[SymmMem] Increase minimum nthreads to cover sync needs in NVL72#161983
Closed
kwen2501 wants to merge 2 commits intogh/kwen2501/220/basefrom
Closed
[SymmMem] Increase minimum nthreads to cover sync needs in NVL72#161983kwen2501 wants to merge 2 commits intogh/kwen2501/220/basefrom
kwen2501 wants to merge 2 commits intogh/kwen2501/220/basefrom
Conversation
🔗 Helpful Links🧪 See artifacts and rendered test results at hud.pytorch.org/pr/161983
Note: Links to docs will display an error until the docs builds have been completed. ❗ 1 Active SEVsThere are 1 currently active SEVs. If your PR is affected, please view them below: ✅ No FailuresAs of commit a4fc434 with merge base 524b78d ( This comment was automatically generated by Dr. CI and updates every 15 minutes. |
ngimel
reviewed
Sep 2, 2025
| at::ceil_div(numel_per_split, numel_per_thread), | ||
| static_cast<size_t>(at::cuda::warp_size())); | ||
| num_threads = at::ceil_div(numel_per_split, numel_per_thread); | ||
| num_threads = max(num_threads, MAX_CUDA_P2P_DOMAIN_SIZE); |
Collaborator
There was a problem hiding this comment.
Better to pass world_size here and make max equal to world size
ngimel
approved these changes
Sep 2, 2025
Collaborator
Author
|
@pytorchbot merge |
Collaborator
Merge startedYour change will be merged once all checks pass (ETA 0-4 Hours). Learn more about merging in the wiki. Questions? Feedback? Please reach out to the PyTorch DevX Team |
pytorchmergebot
pushed a commit
that referenced
this pull request
Sep 4, 2025
…1984) Added a helper API to tell if the world is entirely within a P2P domain or crosses network. This is mainly for nblocks tuning purpose. (In later PRs) Pull Request resolved: #161984 Approved by: https://github.com/ngimel ghstack dependencies: #161983
markc-614
pushed a commit
to markc-614/pytorch
that referenced
this pull request
Sep 17, 2025
…orch#161983) `sync_remote_blocks` maps threads to peers. Previously min nthreads is warp size, which is too small to cover NVL72. Bumping it. Pull Request resolved: pytorch#161983 Approved by: https://github.com/ngimel
markc-614
pushed a commit
to markc-614/pytorch
that referenced
this pull request
Sep 17, 2025
…orch#161984) Added a helper API to tell if the world is entirely within a P2P domain or crosses network. This is mainly for nblocks tuning purpose. (In later PRs) Pull Request resolved: pytorch#161984 Approved by: https://github.com/ngimel ghstack dependencies: pytorch#161983
mansiag05
pushed a commit
to mansiag05/pytorch
that referenced
this pull request
Sep 22, 2025
…orch#161983) `sync_remote_blocks` maps threads to peers. Previously min nthreads is warp size, which is too small to cover NVL72. Bumping it. Pull Request resolved: pytorch#161983 Approved by: https://github.com/ngimel
mansiag05
pushed a commit
to mansiag05/pytorch
that referenced
this pull request
Sep 22, 2025
…orch#161984) Added a helper API to tell if the world is entirely within a P2P domain or crosses network. This is mainly for nblocks tuning purpose. (In later PRs) Pull Request resolved: pytorch#161984 Approved by: https://github.com/ngimel ghstack dependencies: pytorch#161983
cleonard530
pushed a commit
to cleonard530/pytorch
that referenced
this pull request
Sep 22, 2025
…orch#161984) Added a helper API to tell if the world is entirely within a P2P domain or crosses network. This is mainly for nblocks tuning purpose. (In later PRs) Pull Request resolved: pytorch#161984 Approved by: https://github.com/ngimel ghstack dependencies: pytorch#161983
dsashidh
pushed a commit
to dsashidh/pytorch
that referenced
this pull request
Sep 26, 2025
…orch#161984) Added a helper API to tell if the world is entirely within a P2P domain or crosses network. This is mainly for nblocks tuning purpose. (In later PRs) Pull Request resolved: pytorch#161984 Approved by: https://github.com/ngimel ghstack dependencies: pytorch#161983
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
3 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.
Stack from ghstack (oldest at bottom):
sync_remote_blocksmaps threads to peers. Previously min nthreads is warp size, which is too small to cover NVL72. Bumping it.cc @H-Huang @awgu @wanchaol @fegin @fduwjj @wz337 @wconstab @d4l3k @pragupta