NVIDIA and OpenAI Forge Unprecedented AI Infrastructure Deal, CEO Jensen Huang Hails 'Largest Computing Project Ever'

• NVIDIA commits up to $100 billion in a phased investment to build at least 10 gigawatts of AI datacenters for OpenAI.
• The partnership, described by CEO Jensen Huang as the start of an 'AI industrial revolution,' will begin deployment in 2026.
• The scale of the project, unprecedented in the history of computing, is a direct response to soaring demand driven by ChatGPT and next-generation AI models.

NVIDIA Corp. and OpenAI have finalized a strategic partnership to construct what NVIDIA Chief Executive Jensen Huang termed "the largest computing project in history," a multi-year effort to build a massive AI infrastructure network. The deal involves NVIDIA committing up to $100 billion of phased investment to support OpenAI’s next-generation AI compute needs, with the first phase of deployment targeting 2026.

The collaboration aims to build at least 10 gigawatts of dedicated AI datacenter capacity, an unprecedented scale that underscores the explosive demand for computational power fueled by advanced AI models like GPT. Huang, in a statement, framed the endeavor as a pivotal moment that will move AI technology from research labs into widespread real-world application, marking the beginning of what he calls an "AI industrial revolution."

Efforts to secure the necessary power allocations and real estate for the first gigawatt of infrastructure are already underway, according to people familiar with the matter. The project will leverage NVIDIA's upcoming Vera Rubin platform and is seen as a critical step towards supporting the development of artificial general intelligence (AGI). OpenAI’s ChatGPT now boasts 700 million weekly active users, creating immense and sustained pressure on computing resources.

This partnership deepens a long-standing relationship where NVIDIA has been the primary compute supplier for OpenAI's model training. The sheer magnitude of the investment, however, sets a new benchmark for private sector infrastructure projects and is likely to stimulate adjacent industries, from chip fabrication to renewable energy, given the massive power requirements. A spokesperson for OpenAI did not immediately respond to a request for further comment on the deployment timeline.

This article was updated to clarify that the $100 billion investment from NVIDIA is phased over the course of the project.