OSSForks/pytorch
1
0
Fork 0
mirror of https://github.com/zebrajr/pytorch.git synced 2025-12-06 12:20:52 +01:00
Code Issues Actions 38 Packages Projects Releases Wiki Activity

[inductor] Change minimum number of SMs to 60 to let Ada use Triton GEMM backend (#150888)

Browse Source 
context: https://github.com/pytorch/pytorch/issues/150390#issuecomment-2790272814

Pull Request resolved: https://github.com/pytorch/pytorch/pull/150888
Approved by: https://github.com/jansel
This commit is contained in:
henrylhtsang 2025-04-08 16:56:30 -07:00 committed by PyTorch MergeBot
parent 4161c752bb
commit 115a165f9b
1 changed files with 1 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
torch/_inductor/utils.py
View File

@ -1275,7 +1275,7 @@ def is_big_gpu(index_or_device: Union[int, torch.device] = 0) -> bool:
return False
return True
min_sms = 16 if device.type == "xpu" else 68 # 3080
min_sms = 16 if device.type == "xpu" else 60 # 3080
avail_sms = prop.multi_processor_count
if avail_sms < min_sms:
log.warning(