YC should give an A100 to any machine learning team that gets in.
@johncoogan We definitely would have applied if that were the case.
@johncoogan To be fair, a lot of YC funding probably goes to A100s indirectly 😂
@johncoogan Agree! If possible, it would be good to give access to a cluster or Amazon/GCP credits! Training on multiple GPUs may reduce a ton the time for training models!
@johncoogan it’s kind of hard to build an a100 into an existing computer FWIW, you’d be much better off with two of the new generation A6000 ADAs for the same money
@johncoogan Pretty wasteful, only reason to overpay for a100 is for massive scaling on a GPU cluster, much better off with a few 4090s or a6000s
@johncoogan If I were @stephenbalaban that is exactly what I’d do.