Optimize lxu_cache_lookup. #952
Closed
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
Differential Revision: D34353556 fbshipit-source-id: e61a36d0affba2d0dc976a28ed3fe03859cf2885
Differential Revision: D34491026 fbshipit-source-id: ea68c8f1cbb28681d743545ceab0e3aada39a852
Summary: Optimize lxu_cache_lookup kernel by doing: (1) write coalescing (2) early return 35% improvement when UVM cache is fully used, and almost 10x improvement when UVM cache is partially used. Differential Revision: D34491095 fbshipit-source-id: 87faaf347883823c728b27f3dd54c19cc7570cc0
|
This pull request was exported from Phabricator. Differential Revision: D34491095 |
|
This pull request has been reverted by abc7b74. |
q10
pushed a commit
to q10/FBGEMM
that referenced
this pull request
Apr 10, 2025
Summary: X-link: pytorch#3862 Pull Request resolved: facebookresearch/FBGEMM#952 Applies the same optimization used in D71510967 to cutlass fp8 grouped gemm. This should help performance for cases where G > M. Reviewed By: jiawenliu64 Differential Revision: D71582782 fbshipit-source-id: 05a86398164b1a4bd6af46e9af2ec7f5faabdeb0
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
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Outdated suggestions cannot be applied.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.
Summary:
Optimize lxu_cache_lookup kernel by doing:
(1) write coalescing
(2) early return
35% improvement when UVM cache is fully used, and almost 10x improvement when UVM cache is partially used.
Differential Revision: D34491095