OpenAI reportedly plans to buy Google Cloud TPUs

On October 27, 2026, The Information reported, citing people familiar with the matter, that OpenAI plans to buy Google TPU chips. The report pointed to a job listing showing OpenAI is seeking an engineer to lead optimization of its models on TPU, suggesting the initiative has moved into concrete execution. The move would be a breakthrough for Google's chip business: the TPU has long been viewed as a potential challenger to Nvidia GPUs in the AI compute market, while OpenAI is among Nvidia's most valuable customers. One report noted this would benefit a rising competitor to OpenAI backer Nvidia.

OpenAI's ties with Nvidia are deep. In August 2026, Nvidia agreed to provide more than $100 billion in credit support for a large data center project OpenAI is planning in Ohio, which would use Nvidia chip architecture exclusively. Nvidia CEO Jensen Huang wrote in a blog post at the time that OpenAI had committed to spending roughly $600 billion on Nvidia servers through 2030. That commitment has not stopped OpenAI from seeking supply diversification. According to a prior The Information report, OpenAI had already begun using Google TPU to reduce dependence on Nvidia chips. On October 28, Bloomberg, also citing a person familiar with the matter, reported that OpenAI intends to purchase AI computing chips from Google, a unit of Alphabet, as part of its broader AI infrastructure buildout; the report did not disclose purchase volume, chip model, or timing. It is worth noting that OpenAI said in July 2025 that it was early-testing Google TPU and had no plans for large-scale deployment at the time, five quarters ago.

OpenAI's chip path runs as follows: Google TPU played an important role in its early startup days; the generative AI company later shifted toward Nvidia compute, and subsequently purchased chips from AMD, Cerebras, and AWS, while developing its own Jalapeño. As of October 28, 2026, neither OpenAI nor Google had officially confirmed the procurement plan; all reporting rests on people familiar with the matter and the job listing.