CUDA/HIP: partial-top-k argsort — up to 6x faster TOP_K on RDNA3.5/RDNA4#13
Closed
DeanoC wants to merge 1 commit into
Closed
CUDA/HIP: partial-top-k argsort — up to 6x faster TOP_K on RDNA3.5/RDNA4#13DeanoC wants to merge 1 commit into
DeanoC wants to merge 1 commit into