Google’s ‘secret weapon’ in AI race against OpenAI is what has been giving nightmares to Nvidia – The Times of India

Nvidia recently got so spooked by a report on Google that the company made a public post on X, formerly Twitter, to defend itself. As for the news: A report in Information claimed that one of Nvidia’s biggest customers, Meta, is considering shifting part of its AI infrastructure to Google’s in-house chips, called TPUs or Tensor Processing Units. These TPUs are what analysts consider Google’s secret weapon in the artificial intelligence race against OpenAI. These are chips that have helped Google models leap ahead of OpenAI, prompting tech investors to reassess a new threat to OpenAI as well as Nvidia’s dominance. Google’s “tensor processing unit” has been central to the company’s efforts to boost the performance of its all-new Gemini 3 AI models, which are reported to have outperformed OpenAI’s GPT-5 in independent benchmarking tests and impressed both analysts and reviewers. These TPUs also power Google Gemini 3 AI models that reportedly made OpenAI CEO Sam Altman send Code Red memos to employees across the company.

How Google has become a ‘worry’ for Nvidia

For almost the past decade, Google’s TPUs were viewed by the industry as a potent but insular tool — highly efficient custom silicon designed specifically for Google’s internal ecosystem. Now, the tech giant appears ready to break that isolation, launching a strategic offensive to challenge Nvidia’s near-monopoly on the AI hardware…