The completely different software stack is a killer. It’s not that you can’t find versions of a model to run, but almost everything that hits the GPU for compute is going to be targeting CUDA, not RocM. From a compatibility standpoint alone this killed AMD for me. I just do not want to spend my time fighting the stack to get these models running.
It’s utterly disturbing how easy it is to get into so many networks and devices just by looking up their default password or plugging in the default password on the side of the box. Absolutely the first thing you need to change on a new device if it comes with a preset PW.
Somebody’s gonna make a federated TikTok, aren’t they? We’re gonna have TikTokers flooding the Fediverse. We’re so fucked.
Yeah, it’s not that I like this state of affairs, but right now the vendor lock-in is so one-sided that it’s hard to say there’s a viable alternative to CUDA. I hope that changes one day.