alt
Hacker News
throwdbaaway
•
yesterday at 9:38 PM
•
0 replies
•
view on HN
Exactly. Coding for inference is solved. CUDA is no longer a moat.