Nvidia-Google AI chip war heats up as world’s most valuable company ‘delighted by Google’s success’ despite its own stock fall after report of Meta-Google chip deal

The battle for dominance in AI hardware is intensifying, and for once, industry titan Nvidia appears to be on the defensive. The $4 trillion chipmaker, long seen as the undisputed leader powering the AI boom, took the unusual step of publicly responding to a report suggesting that Meta may shift part of its future AI infrastructure to Google’s in-house chips.

On Tuesday, Nvidia posted a statement on X after shares dropped more than 2.5 per cent following a report from The Information. The report claimed Google has begun aggressively pitching its AI Tensor Processing Units (TPUs) to major companies, including Meta and several major financial institutions, as alternatives to Nvidia’s GPUs. While Google already offers TPUs through its cloud, expanding them into customers’ own data centers would significantly escalate the competition between the two tech giants.

Meanwhile, Alphabet stock climbed for a third consecutive day, driven by the strong reception of Google’s new Gemini 3 AI model, which earned praise from high-profile figures such as Salesforce CEO Marc Benioff.


In a defensive tone, Nvidia acknowledged the success of one of its biggest customers, writing, X: “We’re delighted by Google’s success—they’ve made great advances in AI, and we continue to supply to Google. Nvidia is a generation ahead of the industry—it’s the only platform that runs every AI model and does it everywhere computing is done.”


Google’s AI chips have long been seen as powerful but limited, highly efficient for Google’s own systems yet no real threat to Nvidia’s dominant GPUs, which power more than 90% of the global AI accelerator market.

That gap was largely due to design differences. In simple words Google’s TPUs are ASICs built for specific AI tasks, while Nvidia’s GPUs are versatile, general-purpose workhorses.

Nvidia underscored that distinction in its X post, stressing that its GPUs deliver “greater performance, versatility, and fungibility” than ASICs and highlighting its new Blackwell architecture as still a generation ahead.