NVIDIA / NVIDIA/cccl

[FEA]: Design generic segmented sort tuning API

Open
#10,779 4 comments 0 reactions 1 assignee Claimed by @anikaj-eng View on GitHub
Dominant language
C++
Stars
2.5k
Forks
486
Avg merge
2d 6h
Merged PRs (30d)
295

Description

### Is this a duplicate?

- [x] I confirmed there appear to be no [duplicate issues](https://github.com/NVIDIA/cccl/issues) for this request and that I agree to the [Code of Conduct](CODE_OF_CONDUCT.md)

### Area

CUB

### Is your feature request related to a problem? Please describe.

We don't know how to express a tuning for segmented sort that'd allow arbitrary number of buckets etc.

### Describe the solution you'd like

This issue can be closed with a comment illustrating a design of a generic segmented sort tuning API that:
- allows specifying number of buckets
- algorithm and tuning parameters for a given bucket
- partitioning algorithm
- in-place sorting of segments in partitioning algorithm and threshold for it
- doesn't break code that passes existing tuning policy into segmented sort

Add a branch that shows backend changes needed to support this interface.
### Describe alternatives you've considered

_No response_

### Additional context

_No response_

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.