3 ms·
Dedicated neural net silicon is faster than GPU, at the cost of being less general purpose. NVIDIA GPUs for example dedicate a modest fraction of their die to
by dannyw 5y ago
Dedicated neural net silicon is faster than GPU, at the cost of being less general purpose.
NVIDIA GPUs for example dedicate a modest fraction of their die to Tensor cores, at the cost of less general-purpose CUDA cores.
- martinald 5y agoThanks. I didn't realise how much faster. Probably you can do 10x the FLOPs in similar die space on dedicated silicion vs GPU. I thought it would be faster, but not an order of magnitude more.