NVIDIA / NVIDIA/cccl

[CUDAX] [CUCO] Use cuda::launch and thread hierarchies in cuco

Open
#9,656 1 comment 0 reactions 1 assignee Claimed by @davebayer View on GitHub
cuco
Dominant language
C++
Stars
2.5k
Forks
486
Avg merge
2d 6h
Merged PRs (30d)
295

Description

Depends on #9595

cuCollections data structures rely extensively on the Cooperative Groups APIs.
Although the interface is flexible, the underlying implementation is not always optimal. This can result in suboptimal code generation and increased register pressure.

As we migrate cuco into cudax, we have an excellent opportunity to evaluate the redesigned thread hierarchy APIs, which are expected to address many of the code generation issues found in the previous CG implementation.

Contributor guide

Open the contributing guide

Assessment

This issue has not been assessed yet.

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.