What serious Linux users buy GPUs based on raw gaming performance on release week?

I personally buy based on open-source driver support. And this includes long-term active support, AND developer approachability.

My current GPU is an AMD/Radeon one because of that. But I'm reconsidering my position when my next hardware upgrade comes.

I reported an AMD GPU driver issue to mesa once. It was tested, confirmed, and patched by a competent AMD developer within a few days. Now you have easily reproducible issues like this not even going past the testing phase after many months. And there are similar issues across all model generations.

If I were to upgrade my workstation next year, I would probably go with an AMD CPU and an Intel GPU, which is the exact opposite of my current setup 🙃. One should never rely on outdated perceptions.

Not for nothing, but:

...but support for Linux can still be fairly patchy.

...

And if you’re thinking about Nvidia, their closed-source drivers have historically been problematic for gaming on Linux.

These aren't technically false statements, but they aren't exactly indicative of the current state of Linux gaming. The Nvidia one, especially, is kind of overstating how problematic Nvidia is; I don't think the author plays games on Linux much or talks to other people who play games on non-AMD systems.

Also, five games isn't much. It hit ≈60fps for two of the titles (what you'd want in a budget GPU), and Cyberpunk 2077 isn't an especially great game to use as a metric baseline, given its history of bugs.

They could be entirely right, but I would want to see more data on this, as well as better and real world tests before sounding such a note of finality.

Likely driver issues, hopefully they can get it fixed!

Intel did a great job with the drivers in windows last time, time will tell if they fix up the Linux ones

This was what I was waiting for. Most reviews were using arrow lake cpus and on windows. I was waiting for amd CPU on Linux to see if it was going to work. Also, if anyone sees any reviews for how it would do in a server for machine learning specifically in TrueNAS Scale, that'd be great.