Really telling they have to ask us for what cards we want as opposed to supporting all cards by default from day 1 like Nvidia.
All because they went with a boneheaded decision to require per-device code compilation (gfx1030, gfx1031...) instead of compiling to an intermediate representation like CUDA's PTX. Doubly boneheaded considering the graphics API they developed, Vulkan, literally does that via SPIR-V!
The author of the issue comments that they'll eventually support all cards. What he really is asking for, is what cards people want them to prioritize, not just support.
I read it fully. Whole point of my post is that, based on their track record so far plus the technical limitations, it is impossible for AMD to provide the same day 1 drop in compatibility that the CUDA ecosystem offers.
Edit:
> No guarantees of future support but we will try hard to add support.
yes. We are behind on software support for all consumer cards and would love to support all cards. But are looking for guidance / feedback so we can prioritize.
> No guarantees of future support but we will try hard to add support.
AMD reps told me exactly the same thing years ago about how they'd love to support all cards, when RDNA2 had just launched. Fast forward, only W6800 is properly supported from that gen. The last time I tried, it had tons of kernel bugs that caused hard freezes outside most basic cases.
You need to come out and say that you will support all cards, no ifs or buts, by a hard deadline.
All because they went with a boneheaded decision to require per-device code compilation (gfx1030, gfx1031...) instead of compiling to an intermediate representation like CUDA's PTX. Doubly boneheaded considering the graphics API they developed, Vulkan, literally does that via SPIR-V!