OpenAI has announced that it will not use Google’s AI chips. The company confirmed that it has conducted some tests on TPU chips developed by Google, but there are no plans to include these hardware in its products.

OpenAI does not use Google’s chips

The statement came after some claims in some news reports that have been reflected in the public in recent days. These reports suggested that OpenAI has started using Google Cloud infrastructure and will also benefit from TPU chips in this context.

The company’s spokesperson stated that Google TPUs are being evaluated in the early testing phase, but there is no concrete step for scaled use. Google has not made any statement on this issue. This silence increases the uncertainty about clarifying the boundaries of the relationship between the two companies, and continues to be closely monitored in the industry.

OpenAI is currently working with chips manufactured by Nvidia and AMD. In addition, the company has been developing its own AI chip for a long time. The special chip that OpenAI is working on is expected to reach the “tape-out” stage by the end of the year. This stage refers to the process in which the chip design is completed and sent to production. The company continues its efforts to develop its own hardware without slowing down.

It is true that OpenAI continues to use Google Cloud services. However, this collaboration only covers public cloud infrastructure. OpenAI and Google are in direct competition in the field of artificial intelligence. Despite this, OpenAI’s use of Google Cloud resources has created a remarkable situation in the technology world.

Google, on the other hand, has long used TPU chips only in its own internal projects. However, it has recently opened these chips to external customers. Currently, companies such as Apple, Anthropic and Safe Superintelligence use Google’s chips. It has been confirmed that OpenAI is not among these companies.