perf(kernel): optimize SM120 FP8 GEMM with FP16 accumulation - #1479
Merged
Conversation
STwangyingrui
force-pushed
the
yr/h3-fp8-f16-accum
branch
from
September 8, 2026 15:01
6e1aff7 to
a527461
Compare
llmc-reviewer
approved these changes
Sep 10, 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
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Optimizes MiniMax-H3 inference on RTX 5090/SM120 with an FP8 GEMM path using FP16 accumulation. It includes exact-shape CUTLASS autotuning with persistent caching, qmax-aware checkpoint conversion and validation, and safe FP8-SGL fallback. The optimized path covers selected DiT and Video VAE decoder projections, with configuration and documentation included.