Skip to content
CI

For CUDA fro-norm kernels, move syncthreads out of loop. Problem discovered under dpct-generated-sycl kernels. #590

Sign in to view logs

For CUDA fro-norm kernels, move syncthreads out of loop. Problem discovered under dpct-generated-sycl kernels.

For CUDA fro-norm kernels, move syncthreads out of loop. Problem discovered under dpct-generated-sycl kernels. #590

Run time

Learn about OS pricing on GitHub Actions
Job Run time
icl_slate (make, cpu)
7m 14s
icl_slate (make, gpu_nvidia)
10m 27s
icl_slate (make, gpu_amd)
11m 6s
icl_slate (make, gpu_intel)
5m 58s
icl_slate (cmake, cpu)
3m 44s
icl_slate (cmake, gpu_nvidia)
5m 37s
icl_slate (cmake, gpu_amd)
3m 57s
icl_slate (cmake, gpu_intel)
5m 21s
53m 24s