Hacker News Reader

Look into groq.com guys. some good models at similar speed to inception labs

It's faster inference because of the Hardware (LPUs), here the question is about architectures (AR or Diffusions)

I realize that, but it can be used now with many models in real-life situations. I just wanted to mention it if someone doesn't know it.

SRAM doesn't scale with advanced semiconductor node.

Groq is heading to a dead end.