carpentries-incubator / lesson-gpu-programming

GPU Programming with Python and CUDA.
https://carpentries-incubator.github.io/lesson-gpu-programming/
Other
19 stars 12 forks source link

Update setup.md #84

Closed suvayu closed 1 year ago

suvayu commented 1 year ago

Add instructions to setup a local environment using pip or conda

I chose this way of installation, because it takes care of installing cudatoolkit automatically. The downside is, this way nvcc is not available. For nvcc, you need to install cudatoolkit manually. Should I add this as a note in the end? Or would that be too much information?

isazi commented 1 year ago

We do not use nvcc, so it is not much of an issue, but I didn't know this was the case.

isazi commented 1 year ago

This is in the package description:

The CUDA Toolkit from NVIDIA provides everything you need to develop GPU-accelerated applications. This CUDA Toolkit includes GPU-accelerated libraries, and the CUDA runtime for the Conda ecosystem. For the full CUDA Toolkit with a compiler and development tools visit https://developer.nvidia.com/cuda-downloads

Something similar could work, saying that if you want all development tools you can download them yourself from https://developer.nvidia.com/cuda-downloads but that we do not need them for this lesson.

suvayu commented 1 year ago

I added a note. Hope that addresses the issue.