alt Hacker NewsI thought these TPUs were primarily used for inference?
Replies
knowaveragejoe • yesterday at 2:11 PM
As the article states, there's both training and inference dedicated chips.
I thought these TPUs were primarily used for inference?
As the article states, there's both training and inference dedicated chips.
TPU8t is for training. But even still, once you’ve trained, you need to run the model too. And these kinds of models already have a huge latency hit so there’s not much hurting running it away from the trading switches.