feat: add nccl-extensions dependency - #3973
Open
kingchc wants to merge 1 commit into
Open
Conversation
Signed-off-by: Ching-Hsiang Chu <chichu@nvidia.com>
kingchc
marked this pull request as ready for review
September 2, 2026 22:21
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
What does this PR do ?
Adds
nccl-extensions==0.1.0as a Linux dependency for x86_64 and aarch64 systems, and updatesuv.lockwith the published wheels.This makes the native
nccl.m2n.reshardimplementation available to thenccl_reshardrefit transport without a separate package installation.Issues
N/A
Usage
No usage change is required. Configure the existing transport as usual:
Before your PR is "Ready for review"
Pre checks:
uv lock --checkwith Python 3.13.14.nccl_reshardinterface is unchanged.Additional Information
Validated manually with four GB200 GPUs using Python 3.13.14, Torch 2.11.0+cu130, CUDA 13.0, NCCL 2.30.7,
nccl4py==0.3.1, andnccl-extensions==0.1.0.The four-rank cross-dimension
nccl.m2n.reshardbenchmark completed successfully with validation enabled.