Woofun AI reports that Google is launching a competitive offensive against NVIDIA by expanding the commercial deployment of its Tensor Processing Units (TPUs). Previously restricted to internal data centers and Google Cloud rentals, TPUs are now being sold directly to emerging cloud service providers, known as Neoclouds. This strategic pivot marks a transition from specialized internal infrastructure to a broader commercial platform, targeting the same customer segment driving NVIDIA GPU demand. The move aims to diminish AI companies' dependence on the NVIDIA ecosystem while intensifying the shift toward a diversified chip landscape.