alt Hacker News> Frontier AI companies are selling at a loss.
There are huge economies to be had by batching requests and using lots of RAM for MoE (sparse models). You can't achieve that efficiency at batch size 1 on a single node.
Replies
eikenberry • yesterday at 7:58 PM
Inference isn't the problem.
Exactly, they put a lot of money into engineering and it does give results