Skip to content
Windows_CI

[aarch64] Add Sbgemm kernel to accelerate fp32 tensor matmul with bfloat16 #17143

Sign in to view logs

[aarch64] Add Sbgemm kernel to accelerate fp32 tensor matmul with bfloat16

[aarch64] Add Sbgemm kernel to accelerate fp32 tensor matmul with bfloat16 #17143

Run time

Learn about OS pricing on GitHub Actions
Job Run time
Onnxruntime-TVM
1h 35m 31s
Windows-CUDA-12
17m 7s
1h 52m 38s