search

ROCm / Tensile

Stretching GPU performance for GEMMs and tensor contractions.
MIT License
218 stars 147 forks source link

Ignore asm cap check for kernel arg preload for rocm6.0 #1898

Closed nakajee closed 7 months ago

nakajee commented 7 months ago

According to #1757, asm cap check fails with rocm6.0.0. I changed the if condition to ignore asm cap check for preload kernel arg for rocm6.0.

GZGavinZhao commented 7 months ago

This fixed the issue for me on ROCm 6.0.0