OpenAI and NVIDIA have announced plans for a strategic partnership to build at least 10 gigawatts of AI data center capacity using NVIDIA systems. This infrastructure will include millions of GPUs to support the training and deployment of OpenAI’s future AI models. To enable this, NVIDIA intends to invest up to $100 billion in OpenAI, tied to the rollout of each gigawatt.
The first stage of deployment is expected in the second half of 2026 on the NVIDIA Vera Rubin platform. Both companies emphasized their long-standing collaboration, noting how past work laid the foundation for widely used AI systems today. OpenAI leaders highlighted the importance of compute infrastructure as the backbone of future economies, while NVIDIA sees this step as central to scaling AI capabilities.
As part of the agreement, OpenAI will treat NVIDIA as its preferred partner for compute and networking. The companies will align their hardware and software roadmaps to improve performance and efficiency. This initiative complements existing collaborations with Microsoft, Oracle, SoftBank, and others working to build advanced AI infrastructure. Details of the agreement are expected to be finalized soon.
Leave a comment