Yeah I defend stuff that works :p AMD is what you get without telemetry; crap API support on Windows (they took their already mangled stack and Jay Wilson'd it

Yeah I defend stuff that works :p AMD is what you get without telemetry; crap API support on Windows (they took their already mangled stack and Jay Wilson'd it May 2022), next to no innovation with AI, flopped OpenCL, and have to concede with HIP since CUDA is too good. They can't even fix their botched TPM support that's so bad it had to be disabled globally on Linux.