fix(rocm): avoid SIMT workgroup-count resonance#2012
Open
sstamenk wants to merge 1 commit into
Open
Conversation
Add one bounds-checked workgroup when a large HIP SIMT grid is an exact multiple of 256, avoiding scheduler underutilization observed on RDNA3 and RDNA4.\n\nCo-Authored-By: Claude <noreply@anthropic.com>
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Add one bounds-checked workgroup when a large HIP SIMT grid is an exact multiple of 256, avoiding scheduler underutilization observed on RDNA3 and RDNA4.
The extra workgroup is guarded and does no work:
csrc/gemm_4bit_simt.cu#L196-L197
This addresses the suboptimal performance on certain shapes that was originally noticed in #1979 and achieves up to 3.5x performance improvement at M=1 for specific shapes across fp16, bf16 and fp32.
Not exactly sure of the mechanism that causes the lower-than-expected performance.
gfx1201
fp16
bf16
fp32