Elon Musk

@elonmusk ↗

Deleted No

Hibernated Yes

Last Checked April 2, 2021

Created

Tue Aug 25 01:55:43 +0000 2020

Likes

4,053

Retweets

197

Source

Twitter for iPhone

View Raw Data

JSON Data

View on Twitter

Likely Available

Elon Musk @elonmusk

@jwangARK @wintonARK Not exactly GPUs for NN. Lots of cores optimized for dot products with fast local memory & super high bandwidth between cores & efficient liquid cooling. Compiler needs to be optimized too. — PolitiTweet.org

Posted Aug. 25, 2020 Hibernated

Preceded By

Elon Musk @elonmusk

@tesla_addicted @Erdayastronaut No & not possible at this scale — PolitiTweet.org

Posted Aug. 24, 2020 Hibernated

Followed By

Elon Musk @elonmusk

@jwangARK @wintonARK Very different computers are needed for training vs inference. FSD is mostly Int8, but Dojo is de facto FP32 (feels like FP32, but actually FP16 with boring bits truncated) so we don’t need to rewrite all the training software. It should work almost immediately. — PolitiTweet.org

Posted Aug. 25, 2020 Hibernated