4 ms·
Those are all good points - I was focusing on my use case (i.e. small deep learning jobs), but definitely locking into pre-built configurations can be wasteful.
by singhrac 9y ago
Those are all good points - I was focusing on my use case (i.e. small deep learning jobs), but definitely locking into pre-built configurations can be wasteful. I'll likely use GC anyway, but appreciate the extra reasons given.
- boulos 9y agoEither way, I'd encourage trying out multi-GPU training. The latest NVLINK is pretty impressive, and reducing your iteration time for free (because linear pricing) is pure goodness.