A friend of mine is doing heavy CUDA work at the Helmholtz research center in Dresden, and at the same time he's desperately waiting for AMD Vega cards to be in stock somewhere so that he can build a desktop without shitty drivers.

Reference Vega cards are available in stock. Check mindfactory.de for example.

Yeah, but the reference coolers are apparently not very good. Yesterday evening, my friend was pondering if he could remove the reference cooler and install an aftermarket CPU cooler.

>I do CUDA work (OpenCL is close, but not nearly as powerful)

Do you mean cuda as a language v/s the latest opencl spec as a language, or implementations ? No one other than Intel is even implementing and updating OpenCL really, and nvidia will always push CUDA over opencl for lock ins. Which is kind of the same thing as every other argument that you can make for nvidia.

Depends on how you value Wayland support.

That said, if I were Nvidia, and the post was an example of "cooperating" and "working" with the Wayland developers, I don't think I'd go out of my way to cooperate.

You can swap "Nvidia" and "Wayland developers" and still end up with an equally valid argument.

How so? The Wayland developers are pushing for open, unified standards that work across all chipsets. Nvidia is pushing for their own special snowflake implementation/interface, that no one else would or could use.

You can certainly make the swap and the sentence still makes sense and is factually correct, but the context matters.

I think he means that

>That said, if I were (the Wayland developers), and (witnessed how Nvidia is) "cooperating" and "working" with (us), I don't think I'd go out of my way to cooperate.

Is also a valid point.

Gah, I read it as the opposite, even though that was what the grandparent had already stated.

How does using Vulkan for GPGPU today compare to using CUDA, I mean performance wise?