OpenAI plans to buy Google AI chips, according to a report published on Tuesday, in a move that would turn a rented computing relationship into owned hardware and deepen commercial ties between two companies that compete in artificial intelligence and search.
The Information reported the plan, citing a person familiar with the matter. The report did not disclose financial terms or other details. Google did not immediately respond to a Reuters request for comment outside regular business hours, while OpenAI declined to comment.
From rented computing to owned hardware
OpenAI already uses Google's processors. Reuters reported in June 2025 that the ChatGPT maker had begun renting Tensor Processing Units through Google Cloud, and last week OpenAI said it would make its frontier models available through Google Cloud. Buying chips outright would change that arrangement, shifting OpenAI from paying for access to owning part of the hardware its models run on.
OpenAI has also advertised engineering roles focused on optimising its models for Google's processors, according to job listings cited in reports of the plan. Earlier this year, the company unveiled its first custom AI chip, designed with Broadcom, and said it planned to deploy the processor by the end of this year as part of a multi-generation chip programme.
Tensor Processing Units are Google's in-house processors for machine learning, built to train large models and serve answers to users. Unlike general-purpose graphics processors, they are tuned for the calculations behind neural networks, which is why several AI labs now test them alongside Nvidia hardware.
Why the chip supply matters
The shift reflects a wider change in the chip market. For years, Nvidia's graphics processors were the default choice for training large AI models, and its hardware remains central to most AI data centres. Google's decision to sell its processors directly, rather than offering them only through its own cloud, has opened another supply route for labs that want more bargaining power over price and availability.
OpenAI's hardware plans already span rented capacity, a custom design built with Broadcom and purchases from Nvidia. Adding Google's processors would give the company a further option when it negotiates supply, and it would put a rival's chips inside the data centres that run ChatGPT.
OpenAI is among the largest buyers of Nvidia's graphics processing units, the chips it uses to train and run its models. Google's Tensor Processing Units have gained traction as an alternative as companies look for cheaper options and processors suited to running models. Google began selling the processors directly to some customers this year.
Deals with customers including rival Anthropic have fuelled demand for Google's processors. Alphabet, Google's parent company, expects to begin recognising a small percentage of revenue from processor sales agreements by the end of this year, with most of it expected the following year, chief financial officer Anat Ashkenazi previously said. Google and Oracle also announced a partnership on Tuesday to deploy the processors on Oracle Cloud Infrastructure starting in 2027, with Oracle providing data centre power, cooling and network infrastructure.
Demand for AI computing has surged as more people and businesses use chatbots and AI tools each day. Amazon, Microsoft and Alphabet plan to spend as much as six hundred billion dollars combined this year on AI infrastructure, according to Reuters, even as investors increasingly question when such spending will turn into meaningful returns. Alphabet reports quarterly results on Wednesday, and investors will look for signs that heavy AI investment keeps driving growth in its cloud business.
What remains unknown
No price, quantity or timetable has been disclosed, and neither company has confirmed the plan publicly. The report cites a person familiar with the matter, so the shape of any deal could still change. Alphabet's results on Wednesday may offer the next public signal on cloud growth and chip sales.
Further detail may also come from Reuters reporting on the plan and from additional coverage of the reported purchase, which notes the engineering roles and the Oracle partnership. Related coverage on genznewz includes Oracle AI chip shakeup widens beyond Nvidia and Nuix Neo AI agent brings audit trail to investigations.
Comments 0
No comments yet. Be the first to share your thoughts!
Leave a comment
Share your thoughts. Your email will not be published.