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

>You can have all the talent in the world but if you can't get 100k H100s and a dedicated power plant, you're out.

I really have to wonder, how long will it be before the competition moves into who has the most wafer-scale engines. I mean, surely the GPU is a more inefficient packaging form factor than large dies with on-board HBM, with a massive single block cooler?



Sentiment I have heard is manufactories do not want to increase die size because defects per die increases at the same time.


Meanwhile at Cerebras...heh

But I do believe that their cost per compute is still far more than disparate chips.


This is why chiplets are used.




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

Search: