Skip to content

WIP: Prepare cagra::optimize for use with pinned memory#1830

Open
mfoerste4 wants to merge 14 commits intorapidsai:mainfrom
mfoerste4:cagra_optimize
Open

WIP: Prepare cagra::optimize for use with pinned memory#1830
mfoerste4 wants to merge 14 commits intorapidsai:mainfrom
mfoerste4:cagra_optimize

Conversation

@mfoerste4
Copy link
Contributor

@mfoerste4 mfoerste4 commented Feb 20, 2026

In preparation for large scale graph creation this PR adds support for pinned memory usage in cagra:optimize by:

  • adding GPU path for pruning and merging (done)
  • restructure pruning batching to minimize device memory requirements (done)
  • Improve/simplify memory strategy to always use GPU path
  • support pinned memory arguments in API

New kernels are based on experiments by @bpark-nvidia

CC @tfeher

@copy-pr-bot
Copy link

copy-pr-bot bot commented Feb 20, 2026

Auto-sync is disabled for draft pull requests in this repository. Workflows must be run manually.

Contributors can view more details about this message here.

@aamijar aamijar added non-breaking Introduces a non-breaking change feature request New feature or request labels Feb 24, 2026
@mfoerste4 mfoerste4 marked this pull request as ready for review March 2, 2026 23:35
@mfoerste4 mfoerste4 requested a review from a team as a code owner March 2, 2026 23:35
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

feature request New feature or request non-breaking Introduces a non-breaking change

Projects

Development

Successfully merging this pull request may close these issues.

2 participants