alt
Hacker News
olalonde
•
today at 5:45 PM
•
0 replies
•
view on HN
That's incorrect. TPUs can support many ML workloads, they're not exclusive to LLMs.
alt Hacker News