Google unveils chips for AI training and inference in latest shot at Nvidia

Google is packing ample amounts of static random access memory into a dedicated chip for running artificial intelligence models, following Nvidia’s plans.