NVIDIA and OpenAI Partner on Record-Breaking Multi-Gigawatt AI Infrastructure
NVIDIA and OpenAI announce a historic partnership to build million-GPU data centers capable of handling 10 gigawatts of power. NVIDIA plans to invest up to $100 billion in OpenAI as the companies construct the largest AI infrastructure project in history.
NVIDIA and OpenAI announce a historic partnership to build the largest AI infrastructure project in history. The agreement involves deploying multi-gigawatt data centers powered by millions of NVIDIA GPUs, specifically utilizing the upcoming NVIDIA Vera Rubin platform. NVIDIA founder and CEO Jensen Huang states this massive undertaking aims to move AI out of laboratories and into real-world applications across every industry.
To support this enormous scale, OpenAI plans to deploy at least 10 gigawatts of NVIDIA systems to handle the intense training and inference demands of its next-generation AI models. As part of the deal, NVIDIA intends to invest up to $100 billion in OpenAI progressively as each gigawatt of computing capacity comes online. OpenAI CEO Sam Altman emphasizes that no other company possesses the speed and scale required for such an immense build-out.
This massive expansion is critical for OpenAI to support its explosive growth, as the company currently serves over 700 million weekly active users. The new million-GPU AI factories provide the necessary fuel to power advanced capabilities like agentic AI, complex reasoning, and multimodal data processing. Ultimately, this infrastructure ensures OpenAI continues to drive down the cost per unit of intelligence while delivering more powerful models to users worldwide.