Comment by seanmcdirmid

Comment by seanmcdirmid 3 days ago

1 reply

64GB is what you need to run some decent quantized mid-sized LLMs locally…with unified memory on Apple silicon. Should be standard, that would open up a lot if new applications. Incidentally, even high DPI monitors aren’t standard yet for non-mobile devices. Sad how slowly things move.