Most people don't run their own LLMs, let alone install GPUs themselves.

I'm one of those Mac users, but I do have a linux tower for remote dev work and LLMs.