OpenAI reportedly plans to buy Google’s artificial intelligence chips, potentially deepening an unexpected partnership between two rivals competing for dominance in AI. The reported move could also strengthen Google’s challenge to Nvidia, whose processors power much of the world’s advanced AI infrastructure.
According to The Information, the proposed purchase involves Google’s Tensor Processing Units (TPUs). Financial terms, order size and delivery timelines have not been disclosed, and neither company confirmed the reported plan.
Why Does OpenAI Want Google’s AI Chips?
OpenAI relies heavily on Nvidia’s graphics processing units (GPUs) to train and operate AI models. However, running services such as ChatGPT requires enormous computing capacity, making chip availability, energy efficiency and operating costs increasingly important.
Google’s TPUs are designed specifically for AI workloads. They offer an alternative to Nvidia’s general-purpose AI computing platform and could help OpenAI diversify its hardware supply.
OpenAI had already begun renting Google’s AI chips in 2025. The reported purchase plan would represent a potentially deeper relationship.
Google TPUs vs Nvidia GPUs
Nvidia dominates the AI accelerator market, supported by its hardware ecosystem and widely used CUDA software platform. Google, meanwhile, designs its own TPUs and has expanded their availability to external customers.
The competition is no longer just about which chip is fastest. AI companies must balance performance, electricity consumption, infrastructure costs and the ability to secure enough processors to serve millions of users.
A major OpenAI order could validate Google’s strategy of selling its in-house AI hardware beyond its own data centres.
OpenAI Is Also Building Its Own AI Chips
OpenAI announced a custom AI processor developed with Broadcom in June 2026, with deployment planned by the end of the year. The initiative aims to improve the economics of running large language models, known as inference.
This suggests OpenAI is pursuing multiple hardware options rather than relying entirely on one supplier.
The broader stakes are enormous: Amazon, Microsoft and Alphabet are expected to spend a combined $600 billion on AI infrastructure in 2026, according to Reuters.
What Does This Mean for the AI Industry?
If the reported deal proceeds, Google could gain a major customer, OpenAI could gain more computing flexibility, and Nvidia could face greater competition.
The bigger story is the race to control AI computing—not simply the race to build smarter models. As demand grows, access to affordable chips, data centres and electricity may increasingly determine which companies can scale their AI services profitably.
