Skip to content
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] Make fetching indices optional for NS. #7662

Merged
merged 2 commits into from
Aug 7, 2024

Conversation

mfbalin
Copy link
Collaborator

@mfbalin mfbalin commented Aug 6, 2024

Description

When indices is pinned, its fetch takes long. We want to actually overlap its fetch operation. We already have a overlap_graph_fetch option that fetches the full InSubgraph. However, this is wasteful for NeighborSampling as it only requires fetching the indices for the sampled edges. Thus, we make it optional so that it can be fetched in another stage during sampling in an overlapped way.

Checklist

Please feel free to remove inapplicable items for your PR.

  • The PR title starts with [$CATEGORY] (such as [NN], [Model], [Doc], [Feature]])
  • I've leverage the tools to beautify the python and c++ code.
  • The PR is complete and small, read the Google eng practice (CL equals to PR) to understand more about small PR. In DGL, we consider PRs with less than 200 lines of core code change are small (example, test and documentation could be exempted).
  • All changes have test coverage
  • Code is well-documented
  • To the best of my knowledge, examples are either not affected by this change, or have been fixed to be compatible with this change
  • Related issue is referred in this PR
  • If the PR is for a new model/paper, I've updated the example index here.

Changes

@mfbalin mfbalin added the expedited if it doesn't affect the main path approve first to unblock related projects, and review later label Aug 6, 2024
@mfbalin mfbalin requested a review from frozenbugs August 6, 2024 23:46
@dgl-bot
Copy link
Collaborator

dgl-bot commented Aug 6, 2024

To trigger regression tests:

  • @dgl-bot run [instance-type] [which tests] [compare-with-branch];
    For example: @dgl-bot run g4dn.4xlarge all dmlc/master or @dgl-bot run c5.9xlarge kernel,api dmlc/master

@dgl-bot
Copy link
Collaborator

dgl-bot commented Aug 7, 2024

Commit ID: 9c09380

Build ID: 1

Status: ⚪️ CI test cancelled due to overrun.

Report path: link

Full logs path: link

@dgl-bot
Copy link
Collaborator

dgl-bot commented Aug 7, 2024

Commit ID: 93a519e

Build ID: 2

Status: ✅ CI test succeeded.

Report path: link

Full logs path: link

@mfbalin mfbalin merged commit d6f0771 into dmlc:master Aug 7, 2024
2 checks passed
@mfbalin mfbalin deleted the gb_cuda_optimize_ns_indices_fetch_1 branch August 7, 2024 00:37
output_indices = Gather(indices, picked_eids);
// TODO @mfbalin: remove true from here once fetching indices later is
// setup.
if (true || layer || utils::is_on_gpu(indices)) {
Copy link
Collaborator

Choose a reason for hiding this comment

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

I didn't catch it in previous review, but layer is not a good name for boolean.

This pull request was closed.
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
expedited if it doesn't affect the main path approve first to unblock related projects, and review later
Projects
None yet
Development

Successfully merging this pull request may close these issues.

3 participants