2023 was the year that GPUs stood still

Re:2023 was the year that GPUs stood still by brucethemoose2 on Friday December 29, 2023 @12:51PM (#38807930)

I don't see the distinction. Tons of popular PyTorch projects run hand made CUDA kernels, whether through a library or one written themselves.

In some cases one can replace rocm/openvino incompatible libraries with pure PyTorch, but the performance hit is often severe, and not many inference projects are using torch.compile or triton kernels to compensate.

[ Parent | on HN ]

Re:2023 was the year that GPUs stood still by sangnoir on Friday December 29, 2023 @01:18PM (#38808280)

> I don't see the distinction

Performant Pytorch compatibility can be achieved without CUDA compatibility.

> In some cases one can replace rocm/openvino incompatible libraries with pure PyTorch

This is the crux of my point: AMD ought to focus on fixing these Pytorch incompatibilities instead of splitting focus with chasing CUDA translations.

[ Parent | on HN ]