Google is taking direct aim at Nvidia's dominance in AI hardware, and according to a Wall Street Journal report it is borrowing the chip leader's own strategy to do it.
The Journal reports that Google is wielding its enormous cash reserves to win data-center customers for its in-house silicon, known as TPUs (Tensor Processing Units). As the livemint summary of the WSJ reporting puts it, the world's second-biggest company is "taking a page" from the world's biggest.
A centerpiece of that push: according to sources cited by the Wall Street Journal, Google is providing $3.2 billion to fund a New York data center that will rent its TPUs to Anthropic, the AI startup. Funding the infrastructure that runs your chips, then lining up a marquee customer to use it, echoes the kind of ecosystem-building that helped Nvidia lock in its lead.
Google is also looking beyond its own factories. The Motley Fool reported on June 8 that Intel shares surged on a reported Google AI chip foundry order — meaning Google may tap Intel to manufacture some of its chips. That would be a notable win for Intel, which has been trying to build up its contract-manufacturing business.
Taken together, the moves sketch a company attacking the AI chip market on multiple fronts at once: designing its own silicon, bankrolling the data centers that deploy it, securing high-profile customers like Anthropic, and spreading its manufacturing across foundry partners.
Why it matters: Nvidia's chips currently power most of the AI boom, and a credible challenge from a rival as deep-pocketed as Google could reshape who controls the hardware — and the economics — behind artificial intelligence.