AI-powered
podcast player
Listen to all your favourite podcasts with AI-powered features
Microsoft and Google Cloud Tensor Processing Units
Google has been doing their own chips for seven or eight years it is not new right and but they have what they call a TPU and they use it extensively for their own internal workloads absolutely Amazon has their own chips. Microsoft apparently is working on their own I mean my feeling is every hyperscaler is working ontheir own chip particularly for their owninternal workloads so that's an area when we talk about NVIDIA software mode like Google doesn't need CUDA anything however that is facing an end customer in the public cloud.