Skip to content

[AUTOGENERATED] [release/2.8] [release/2.7] [SWDEV-543214] Reland #2416 Fix warps runtime part 2 #2455

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
merged 1 commit into from
Aug 4, 2025
Merged

[AUTOGENERATED] [release/2.8] [release/2.7] [SWDEV-543214] Reland #2416 Fix warps runtime part 2 #2455

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
11 changes: 8 additions & 3 deletions torch/_inductor/runtime/coordinate_descent_tuner.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -63,9 +63,14 @@ def get_config_max(self, prefix: str) -> int:

@lru_cache(maxsize=1)
def get_warpsmax(self):
# Currently, CUDA has a maximum of 1024 threads, so 32 is the max
# number of warps.
return 1024 // 32
# CUDA/ROCm has a maximum of 1024 threads per block
from torch.cuda import current_device, get_device_properties, is_available

warp_size = (
get_device_properties(current_device()).warp_size if is_available() else 32
)

return 1024 // warp_size

def cache_benchmark_result(self, config, timing):
self.cached_benchmark_results[triton_config_to_hashable(config)] = timing
Expand Down