58
Apple Develops Breakthrough Method for Running LLMs on iPhones
(www.macrumors.com)
This is a most excellent place for technology news and articles.
Whispercpp works off the ML cores on the m series chips. It's faster than my 1080ti that I have in a server doing the same things--by orders of magnitude. And it sips power.
Purpose built chips can be super powerful for their specific purposes.