Re:New chips for machine intelligence
by alexhutcheson on Monday October 07, 2019 @01:01PM
(#21183175)
I've never heard anyone suggest that WebGPU would be appropriate for ML training workloads. Maybe inference, but not training.
[ Parent
| on HN ]
Re:New chips for machine intelligence
by trust07007707 on Monday October 07, 2019 @04:00PM
(#21185315)
Well, the memory limit of a WebGPU process would be the limiting factor for training. In addition, the bandwidth between the nodes and the parameter server, if doing training in data-parallel fashion, is another limiting factor.