The methods for splitting weights across multiple chips are well established. Gr... | Hacker News

Hacker Newsnew | past | comments | ask | show | jobs | submit

wmf 15 days ago | parent | context | favorite | on: AMD acquires Taalas to boost inference performance...

The methods for splitting weights across multiple chips are well established. Groq/Cerebras can't hold a model on one chip either.

pyrolistical 15 days ago [–]

Umm I have an extra 35, do you have layer 6?

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact