HN
Hacker News
Top
New
Best
Ask
Show
Jobs
Comment by whatever1 | Hacker News Reader
Full thread
whatever1
·
How do they parallelize the sequential actions ?
View on HN
erwincoumans
·
The primary use case of Newton Physics is reinforcement learning, with 1000s of similar environments. Even if each environment would have sequential actions, you run many envs in parallel.
Reply on news.ycombinator.com