Google is adopting Nvidia's proven playbook to expand its AI chip business, according to sources cited by the Wall Street Journal. The world's second-biggest company is wielding its considerable war chest to win data-center customers for its custom silicon.
The strategy involves providing $3.2 billion to fund a New York data center that will rent Google's Tensor Processing Units (TPUs) to Anthropic, the AI safety startup. It mirrors how Nvidia has invested in early-stage AI companies to secure demand for its graphics processors.
By funding infrastructure that locks clients into its TPUs, Google aims to challenge Nvidia's dominance in the AI chip market. The move signals an intensifying battle between cloud giants and chipmakers for control over the hardware powering artificial intelligence.
The arrangement could constrain Anthropic's flexibility if it wants to switch to other chips or cloud providers. Critics argue such deals risk creating vendor lock-in, potentially stifling competition and innovation in the rapidly evolving AI sector.
Neither Google nor Anthropic have publicly commented on the reported financial terms. The data center's construction timeline and operational status remain unclear.