[Autotune] Reduce XLBOCK for outer reduction #114284
Closed
Conversation
This file contains 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
🔗 Helpful Links🧪 See artifacts and rendered test results at hud.pytorch.org/pr/114284
Note: Links to docs will display an error until the docs builds have been completed. ✅ No FailuresAs of commit af820b4 with merge base 9efbb4e ( This comment was automatically generated by Dr. CI and updates every 15 minutes. |
jansel
approved these changes
Nov 21, 2023
|
@pytorchbot label "topic: not user facing" |
|
@pytorchbot merge |
Merge startedYour change will be merged once all checks pass (ETA 0-4 Hours). Learn more about merging in the wiki. Questions? Feedback? Please reach out to the PyTorch DevX Team |
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Outdated suggestions cannot be applied.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.
I have observed that quite a few Reduction.Outer kernels have potential for performance improvement by reducing register pressure. This is due to our current register pressure reduction logics, which only reduces RBLOCK, doesn't work for outer reductions. While we can tighten up there, which will likely increase compile time, I found a better workaround to tune down XBLOCK in the first place.
Perf job: main 9efbb4e (11/16) vs hoy/autotune/reduction
Slight compile time and perf improvement seen.
I also saw perf improvement locally for the few kernels being investigated.
cc @voznesenskym @penguinwu @EikanWang @jgong5 @Guobing-Chen @XiaobingSuper @zhuhaozhe @blzheng @wenzhe-nrv @jiayisunx @peterbell10 @ipiszy @yf225 @chenyang78 @kadeng @muchulee8 @aakhundov @ColinPeppler