Nvidia and OpenAI have officially embarked on a quest to build the world’s most powerful and advanced AI, backed by a formidable $100 billion war chest from the chipmaker. The partnership’s stated goal is to create the infrastructure necessary to train and run models on the path to “super-intelligence,” a feat that requires an unprecedented concentration of computing power.
The physical manifestation of this quest will be a new network of data centers deploying a total of 10 gigawatts of Nvidia’s cutting-edge systems. This colossal infrastructure will be OpenAI’s exclusive proving ground for its most ambitious research projects, moving beyond the capabilities of existing systems and tackling entirely new classes of problems.
This ambitious endeavor is financed through a staged investment from Nvidia, which will also become an equity holder in OpenAI. The first $10 billion of capital is tied to the first gigawatt of the infrastructure becoming operational, a milestone targeted for the latter half of 2026. This links the grand vision to a concrete and accountable execution plan.
The leaders of both companies, Jensen Huang and Sam Altman, have framed this as the next chapter in their decade-long collaboration. Huang’s reference to a “new era of intelligence” and Altman’s focus on creating “new AI breakthroughs” underscores a shared commitment to pushing the absolute limits of what is technologically possible.
For a world that has already been transformed by OpenAI’s ChatGPT, this deal promises a future where AI’s capabilities are orders of magnitude greater. It is a focused, heavily funded, and unapologetically ambitious effort to solve intelligence, powered by the combined might of the world’s leading AI software and hardware companies.