Multi-thread Performance Improvement of GEMM on NeoverseV1 with DIVIDE_RATE=1#5407
Merged
martin-frbg merged 1 commit intoOpenMathLib:developfrom Jul 30, 2025
Conversation
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
2 participants
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
This pull request provides a performance improvement for Neoverse V1, addressing Issue #5347.
It differs from the fix in pull request #5353 for A64FX, focusing on matrix size N=2.
While this change primarily enhances performance for N=2, there's potential for further gains up to N=6 on certain architectures. To support this, a new macro,
GEMM_DIVIDE_LIMIT, has been introduced to manage theDIVIDE_RATEthreshold.This modification has shown performance improvements for GEMM operations on AWS Graviton3E (Neoverse V1) when N=2, as illustrated in the graph below.