NVIDIA / NVIDIA/cuvs

Replace Vamana build scatter kernel with RAFT primitives

Open
#2,493 0 comments 0 reactions 0 assignees View on GitHub

Nobody has claimed this yet.

Dominant language
Cuda
Stars
854
Forks
236
Avg merge
3d 3h
Merged PRs (30d)
62

Description

Currently the GPU Vamana build uses a custom kernel here.
This needs to be replaced with a RAFT primitive to avoid excessive kernel definitions and simplify the codebase. However, it's not a standard access pattern so needs to be done with a strided view or something similar.

Contributor guide

Open the contributing guide

First steps

  1. Read the whole issue, then the project's contributing guide.
  2. Comment on the issue to say you are picking it up — it saves two people doing the same work.
  3. Fork the repository and make your change on a branch.
  4. Open a pull request that references the issue number.

Research direction

The linked custom GPU Vamana scatter kernel is the entry point; inspect its access pattern and the surrounding build path first. Replace that kernel with an equivalent RAFT primitive using a strided view where needed, then verify the Vamana build behavior and relevant existing checks.

Written by the indexing model from the issue text.

Assessment

Domain
machine-learning
Issue type
Refactor
Difficulty
4/5
Estimated time
3-5 days
Activity status
Quiet
Clarity
Mostly clear
Newbie friendliness
48/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.