“The No. 1 thing we’re hearing from people right now is that on LLMs, latency is the problem,” Ross said. Read more in this week's article in the EE Times by Sally Ward Foxton.

→ View original post on X — @groqinc