Open
Conversation
…ersion) Requires CUDA 13.0+ ptxas. Gives ~10% speedup on activation quantization kernels, translating to 3-5% end-to-end improvement at M=1024. Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
…1x2 conversion)" This reverts commit 4631e2f.
Adds GEMLITE_ENABLE_PTX_PACK global (default False) and gemlite.set_ptx_pack(True/False) API. When enabled, MXFP4/NVFP4 activation quantization kernels use hardware cvt.rn.satfinite.e2m1x2 PTX instruction instead of threshold comparisons. Requires CUDA 13.0+ ptxas to be installed in the Triton backends directory. Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
This PR is makes major changes to gemlite to address various bugs and performance improvements, especially for
sm_120gpus (RTX PRO 6000 more specifically)cvt.rn.satfinite.e2m1x2.f16x2with ptxas 13 but also supports a fallback implementation for older ptxas versions.