3 ms·
For training, yes, but no indications on inference workloads. Apple has said they would use their own silicon for inference in the cloud.
by dlewis1788 2y ago
For training, yes, but no indications on inference workloads. Apple has said they would use their own silicon for inference in the cloud.
- walterbell 2y agoPlus the Apple "Neural Engine" which has shipped on millions of iPhones for local inference.