Skip to content

Fix #423: Julia backend no longer installs all of CUDA toolkit - #424

Merged
fjwillemsen merged 5 commits into
masterfrom
julia
Nov 13, 2026
Merged

fjwillemsen merged 5 commits into
masterfrom
julia

Conversation

@fjwillemsen

Copy link
Copy Markdown
Member

Fixes the issue described in #423 by having the Julia CUDA backend install and loading CUDACore and NVML instead of CUDA by default. If CUDA.jl is installed, it reuses that. When the backend has to install CUDACore and finds a CUDA toolkit already on the system that works with the driver, it tells CUDA.jl to use that toolkit instead of downloading one. "Works" here means CUDA 12 or newer, a major version no higher than the driver supports, and the parts CUDA.jl needs all present. It looks in the same places CUDA.jl does (CUDA_HOME/CUDA_PATH/CUDA_ROOT, ptxas on PATH, /usr/local/cuda*, /opt/cuda*) and prints a warning saying which toolkit it chose. The setting is written into the project before installation, because CUDA.jl decides what to download at install time.
If so desired, a new environment variable can be used, KERNEL_TUNER_JULIA_LOCAL_CUDA_TOOLKIT, to control this: auto (default), yes or no.
This PR has been tested on CUDA & Metal with both nox and the tests of KernelTuner.jl as well as the upcoming SC notebook.

@sonarqubecloud

Copy link
Copy Markdown

@fjwillemsen
fjwillemsen merged commit 65e8dd6 into master Nov 13, 2026
4 checks passed
@fjwillemsen
fjwillemsen deployed to dev_environment November 13, 2026 19:28 — with GitHub Actions Active

This branch was successfully deployed

2 active deployments
dev_environment — 65e8dd69 Deployed Nov 13, 2026 by fjwillemsen via build #27
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants