Skip to content

Comment on Neural Boidsparent

Comments

based on the article the noida approach has better performance and is able to run on a GPU while the Sterling implementation must run on a cpu.

Potentially though there's nothing stopping someone from writing the original three rules using cuda

AboutSource Built by g1lg1l

Hackerly is an independent reader for Hacker News, built on the public HN API. Not affiliated with Y Combinator.