Open
Description
I am using Triton==3.2.0, but the GEMM ops works as expected only on H100 and few high-end GPUs, and they all fails on normal GPUs like 2080Ti and MI50, by throwing the reason:
File "/home/pi-user/trifast/.venv/lib/python3.11/site-packages/triton/backends/nvidia/compiler.py", line 341, in make_llir
pm.run(mod)
RuntimeError: PassManager::run failed
Metadata
Metadata
Assignees
Labels
No labels