Using multi-cores on a GPU


(Tony O'Dowd) #1


I’m running OpenNMT on the AWS P2 machines. These have a single GPU (with four cores),

When training a model only one of the cores is being used, any one figure out how to use all four cores to speed the training up?


(Guillaume Klein) #2


Do you mean 4 CPU cores? If the GPU is fully used (i.e. close to 100% utilization in nvidia-smi), there will be very little gains in using additional CPU cores.

(Tony O'Dowd) #3

That explains it then. Thanks for the update.