<ul data-eligibleForWebStory="true"><li>OpenAI is testing Google's Tensor Processing Unit AI chips but has no active plans to deploy them at scale.</li><li>Nvidia AI accelerators are primarily used by OpenAI, creating a barrier to adapting workloads to run on other GPUs like AMD's.</li><li>OpenAI is utilizing Google Cloud data center infrastructure to meet increased demand, while also developing its own in-house chip set to be finalized for manufacturing this year.</li><li>Amazon, a major player in the cloud services sector, is notably absent from the cloud providers being used by OpenAI, which spans across Google, Microsoft Azure, Oracle, and CoreWeave.</li></ul>

OpenAI Says No Plans To Deploy Google AI Chips

Discover more