← Back to context

Comment by mft_

3 days ago

Apple’s article isn’t targeting people who bought multiple M3 Ultras before the RAMpocalypse - those folks can be safely assumed to already be on board with running large models locally.

I thought the article was about showing this specific model can be run locally, not convincing people about running models locally in the first place.

  • Indeed; my point is that people who spent $40k on an M3 Ultra cluster are mostly already doing this or similar, and will likely be ahead of Apple’s article.