Skip to content

NCCL UB fix: reduce memory cost and correctly deregister NCCL mem pool - #4492

Merged
Phlip79 merged 8 commits into
NVIDIA:mainfrom
xrennvidia:xren/nccl_ub_fix
May 14, 2026
Merged

NCCL UB fix: reduce memory cost and correctly deregister NCCL mem pool#4492
Phlip79 merged 8 commits into
NVIDIA:mainfrom
xrennvidia:xren/nccl_ub_fix

Conversation

@xrennvidia

@xrennvidia xrennvidia commented Apr 28, 2026

Copy link
Copy Markdown
Contributor

What does this PR do ?

  • Fix a post-init check so that --use-nccl-ub can work with export PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True. This is because Pytorch has made them work with each other (refer this PR)
  • Manually deregister NCCL mem pool at the end of training step. We should not need this after this Pytorch PR is landed, but it's suggested to keep the code because it's the user's responsibility to manage the lifecycle of the mem pool.

Issue tracking

For PRs from open-source community contributors:

  • New features: a linked issue is required. Please open a feature request and reference it here before submitting the PR.
  • Small updates (bug fixes, minor improvements): a linked issue is recommended and will accelerate the PR review process.

Linked issue:

Contribution process

Pre-checks

  • I have added relevant unit tests
  • I have added relevant functional tests
  • I have added proper typing to my code Typing guidelines
  • I have added relevant documentation
  • I have run the autoformatter.sh on my PR

Code review

Feel free to message or comment the @mcore-oncall to help accelerate your merge into main. The less complex your PR is, the faster it will be approved and merged!

All PRs start as draft. If you open a non-draft PR, it will be automatically converted to draft.

Step 1: Mark PR as "Ready for Review"

  1. When your PR is ready, click Ready for Review.
  2. An oncall reviewer is auto-assigned and expert reviewers are notified based on your changes.
    • Some PRs may jump straight to step 2. This is determined by .github/CODEOWNERS.

⚠️ Only mark as ready once merge-conflicts are resolved and the CI is passing.
Final Review might get declined if these requirements are not fulfilled.

Step 2: Final Review

For PRs that change megatron/core, once all expert reviewers have approved, the Final Review label is applied automatically and final reviewers are assigned.

For PRs outside megatron/core, this step is skipped.

Step 3: Approved

Once all required reviewers have approved, the Approved label is applied automatically.

Merge

Any member of mcore-engineers will be able to merge your PR.

For MRs into `dev` branch The proposed review process for `dev` branch is under active discussion.

MRs are mergable after one approval by either eharper@nvidia.com or zijiey@nvidia.com.

@xrennvidia
fix a post_init check
83466b6 
Signed-off-by: Xiaowei Ren <xren@nvidia.com>
@xrennvidia
manually deregister mem pool
ebd9843 
Signed-off-by: Xiaowei Ren <xren@nvidia.com>
@xrennvidia
fix a typo
708e221 
Signed-off-by: Xiaowei Ren <xren@nvidia.com>
@xrennvidia
add a barrier before deregister nccl mem pool
b8af80b 
Signed-off-by: Xiaowei Ren <xren@nvidia.com>
@xrennvidia
xrennvidia requested review from a team as code owners April 28, 2026 08:45
@copy-pr-bot

copy-pr-bot Bot commented Apr 28, 2026

Copy link
Copy Markdown

This pull request requires additional validation before any workflows can run on NVIDIA's runners.

Pull request vetters can view their responsibilities here.

Contributors can view more details about this message here.

@svcnvidia-nemo-ci
svcnvidia-nemo-ci marked this pull request as draft April 28, 2026 08:46
@github-actions

github-actions Bot commented Apr 28, 2026

Copy link
Copy Markdown
Contributor

This PR has been automatically converted to draft because all PRs must start as drafts.

When you are ready for review, click Ready for Review to begin the review process. This will:

  1. Add the oncall reviewer (optional reviewer)
  2. Add required review teams based on your changes

See the contribution guide for more details.

@xrennvidia
xrennvidia marked this pull request as ready for review April 28, 2026 08:46
@svcnvidia-nemo-ci
svcnvidia-nemo-ci requested a review from a team April 28, 2026 08:46
@xrennvidia xrennvidia changed the title Fix: NCCL UB, reduce memory cost and correctly deregister NCCL mem pool NCCL UB fix: reduce memory cost and correctly deregister NCCL mem pool Apr 28, 2026
@xrennvidia
auto formatting
90981ac 
Signed-off-by: Xiaowei Ren <xren@nvidia.com>
@xrennvidia
Merge branch 'main' into xren/nccl_ub_fix
041e900
cspades
cspades approved these changes Apr 30, 2026
View reviewed changes

@cspades cspades left a comment

Copy link
Copy Markdown
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Never hit this error before but makes sense to me, we also have a memory leak so any kind of better cleanup is welcome.

xrennvidia added 2 commits May 5, 2026 13:31
@xrennvidia
merge with main
8dd5c32 
Signed-off-by: Xiaowei Ren <xren@nvidia.com>
@xrennvidia
Merge branch 'main' into xren/nccl_ub_fix
4b42c08
@svcnvidia-nemo-ci svcnvidia-nemo-ci added the Final Review PR is in the "final review" stage label May 13, 2026
@svcnvidia-nemo-ci svcnvidia-nemo-ci added Approved All necessary approvals have been made and removed Final Review PR is in the "final review" stage labels May 13, 2026
@Phlip79

Phlip79 commented May 13, 2026

Copy link
Copy Markdown
Member

/ok to test 4b42c08

@Phlip79
Phlip79 enabled auto-merge May 13, 2026 21:59
@Phlip79
Phlip79 added this pull request to the merge queue May 13, 2026
@svcnvidia-nemo-ci

svcnvidia-nemo-ci commented May 13, 2026

Copy link
Copy Markdown
Contributor

🔄 Merge queue validation started!

You can track the progress here: https://github.com/NVIDIA/Megatron-LM/actions/runs/25832035771

@svcnvidia-nemo-ci

svcnvidia-nemo-ci commented May 13, 2026

Copy link
Copy Markdown
Contributor

🔄 Merge queue validation started!

You can track the progress here: https://github.com/NVIDIA/Megatron-LM/actions/runs/25833058079

Merged via the queue into NVIDIA:main with commit e35d4e5 May 14, 2026
66 of 70 checks passed
@xrennvidia
xrennvidia deleted the xren/nccl_ub_fix branch May 14, 2026 08:02
@cspades

cspades commented May 14, 2026 •
edited
Loading

Copy link
Copy Markdown
Member

@Phlip79 @yaoyu-33 @ko3n1g FYI seems like this broke Megatron-FSDP PyPI publishing since we're importing a utility from MCore!

Just noticed that the PyPI package flow is not required to merge.

janEbert pushed a commit to janEbert/Megatron-LM that referenced this pull request Jun 2, 2026
@xrennvidia @janEbert
NCCL UB fix: reduce memory cost and correctly deregister NCCL mem pool (
59b115b 
NVIDIA#4492)

Signed-off-by: Xiaowei Ren <xren@nvidia.com>
yhgalaxy pushed a commit to yhgalaxy/Megatron-LM that referenced this pull request Jun 17, 2026
@xrennvidia @yhgalaxy
NCCL UB fix: reduce memory cost and correctly deregister NCCL mem pool (
80c0fe5 
NVIDIA#4492)

Signed-off-by: Xiaowei Ren <xren@nvidia.com>
Signed-off-by: yhgalaxy <yhgalaxy@outlook.com>
jon-barker pushed a commit to jon-barker/Megatron-LM that referenced this pull request Jul 10, 2026
@xrennvidia Jon Barker
NCCL UB fix: reduce memory cost and correctly deregister NCCL mem pool (
8605c72 
NVIDIA#4492)

Signed-off-by: Xiaowei Ren <xren@nvidia.com>
Signed-off-by: Jon Barker <jbarker@aws-cmh-slurm-1-vscode-02.cm.cluster>
terminator123 pushed a commit to 021ai/Megatron-LM that referenced this pull request Aug 3, 2026
@xrennvidia @terminator123
NCCL UB fix: reduce memory cost and correctly deregister NCCL mem pool (
b062387 
NVIDIA#4492)

Signed-off-by: Xiaowei Ren <xren@nvidia.com>
svcnvidia-nemo-ci pushed a commit to dimapihtar/Megatron-LM that referenced this pull request Aug 4, 2026
@xrennvidia @dimapihtar
NCCL UB fix: reduce memory cost and correctly deregister NCCL mem pool (
c8f94f8 
NVIDIA#4492)

Signed-off-by: Xiaowei Ren <xren@nvidia.com>
Signed-off-by: Dmytro Pykhtar <dpykhtar@nvidia.com>
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Reviewers

@deepakn94 deepakn94 deepakn94 approved these changes
@cspades cspades cspades approved these changes
@yaoyu-33 yaoyu-33 yaoyu-33 approved these changes
@asolergi-nv asolergi-nv asolergi-nv approved these changes
@youngeunkwon0405 youngeunkwon0405 Awaiting requested review from youngeunkwon0405

Assignees

No one assigned

Labels

26.06 Approved All necessary approvals have been made complexity: low nemotron

Projects

None yet

Milestone

No milestone

Development

Successfully merging this pull request may close these issues.

8 participants

@xrennvidia @Phlip79 @svcnvidia-nemo-ci @cspades @deepakn94 @yaoyu-33 @asolergi-nv @erhoo82