Meta in Talks to Buy Google Chips, Challenging Nvidia’s AI Dominance

Meta Platforms is reportedly in discussions with Google to spend billions of dollars on the Alphabet-owned company’s custom chips for use in its data centers starting in 2027, according to The Information. The move, if finalized, could position Google as a serious competitor to semiconductor giant Nvidia in the booming market for AI-focused processors.

The talks reportedly include Meta renting Google Cloud chips as early as next year, part of a broader strategy by Google to encourage enterprise adoption of its tensor processing units (TPUs), which are designed specifically for AI workloads. Traditionally, Google has limited TPUs to its own data centers, but opening the technology to other companies could significantly expand its market share.

“Securing a partnership with Meta would mark a major step in Google’s effort to challenge Nvidia’s dominance in AI chips,” said an industry analyst familiar with the sector. “It’s a high-stakes move in a market expected to see hundreds of billions in data-center processor spending over the next several years.”

Market Reaction

Following the report, Alphabet’s shares rose more than 4% in premarket trading, putting the company on track to reach a historic $4 trillion valuation if the gains hold. Broadcom, which manufactures components for Google’s AI chips, also gained 2%, while Nvidia shares fell 3.2%.

Meta is one of Nvidia’s largest customers, with plans to spend up to $72 billion this year on AI infrastructure. A deal with Google could signal a shift in the competitive landscape, potentially redirecting a significant portion of AI chip revenue from Nvidia to its longtime cloud rival.

Why TPUs Matter

Custom AI chips like Google’s TPUs have seen surging demand as businesses seek alternatives to Nvidia’s graphics processors, which are both expensive and supply-constrained. In recent months, Anthropic expanded its own deal with Google to use up to one million TPUs, worth tens of billions of dollars.

Google has strengthened its cloud business through high-profile investments and partnerships, including backing from Warren Buffett’s Berkshire Hathaway. Its Gemini 3 AI model has earned positive early reviews, helping the company turn its cloud division into a major growth engine.

However, challenging Nvidia is no small task. The company’s CUDA software platform has been the foundation of AI and other applications for nearly two decades, relied on by more than 4 million developers worldwide. Transitioning to Google’s TPUs would require significant adjustments for companies heavily embedded in Nvidia’s ecosystem.

Looking Ahead

If the Meta-Google partnership proceeds, it could reshape the AI hardware landscape, giving Google a foothold in a market long dominated by Nvidia. The move may also encourage other enterprises to explore alternatives to Nvidia, potentially driving innovation and competition in the AI chip sector.

Neither Meta, Google, nor Nvidia have confirmed the discussions publicly, and the details of the agreement remain unverified. Analysts say the next 12–18 months could determine whether Google’s push to expand TPU adoption can successfully challenge Nvidia’s entrenched position.