Meta Platforms is in talks with Google to spend billions of dollars on the Alphabet-owned company’s chips for use in its data centres starting from 2027, The Information reported, a move that would cast Google as a serious rival to semiconductor giant Nvidia.
The talks also involve Meta renting chips from Google Cloud as early as next year and are part of Google’s broader push to get customers to adopt its tensor processing units (TPUs) – used for AI workloads – in their own data centres, the report said, citing people involved in the talks.
The move would mark a departure from Google’s current strategy of using TPUs only in its own data centres and could sharply expand the market for its chips, putting the company in direct competition for the hundreds of billions being spent on data centre processors to power AI services.
Some Google Cloud executives have suggested the strategy could help it capture as much as 10 percent of Nvidia’s annual revenue, a slice worth billions of dollars, according to the report.
Clinching a chip deal with Meta, one of the biggest Nvidia customers with up to US$72 billion ($111.3 million) planned in spending this year, would mark a major coup for Google — already one of the biggest winners of the generative AI boom thanks to a surge in demand for its cloud services from businesses adopting the technology.
Alphabet, Meta and Nvidia did not immediately respond to requests for comment. Reuters could not verify the report.
Demand has surged for custom chips such as TPUs in recent years as businesses look for alternatives to Nvidia’s pricey and supply-constrained graphics processors.
Anthropic said last month it was expanding its Google deal to use up to one million of the tech giant’s AI chips, worth tens of billions of dollars.
Google has built momentum in recent months by drawing Warren Buffett’s Berkshire Hathaway as an investor, turning its once-marginal cloud unit into a growth engine and earning strong early reviews for its latest Gemini 3 model.
Renting Nvidia chips to customers is a big revenue source for its cloud unit.
Taking on Nvidia’s dominance would require Google to overcome nearly two decades of proprietary Nvidia code that has made the company’s ecosystem hard to dislodge.
More than 4 million developers worldwide rely on Nvidia’s CUDA software platform to build AI and other applications.