As the tech sector faces surging demand for computational power, OpenAI is reportedly exploring the purchase of Google’s artificial intelligence chips. Citing an individual with knowledge of the situation, The Information broke the news on Tuesday. This potential agreement would allow OpenAI to broaden its computing resources and decrease its reliance on NVIDIA.
Should the deal go through, it would forge closer commercial links between Google and OpenAI, even as the two compete in search and AI applications. Google builds Tensor Processing Units (TPUs), which are custom processors built specifically for training and executing AI models. At this time, neither organization has verified the proposed transaction, and the initial report omitted any specifics regarding pricing, quantities, or timelines.
Google TPUs Offer an Alternative to NVIDIA
At present, OpenAI depends heavily on NVIDIA graphics processing units (GPUs) to run and build its AI systems. Nevertheless, past reporting by Reuters revealed that OpenAI started renting Google’s AI chips in 2025. Buying them outright would strengthen this connection and secure an additional avenue for processing power.
For AI firms looking for alternatives to ubiquitous NVIDIA hardware, Google’s TPUs have become an appealing option. Amid soaring demand for AI infrastructure, Google has broadened the direct commercial sales of its processors. Sales to clients such as competing AI firm Anthropic have further boosted the market for Google’s hardware.
OpenAI Continues Building its Chip Strategy
This news surfaces as leading tech corporations pour immense capital into AI infrastructure. Alphabet (Google’s parent company), Amazon, and Microsoft together anticipate spending up to USD 600 billion this year alone. Meanwhile, investors are pushing for proof that these massive outlays will yield tangible financial returns.
Simultaneously, OpenAI is collaborating with Broadcom to create its own custom silicon. After unveiling its inaugural custom AI chip project in June, the company anticipates rolling out the processor before the year concludes. Utilizing multiple vendors and proprietary hardware should help OpenAI maintain adequate computing power to support its growing services.
Because the potential Google chip acquisition is still unconfirmed, details regarding financial figures and schedules are unavailable. Any official accord would represent another milestone in OpenAI’s strategy to diversify the hardware driving its AI architecture.
Also Read: Google’s Project Suncatcher Tests AI Chips in Space



