OpenAI reportedly plans to buy Google’s AI chips to expand its computing capacity, deepening a business relationship between two rivals in AI products and search. The planned purchase of Tensor Processing Units, or TPUs, comes as OpenAI develops its own processor with Broadcom for deployment by the end of 2026.
Key takeaways
OpenAI’s reported TPU purchase plan follows earlier rentals of Google chips.
OpenAI targets the end of 2026 for its Broadcom-designed Jalapeno processor.
Oracle plans to offer Google TPUs through its cloud infrastructure.
Alphabet expects most revenue from TPU sales agreements in 2027.
OpenAI’s plans for Google’s AI chips
OpenAI plans to purchase Google TPUs , according to a person familiar with the matter. It has already rented the processors and remains one of the largest buyers of Nvidia graphics processing units , which it uses to train and run AI models.
Google began developing TPUs more than a decade ago, initially to make its web search engine faster and more efficient. The processors later proved useful for AI workloads and became an alternative to Nvidia’s GPUs. Google started selling TPUs directly to some customers in 2026.
OpenAI had begun renting Google’s AI chips by June 2025. Buying the processors would extend that relationship alongside the companies’ other business agreements.
The two companies compete in AI products and search, but have signed more deals in recent months. Last week, OpenAI said it would offer its frontier AI models through Google Cloud .
Jalapeno gives OpenAI a separate chip project
OpenAI is also developing Jalapeno , its first custom AI processor, with Broadcom. The company has unveiled the chip and said it plans to deploy it by the end of 2026 as part of a multi-generation chip program.
The effort has a direct connection to Google’s processor development. Richard Ho , who leads OpenAI’s internal chip team, previously worked on Google’s TPUs, and Google veterans helped create the in-house chips.
OpenAI’s planned purchase of Google AI chips sits alongside that internal development effort and its existing Nvidia purchases. OpenAI is among a growing group of companies building custom AI processors, many of which hope to challenge Nvidia’s lead.
Oracle expands TPU access as Alphabet awaits revenue
Oracle said on Tuesday that Google TPUs will become available through its cloud computing infrastructure. Customer agreements, including deals with Anthropic, have increased demand for Google’s processors.
“The next generation of AI will require a diverse range of compute,” Google executive Amin Vahdat said in a statement about the Oracle agreement.
Alphabet CFO Anat Ashkenazi has said the company expects to recognize a small portion of revenue from TPU sales agreements by the end of 2026. Most of that revenue is expected in 2027 .
The chip deals come amid plans by Amazon, Microsoft and Google parent Alphabet to spend as much as $600 billion combined on AI infrastructure in 2026. Investors are asking when those investments will produce meaningful returns.
Alphabet is scheduled to report quarterly results later on Wednesday, October 28, with investors looking for evidence that AI spending is driving growth in its cloud business.
Article produced with the assistance of artificial intelligence and reviewed by the editorial team.
