[Kernel][SM70] Integrate copy-only GDN projection split from #504 - #549
Merged
yangzhuxinyzx merged 1 commit intoSep 7, 2026
Merged
Conversation
Extract the independently verified copy kernel from #504; preserve main GEMV roles and unrelated defaults. Co-authored-by: OpenAI Codex <noreply@openai.com> Signed-off-by: yangzhuxinyzx <153831768+yangzhuxinyzx@users.noreply.github.com>
yangzhuxinyzx
marked this pull request as ready for review
September 7, 2026 10:23
This was referenced Sep 7, 2026
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
AI-assisted integration requested by the maintainer. Narrow extraction from #504; the parent experimental HC/QSA bundle stays open. Base: 8d9c351.
Purpose: fuse only the four existing batched GDN output copies without changing either GEMM or latest-main role-specific plans. Copy flag defaults on; unrelated arithmetic flags unchanged.
Test Plan/Result: 61 targeted CPU/GPU tests passed on owned V100; changing-input CUDA Graph replays and exhaustive 65536 half-payload checks passed. Actual layer weights, synthetic activations, paired complete-projection graph latency improves 11–19% at M2/4/8/16/32/64. No E2E suite or model-throughput claim. Full evidence in docs/design/sm70_gdn_projection_copy_audit.md. Local scoped pre-commit passed. Draft pending hosted checks and final SHA review.