3 ms·
Yes exactly. The instances with 8 or 16 GPUs. Does the training time reduce linearly, is the GPU utilisation 100%, is it plug and play with TF
by dharma1 9y ago
Yes exactly. The instances with 8 or 16 GPUs. Does the training time reduce linearly, is the GPU utilisation 100%, is it plug and play with TF