perf(ep): dispatch intranode opt within small token#333
Open
kawhil-amd wants to merge 3 commits into
Open
Conversation
…rnel - Explicitly use Unroll=2 for token data WarpCopy in dispatch kernel to reduce loop iterations from 14 to 7 for hiddenDim=7168 - Add __launch_bounds__(512, 1) to help compiler optimize register allocation Co-Authored-By: Claude Opus 4 <noreply@anthropic.com>
Collaborator
|
Maybe you could record the data before and after optimization in this PR, as well as block/warp, for 128 and 4096 tokens, respectively? |
Contributor
Author
Sure, will do it later. |
kawhil-amd
commented
May 21, 2026
| FlatTokenIndex(config, myPe, srcTokId); | ||
|
|
||
| // Wait for previous round to be consumed (use atomicAdd for shared memory) | ||
| while (atomicAdd(&groupCounters[warpGroupIdInBlock], 0) != 0) { |
Contributor
Author
There was a problem hiding this comment.
should move forward?
kawhil-amd
commented
May 21, 2026
|
|
||
| core::WarpCopy(args.intraNodeTokBufs.dispatchOut->template GetAs<T*>(destPe) + destTokOffset, | ||
| args.inpTokenBuf + srcTokOffset, hiddenDim); | ||
| if (myDimChunk > 0) { |
Contributor
Author
There was a problem hiding this comment.
remove branch within ensure divisibility condition?
kawhil-amd
commented
May 22, 2026
| // Use atomicAdd for shared memory atomic load (workgroup scope) | ||
| uint64_t val; | ||
| do { | ||
| val = atomicAdd((unsigned long long*)&groupData[warpGroupIdInBlock], 0ULL); |
kawhil-amd
changed the title
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
No description provided.