Re:Tesla Project Dojo Overview
by inciampati on Monday August 23, 2021 @05:49AM
(#28273850)
Actually, nope! SGD is embarrassingly parallel when it's HOGWILD! https://arxiv.org/abs/1106.5730. Formally, the hogwild and serial implementations are indistinguishable. This has a lot of citations but it's such an easy trick that no door path beating seems to have happened. The "trick" is just to make model updates atomic, and then each SGD step can happen in parallel without coordination or locks.
[ Parent
| on HN ]
Sorry, all Hogwild does is remove locks. It still needs constant global communication (global shared memory is assumed in the paper). That's not embarrassingly parallel.
Re:Tesla Project Dojo Overview
by kettleballroll on Monday August 23, 2021 @11:20AM
(#28277169)
As far as I know, hogwild doesn't really scale to very large systems and doesn't get used in practice, but maybe that's just because it's most useful on CPUs, which are inefficient for other reasons