OpenAI has no active plans to use Google's TPUs (Tensor Processing Units) for its AI products, despite early testing with some of Google's TPUs.
OpenAI is currently one of NVIDIA's largest GPU customers for training AI models, which has led NVIDIA to top the US stock markets due to the surge in demand for its GPUs.
Google uses its in-house hardware systems, TPUs, to train and deploy AI models. Amazon Web Services (AWS) is working on Project Rainer, which includes half a million Trainium2 chips rather than NVIDIA chips for cost efficiency.
OpenAI is also developing its own in-house chips with assistance from Broadcom and Taiwan Semiconductor Manufacturing Company, expected to launch in the fourth quarter of this year.