[None][fix] Bump DeepGEMM to nv_dev HEAD for SM120 attention support#16217
Open
lucifer1004 wants to merge 1 commit into
Open
[None][fix] Bump DeepGEMM to nv_dev HEAD for SM120 attention support#16217lucifer1004 wants to merge 1 commit into
lucifer1004 wants to merge 1 commit into
Conversation
The pinned DeepGEMM (245dc5d, 2026-06-04) predates the SM120 support merged on nv_dev (NVIDIA#324, 2026-06-24): its attention APIs branch on arch_major 9/10 only and DeepSeek-V4 serving on SM120 dies at engine warmup with 'Unsupported architecture' from get_paged_mqa_logits_metadata. The current nv_dev head carries arch_major 12 branches across the attention interface (fp8/fp4 MQA logits, paged logits metadata, SM120 block_kv sizing) plus the indexer_n_heads=16 support, enabling the DSv4 sparse indexer on SM120 without any TensorRT-LLM-side workaround. Co-authored-by: Claude Fable 5 <noreply@anthropic.com> Signed-off-by: Zihua Wu <13583761+lucifer1004@users.noreply.github.com>
Contributor
|
No actionable comments were generated in the recent review. 🎉 ℹ️ Recent review info⚙️ Run configurationConfiguration used: Path: .coderabbit.yaml Review profile: CHILL Plan: Enterprise Run ID: 📒 Files selected for processing (1)
📝 WalkthroughWalkthroughThe DeepGEMM dependency pin in ChangesDeepGEMM Dependency Update
Estimated code review effort: 1 (Trivial) | ~2 minutes 🚥 Pre-merge checks | ✅ 4 | ❌ 1❌ Failed checks (1 warning)
✅ Passed checks (4 passed)
✨ Finishing Touches🧪 Generate unit tests (beta)
Comment |
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
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
The pinned DeepGEMM commit predates the SM120 implementations of the DSA/DSv4 indexer attention kernels (
sm120_fp8_mqa_logits/sm120_fp8_paged_mqa_logits), so on SM120/SM121 the sparse-attention indexer fails with an unsupported-architecture error even though current DeepGEMMnv_devsupports the arch.Bump the 3rdparty fetch pin to nv_dev
a6b593d2826719dcf4892609af7b84ee23aaf32a, which carries the SM120 kernels. Validated as part of DeepSeek-V4 bring-up on RTX PRO 6000 Blackwell (the indexer dispatches and the model passes gsm8k gating at tp2/tp4).🤖 Generated with Claude Code
Summary by CodeRabbit