, 8.11%↑) 76.922 0.198

Improved Latencies 🎉

Benchmark Name Average Latency (ms) Median Latency (ms) Latency Standard Deviation (ms)
MobileBertSquad\_int8(tflite) [armv8.2-a-generic-linux\_android29-llvm\_cpu][default-flags,dt-uk] local\_task(embedded\_elf)[2-thread,full-inference,system-scheduling] with default @ pixel-6-pro[big-cores] 157.294 (vs. 191.253, 17.76%↓) 158.182 4.144
MobileBertSquad\_int8(tflite) [armv8.2-a-generic-linux\_android29-llvm\_cpu][default-flags,dt-uk] local\_sync(embedded\_elf)[full-inference,default-flags] with default @ pixel-6-pro[big-cores] 266.730 (vs. 322.032, 17.17%↓) 267.317 2.030
MobileBertSquad\_int8(tflite) [armv8.2-a-generic-linux\_android29-llvm\_cpu][default-flags,dt-uk] local\_task(embedded\_elf)[1-thread,full-inference,system-scheduling] with default @ pixel-6-pro[big-cores] 265.371 (vs. 313.399, 15.32%↓) 266.121 2.363
GPT2\_117M\_TF\_1X4XI32(stablehlo) [armv8.2-a-generic-linux\_android29-llvm\_cpu][experimental-flags,no-dt] local\_task(embedded\_elf)[1-thread,full-inference,system-scheduling] with default @ pixel-6-pro[big-cores] 124.005 (vs. 134.159, 7.57%↓) 123.973 0.182
MobileBertSquad\_int8(tflite) [armv8.2-a-generic-linux\_android29-llvm\_cpu][experimental-flags,no-dt] local\_task(embedded\_elf)[2-thread,full-inference,system-scheduling] with default @ pixel-6-pro[big-cores] 552.999 (vs. 596.173, 7.24%↓) 554.962 8.253
MobileBertSquad\_fp32(tflite) [armv8.2-a-generic-linux\_android29-llvm\_cpu][default-flags,dt-uk] local\_task(embedded\_elf)[2-thread,full-inference,system-scheduling] with default @ pixel-6-pro[big-cores] 216.870 (vs. 229.792, 5.62%↓) 219.486 6.749
GPT2\_117M\_TF\_1X4XI32(stablehlo) [armv8.2-a-generic-linux\_android29-llvm\_cpu][default-flags,dt-uk] local\_task(embedded\_elf)[1-thread,full-inference,system-scheduling] with default @ pixel-6-pro[big-cores] 29.448 (vs. 31.189, 5.58%↓) 29.555 0.794
GPT2\_117M\_TF\_1X4XI32(stablehlo) [armv8.2-a-generic-linux\_android29-llvm\_cpu][default-flags,dt-uk] local\_sync(embedded\_elf)[full-inference,default-flags] with default @ pixel-6-pro[big-cores] 27.606 (vs. 29.228, 5.55%↓) 27.130 2.052
MobileBertSquad\_int8(tflite) [armv8.2-a-generic-linux\_android29-llvm\_cpu][experimental-flags,no-dt] local\_sync(embedded\_elf)[full-inference,default-flags] with default @ pixel-6-pro[big-cores] 1059.008 (vs. 1116.992, 5.19%↓) 1058.844 0.980
DeepLabV3\_fp32(tflite) [armv8.2-a-generic-linux\_android29-llvm\_cpu][default-flags,dt-uk] local\_task(embedded\_elf)[2-thread,full-inference,system-scheduling] with default @ pixel-6-pro[big-cores] 27.873 (vs. 29.399, 5.19%↓) 28.124 0.810
Vit\_int8(tflite) [armv8.2-a-generic-linux\_android29-llvm\_cpu][default-flags,dt-uk] local\_task(embedded\_elf)[2-thread,full-inference,system-scheduling] with default @ pixel-6-pro[big-cores] 185.222 (vs. 195.244, 5.13%↓) 186.075 4.428
MobileBertSquad\_int8(tflite) [armv8.2-a-generic-linux\_android29-llvm\_cpu][experimental-flags,no-dt] local\_task(embedded\_elf)[1-thread,full-inference,system-scheduling] with default @ pixel-6-pro[big-cores] 1058.553 (vs. 1115.685, 5.12%↓) 1058.330 0.942

Similar Latencies

Benchmark Name Average Latency (ms) Median Latency (ms) Latency Standard Deviation (ms)
MobileNetV2\_fp32(tflite) [x86\_64-cascadelake-linux\_gnu-llvm\_cpu][experimental-flags,no-dt] local\_task(embedded\_elf)[1-thread,full-inference,default-flags] with default @ c2-standard-60[cpu] 11.550 (vs. 12.437, 7.13%↓) 11.560 0.073
MobileBertSquad\_int8(tflite) [x86\_64-cascadelake-linux\_gnu-llvm\_cpu][experimental-flags,no-dt] local\_task(embedded\_elf)[1-thread,full-inference,default-flags] with default @ c2-standard-60[cpu] 517.199 (vs. 488.208, 5.94%↑) 516.272 2.223
GPT2\_117M\_TF\_1X1XI32(stablehlo) [armv8.2-a-generic-linux\_android29-llvm\_cpu][default-flags,dt-uk] local\_task(embedded\_elf)[2-thread,full-inference,system-scheduling] with default @ pixel-6-pro[big-cores] 22.052 (vs. 23.165, 4.81%↓) 22.082 0.197
Vit\_int8(tflite) [armv8.2-a-generic-linux\_android29-llvm\_cpu][default-flags,dt-uk] local\_sync(embedded\_elf)[full-inference,default-flags] with default @ pixel-6-pro[big-cores] 305.475 (vs. 320.020, 4.55%↓) 306.750 3.685
GPT2\_117M\_TF\_1X4XI32(stablehlo) [armv8.2-a-generic-linux\_android29-llvm\_cpu][default-flags,dt-uk] local\_task(embedded\_elf)[2-thread,full-inference,system-scheduling] with default @ pixel-6-pro[big-cores] 26.520 (vs. 27.771, 4.50%↓) 26.585 0.333
MobileBertSquad\_fp32(tflite) [armv8.2-a-generic-linux\_android29-llvm\_cpu][default-flags,dt-uk] local\_task(embedded\_elf)[1-thread,full-inference,system-scheduling] with default @ pixel-6-pro[big-cores] 361.714 (vs. 377.610, 4.21%↓) 364.581 6.481
MobileBertSquad\_fp32(tflite) [armv8.2-a-generic-linux\_android29-llvm\_cpu][experimental-flags,no-dt] local\_task(embedded\_elf)[1-thread,full-inference,system-scheduling] with default @ pixel-6-pro[big-cores] 704.307 (vs. 734.843, 4.16%↓) 705.353 7.428
GPT2\_117M\_TF\_1X1XI32(stablehlo) [armv8.2-a-generic-linux\_android29-llvm\_cpu][default-flags,dt-uk] local\_sync(embedded\_elf)[full-inference,default-flags] with default @ pixel-6-pro[big-cores] 21.878 (vs. 22.804, 4.06%↓) 21.803 0.233
DeepLabV3\_fp32(tflite) [armv8.2-a-generic-linux\_android29-llvm\_cpu][default-flags,dt-uk] local\_task(embedded\_elf)[1-thread,full-inference,system-scheduling] with default @ pixel-6-pro[big-cores] 44.950 (vs. 46.780, 3.91%↓) 45.176 0.515
MobileBertSquad\_fp32(tflite) [armv8.2-a-generic-linux\_android29-llvm\_cpu][experimental-flags,no-dt] local\_task(embedded\_elf)[2-thread,full-inference,system-scheduling] with default @ pixel-6-pro[big-cores] 397.287 (vs. 412.555, 3.70%↓) 397.493 9.938
DeepLabV3\_fp32(tflite) [armv8.2-a-generic-linux\_android29-llvm\_cpu][experimental-flags,no-dt] local\_task(embedded\_elf)[2-thread,full-inference,system-scheduling] with default @ pixel-6-pro[big-cores] 30.156 (vs. 31.282, 3.60%↓) 30.344 0.834
Vit\_int8(tflite) [armv8.2-a-generic-linux\_android29-llvm\_cpu][experimental-flags,no-dt] local\_task(embedded\_elf)[2-thread,full-inference,system-scheduling] with default @ pixel-6-pro[big-cores] 1134.590 (vs. 1176.617, 3.57%↓) 1138.837 17.873
MiniLML12H384Uncased(stablehlo) [cuda-sm\_80-linux\_gnu-cuda][default-flags] cuda(none)[full-inference,default-flags] with default @ a2-highgpu-1g[gpu] 1.655 (vs. 1.600, 3.48%↑) 1.653 0.012
MobileBertSquad\_int8(tflite) [x86\_64-cascadelake-linux\_gnu-llvm\_cpu][experimental-flags,no-dt] local\_task(embedded\_elf)[15-thread,full-inference,default-flags] with default @ c2-standard-60[cpu] 68.484 (vs. 66.257, 3.36%↑) 68.239 0.497
GPT2\_117M\_TF\_1X1XI32(stablehlo) [armv8.2-a-generic-linux\_android29-llvm\_cpu][default-flags,dt-uk] local\_task(embedded\_elf)[1-thread,full-inference,system-scheduling] with default @ pixel-6-pro[big-cores] 22.881 (vs. 23.672, 3.34%↓) 22.873 0.197
EfficientNetV2STF(stablehlo) [x86\_64-cascadelake-linux\_gnu-llvm\_cpu][default-flags,dt-uk] local\_task(embedded\_elf)[1-thread,full-inference,default-flags] with default @ c2-standard-60[cpu] 236.681 (vs. 229.189, 3.27%↑) 234.568 4.593
matmul\_128x256x8192\_f16t\_tile\_config\_default(linalg) [cuda-sm\_80-linux\_gnu-cuda][ukernel,matmul,splitk] cuda(none)[full-inference,default-flags] with default @ a2-highgpu-1g[gpu] 0.026 (vs. 0.025, 3.22%↑) 0.026 0.000
MobileBertSquad\_int8(tflite) [x86\_64-cascadelake-linux\_gnu-llvm\_cpu][default-flags,dt-uk] local\_task(embedded\_elf)[15-thread,full-inference,default-flags] with default @ c2-standard-60[cpu] 63.960 (vs. 61.972, 3.21%↑) 63.876 0.373
GPT2\_117M\_TF\_1X4XI32(stablehlo) [x86\_64-cascadelake-linux\_gnu-llvm\_cpu][default-flags,dt-uk] local\_task(embedded\_elf)[1-thread,full-inference,default-flags] with default @ c2-standard-60[cpu] 39.165 (vs. 40.421, 3.11%↓) 38.727 1.224
MobileBertSquad\_fp16(tflite) [x86\_64-cascadelake-linux\_gnu-llvm\_cpu][experimental-flags,no-dt] local\_task(embedded\_elf)[15-thread,full-inference,default-flags] with default @ c2-standard-60[cpu] 34.045 (vs. 33.040, 3.04%↑) 33.652 0.852
PoseNet\_fp32(tflite) [x86\_64-cascadelake-linux\_gnu-llvm\_cpu][default-flags,dt-uk] local\_task(embedded\_elf)[1-thread,full-inference,default-flags] with default @ c2-standard-60[cpu] 19.645 (vs. 19.077, 2.98%↑) 19.648 0.095
MiniLML12H384Uncased(stablehlo) [x86\_64-cascadelake-linux\_gnu-llvm\_cpu][experimental-flags,no-dt] local\_task(embedded\_elf)[15-thread,full-inference,default-flags] with default @ c2-standard-60[cpu] 12.215 (vs. 12.574, 2.86%↓) 12.027 0.303
Vit\_int8(tflite) [armv8.2-a-generic-linux\_android29-llvm\_cpu][default-flags,dt-uk] local\_task(embedded\_elf)[1-thread,full-inference,system-scheduling] with default @ pixel-6-pro[big-cores] 308.456 (vs. 316.980, 2.69%↓) 308.289 2.314
DeepLabV3\_fp32(tflite) [armv8.2-a-generic-linux\_android29-llvm\_cpu][experimental-flags,no-dt] local\_task(embedded\_elf)[1-thread,full-inference,system-scheduling] with default @ pixel-6-pro[big-cores] 50.220 (vs. 51.555, 2.59%↓) 50.336 0.721
MobileSSD\_fp32(tflite) [x86\_64-cascadelake-linux\_gnu-llvm\_cpu][experimental-flags,no-dt] local\_task(embedded\_elf)[1-thread,full-inference,default-flags] with default @ c2-standard-60[cpu] 34.376 (vs. 35.254, 2.49%↓) 34.349 0.242
MobileNetV1\_fp32(tflite) [x86\_64-cascadelake-linux\_gnu-llvm\_cpu][experimental-flags,no-dt] local\_task(embedded\_elf)[1-thread,full-inference,default-flags] with default @ c2-standard-60[cpu] 24.879 (vs. 25.500, 2.43%↓) 24.847 0.178
GPT2\_117M\_TF\_1X1XI32(stablehlo) [x86\_64-cascadelake-linux\_gnu-llvm\_cpu][default-flags,dt-uk] local\_task(embedded\_elf)[1-thread,full-inference,default-flags] with default @ c2-standard-60[cpu] 37.586 (vs. 38.464, 2.28%↓) 37.102 1.317
MobileBertSquad\_fp32(tflite) [armv8.2-a-generic-linux\_android29-llvm\_cpu][experimental-flags,no-dt] local\_sync(embedded\_elf)[full-inference,default-flags] with default @ pixel-6-pro[big-cores] 708.956 (vs. 725.372, 2.26%↓) 713.314 11.493
MobileBertSquad\_fp32(tflite) [arm-valhall-vulkan\_android31-vulkan\_spirv][experimental-flags,fuse-padding,max-concurrency,repeated-kernel] vulkan(none)[full-inference,experimental-flags] with default @ pixel-6-pro[gpu] 123.126 (vs. 125.648, 2.01%↓) 123.367 0.508
matmul\_256x256x2048\_i8\_i4\_i32\_tile\_config\_default(linalg) [armv8.2-a-generic-linux\_android29-llvm\_cpu][default-flags,dt-uk] local\_sync(embedded\_elf)[full-inference,default-flags] with default @ pixel-6-pro[big-cores] 1.323 (vs. 1.350, 1.99%↓) 1.329 0.014
BertForMaskedLMTF(stablehlo) [cuda-sm\_80-linux\_gnu-cuda][default-flags] cuda(none)[full-inference,default-flags] with default @ a2-highgpu-1g[gpu] 7.063 (vs. 6.930, 1.92%↑) 7.045 0.078
BertLargeTF(stablehlo) [x86\_64-cascadelake-linux\_gnu-llvm\_cpu][experimental-flags,no-dt] local\_task(embedded\_elf)[30-thread,full-inference,default-flags] with default @ c2-standard-60[cpu] 780.636 (vs. 766.026, 1.91%↑) 773.562 40.203
MobileNetV3Small\_fp32(tflite) [x86\_64-cascadelake-linux\_gnu-llvm\_cpu][default-flags,dt-uk] local\_sync(embedded\_elf)[full-inference,default-flags] with default @ c2-standard-60[cpu] 2.724 (vs. 2.673, 1.89%↑) 2.720 0.027
GPT2\_117M\_TF\_1X1XI32(stablehlo) [armv8.2-a-generic-linux\_android29-llvm\_cpu][experimental-flags,no-dt] local\_task(embedded\_elf)[1-thread,full-inference,system-scheduling] with default @ pixel-6-pro[big-cores] 91.902 (vs. 93.605, 1.82%↓) 91.935 0.093
DeepLabV3\_fp32(tflite) [armv8.2-a-generic-linux\_android29-llvm\_cpu][default-flags,dt-uk] local\_sync(embedded\_elf)[full-inference,default-flags] with default @ pixel-6-pro[big-cores] 44.412 (vs. 45.213, 1.77%↓) 44.436 0.359
matmul\_256x256x2048\_i8\_i4\_i32\_tile\_config\_default(linalg) [armv8.2-a-generic-linux\_android29-llvm\_cpu][experimental-flags,no-dt] local\_sync(embedded\_elf)[full-inference,default-flags] with default @ pixel-6-pro[big-cores] 12.274 (vs. 12.485, 1.69%↓) 12.300 0.073
MobileBertSquad\_fp16(tflite) [arm-valhall-vulkan\_android31-vulkan\_spirv][default-flags,demote-f32-to-f16] vulkan(none)[full-inference,default-flags] with default @ pixel-6-pro[gpu] 86.252 (vs. 84.859, 1.64%↑) 86.371 0.767
MobileBertSquad\_fp32(tflite) [armv8.2-a-generic-linux\_android29-llvm\_cpu][default-flags,dt-uk] local\_sync(embedded\_elf)[full-inference,default-flags] with default @ pixel-6-pro[big-cores] 356.168 (vs. 362.070, 1.63%↓) 357.268 5.649
MiniLML12H384Uncased(stablehlo) [x86\_64-cascadelake-linux\_gnu-llvm\_cpu][experimental-flags,no-dt] local\_task(embedded\_elf)[1-thread,full-inference,default-flags] with default @ c2-standard-60[cpu] 80.271 (vs. 79.037, 1.56%↑) 79.920 1.089
EfficientNetV2STF(stablehlo) [cuda-sm\_80-linux\_gnu-cuda][default-flags] cuda(none)[full-inference,default-flags] with default @ a2-highgpu-1g[gpu] 4.375 (vs. 4.308, 1.56%↑) 4.323 0.114
MobileNetV1\_fp32(tflite) [x86\_64-cascadelake-linux\_gnu-llvm\_cpu][experimental-flags,no-dt] local\_task(embedded\_elf)[8-thread,full-inference,default-flags] with default @ c2-standard-60[cpu] 4.830 (vs. 4.757, 1.52%↑) 4.813 0.080
BertForMaskedLMTF(stablehlo) [x86\_64-cascadelake-linux\_gnu-llvm\_cpu][experimental-flags,no-dt] local\_task(embedded\_elf)[30-thread,full-inference,default-flags] with default @ c2-standard-60[cpu] 218.814 (vs. 221.925, 1.40%↓) 213.264 15.249
DeepLabV3\_fp32(tflite) [x86\_64-cascadelake-linux\_gnu-llvm\_cpu][experimental-flags,no-dt] local\_task(embedded\_elf)[8-thread,full-inference,default-flags] with default @ c2-standard-60[cpu] 6.905 (vs. 7.001, 1.37%↓) 6.906 0.026
EfficientNet\_int8(tflite) [x86\_64-cascadelake-linux\_gnu-llvm\_cpu][default-flags,dt-uk] local\_task(embedded\_elf)[1-thread,full-inference,default-flags] with default @ c2-standard-60[cpu] 13.231 (vs. 13.053, 1.36%↑) 13.213 0.037
matmul\_128x256x8192\_f32t\_tile\_config\_default(linalg) [cuda-sm\_80-linux\_gnu-cuda][ukernel,matmul,splitk] cuda(none)[full-inference,default-flags] with default @ a2-highgpu-1g[gpu] 0.048 (vs. 0.047, 1.36%↑) 0.048 0.000
GPT2\_117M\_TF\_1X1XI32(stablehlo) [x86\_64-cascadelake-linux\_gnu-llvm\_cpu][experimental-flags,no-dt] local\_task(embedded\_elf)[15-thread,full-inference,default-flags] with default @ c2-standard-60[cpu] 9.085 (vs. 9.207, 1.32%↓) 8.905 0.426
EfficientNetV2STF(stablehlo) [x86\_64-cascadelake-linux\_gnu-llvm\_cpu][experimental-flags,no-dt] local\_task(embedded\_elf)[1-thread,full-inference,default-flags] with default @ c2-standard-60[cpu] 276.144 (vs. 272.673, 1.27%↑) 275.898 3.682
MobileNetV2\_fp32(tflite) [x86\_64-cascadelake-linux\_gnu-llvm\_cpu][default-flags,dt-uk] local\_task(embedded\_elf)[1-thread,full-inference,default-flags] with default @ c2-standard-60[cpu] 11.545 (vs. 11.406, 1.22%↑) 11.554 0.114
MobileBertSquad\_fp32(tflite) [x86\_64-cascadelake-linux\_gnu-llvm\_cpu][experimental-flags,no-dt] local\_task(embedded\_elf)[15-thread,full-inference,default-flags] with default @ c2-standard-60[cpu] 34.229 (vs. 33.850, 1.12%↑) 33.741 0.912
GPT2\_117M\_TF\_1X1XI32(stablehlo) [armv8.2-a-generic-linux\_android29-llvm\_cpu][experimental-flags,no-dt] local\_task(embedded\_elf)[2-thread,full-inference,system-scheduling] with default @ pixel-6-pro[big-cores] 52.074 (vs. 52.659, 1.11%↓) 52.065 0.046
GPT2\_117M\_TF\_1X1XI32(stablehlo) [x86\_64-cascadelake-linux\_gnu-llvm\_cpu][experimental-flags,no-dt] local\_task(embedded\_elf)[1-thread,full-inference,default-flags] with default @ c2-standard-60[cpu] 70.114 (vs. 70.900, 1.11%↓) 70.326 2.694
MobileBertSquad\_fp32(tflite) [arm-valhall-vulkan\_android31-vulkan\_spirv][experimental-flags,fuse-padding,max-concurrency] vulkan(none)[full-inference,default-flags] with default @ pixel-6-pro[gpu] 105.623 (vs. 106.746, 1.05%↓) 105.751 0.403
matmul\_3456x1024x2048\_f32t\_tile\_config\_default(linalg) [cuda-sm\_80-linux\_gnu-cuda][ukernel,matmul] cuda(none)[full-inference,default-flags] with default @ a2-highgpu-1g[gpu] 0.166 (vs. 0.164, 1.03%↑) 0.166 0.000
MobileNetV2\_fp32(tflite) [x86\_64-cascadelake-linux\_gnu-llvm\_cpu][default-flags,dt-uk] local\_task(embedded\_elf)[8-thread,full-inference,default-flags] with default @ c2-standard-60[cpu] 4.880 (vs. 4.930, 1.02%↓) 4.878 0.022
GPT2\_117M\_TF\_1X4XI32(stablehlo) [x86\_64-cascadelake-linux\_gnu-llvm\_cpu][experimental-flags,no-dt] local\_task(embedded\_elf)[15-thread,full-inference,default-flags] with default @ c2-standard-60[cpu] 10.986 (vs. 11.098, 1.01%↓) 10.620 0.614
MobileSSD\_fp32(tflite) [x86\_64-cascadelake-linux\_gnu-llvm\_cpu][default-flags,dt-uk] local\_task(embedded\_elf)[8-thread,full-inference,default-flags] with default @ c2-standard-60[cpu] 9.823 (vs. 9.922, 1.00%↓) 9.819 0.048
EfficientNetV2STF(stablehlo) [x86\_64-cascadelake-linux\_gnu-llvm\_cpu][experimental-flags,no-dt] local\_task(embedded\_elf)[15-thread,full-inference,default-flags] with default @ c2-standard-60[cpu] 35.762 (vs. 36.114, 0.98%↓) 35.226 0.952
MobileNetV3Small\_fp32(tflite) [x86\_64-cascadelake-linux\_gnu-llvm\_cpu][experimental-flags,no-dt] local\_sync(embedded\_elf)[full-inference,default-flags] with default @ c2-standard-60[cpu] 2.843 (vs. 2.818, 0.91%↑) 2.839 0.012
MobileNetV2\_int8(tflite) [x86\_64-cascadelake-linux\_gnu-llvm\_cpu][experimental-flags,no-dt] local\_task(embedded\_elf)[8-thread,full-inference,default-flags] with default @ c2-standard-60[cpu] 5.832 (vs. 5.885, 0.90%↓) 5.828 0.016
MobileNetV3Small\_fp32(tflite) [x86\_64-cascadelake-linux\_gnu-llvm\_cpu][experimental-flags,no-dt] local\_task(embedded\_elf)[1-thread,full-inference,default-flags] with default @ c2-standard-60[cpu] 2.848 (vs. 2.873, 0.89%↓) 2.844 0.011
matmul\_123x2561x2561\_f32t\_f32t\_f32t\_tile\_config\_default(linalg) [cuda-sm\_80-linux\_gnu-cuda][ukernel,matmul] cuda(none)[full-inference,default-flags] with default @ a2-highgpu-1g[gpu] 0.202 (vs. 0.200, 0.85%↑) 0.202 0.001
matmul\_2560x2560x2560\_f16t\_tile\_config\_default(linalg) [cuda-sm\_80-linux\_gnu-cuda][ukernel,matmul] cuda(none)[full-inference,default-flags] with default @ a2-highgpu-1g[gpu] 0.142 (vs. 0.141, 0.81%↑) 0.142 0.000
matmul\_3456x1024x2048\_f16t\_tile\_config\_default(linalg) [cuda-sm\_80-linux\_gnu-cuda][ukernel,matmul] cuda(none)[full-inference,default-flags] with default @ a2-highgpu-1g[gpu] 0.063 (vs. 0.063, 0.81%↑) 0.063 0.000
MobileBertSquad\_fp16(tflite) [x86\_64-cascadelake-linux\_gnu-llvm\_cpu][experimental-flags,no-dt] local\_task(embedded\_elf)[1-thread,full-inference,default-flags] with default @ c2-standard-60[cpu] 181.271 (vs. 179.845, 0.79%↑) 180.226 3.010
Vit\_int8(tflite) [armv8.2-a-generic-linux\_android29-llvm\_cpu][experimental-flags,no-dt] local\_sync(embedded\_elf)[full-inference,default-flags] with default @ pixel-6-pro[big-cores] 2104.567 (vs. 2120.303, 0.74%↓) 2105.458 1.938
PoseNet\_fp32(tflite) [x86\_64-cascadelake-linux\_gnu-llvm\_cpu][experimental-flags,no-dt] local\_task(embedded\_elf)[8-thread,full-inference,default-flags] with default @ c2-standard-60[cpu] 4.128 (vs. 4.098, 0.73%↑) 4.127 0.015
BertForMaskedLMTF(stablehlo) [x86\_64-cascadelake-linux\_gnu-llvm\_cpu][default-flags,dt-uk] local\_task(embedded\_elf)[30-thread,full-inference,default-flags] with default @ c2-standard-60[cpu] 106.242 (vs. 107.024, 0.73%↓) 105.823 1.043
MobileNetV3Small\_fp32(tflite) [x86\_64-cascadelake-linux\_gnu-llvm\_cpu][default-flags,dt-uk] local\_task(embedded\_elf)[1-thread,full-inference,default-flags] with default @ c2-standard-60[cpu] 2.815 (vs. 2.795, 0.70%↑) 2.809 0.029
MobileNetV3Small\_fp32(tflite) [vmvx-generic-vmvx-vmvx][experimental-flags] local\_task(vmvx\_module)[8-thread,full-inference,default-flags] with default @ c2-standard-60[cpu] 1818.428 (vs. 1805.802, 0.70%↑) 1818.262 1.056
GPT2\_117M\_TF\_1X4XI32(stablehlo) [x86\_64-cascadelake-linux\_gnu-llvm\_cpu][experimental-flags,no-dt] local\_task(embedded\_elf)[1-thread,full-inference,default-flags] with default @ c2-standard-60[cpu] 87.114 (vs. 87.722, 0.69%↓) 86.453 3.834
DeepLabV3\_fp32(tflite) [x86\_64-cascadelake-linux\_gnu-llvm\_cpu][default-flags,dt-uk] local\_task(embedded\_elf)[1-thread,full-inference,default-flags] with default @ c2-standard-60[cpu] 30.118 (vs. 29.918, 0.67%↑) 30.081 0.124
EfficientNetV2STF(stablehlo) [x86\_64-cascadelake-linux\_gnu-llvm\_cpu][default-flags,dt-uk] local\_task(embedded\_elf)[15-thread,full-inference,default-flags] with default @ c2-standard-60[cpu] 34.370 (vs. 34.171, 0.58%↑) 34.290 0.319
PersonDetect\_int8(tflite) [x86\_64-cascadelake-linux\_gnu-llvm\_cpu][experimental-flags,no-dt] local\_task(embedded\_elf)[1-thread,full-inference,default-flags] with default @ c2-standard-60[cpu] 0.767 (vs. 0.772, 0.57%↓) 0.767 0.007
Vit\_int8(tflite) [armv8.2-a-generic-linux\_android29-llvm\_cpu][experimental-flags,no-dt] local\_task(embedded\_elf)[1-thread,full-inference,system-scheduling] with default @ pixel-6-pro[big-cores] 2105.228 (vs. 2117.173, 0.56%↓) 2105.014 2.040
BertLargeTF(stablehlo) [cuda-sm\_80-linux\_gnu-cuda][default-flags] cuda(none)[full-inference,default-flags] with default @ a2-highgpu-1g[gpu] 10.771 (vs. 10.715, 0.52%↑) 10.766 0.029
matmul\_256x256x2048\_i8\_i4\_i32\_tile\_config\_default(linalg) [x86\_64-cascadelake-linux\_gnu-llvm\_cpu][experimental-flags,no-dt] local\_sync(embedded\_elf)[full-inference,default-flags] with default @ c2-standard-60[cpu] 7.556 (vs. 7.596, 0.52%↓) 7.556 0.011
MiniLML12H384Uncased(stablehlo) [x86\_64-cascadelake-linux\_gnu-llvm\_cpu][default-flags,dt-uk] local\_task(embedded\_elf)[15-thread,full-inference,default-flags] with default @ c2-standard-60[cpu] 13.035 (vs. 13.103, 0.52%↓) 13.059 0.184
matmul\_2560x2560x2560\_f32t\_tile\_config\_default(linalg) [cuda-sm\_80-linux\_gnu-cuda][ukernel,matmul] cuda(none)[full-inference,default-flags] with default @ a2-highgpu-1g[gpu] 0.301 (vs. 0.299, 0.50%↑) 0.301 0.001
MobileNetV2\_int8(tflite) [x86\_64-cascadelake-linux\_gnu-llvm\_cpu][default-flags,dt-uk] local\_task(embedded\_elf)[1-thread,full-inference,default-flags] with default @ c2-standard-60[cpu] 11.802 (vs. 11.861, 0.49%↓) 11.811 0.087
GPT2\_117M\_TF\_1X1XI32(stablehlo) [x86\_64-cascadelake-linux\_gnu-llvm\_cpu][default-flags,dt-uk] local\_task(embedded\_elf)[15-thread,full-inference,default-flags] with default @ c2-standard-60[cpu] 8.749 (vs. 8.790, 0.47%↓) 8.696 0.125
matmul\_2562x2564x2562\_f32t\_f32t\_f32t\_tile\_config\_default(linalg) [cuda-sm\_80-linux\_gnu-cuda][ukernel,matmul] cuda(none)[full-inference,default-flags] with default @ a2-highgpu-1g[gpu] 1.116 (vs. 1.121, 0.46%↓) 1.113 0.009
MobileBertSquad\_fp16(tflite) [arm-valhall-vulkan\_android31-vulkan\_spirv][experimental-flags,fuse-padding,max-concurrency,repeated-kernel,demote-f32-to-f16] vulkan(none)[full-inference,experimental-flags] with default @ pixel-6-pro[gpu] 79.209 (vs. 79.576, 0.46%↓) 79.246 0.617
GPT2\_117M\_TF\_1X4XI32(stablehlo) [armv8.2-a-generic-linux\_android29-llvm\_cpu][experimental-flags,no-dt] local\_sync(embedded\_elf)[full-inference,default-flags] with default @ pixel-6-pro[big-cores] 138.227 (vs. 137.617, 0.44%↑) 138.109 0.463
DeepLabV3\_fp32(tflite) [x86\_64-cascadelake-linux\_gnu-llvm\_cpu][experimental-flags,no-dt] local\_task(embedded\_elf)[1-thread,full-inference,default-flags] with default @ c2-standard-60[cpu] 32.027 (vs. 31.899, 0.40%↑) 32.011 0.176
BertLargeTF(stablehlo) [x86\_64-cascadelake-linux\_gnu-llvm\_cpu][default-flags,dt-uk] local\_task(embedded\_elf)[30-thread,full-inference,default-flags] with default @ c2-standard-60[cpu] 222.751 (vs. 223.640, 0.40%↓) 222.523 2.028
MobileBertSquad\_fp32(tflite) [x86\_64-cascadelake-linux\_gnu-llvm\_cpu][experimental-flags,no-dt] local\_task(embedded\_elf)[1-thread,full-inference,default-flags] with default @ c2-standard-60[cpu] 181.289 (vs. 180.592, 0.39%↑) 180.203 2.643
MobileBertSquad\_fp32(tflite) [x86\_64-cascadelake-linux\_gnu-llvm\_cpu][default-flags,dt-uk] local\_task(embedded\_elf)[15-thread,full-inference,default-flags] with default @ c2-standard-60[cpu] 61.197 (vs. 61.431, 0.38%↓) 60.888 0.650
matmul\_2564x2564x2564\_f32t\_f32t\_f32t\_tile\_config\_default(linalg) [cuda-sm\_80-linux\_gnu-cuda][ukernel,matmul] cuda(none)[full-inference,default-flags] with default @ a2-highgpu-1g[gpu] 0.945 (vs. 0.942, 0.35%↑) 0.945 0.000
MobileBertSquad\_fp16(tflite) [x86\_64-cascadelake-linux\_gnu-llvm\_cpu][default-flags,dt-uk] local\_task(embedded\_elf)[1-thread,full-inference,default-flags] with default @ c2-standard-60[cpu] 185.715 (vs. 185.082, 0.34%↑) 184.513 3.075
PoseNet\_fp32(tflite) [x86\_64-cascadelake-linux\_gnu-llvm\_cpu][experimental-flags,no-dt] local\_task(embedded\_elf)[1-thread,full-inference,default-flags] with default @ c2-standard-60[cpu] 17.644 (vs. 17.703, 0.33%↓) 17.629 0.096
EfficientNet\_int8(tflite) [x86\_64-cascadelake-linux\_gnu-llvm\_cpu][experimental-flags,no-dt] local\_task(embedded\_elf)[8-thread,full-inference,default-flags] with default @ c2-standard-60[cpu] 5.804 (vs. 5.823, 0.33%↓) 5.790 0.037
MobileBertSquad\_fp32(tflite) [qualcomm-adreno-vulkan\_android31-vulkan\_spirv][experimental-flags,fuse-padding] vulkan(none)[full-inference,default-flags] with default @ moto-edge-x30[gpu] 259.405 (vs. 260.199, 0.31%↓) 259.441 0.639
EfficientNet\_int8(tflite) [x86\_64-cascadelake-linux\_gnu-llvm\_cpu][default-flags,dt-uk] local\_task(embedded\_elf)[8-thread,full-inference,default-flags] with default @ c2-standard-60[cpu] 5.033 (vs. 5.020, 0.25%↑) 5.039 0.028
DeepLabV3\_fp32(tflite) [armv8.2-a-generic-linux\_android29-llvm\_cpu][experimental-flags,no-dt] local\_sync(embedded\_elf)[full-inference,default-flags] with default @ pixel-6-pro[big-cores] 49.011 (vs. 48.894, 0.24%↑) 49.380 0.886
PersonDetect\_int8(tflite) [x86\_64-cascadelake-linux\_gnu-llvm\_cpu][experimental-flags,no-dt] local\_sync(embedded\_elf)[full-inference,default-flags] with default @ c2-standard-60[cpu] 0.699 (vs. 0.698, 0.23%↑) 0.699 0.001
DeepLabV3\_fp32(tflite) [x86\_64-cascadelake-linux\_gnu-llvm\_cpu][default-flags,dt-uk] local\_task(embedded\_elf)[8-thread,full-inference,default-flags] with default @ c2-standard-60[cpu] 8.525 (vs. 8.507, 0.22%↑) 8.522 0.039
MobileSSD\_fp32(tflite) [x86\_64-cascadelake-linux\_gnu-llvm\_cpu][default-flags,dt-uk] local\_task(embedded\_elf)[1-thread,full-inference,default-flags] with default @ c2-standard-60[cpu] 31.267 (vs. 31.204, 0.20%↑) 31.193 0.205
GPT2\_117M\_TF\_1X1XI32(stablehlo) [armv8.2-a-generic-linux\_android29-llvm\_cpu][experimental-flags,no-dt] local\_sync(embedded\_elf)[full-inference,default-flags] with default @ pixel-6-pro[big-cores] 92.587 (vs. 92.777, 0.20%↓) 92.569 0.108
MobileNetV1\_fp32(tflite) [x86\_64-cascadelake-linux\_gnu-llvm\_cpu][default-flags,dt-uk] local\_task(embedded\_elf)[1-thread,full-inference,default-flags] with default @ c2-standard-60[cpu] 17.823 (vs. 17.787, 0.20%↑) 17.717 0.179
MobileNetV2\_int8(tflite) [x86\_64-cascadelake-linux\_gnu-llvm\_cpu][experimental-flags,no-dt] local\_task(embedded\_elf)[1-thread,full-inference,default-flags] with default @ c2-standard-60[cpu] 21.571 (vs. 21.615, 0.20%↓) 21.517 0.120
MobileNetV2\_int8(tflite) [x86\_64-cascadelake-linux\_gnu-llvm\_cpu][default-flags,dt-uk] local\_task(embedded\_elf)[8-thread,full-inference,default-flags] with default @ c2-standard-60[cpu] 5.419 (vs. 5.410, 0.16%↑) 5.421 0.027
MobileNetV2\_fp32(tflite) [x86\_64-cascadelake-linux\_gnu-llvm\_cpu][experimental-flags,no-dt] local\_task(embedded\_elf)[8-thread,full-inference,default-flags] with default @ c2-standard-60[cpu] 3.759 (vs. 3.753, 0.15%↑) 3.756 0.026
PersonDetect\_int8(tflite) [x86\_64-cascadelake-linux\_gnu-llvm\_cpu][default-flags,dt-uk] local\_task(embedded\_elf)[1-thread,full-inference,default-flags] with default @ c2-standard-60[cpu] 0.631 (vs. 0.632, 0.14%↓) 0.631 0.001
MobileBertSquad\_fp32(tflite) [qualcomm-adreno-vulkan\_android31-vulkan\_spirv][default-flags] vulkan(none)[full-inference,default-flags] with default @ moto-edge-x30[gpu] 260.440 (vs. 260.121, 0.12%↑) 260.438 0.765
EfficientNet\_int8(tflite) [x86\_64-cascadelake-linux\_gnu-llvm\_cpu][experimental-flags,no-dt] local\_task(embedded\_elf)[1-thread,full-inference,default-flags] with default @ c2-standard-60[cpu] 26.879 (vs. 26.848, 0.12%↑) 26.867 0.071
PersonDetect\_int8(tflite) [x86\_64-cascadelake-linux\_gnu-llvm\_cpu][default-flags,dt-uk] local\_sync(embedded\_elf)[full-inference,default-flags] with default @ c2-standard-60[cpu] 0.568 (vs. 0.568, 0.11%↓) 0.568 0.001
MobileNetV2\_fp32(tflite) [vmvx-generic-vmvx-vmvx][experimental-flags] local\_task(vmvx\_module)[8-thread,full-inference,default-flags] with default @ c2-standard-60[cpu] 7900.089 (vs. 7908.588, 0.11%↓) 7897.414 5.220
MobileNetV1\_fp32(tflite) [x86\_64-cascadelake-linux\_gnu-llvm\_cpu][default-flags,dt-uk] local\_task(embedded\_elf)[8-thread,full-inference,default-flags] with default @ c2-standard-60[cpu] 4.569 (vs. 4.574, 0.10%↓) 4.563 0.022
matmul\_2562x2561x2561\_f32t\_f32t\_f32t\_tile\_config\_default(linalg) [cuda-sm\_80-linux\_gnu-cuda][ukernel,matmul] cuda(none)[full-inference,default-flags] with default @ a2-highgpu-1g[gpu] 1.370 (vs. 1.369, 0.08%↑) 1.370 0.000
MobileSSD\_fp32(tflite) [x86\_64-cascadelake-linux\_gnu-llvm\_cpu][experimental-flags,no-dt] local\_task(embedded\_elf)[8-thread,full-inference,default-flags] with default @ c2-standard-60[cpu] 8.462 (vs. 8.456, 0.07%↑) 8.427 0.063
GPT2\_117M\_TF\_1X4XI32(stablehlo) [x86\_64-cascadelake-linux\_gnu-llvm\_cpu][default-flags,dt-uk] local\_task(embedded\_elf)[15-thread,full-inference,default-flags] with default @ c2-standard-60[cpu] 8.569 (vs. 8.567, 0.03%↑) 8.534 0.087
MiniLML12H384Uncased(stablehlo) [x86\_64-cascadelake-linux\_gnu-llvm\_cpu][default-flags,dt-uk] local\_task(embedded\_elf)[1-thread,full-inference,default-flags] with default @ c2-standard-60[cpu] 56.198 (vs. 56.188, 0.02%↑) 55.803 0.957
PoseNet\_fp32(tflite) [x86\_64-cascadelake-linux\_gnu-llvm\_cpu][default-flags,dt-uk] local\_task(embedded\_elf)[8-thread,full-inference,default-flags] with default @ c2-standard-60[cpu] 5.156 (vs. 5.156, 0.02%↓) 5.149 0.025
MobileBertSquad\_fp16(tflite) [x86\_64-cascadelake-linux\_gnu-llvm\_cpu][default-flags,dt-uk] local\_task(embedded\_elf)[15-thread,full-inference,default-flags] with default @ c2-standard-60[cpu] 61.171 (vs. 61.162, 0.01%↑) 60.917 0.506
matmul\_256x256x2048\_i8\_i4\_i32\_tile\_config\_default(linalg) [x86\_64-cascadelake-linux\_gnu-llvm\_cpu][default-flags,dt-uk] local\_sync(embedded\_elf)[full-inference,default-flags] with default @ c2-standard-60[cpu] 7.571 (vs. 7.570, 0.01%↑) 7.570 0.011
MobileBertSquad\_fp32(tflite) [arm-valhall-vulkan\_android31-vulkan\_spirv][default-flags] vulkan(none)[full-inference,default-flags] with default @ pixel-6-pro[gpu] 97.948 (vs. 97.945, 0.00%↑) 97.926 0.371
MobileBertSquad\_fp32(tflite) [x86\_64-cascadelake-linux\_gnu-llvm\_cpu][default-flags,dt-uk] local\_task(embedded\_elf)[1-thread,full-inference,default-flags] with default @ c2-standard-60[cpu] 189.493 (vs. 189.493, 0.00%↑) 187.878 2.954

Regressed Total Dispatch Sizes 🚩

Benchmark Name Total Dispatch Size (bytes)
MobileBertSquad\_int8(tflite) [riscv\_32-generic-linux\_gnu-llvm\_cpu][default-flags,compile-stats] 3810668 (vs. 2116860, 80.02%↑)
MobileBertSquad\_int8(tflite) [riscv\_64-generic-linux\_gnu-llvm\_cpu][default-flags,compile-stats] 3602552 (vs. 2108296, 70.88%↑)
MobileBertSquad\_int8(tflite) [armv8.2-a-generic-linux\_android29-llvm\_cpu][default-flags,dt-uk,compile-stats] 1807936 (vs. 1372976, 31.68%↑)
MobileBertSquad\_int8(tflite) [armv8.2-a-generic-linux\_android29-llvm\_cpu][experimental-flags,dt-only,compile-stats] 1814848 (vs. 1379904, 31.52%↑)
MobileBertSquad\_int8(tflite) [x86\_64-cascadelake-linux\_gnu-llvm\_cpu][default-flags,dt-uk,compile-stats] 3405104 (vs. 2599872, 30.97%↑)
MobileBertSquad\_int8(tflite) [x86\_64-cascadelake-linux\_gnu-llvm\_cpu][experimental-flags,dt-only,compile-stats] 3415536 (vs. 2610240, 30.85%↑)
MobileBertSquad\_int8(tflite) [x86\_64-cascadelake-linux\_gnu-llvm\_cpu][experimental-flags,no-dt,compile-stats] 6648592 (vs. 5843296, 13.78%↑)
MobileBertSquad\_int8(tflite) [armv8.2-a-generic-linux\_android29-llvm\_cpu][experimental-flags,no-dt,compile-stats] 4822752 (vs. 4387712, 9.91%↑)

Regressed Total Artifact Sizes 🚩

Benchmark Name Total Artifact Size (bytes)
MobileBertSquad\_int8(tflite) [riscv\_32-generic-linux\_gnu-llvm\_cpu][default-flags,compile-stats] 30105159 (vs. 28415879, 5.94%↑)
MobileBertSquad\_int8(tflite) [riscv\_64-generic-linux\_gnu-llvm\_cpu][default-flags,compile-stats] 29897031 (vs. 28407303, 5.24%↑)

Improved Stream IR Dispatch Count (# of cmd.dispatch ops) 🎉

Benchmark Name Stream IR Dispatch Count (# of cmd.dispatch ops)
MobileBertSquad\_int8(tflite) [x86\_64-cascadelake-linux\_gnu-llvm\_cpu][experimental-flags,no-dt,compile-stats] 1078 (vs. 1102, 2.18%↓)
MobileBertSquad\_int8(tflite) [riscv\_64-generic-linux\_gnu-llvm\_cpu][default-flags,compile-stats] 1078 (vs. 1102, 2.18%↓)
MobileBertSquad\_int8(tflite) [riscv\_32-generic-linux\_gnu-llvm\_cpu][default-flags,compile-stats] 1078 (vs. 1102, 2.18%↓)
MobileBertSquad\_int8(tflite) [armv8.2-a-generic-linux\_android29-llvm\_cpu][experimental-flags,no-dt,compile-stats] 1078 (vs. 1102, 2.18%↓)
MobileBertSquad\_int8(tflite) [arm-valhall-vulkan\_android31-vulkan\_spirv][default-flags,compile-stats] 1078 (vs. 1102, 2.18%↓)
MobileBertSquad\_int8(tflite) [arm-valhall-vulkan\_android31-vulkan\_spirv][experimental-flags,fuse-padding,max-concurrency,compile-stats] 1078 (vs. 1102, 2.18%↓)
MobileBertSquad\_int8(tflite) [arm-valhall-vulkan\_android31-vulkan\_spirv][experimental-flags,fuse-padding,max-concurrency,repeated-kernel,compile-stats] 1078 (vs. 1102, 2.18%↓)
MobileBertSquad\_int8(tflite) [x86\_64-cascadelake-linux\_gnu-llvm\_cpu][experimental-flags,dt-only,compile-stats] 2136 (vs. 2160, 1.11%↓)
MobileBertSquad\_int8(tflite) [x86\_64-cascadelake-linux\_gnu-llvm\_cpu][default-flags,dt-uk,compile-stats] 2136 (vs. 2160, 1.11%↓)
MobileBertSquad\_int8(tflite) [armv8.2-a-generic-linux\_android29-llvm\_cpu][default-flags,dt-uk,compile-stats] 2136 (vs. 2160, 1.11%↓)
MobileBertSquad\_int8(tflite) [armv8.2-a-generic-linux\_android29-llvm\_cpu][experimental-flags,dt-only,compile-stats] 2136 (vs. 2160, 1.11%↓)

All Compilation Metrics

Benchmark Name Compilation Time (ms) Total Dispatch Size (bytes) Total Artifact Size (bytes) Stream IR Dispatch Count (# of cmd.dispatch ops)
matmul_256x256x2048_i8_i4_i32_tile_config_default(linalg) [x86_64-cascadelake-linux_gnu-llvm_cpu][experimental-flags,no-dt,compile-stats] 1545 (vs. 1386, 11.47%↑) 9664 (vs. 9664, 0.00%) 277497 (vs. 277497, 0.00%) 1 (vs. 1, 0.00%)
PersonDetect_int8(tflite) [x86_64-cascadelake-linux_gnu-llvm_cpu][experimental-flags,no-dt,compile-stats] 22825 (vs. 21757, 4.91%↑) 144032 (vs. 144032, 0.00%) 399877 (vs. 399877, 0.00%) 60 (vs. 60, 0.00%)
MobileNetV3Small_fp32(tflite) [x86_64-cascadelake-linux_gnu-llvm_cpu][experimental-flags,no-dt,compile-stats] 38219 (vs. 36975, 3.36%↑) 238720 (vs. 238720, 0.00%) 10458245 (vs. 10458245, 0.00%) 97 (vs. 97, 0.00%)
DeepLabV3_fp32(tflite) [x86_64-cascadelake-linux_gnu-llvm_cpu][experimental-flags,no-dt,compile-stats] 32388 (vs. 34570, 6.31%↓) 177680 (vs. 177680, 0.00%) 2959045 (vs. 2959045, 0.00%) 79 (vs. 79, 0.00%)
EfficientNet_int8(tflite) [x86_64-cascadelake-linux_gnu-llvm_cpu][experimental-flags,no-dt,compile-stats] 61007 (vs. 61875, 1.40%↓) 680000 (vs. 680000, 0.00%) 5603397 (vs. 5603397, 0.00%) 89 (vs. 89, 0.00%)
MobileNetV1_fp32(tflite) [x86_64-cascadelake-linux_gnu-llvm_cpu][experimental-flags,no-dt,compile-stats] 24485 (vs. 22303, 9.78%↑) 174400 (vs. 174400, 0.00%) 17094405 (vs. 17094405, 0.00%) 51 (vs. 51, 0.00%)
MobileNetV2_fp32(tflite) [x86_64-cascadelake-linux_gnu-llvm_cpu][experimental-flags,no-dt,compile-stats] 34471 (vs. 30354, 13.56%↑) 190096 (vs. 190096, 0.00%) 14173189 (vs. 14173189, 0.00%) 74 (vs. 74, 0.00%)
MobileNetV2_int8(tflite) [x86_64-cascadelake-linux_gnu-llvm_cpu][experimental-flags,no-dt,compile-stats] 71542 (vs. 60367, 18.51%↑) 568928 (vs. 568928, 0.00%) 4219653 (vs. 4219653, 0.00%) 144 (vs. 144, 0.00%)
MobileSSD_fp32(tflite) [x86_64-cascadelake-linux_gnu-llvm_cpu][experimental-flags,no-dt,compile-stats] 47859 (vs. 43270, 10.61%↑) 287600 (vs. 287600, 0.00%) 18229253 (vs. 18229253, 0.00%) 124 (vs. 124, 0.00%)
PoseNet_fp32(tflite) [x86_64-cascadelake-linux_gnu-llvm_cpu][experimental-flags,no-dt,compile-stats] 24115 (vs. 20950, 15.11%↑) 142288 (vs. 142288, 0.00%) 5195653 (vs. 5195653, 0.00%) 48 (vs. 48, 0.00%)
MobileBertSquad_fp16(tflite) [x86_64-cascadelake-linux_gnu-llvm_cpu][experimental-flags,no-dt,compile-stats] 75134 (vs. 64999, 15.59%↑) 84496 (vs. 84496, 0.00%) 99926661 (vs. 99926661, 0.00%) 728 (vs. 728, 0.00%)
MobileBertSquad_fp32(tflite) [x86_64-cascadelake-linux_gnu-llvm_cpu][experimental-flags,no-dt,compile-stats] 68573 (vs. 63600, 7.82%↑) 93056 (vs. 93056, 0.00%) 98443077 (vs. 98443077, 0.00%) 704 (vs. 704, 0.00%)
MobileBertSquad_int8(tflite) [x86_64-cascadelake-linux_gnu-llvm_cpu][experimental-flags,no-dt,compile-stats] 313944 (vs. 287315, 9.27%↑) 6648592 (vs. 5843296, 13.78%↑) 32943045 (vs. 32142341, 2.49%↑) 1078 (vs. 1102, 2.18%↓)
EfficientNetV2STF(stablehlo) [x86_64-cascadelake-linux_gnu-llvm_cpu][experimental-flags,no-dt,compile-stats] 105235 (vs. 100709, 4.49%↑) 216304 (vs. 216304, 0.00%) 164497708 (vs. 164497708, 0.00%) 276 (vs. 276, 0.00%)
MiniLML12H384Uncased(stablehlo) [x86_64-cascadelake-linux_gnu-llvm_cpu][experimental-flags,no-dt,compile-stats] 30788 (vs. 28442, 8.25%↑) 59344 (vs. 59344, 0.00%) 133996207 (vs. 133996207, 0.00%) 209 (vs. 209, 0.00%)
GPT2_117M_TF_1X4XI32(stablehlo) [x86_64-cascadelake-linux_gnu-llvm_cpu][experimental-flags,no-dt,compile-stats] 37176 (vs. 33050, 12.48%↑) 27920 (vs. 27920, 0.00%) 652755092 (vs. 652755092, 0.00%) 233 (vs. 233, 0.00%)
GPT2_117M_TF_1X1XI32(stablehlo) [x86_64-cascadelake-linux_gnu-llvm_cpu][experimental-flags,no-dt,compile-stats] 34670 (vs. 32353, 7.16%↑) 14736 (vs. 14736, 0.00%) 652736084 (vs. 652736084, 0.00%) 246 (vs. 246, 0.00%)
BertForMaskedLMTF(stablehlo) [x86_64-cascadelake-linux_gnu-llvm_cpu][experimental-flags,no-dt,compile-stats] 36157 (vs. 35895, 0.73%↑) 68608 (vs. 68608, 0.00%) 533841087 (vs. 533841087, 0.00%) 212 (vs. 212, 0.00%)
BertLargeTF(stablehlo) [x86_64-cascadelake-linux_gnu-llvm_cpu][experimental-flags,no-dt,compile-stats] 33122 (vs. 34753, 4.69%↓) 50128 (vs. 50128, 0.00%) 1336025023 (vs. 1336025023, 0.00%) 413 (vs. 413, 0.00%)
matmul_256x256x2048_i8_i4_i32_tile_config_default(linalg) [x86_64-cascadelake-linux_gnu-llvm_cpu][experimental-flags,dt-only,compile-stats] 1531 (vs. 1322, 15.81%↑) 9664 (vs. 9664, 0.00%) 277497 (vs. 277497, 0.00%) 1 (vs. 1, 0.00%)
PersonDetect_int8(tflite) [x86_64-cascadelake-linux_gnu-llvm_cpu][experimental-flags,dt-only,compile-stats] 24770 (vs. 23026, 7.57%↑) 107248 (vs. 107248, 0.00%) 371397 (vs. 371397, 0.00%) 87 (vs. 87, 0.00%)
MobileNetV3Small_fp32(tflite) [x86_64-cascadelake-linux_gnu-llvm_cpu][experimental-flags,dt-only,compile-stats] 33492 (vs. 31567, 6.10%↑) 96944 (vs. 96944, 0.00%) 10398725 (vs. 10398725, 0.00%) 147 (vs. 147, 0.00%)
DeepLabV3_fp32(tflite) [x86_64-cascadelake-linux_gnu-llvm_cpu][experimental-flags,dt-only,compile-stats] 31327 (vs. 28878, 8.48%↑) 113392 (vs. 113392, 0.00%) 2921349 (vs. 2921349, 0.00%) 144 (vs. 144, 0.00%)
EfficientNet_int8(tflite) [x86_64-cascadelake-linux_gnu-llvm_cpu][experimental-flags,dt-only,compile-stats] 47870 (vs. 45837, 4.44%↑) 269104 (vs. 269104, 0.00%) 5219205 (vs. 5219205, 0.00%) 148 (vs. 148, 0.00%)
MobileNetV1_fp32(tflite) [x86_64-cascadelake-linux_gnu-llvm_cpu][experimental-flags,dt-only,compile-stats] 18351 (vs. 18654, 1.62%↓) 60288 (vs. 60288, 0.00%) 17017925 (vs. 17017925, 0.00%) 79 (vs. 79, 0.00%)
MobileNetV2_fp32(tflite) [x86_64-cascadelake-linux_gnu-llvm_cpu][experimental-flags,dt-only,compile-stats] 29213 (vs. 26987, 8.25%↑) 95248 (vs. 95248, 0.00%) 14133765 (vs. 14133765, 0.00%) 136 (vs. 136, 0.00%)
MobileNetV2_int8(tflite) [x86_64-cascadelake-linux_gnu-llvm_cpu][experimental-flags,dt-only,compile-stats] 56185 (vs. 51044, 10.07%↑) 330048 (vs. 330048, 0.00%) 4003653 (vs. 4003653, 0.00%) 213 (vs. 213, 0.00%)
MobileSSD_fp32(tflite) [x86_64-cascadelake-linux_gnu-llvm_cpu][experimental-flags,dt-only,compile-stats] 39678 (vs. 40612, 2.30%↓) 136080 (vs. 136080, 0.00%) 18361413 (vs. 18361413, 0.00%) 223 (vs. 223, 0.00%)
PoseNet_fp32(tflite) [x86_64-cascadelake-linux_gnu-llvm_cpu][experimental-flags,dt-only,compile-stats] 15181 (vs. 14999, 1.21%↑) 44208 (vs. 44208, 0.00%) 5147973 (vs. 5147973, 0.00%) 79 (vs. 79, 0.00%)
MobileBertSquad_fp16(tflite) [x86_64-cascadelake-linux_gnu-llvm_cpu][experimental-flags,dt-only,compile-stats] 91665 (vs. 83762, 9.44%↑) 52400 (vs. 52400, 0.00%) 100083141 (vs. 100083141, 0.00%) 1786 (vs. 1786, 0.00%)
MobileBertSquad_fp32(tflite) [x86_64-cascadelake-linux_gnu-llvm_cpu][experimental-flags,dt-only,compile-stats] 82208 (vs. 83380, 1.41%↓) 52832 (vs. 52832, 0.00%) 98588357 (vs. 98588357, 0.00%) 1762 (vs. 1762, 0.00%)
MobileBertSquad_int8(tflite) [x86_64-cascadelake-linux_gnu-llvm_cpu][experimental-flags,dt-only,compile-stats] 282699 (vs. 248492, 13.77%↑) 3415536 (vs. 2610240, 30.85%↑) 29889029 (vs. 29088325, 2.75%↑) 2136 (vs. 2160, 1.11%↓)
EfficientNetV2STF(stablehlo) [x86_64-cascadelake-linux_gnu-llvm_cpu][experimental-flags,dt-only,compile-stats] 108785 (vs. 103428, 5.18%↑) 154864 (vs. 154864, 0.00%) 169910060 (vs. 169910060, 0.00%) 439 (vs. 439, 0.00%)
MiniLML12H384Uncased(stablehlo) [x86_64-cascadelake-linux_gnu-llvm_cpu][experimental-flags,dt-only,compile-stats] 38542 (vs. 36581, 5.36%↑) 37936 (vs. 37936, 0.00%) 219509935 (vs. 219509935, 0.00%) 330 (vs. 330, 0.00%)
GPT2_117M_TF_1X4XI32(stablehlo) [x86_64-cascadelake-linux_gnu-llvm_cpu][experimental-flags,dt-only,compile-stats] 50314 (vs. 43137, 16.64%↑) 19328 (vs. 19328, 0.00%) 992542612 (vs. 992542612, 0.00%) 318 (vs. 318, 0.00%)
GPT2_117M_TF_1X1XI32(stablehlo) [x86_64-cascadelake-linux_gnu-llvm_cpu][experimental-flags,dt-only,compile-stats] 44384 (vs. 40923, 8.46%↑) 12864 (vs. 12864, 0.00%) 992543572 (vs. 992543572, 0.00%) 367 (vs. 367, 0.00%)
BertForMaskedLMTF(stablehlo) [x86_64-cascadelake-linux_gnu-llvm_cpu][experimental-flags,dt-only,compile-stats] 46831 (vs. 43156, 8.52%↑) 40144 (vs. 40144, 0.00%) 875939327 (vs. 875939327, 0.00%) 334 (vs. 334, 0.00%)
BertLargeTF(stablehlo) [x86_64-cascadelake-linux_gnu-llvm_cpu][experimental-flags,dt-only,compile-stats] 52359 (vs. 49435, 5.91%↑) 43840 (vs. 43840, 0.00%) 1336053183 (vs. 1336053183, 0.00%) 654 (vs. 654, 0.00%)
matmul_256x256x2048_i8_i4_i32_tile_config_default(linalg) [x86_64-cascadelake-linux_gnu-llvm_cpu][default-flags,dt-uk,compile-stats] 1585 (vs. 1320, 20.08%↑) 9664 (vs. 9664, 0.00%) 277497 (vs. 277497, 0.00%) 1 (vs. 1, 0.00%)
PersonDetect_int8(tflite) [x86_64-cascadelake-linux_gnu-llvm_cpu][default-flags,dt-uk,compile-stats] 29016 (vs. 24058, 20.61%↑) 99488 (vs. 99488, 0.00%) 363653 (vs. 363653, 0.00%) 87 (vs. 87, 0.00%)
MobileNetV3Small_fp32(tflite) [x86_64-cascadelake-linux_gnu-llvm_cpu][default-flags,dt-uk,compile-stats] 32674 (vs. 31832, 2.65%↑) 98656 (vs. 98656, 0.00%) 10400453 (vs. 10400453, 0.00%) 147 (vs. 147, 0.00%)
DeepLabV3_fp32(tflite) [x86_64-cascadelake-linux_gnu-llvm_cpu][default-flags,dt-uk,compile-stats] 30324 (vs. 29179, 3.92%↑) 117504 (vs. 117504, 0.00%) 2925445 (vs. 2925445, 0.00%) 144 (vs. 144, 0.00%)
EfficientNet_int8(tflite) [x86_64-cascadelake-linux_gnu-llvm_cpu][default-flags,dt-uk,compile-stats] 49577 (vs. 47587, 4.18%↑) 257952 (vs. 257952, 0.00%) 5208069 (vs. 5208069, 0.00%) 148 (vs. 148, 0.00%)
MobileNetV1_fp32(tflite) [x86_64-cascadelake-linux_gnu-llvm_cpu][default-flags,dt-uk,compile-stats] 21118 (vs. 19233, 9.80%↑) 65376 (vs. 65376, 0.00%) 17023045 (vs. 17023045, 0.00%) 79 (vs. 79, 0.00%)
MobileNetV2_fp32(tflite) [x86_64-cascadelake-linux_gnu-llvm_cpu][default-flags,dt-uk,compile-stats] 30385 (vs. 28228, 7.64%↑) 99104 (vs. 99104, 0.00%) 14137669 (vs. 14137669, 0.00%) 136 (vs. 136, 0.00%)
MobileNetV2_int8(tflite) [x86_64-cascadelake-linux_gnu-llvm_cpu][default-flags,dt-uk,compile-stats] 57488 (vs. 58563, 1.84%↓) 323792 (vs. 323792, 0.00%) 3997381 (vs. 3997381, 0.00%) 213 (vs. 213, 0.00%)
MobileSSD_fp32(tflite) [x86_64-cascadelake-linux_gnu-llvm_cpu][default-flags,dt-uk,compile-stats] 43127 (vs. 37200, 15.93%↑) 122656 (vs. 122656, 0.00%) 18348037 (vs. 18348037, 0.00%) 223 (vs. 223, 0.00%)
PoseNet_fp32(tflite) [x86_64-cascadelake-linux_gnu-llvm_cpu][default-flags,dt-uk,compile-stats] 16535 (vs. 17689, 6.52%↓) 47200 (vs. 47200, 0.00%) 5150981 (vs. 5150981, 0.00%) 79 (vs. 79, 0.00%)
MobileBertSquad_fp16(tflite) [x86_64-cascadelake-linux_gnu-llvm_cpu][default-flags,dt-uk,compile-stats] 92814 (vs. 82457, 12.56%↑) 38880 (vs. 38880, 0.00%) 100069637 (vs. 100069637, 0.00%) 1786 (vs. 1786, 0.00%)
MobileBertSquad_fp32(tflite) [x86_64-cascadelake-linux_gnu-llvm_cpu][default-flags,dt-uk,compile-stats] 90032 (vs. 77528, 16.13%↑) 39296 (vs. 39296, 0.00%) 98574789 (vs. 98574789, 0.00%) 1762 (vs. 1762, 0.00%)
MobileBertSquad_int8(tflite) [x86_64-cascadelake-linux_gnu-llvm_cpu][default-flags,dt-uk,compile-stats] 274308 (vs. 238669, 14.93%↑) 3405104 (vs. 2599872, 30.97%↑) 29878597 (vs. 29077957, 2.75%↑) 2136 (vs. 2160, 1.11%↓)
EfficientNetV2STF(stablehlo) [x86_64-cascadelake-linux_gnu-llvm_cpu][default-flags,dt-uk,compile-stats] 113340 (vs. 103989, 8.99%↑) 140240 (vs. 140240, 0.00%) 169895468 (vs. 169895468, 0.00%) 439 (vs. 439, 0.00%)
MiniLML12H384Uncased(stablehlo) [x86_64-cascadelake-linux_gnu-llvm_cpu][default-flags,dt-uk,compile-stats] 38118 (vs. 33811, 12.74%↑) 33904 (vs. 33904, 0.00%) 219505903 (vs. 219505903, 0.00%) 330 (vs. 330, 0.00%)
GPT2_117M_TF_1X4XI32(stablehlo) [x86_64-cascadelake-linux_gnu-llvm_cpu][default-flags,dt-uk,compile-stats] 51946 (vs. 43196, 20.26%↑) 18640 (vs. 18640, 0.00%) 992541972 (vs. 992541972, 0.00%) 318 (vs. 318, 0.00%)
GPT2_117M_TF_1X1XI32(stablehlo) [x86_64-cascadelake-linux_gnu-llvm_cpu][default-flags,dt-uk,compile-stats] 49232 (vs. 42281, 16.44%↑) 12336 (vs. 12336, 0.00%) 992543060 (vs. 992543060, 0.00%) 367 (vs. 367, 0.00%)
BertForMaskedLMTF(stablehlo) [x86_64-cascadelake-linux_gnu-llvm_cpu][default-flags,dt-uk,compile-stats] 49230 (vs. 43111, 14.19%↑) 34928 (vs. 34928, 0.00%) 875934143 (vs. 875934143, 0.00%) 334 (vs. 334, 0.00%)
BertLargeTF(stablehlo) [x86_64-cascadelake-linux_gnu-llvm_cpu][default-flags,dt-uk,compile-stats] 60208 (vs. 51077, 17.88%↑) 39360 (vs. 39360, 0.00%) 1336048703 (vs. 1336048703, 0.00%) 654 (vs. 654, 0.00%)
EfficientNetV2STF(stablehlo) [cuda-sm_80-linux_gnu-cuda][default-flags,compile-stats] 95560 (vs. 83895, 13.90%↑) 844192 (vs. 844192, 0.00%) 165158932 (vs. 165158932, 0.00%) 276 (vs. 276, 0.00%)
MiniLML12H384Uncased(stablehlo) [cuda-sm_80-linux_gnu-cuda][default-flags,compile-stats] 26026 (vs. 21694, 19.97%↑) 145212 (vs. 145212, 0.00%) 134088431 (vs. 134088431, 0.00%) 209 (vs. 209, 0.00%)
BertForMaskedLMTF(stablehlo) [cuda-sm_80-linux_gnu-cuda][default-flags,compile-stats] 35542 (vs. 32846, 8.21%↑) 227968 (vs. 227968, 0.00%) 534005858 (vs. 534005858, 0.00%) 212 (vs. 212, 0.00%)
BertLargeTF(stablehlo) [cuda-sm_80-linux_gnu-cuda][default-flags,compile-stats] 35393 (vs. 34069, 3.89%↑) 130852 (vs. 130852, 0.00%) 1336110955 (vs. 1336110955, 0.00%) 413 (vs. 413, 0.00%)
matmul_3456x1024x2048_f16t_tile_config_default(linalg) [cuda-sm_80-linux_gnu-cuda][ukernel,matmul,compile-stats] 1682 (vs. 1712, 1.75%↓) 30540 (vs. 30540, 0.00%) 41263 (vs. 41263, 0.00%) 1 (vs. 1, 0.00%)
matmul_3456x1024x2048_f32t_tile_config_default(linalg) [cuda-sm_80-linux_gnu-cuda][ukernel,matmul,compile-stats] 1942 (vs. 1776, 9.35%↑) 43516 (vs. 43516, 0.00%) 54239 (vs. 54239, 0.00%) 1 (vs. 1, 0.00%)
matmul_2560x2560x2560_f16t_tile_config_default(linalg) [cuda-sm_80-linux_gnu-cuda][ukernel,matmul,compile-stats] 1645 (vs. 1390, 18.35%↑) 28492 (vs. 28492, 0.00%) 39151 (vs. 39151, 0.00%) 1 (vs. 1, 0.00%)
matmul_2560x2560x2560_f32t_tile_config_default(linalg) [cuda-sm_80-linux_gnu-cuda][ukernel,matmul,compile-stats] 1953 (vs. 1935, 0.93%↑) 39892 (vs. 39892, 0.00%) 50551 (vs. 50551, 0.00%) 1 (vs. 1, 0.00%)
matmul_2564x2564x2564_f32t_f32t_f32t_tile_config_default(linalg) [cuda-sm_80-linux_gnu-cuda][ukernel,matmul,compile-stats] 2362 (vs. 1961, 20.45%↑) 85344 (vs. 85344, 0.00%) 96003 (vs. 96003, 0.00%) 1 (vs. 1, 0.00%)
matmul_2562x2564x2562_f32t_f32t_f32t_tile_config_default(linalg) [cuda-sm_80-linux_gnu-cuda][ukernel,matmul,compile-stats] 2227 (vs. 2162, 3.01%↑) 88224 (vs. 88224, 0.00%) 98947 (vs. 98947, 0.00%) 1 (vs. 1, 0.00%)
matmul_2562x2561x2561_f32t_f32t_f32t_tile_config_default(linalg) [cuda-sm_80-linux_gnu-cuda][ukernel,matmul,compile-stats] 2646 (vs. 2526, 4.75%↑) 82792 (vs. 82792, 0.00%) 93515 (vs. 93515, 0.00%) 1 (vs. 1, 0.00%)
matmul_123x2561x2561_f32t_f32t_f32t_tile_config_default(linalg) [cuda-sm_80-linux_gnu-cuda][ukernel,matmul,compile-stats] 2013 (vs. 1970, 2.18%↑) 49772 (vs. 49772, 0.00%) 60494 (vs. 60494, 0.00%) 1 (vs. 1, 0.00%)
matmul_128x256x8192_f16t_tile_config_default(linalg) [cuda-sm_80-linux_gnu-cuda][ukernel,matmul,splitk,compile-stats] 884 (vs. 765, 15.56%↑) 8828 (vs. 8828, 0.00%) 26181 (vs. 26181, 0.00%) 2 (vs. 2, 0.00%)
matmul_128x256x8192_f32t_tile_config_default(linalg) [cuda-sm_80-linux_gnu-cuda][ukernel,matmul,splitk,compile-stats] 882 (vs. 945, 6.67%↓) 9784 (vs. 9784, 0.00%) 27137 (vs. 27137, 0.00%) 2 (vs. 2, 0.00%)
MiniLML12H384Uncased(stablehlo) [riscv_64-generic-linux_gnu-llvm_cpu][default-flags,compile-stats] 31915 (vs. 28724, 11.11%↑) 50552 (vs. 50552, 0.00%) 133987441 (vs. 133987441, 0.00%) 209 (vs. 209, 0.00%)
DeepLabV3_fp32(tflite) [riscv_64-generic-linux_gnu-llvm_cpu][default-flags,compile-stats] 20299 (vs. 18361, 10.55%↑) 42840 (vs. 42840, 0.00%) 2824263 (vs. 2824263, 0.00%) 79 (vs. 79, 0.00%)
EfficientNet_int8(tflite) [riscv_64-generic-linux_gnu-llvm_cpu][default-flags,compile-stats] 39589 (vs. 41954, 5.64%↓) 185480 (vs. 185480, 0.00%) 5108871 (vs. 5108871, 0.00%) 89 (vs. 89, 0.00%)
MobileBertSquad_fp32(tflite) [riscv_64-generic-linux_gnu-llvm_cpu][default-flags,compile-stats] 70134 (vs. 64092, 9.43%↑) 49416 (vs. 49416, 0.00%) 98399431 (vs. 98399431, 0.00%) 704 (vs. 704, 0.00%)
MobileBertSquad_int8(tflite) [riscv_64-generic-linux_gnu-llvm_cpu][default-flags,compile-stats] 364311 (vs. 242212, 50.41%↑) 3602552 (vs. 2108296, 70.88%↑) 29897031 (vs. 28407303, 5.24%↑) 1078 (vs. 1102, 2.18%↓)
MobileNetV1_fp32(tflite) [riscv_64-generic-linux_gnu-llvm_cpu][default-flags,compile-stats] 18902 (vs. 18583, 1.72%↑) 50984 (vs. 50984, 0.00%) 16976263 (vs. 16976263, 0.00%) 65 (vs. 65, 0.00%)
MobileNetV2_int8(tflite) [riscv_64-generic-linux_gnu-llvm_cpu][default-flags,compile-stats] 53288 (vs. 53138, 0.28%↑) 217032 (vs. 217032, 0.00%) 3867719 (vs. 3867719, 0.00%) 144 (vs. 144, 0.00%)
PersonDetect_int8(tflite) [riscv_64-generic-linux_gnu-llvm_cpu][default-flags,compile-stats] 18526 (vs. 18332, 1.06%↑) 58936 (vs. 58936, 0.00%) 314759 (vs. 314759, 0.00%) 60 (vs. 60, 0.00%)
EfficientNet_int8(tflite) [riscv_32-generic-linux_gnu-llvm_cpu][default-flags,compile-stats] 56710 (vs. 52638, 7.74%↑) 390748 (vs. 390748, 0.00%) 5314183 (vs. 5314183, 0.00%) 89 (vs. 89, 0.00%)
MobileBertSquad_int8(tflite) [riscv_32-generic-linux_gnu-llvm_cpu][default-flags,compile-stats] 427744 (vs. 279862, 52.84%↑) 3810668 (vs. 2116860, 80.02%↑) 30105159 (vs. 28415879, 5.94%↑) 1078 (vs. 1102, 2.18%↓)
PersonDetect_int8(tflite) [riscv_32-generic-linux_gnu-llvm_cpu][default-flags,compile-stats] 21135 (vs. 19355, 9.20%↑) 124428 (vs. 124428, 0.00%) 380231 (vs. 380231, 0.00%) 60 (vs. 60, 0.00%)
MobileNetV2_int8(tflite) [riscv_32-generic-linux_gnu-llvm_cpu][default-flags,compile-stats] 58037 (vs. 55570, 4.44%↑) 328664 (vs. 328664, 0.00%) 3979399 (vs. 3979399, 0.00%) 144 (vs. 144, 0.00%)
DeepLabV3_fp32(tflite) [armv8.2-a-generic-linux_android29-llvm_cpu][experimental-flags,no-dt,compile-stats] 18299 (vs. 18476, 0.96%↓) 56272 (vs. 56272, 0.00%) 2837637 (vs. 2837637, 0.00%) 79 (vs. 79, 0.00%)
MobileBertSquad_fp32(tflite) [armv8.2-a-generic-linux_android29-llvm_cpu][experimental-flags,no-dt,compile-stats] 45904 (vs. 45877, 0.06%↑) 32912 (vs. 32912, 0.00%) 98382917 (vs. 98382917, 0.00%) 704 (vs. 704, 0.00%)
GPT2_117M_TF_1X4XI32(stablehlo) [armv8.2-a-generic-linux_android29-llvm_cpu][experimental-flags,no-dt,compile-stats] 39778 (vs. 35663, 11.54%↑) 20320 (vs. 20320, 0.00%) 652747476 (vs. 652747476, 0.00%) 233 (vs. 233, 0.00%)
GPT2_117M_TF_1X1XI32(stablehlo) [armv8.2-a-generic-linux_android29-llvm_cpu][experimental-flags,no-dt,compile-stats] 36998 (vs. 34666, 6.73%↑) 9024 (vs. 9024, 0.00%) 652730324 (vs. 652730324, 0.00%) 246 (vs. 246, 0.00%)
MobileBertSquad_int8(tflite) [armv8.2-a-generic-linux_android29-llvm_cpu][experimental-flags,no-dt,compile-stats] 295610 (vs. 279221, 5.87%↑) 4822752 (vs. 4387712, 9.91%↑) 31117253 (vs. 30686725, 1.40%↑) 1078 (vs. 1102, 2.18%↓)
Vit_int8(tflite) [armv8.2-a-generic-linux_android29-llvm_cpu][experimental-flags,no-dt,compile-stats] 106374 (vs. 102949, 3.33%↑) 694624 (vs. 694624, 0.00%) 88728645 (vs. 88728645, 0.00%) 268 (vs. 268, 0.00%)
DeepLabV3_fp32(tflite) [armv8.2-a-generic-linux_android29-llvm_cpu][default-flags,dt-uk,compile-stats] 25933 (vs. 25510, 1.66%↑) 49984 (vs. 49984, 0.00%) 2848965 (vs. 2848965, 0.00%) 144 (vs. 144, 0.00%)
MobileBertSquad_fp32(tflite) [armv8.2-a-generic-linux_android29-llvm_cpu][default-flags,dt-uk,compile-stats] 85552 (vs. 79697, 7.35%↑) 21600 (vs. 21600, 0.00%) 98557061 (vs. 98557061, 0.00%) 1762 (vs. 1762, 0.00%)
GPT2_117M_TF_1X4XI32(stablehlo) [armv8.2-a-generic-linux_android29-llvm_cpu][default-flags,dt-uk,compile-stats] 46853 (vs. 43460, 7.81%↑) 11040 (vs. 11040, 0.00%) 992509844 (vs. 992509844, 0.00%) 318 (vs. 318, 0.00%)
GPT2_117M_TF_1X1XI32(stablehlo) [armv8.2-a-generic-linux_android29-llvm_cpu][default-flags,dt-uk,compile-stats] 46435 (vs. 40783, 13.86%↑) 9232 (vs. 9232, 0.00%) 992515476 (vs. 992515476, 0.00%) 367 (vs. 367, 0.00%)
MobileBertSquad_int8(tflite) [armv8.2-a-generic-linux_android29-llvm_cpu][default-flags,dt-uk,compile-stats] 235886 (vs. 215325, 9.55%↑) 1807936 (vs. 1372976, 31.68%↑) 28281413 (vs. 27851077, 1.55%↑) 2136 (vs. 2160, 1.11%↓)
Vit_int8(tflite) [armv8.2-a-generic-linux_android29-llvm_cpu][default-flags,dt-uk,compile-stats] 79198 (vs. 80880, 2.08%↓) 125600 (vs. 125600, 0.00%) 88155333 (vs. 88155333, 0.00%) 386 (vs. 386, 0.00%)
DeepLabV3_fp32(tflite) [armv8.2-a-generic-linux_android29-llvm_cpu][experimental-flags,dt-only,compile-stats] 21901 (vs. 22515, 2.73%↓) 41008 (vs. 41008, 0.00%) 2840005 (vs. 2840005, 0.00%) 144 (vs. 144, 0.00%)
MobileBertSquad_fp32(tflite) [armv8.2-a-generic-linux_android29-llvm_cpu][experimental-flags,dt-only,compile-stats] 78283 (vs. 77547, 0.95%↑) 22096 (vs. 22096, 0.00%) 98557509 (vs. 98557509, 0.00%) 1762 (vs. 1762, 0.00%)
GPT2_117M_TF_1X4XI32(stablehlo) [armv8.2-a-generic-linux_android29-llvm_cpu][experimental-flags,dt-only,compile-stats] 45914 (vs. 40477, 13.43%↑) 10640 (vs. 10640, 0.00%) 992509460 (vs. 992509460, 0.00%) 318 (vs. 318, 0.00%)
GPT2_117M_TF_1X1XI32(stablehlo) [armv8.2-a-generic-linux_android29-llvm_cpu][experimental-flags,dt-only,compile-stats] 41015 (vs. 38440, 6.70%↑) 8928 (vs. 8928, 0.00%) 992515156 (vs. 992515156, 0.00%) 367 (vs. 367, 0.00%)
MobileBertSquad_int8(tflite) [armv8.2-a-generic-linux_android29-llvm_cpu][experimental-flags,dt-only,compile-stats] 234745 (vs. 219168, 7.11%↑) 1814848 (vs. 1379904, 31.52%↑) 28288325 (vs. 27857989, 1.54%↑) 2136 (vs. 2160, 1.11%↓)
Vit_int8(tflite) [armv8.2-a-generic-linux_android29-llvm_cpu][experimental-flags,dt-only,compile-stats] 68466 (vs. 71944, 4.83%↓) 128848 (vs. 128848, 0.00%) 88158533 (vs. 88158533, 0.00%) 386 (vs. 386, 0.00%)
matmul_256x256x2048_i8_i4_i32_tile_config_default(linalg) [armv8.2-a-generic-linux_android29-llvm_cpu][experimental-flags,no-dt,compile-stats] 1096 (vs. 1045, 4.88%↑) 3872 (vs. 3872, 0.00%) 271673 (vs. 271673, 0.00%) 1 (vs. 1, 0.00%)
matmul_256x256x2048_i8_i4_i32_tile_config_default(linalg) [armv8.2-a-generic-linux_android29-llvm_cpu][default-flags,dt-uk,compile-stats] 1942 (vs. 1858, 4.52%↑) 4480 (vs. 4480, 0.00%) 273605 (vs. 273605, 0.00%) 4 (vs. 4, 0.00%)
matmul_256x256x2048_i8_i4_i32_tile_config_default(linalg) [armv8.2-a-generic-linux_android29-llvm_cpu][experimental-flags,dt-only,compile-stats] 1605 (vs. 1538, 4.36%↑) 6512 (vs. 6512, 0.00%) 275653 (vs. 275653, 0.00%) 4 (vs. 4, 0.00%)
MobileBertSquad_fp32(tflite) [qualcomm-adreno-vulkan_android31-vulkan_spirv][default-flags,compile-stats] 67500 (vs. 65392, 3.22%↑) 246320 (vs. 246320, 0.00%) 98597530 (vs. 98597530, 0.00%) 704 (vs. 704, 0.00%)
MobileBertSquad_fp32(tflite) [qualcomm-adreno-vulkan_android31-vulkan_spirv][experimental-flags,fuse-padding,compile-stats] 69153 (vs. 68666, 0.71%↑) 246320 (vs. 246320, 0.00%) 98597530 (vs. 98597530, 0.00%) 704 (vs. 704, 0.00%)
MobileBertSquad_fp32(tflite) [arm-valhall-vulkan_android31-vulkan_spirv][default-flags,compile-stats] 66325 (vs. 65857, 0.71%↑) 143012 (vs. 143012, 0.00%) 98493753 (vs. 98493753, 0.00%) 704 (vs. 704, 0.00%)
MobileBertSquad_fp16(tflite) [arm-valhall-vulkan_android31-vulkan_spirv][default-flags,demote-f32-to-f16,compile-stats] 99910 (vs. 93058, 7.36%↑) 3004452 (vs. 3004452, 0.00%) 52898580 (vs. 52898580, 0.00%) 728 (vs. 728, 0.00%)
MobileBertSquad_int8(tflite) [arm-valhall-vulkan_android31-vulkan_spirv][default-flags,compile-stats] 189540 (vs. 178661, 6.09%↑) 6937016 (vs. 6756400, 2.67%↑) 33232105 (vs. 33056470, 0.53%↑) 1078 (vs. 1102, 2.18%↓)
MobileBertSquad_fp32(tflite) [arm-valhall-vulkan_android31-vulkan_spirv][experimental-flags,fuse-padding,max-concurrency,compile-stats] 70716 (vs. 60351, 17.17%↑) 143012 (vs. 143012, 0.00%) 98493881 (vs. 98493881, 0.00%) 704 (vs. 704, 0.00%)
MobileBertSquad_fp16(tflite) [arm-valhall-vulkan_android31-vulkan_spirv][experimental-flags,fuse-padding,max-concurrency,demote-f32-to-f16,compile-stats] 85336 (vs. 82256, 3.74%↑) 3008480 (vs. 3008480, 0.00%) 52905492 (vs. 52905492, 0.00%) 728 (vs. 728, 0.00%)
MobileBertSquad_int8(tflite) [arm-valhall-vulkan_android31-vulkan_spirv][experimental-flags,fuse-padding,max-concurrency,compile-stats] 192048 (vs. 179199, 7.17%↑) 6919196 (vs. 6736144, 2.72%↑) 33202857 (vs. 33024150, 0.54%↑) 1078 (vs. 1102, 2.18%↓)
MobileBertSquad_fp32(tflite) [arm-valhall-vulkan_android31-vulkan_spirv][experimental-flags,fuse-padding,max-concurrency,repeated-kernel,compile-stats] 109708 (vs. 106112, 3.39%↑) 143012 (vs. 143012, 0.00%) 99716025 (vs. 99716025, 0.00%) 704 (vs. 704, 0.00%)
MobileBertSquad_fp16(tflite) [arm-valhall-vulkan_android31-vulkan_spirv][experimental-flags,fuse-padding,max-concurrency,repeated-kernel,demote-f32-to-f16,compile-stats] 126913 (vs. 124208, 2.18%↑) 3008480 (vs. 3008480, 0.00%) 54169300 (vs. 54169300, 0.00%) 728 (vs. 728, 0.00%)
MobileBertSquad_int8(tflite) [arm-valhall-vulkan_android31-vulkan_spirv][experimental-flags,fuse-padding,max-concurrency,repeated-kernel,compile-stats] 299729 (vs. 298395, 0.45%↑) 6919196 (vs. 6736144, 2.72%↑) 35074217 (vs. 34937238, 0.39%↑) 1078 (vs. 1102, 2.18%↓)
MobileNetV2_fp32(tflite) [vmvx-generic-vmvx-vmvx][experimental-flags,compile-stats] 22864 (vs. 21229, 7.70%↑) 208629 (vs. 208629, 0.00%) 14213630 (vs. 14213630, 0.00%) 172 (vs. 172, 0.00%)
MobileNetV3Small_fp32(tflite) [vmvx-generic-vmvx-vmvx][experimental-flags,compile-stats] 30975 (vs. 30360, 2.03%↑) 311349 (vs. 311349, 0.00%) 10551678 (vs. 10551678, 0.00%) 210 (vs. 210, 0.00%)
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment