Skip to content

Add WarpBitonicTopK#9281

Draft
yong-wang wants to merge 5 commits into
NVIDIA:mainfrom
yong-wang:bitonic-topk
Draft

Add WarpBitonicTopK#9281
yong-wang wants to merge 5 commits into
NVIDIA:mainfrom
yong-wang:bitonic-topk

Conversation

@yong-wang
Copy link
Copy Markdown

Description

Add WarpBitonicTopK which is a warp-wide top-k (WarpTopK), a feature listed in in #5673

It is based on WarpBitonicSort from #8391.

The output guarantees are (1) deterministic (2) unstable sorted.

Tests and benchmarks are also added.

yong-wang added 5 commits June 5, 2026 15:38
- add a new warp-wide sorting primitive that implements a bitonic sorting network
- add tests and benchmarks for it
- add device-side benchmark utilities
- remove anonymous namespace
- implement has_native_shfl_v as variable templates
- replace #pragma unroll with _CCCL_PRAGMA_UNROLL_FULL()
@yong-wang yong-wang self-assigned this Jun 5, 2026
@github-project-automation github-project-automation Bot moved this to Todo in CCCL Jun 5, 2026
@copy-pr-bot
Copy link
Copy Markdown
Contributor

copy-pr-bot Bot commented Jun 5, 2026

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.

@cccl-authenticator-app cccl-authenticator-app Bot moved this from Todo to In Progress in CCCL Jun 5, 2026
@yong-wang yong-wang requested a review from elstehle June 5, 2026 09:26
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

Status: In Progress

Development

Successfully merging this pull request may close these issues.

1 participant