Skip to content

[CuTeDSL] Update atomic_max_float32 to atomic_fmax in blockscaled GEMM amax example #3155

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

Closed
questa-wang wants to merge 1 commit into from
+2 −3
Closed

[CuTeDSL] Update atomic_max_float32 to atomic_fmax in blockscaled GEMM amax example #3155

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
5 changes: 2 additions & 3 deletions examples/python/CuTeDSL/blackwell/dense_blockscaled_gemm_persistent_amax.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -1423,9 +1423,8 @@ def kernel(

# Global atomic max (accumulates across all tiles for final tensor amax)
# Since we compute absolute values, all values are non-negative
# Use wrapper function for atomic max operation
_ = cute.arch.atomic_max_float32(
ptr=mAmax.iterator.llvm_ptr, value=block_amax
_ = cute.arch.atomic_fmax(
mAmax.iterator, block_amax, sign_bit=False
)
#
# Async arrive accumulator buffer empty
Expand Down