In a deal that underscores the immense cost and promise of artificial intelligence, Nvidia has agreed to invest up to $100 billion in OpenAI. The primary objective is to construct an unprecedented AI infrastructure powerful enough to train and operate models on the path to “super-intelligence.” This partnership deepens the ties between the premier chip designer and the leading AI lab, with Nvidia taking an equity stake in the venture.
The financial commitment is colossal and is designed to be unlocked by progress. The first installment of $10 billion will be released when the first gigawatt of computing power—a massive amount in its own right—is brought online. The ultimate goal is to build a system with a staggering 10-gigawatt capacity, providing OpenAI with a computational advantage unlike any other.
Nvidia’s chief, Jensen Huang, positioned the investment as the next frontier in a decade-long journey with OpenAI. “This investment and infrastructure partnership mark the next leap forward,” he said, highlighting the scale of the undertaking. For Nvidia, this ensures its hardware remains the critical engine for the most advanced AI projects in the world.
OpenAI’s chief, Sam Altman, has long argued that compute is the most precious resource in the modern world. “Everything starts with compute,” he stated, framing the new infrastructure as the foundation for the entire future economy. With this new firepower, OpenAI can better meet the demands of its massive user base and accelerate its research into more capable and beneficial AI systems.
The hardware for this ambitious project will be Nvidia’s state-of-the-art Vera Rubin platform, with the first data centers expected to be operational in late 2026. This partnership between OpenAI’s software and Nvidia’s hardware creates a formidable “AI factory,” complementing existing collaborations with Microsoft and Oracle to build a comprehensive ecosystem for AI development.
