Domestic GPUs and AI accelerators, he added, are already being deployed at scale across training and inference workloads as well as edge computing …
http://dlvr.it/TQDbXR