debug onnx benchmark workflow

[Human9000-bit]

No description provided.

[Human9000-bit]

/bench

[Human9000-bit]

/bench

[Human9000-bit]

/bench

[Human9000-bit]

/bench

[Human9000-bit]

/bench

[github-actions]

ONNX CPU benchmark

GitHub-hosted runner signal, advisory only. Base and head are compared within the same architecture job.

arch	case	base p50	head p50	delta
aarch64	gemm_relu_256x512x512	1.22 ms	1.21 ms	-0.7% flat
aarch64	small_conv_relu_64	194 us	474 us	+144.3% slower
aarch64	small_dw_pw_64	168 us	165 us	-1.8% faster
aarch64	small_gemm_relu_1x128	8 us	8 us	+0.0% flat
aarch64	small_residual_conv_64	439 us	1.09 ms	+148.5% slower
aarch64	winograd_3x3_20x20_c256	3.47 ms	36.66 ms	+955.4% slower
aarch64	winograd_3x3_80x80_c64	6.18 ms	76.78 ms	+1142.6% slower
aarch64	yolo11n_bus_640	79.79 ms	415.10 ms	+420.3% slower
aarch64	yolov8n_bus_640	89.37 ms	656.39 ms	+634.4% slower
x86_64	gemm_relu_256x512x512	1.00 ms	987 us	-1.6% faster
x86_64	small_conv_relu_64	383 us	144 us	-62.4% faster
x86_64	small_dw_pw_64	143 us	154 us	+7.7% slower
x86_64	small_gemm_relu_1x128	7 us	7 us	+0.0% flat
x86_64	small_residual_conv_64	421 us	236 us	-43.9% faster
x86_64	winograd_3x3_20x20_c256	2.52 ms	2.75 ms	+9.3% slower
x86_64	winograd_3x3_80x80_c64	6.00 ms	5.95 ms	-0.9% flat
x86_64	yolo11n_bus_640	86.22 ms	85.61 ms	-0.7% flat
x86_64	yolov8n_bus_640	90.12 ms	91.66 ms	+1.7% slower

Head op/kernel summary

arch	case	ops	dispatched kernels
aarch64	gemm_relu_256x512x512	Gemm x1, Relu x1	Gemm via blocked-mr4 x1
aarch64	small_conv_relu_64	Conv_Relu x1	Conv_Relu via indirect-nhwc-3x3 x1
aarch64	small_dw_pw_64	FusedDwPw x1	-
aarch64	small_gemm_relu_1x128	Gemm x1, Relu x1	Gemm via row-gemm x1
aarch64	small_residual_conv_64	Conv_Add_Relu x1	Conv_Add_Relu via indirect-nhwc-3x3 x1
aarch64	winograd_3x3_20x20_c256	Conv_Relu x1	Conv_Relu via indirect-nhwc-3x3 x1
aarch64	winograd_3x3_80x80_c64	Conv_Relu x1	Conv_Relu via indirect-nhwc-3x3 x1
aarch64	yolo11n_bus_640	Conv x85, Mul x83, Sigmoid x78, Concat x39, Add x22, Reshape x17, +18 more	Conv via nhwc-gemm-prepacked/pw-gemm x44, Conv via indirect-nhwc-3x3 x34, Conv via dw-nhwc-padded/dw-neon x6, MatMul via blocked-mr8 x2, Conv via nhwc-padded/first-layer-rgb x1, Conv_Add via dw-nhwc-padded/dw-neon x1
aarch64	yolov8n_bus_640	Conv x64, Mul x61, Sigmoid x58, Concat x33, Add x17, Reshape x14, +15 more	Conv via indirect-nhwc-3x3 x38, Conv via nhwc-gemm-prepacked/pw-gemm x25, Conv via nhwc-padded/first-layer-rgb x1
x86_64	gemm_relu_256x512x512	Gemm x1, Relu x1	Gemm via blocked-mr4 x1
x86_64	small_conv_relu_64	Conv_Relu x1	Conv_Relu via nhwc-padded/im2col-gemm x1
x86_64	small_dw_pw_64	FusedDwPw x1	-
x86_64	small_gemm_relu_1x128	Gemm x1, Relu x1	Gemm via row-gemm x1
x86_64	small_residual_conv_64	Conv_Add_Relu x1	Conv_Add_Relu via nhwc-padded/im2col-gemm x1
x86_64	winograd_3x3_20x20_c256	Conv_Relu x1	Conv_Relu via nhwc-padded/im2col-gemm x1
x86_64	winograd_3x3_80x80_c64	Conv_Relu x1	Conv_Relu via nhwc-padded/im2col-gemm x1
x86_64	yolo11n_bus_640	Conv x85, Mul x83, Sigmoid x78, Concat x39, Add x22, Reshape x17, +18 more	Conv via nhwc-padded/im2col-gemm x34, Conv via nhwc-gemm-prepacked/pw-gemm x24, Conv via nhwc-gemm-prepacked/pw-nx16-direct x20, Conv via dw-nhwc-padded/dw-avx-fma x6, MatMul via blocked-mr4 x2, Conv via nhwc-padded/first-layer-rgb x1, +1 more
x86_64	yolov8n_bus_640	Conv x64, Mul x61, Sigmoid x58, Concat x33, Add x17, Reshape x14, +15 more	Conv via nhwc-padded/im2col-gemm x38, Conv via nhwc-gemm-prepacked/pw-gemm x17, Conv via nhwc-gemm-prepacked/pw-nx16-direct x8, Conv via nhwc-padded/first-layer-rgb x1

Raw JSON artifacts include per-run min/p50/avg/p95/p99, optimized node counts, yscv CPU dispatch report, and per-node runner profile summaries.

[Human9000-bit]

yay