Linux

Get something new enough and continue getting something new enough when AMD pushes them out. The drivers suck for anything older than an RX580, and things like Blender require even newer GPUs despite the hardware being more than capable.

Run Arch and use the ROCm'd PyTorch from the repos. Those packagers know what they're doing.

Other than that, expect everything premade to be made for CUDA (and therefore unusable). There are some tools like https://github.com/ROCm-Developer-Tools/HIPIFY but they aren't "there".

Source: Been running Stable Diffusion on an RX580.

Checkout geohot's latest video about AMD GPUs... Not very favorable

the unfortunate truth is that NVidia has a complete stranglehold on the compute market. They recognized the capabilities of massively parallel compute early on and pushed CUDA super super hard to any organization doing compute. And it worked- CUDA is much easier to implement than openCL, and was released two years earlier too, so everyone ended up standardizing on it. They are currently reaping the benefits of that monopolization through their now huge enterprise GPGPU market and can basically piss down the backs of consumers and competitors alike without repercussions. OpenCL and AMD's implementation was a day late and harder to implement...

Do not buy AMD if you need to do any kind of compute- whether it be rendering ala Blender/AE, accelerated engineering CAD workflows, or big data handling. No tools are designed around anything but CUDA, and it sucks because Jensen is a greedy asshole, but you gotta pay your dues.

Hi, I do not know much about GPUs and ML. My apologies for not being able to answer your question, but I'd like to know what you're trying to achieve running said models. Is ML a hobby of yours?

stable diffusion runs great on a 7900xtx via pytorch and rocm5.5, but you may have to compile pytorch 2.0.1. manually. but with pytorch/rocm:latest docker this is fairly easy, look for instructions to install automatic1111, they can be generally applied to other stacks

I was in the same boat as you, i.e. using the GPU during my studies. My premise is to optimise the most frequent use case, i.e., deep learning.

IMO going with NVIDIA will save you so much worries and frustration that it clearly outweighs the downsides of worse Wayland support compared with AMD.

When you have tough university assignments/projects, you want to focus on the actual problem instead of debugging/compiling libraries for use with AMD. I am sure that with a bit of work many libraries can be made to work with AMD, but apparently it is still a pain oftentimes.

So I strongly suggest choosing NVIDIA. Disclaimer: have not used AMD for deep learning yet, but have monitored the development of AMD support, because I would like to switch to AMD.

Btw. I found Pop!OS to be very nice for both "regular" university work and all computer science tasks.

A note about this, it looks like Hugging Face is partnering with AMD to improve this situation but it looks like they're starting with enterprise first: https://huggingface.co/blog/huggingface-and-amd . No specific platforms listed on that page but https://huggingface.co/amd mentions Linux. But yeah, looks like the situation as of right now is mostly about NVIDIA and I've heard about CUDA going through classes at school.

https://are-we-gfx1100-yet.github.io