JuliaGPU / JuliaGPU/AcceleratedKernels.jl

Use heuristic to determine block size instead of fixed value

Open
#60 0 comments 0 reactions 0 assignees View on GitHub
Dominant language
Julia
Stars
204
Forks
15
Avg merge
10d 5h
Merged PRs (30d)
6

Description

See https://github.com/JuliaGPU/CUDA.jl/pull/2813#issuecomment-3148696577 for more details. Some algorithms/array shapes are severely hindered by the constant block size of 256.

Maybe this would be better implemented in GPUArrays. Open to discussion.

Contributor guide

No contributing guide indexed for this repository

Research direction

Start with the linked CUDA.jl pull-request comment and inspect how the fixed block size of 256 is used. Determine whether the heuristic belongs in AcceleratedKernels.jl or GPUArrays; done means an agreed approach addresses the affected algorithms and array shapes.

Written by the indexing model from the issue text.

Assessment

Tech stack
julia
Domain
performance
Issue type
Feature
Difficulty
5/5
Estimated time
Over a week
Activity status
Stale
Clarity
Needs clarification
Newbie friendliness
30/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.