Full Breakdown
OpenAI Plans to Purchase Google’s Tensor Processing Units
By Drooid · · How we work
In this breakdown
01
OpenAI Moves to Purchase Google’s AI Chips
OpenAI is preparing to buy Google’s Tensor Processing Units (TPUs) for use with its frontier models, according to a person familiar with the private discussions. The deal’s financial terms have not been disclosed, and both companies declined to comment. The move follows OpenAI’s broader effort to diversify its compute supply beyond Nvidia graphics processing units (GPUs) and its own in-house “Jalapeno” chips, which are slated for deployment later in 2026.
02
Background: Chip Landscape and OpenAI’s Prior Strategies
Google’s TPUs, originally built to accelerate the company’s search engine, have become a prominent alternative to Nvidia GPUs for AI workloads. Alphabet began selling TPUs directly to customers this year, and Oracle announced on October 27 that TPUs would be available through its cloud-computing infrastructure. OpenAI previously announced plans to start using its custom chips, developed with Broadcom and led by Richard Ho, a former Google TPU engineer. In June, OpenAI began renting Google’s AI chips, marking an earlier step toward deeper integration.
03
Data & Statistics on Partnerships and Spending
- OpenAI is one of Nvidia’s largest GPU customers and has committed to spending roughly $600 billion on Nvidia servers through 2030, according to Nvidia CEO Jensen Huang.
- Alphabet expects to recognize a small portion of TPU-related revenue by the end of this year, with the majority to be recorded in 2027, as noted by CFO Anat Ashkenazi.
- In August, Nvidia agreed to provide more than $100 billion in credit support for an OpenAI data-center project in Ohio that will use Nvidia’s chip architecture exclusively.
04
Official Statements & Responses
Google executive Amin Vahdat said the next generation of AI will require a “diverse range of compute,” underscoring the company’s openness to supplying TPUs to competitors. Jensen Huang reiterated OpenAI’s ongoing commitment to Nvidia hardware while acknowledging the company’s pursuit of supply diversification.
05
Implications for AI Hardware Competition
The prospective OpenAI-Google chip deal signals a shift toward broader hardware sourcing in the AI sector, potentially reducing OpenAI’s reliance on Nvidia GPUs and strengthening Google’s TPU market presence. Analysts note that such diversification may intensify competition among chip makers as AI model training demands continue to surge, while investors will watch Alphabet’s upcoming quarterly results for signs of revenue growth from TPU sales.