OpenAI is reportedly planning to buy AI chips from Google, according to The Information. The chips, known as Tensor Processing Units (TPUs), would help the ChatGPT maker expand its computing capacity.
Reuters and Bloomberg reported the development on Tuesday, citing a person familiar with the matter. Neither OpenAI nor Google has confirmed the plan. The report did not disclose the financial terms, the number of chips, or the delivery timeline.
The move would mark a significant change for OpenAI, which has long relied on NVIDIA’s GPUs to train and run its AI models. NVIDIA dominates the market for high-end AI processors, but its chips remain costly and supply-constrained.
TPUs offer OpenAI another route. Google designs these chips specifically for AI workloads, including model training and inference. By adding TPUs to its infrastructure, OpenAI could spread its computing needs across multiple suppliers.
OpenAI already rents Google’s AI chips through a cloud arrangement. A direct purchase would deepen that relationship and give the company greater control over its hardware plans.
The reported deal highlights how competition in artificial intelligence is forcing unusual alliances. OpenAI and Google compete directly in AI assistants, search, and enterprise software. Yet Google could become a key hardware supplier to its rival.
For Google, selling TPUs to OpenAI would expand the reach of its custom silicon business. It would also strengthen Google’s challenge to NVIDIA, which has captured much of the value in the AI computing boom.
For OpenAI, the purchase supports a broader strategy to secure more computing power. The company faces rising demand for ChatGPT and its developer tools, while it also works on its own chips with Broadcom.
Also Read: OpenAI Launches GPT‑6.1 Terra to Put AI to Work Across Apps