Nvidia and OpenAI Partner in Landmark $100 Billion Deal to Deploy 10 GW of AI Power

Nvidia has unveiled a sweeping strategic partnership with OpenAI that will see the chipmaker invest up to $100 billion to help build out new data centres and computation infrastructure essential for scaling artificial intelligence. The announcement comes as demand for AI hardware surges and the race to deploy frontier AI models intensifies.

Central to the agreement is a letter of intent for deploying ten gigawatts (GW) of Nvidia systems through OpenAI’s next-generation AI infrastructure—enough compute to train and run increasingly ambitious models. The first phase will bring in the first gigawatt of computing power using Nvidia’s Vera Rubin platform, scheduled to go live in the second half of 2026.

Under the deal, Nvidia will receive equity (i.e., non-controlling ownership) in OpenAI. The investment will be structured in stages—whereby the first tranche of $10 billion is tied to deployment of the first gigawatt of infrastructure. OpenAI, in turn, will treat Nvidia as its preferred partner for both compute hardware and networking needs, aligning its model and infrastructure roadmap with Nvidia’s hardware developments.

OpenAI’s CEO Sam Altman emphasized that “everything starts with compute,” pointing out that foundational infrastructure is critical not just to run today’s AI models but to power breakthroughs in the future. Nvidia CEO Jensen Huang echoed the sentiment, saying this partnership marks “the next leap forward—deploying 10 gigawatts to power the next era of intelligence.”

Experts suggest this deal cements Nvidia’s position as the backbone of large-scale AI hardware, especially at a time when chips, power, and data centre scale are key constraints facing AI development. There is also recognition that as models grow, so do their energy, cooling, and operational demands—meaning that partnerships like this one are likely to define leadership in the AI era.

Still, the deal will raise important questions around investment risk, regulatory scrutiny, and the economics of deploying massive AI infrastructure. But for now, the industry views this as a transformative step: a deal that not only reflects how central AI is becoming, but also how those building AI are thinking long term—about scale, alignment (between hardware and software), and strategic advantage.