-
Notifications
You must be signed in to change notification settings - Fork 3k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
[GraphBolt][CUDA] Overlap feature fetcher #6954
Merged
mfbalin
merged 29 commits into
dmlc:master
from
mfbalin:gb_cuda_overlap_feature_fetcher
Jan 18, 2024
Merged
[GraphBolt][CUDA] Overlap feature fetcher #6954
mfbalin
merged 29 commits into
dmlc:master
from
mfbalin:gb_cuda_overlap_feature_fetcher
Jan 18, 2024
Conversation
This file contains 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
To trigger regression tests:
|
7aecffc
to
ba52ef2
Compare
16151dc
to
2e00869
Compare
frozenbugs
reviewed
Jan 17, 2024
frozenbugs
reviewed
Jan 17, 2024
frozenbugs
reviewed
Jan 18, 2024
frozenbugs
approved these changes
Jan 18, 2024
27 tasks
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Outdated suggestions cannot be applied.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.
Description
UVA feature fetching is overlapped with other operations. The picture below shows that

IndexSelectMultiKernelAligned
executes simultaneously as the GNN model forward/backward operations. We may need to tune the 6144 default value in a later PR for optimal performance (Might also want to implement autotuning logic). If it is too high, then too much unnecessary GPU resources will be used for the UVA copy, if it is too little, we can't fully utilize the PCI-e connection. However, it is safer for overall performance to set it to a higher value to ensure the PCI-e can be fully utilized in all different systems.NOTE: The memory leak was caused by multiple CUDA streams being created when we created multiple dataloaders. Making uva_stream a global variable resolved the issue.
Checklist
Please feel free to remove inapplicable items for your PR.
Changes