The Numbers Behind the Deal: $100B, 10GW, and the Path to 2026

Beyond the headlines, the Nvidia-OpenAI partnership is defined by a set of staggering numbers that lay out a clear and ambitious roadmap. The three key figures—$100 billion, 10 gigawatts, and the year 2026—tell the story of the scale, scope, and timeline of this historic undertaking.
The headline figure, $100 billion, represents the potential investment from Nvidia, making it one of the largest corporate investments in history. This capital is the fuel for the entire project. It’s structured in stages, with an initial $10 billion payment, providing a massive, long-term financial runway for OpenAI’s research and operational needs.
The technical core is 10 gigawatts (GW), the total computing power the new infrastructure will deploy. This is the raw horsepower that will train and run OpenAI’s next-generation models. To put this in perspective, it’s a quantity of specialized computing power far exceeding what is available in most commercial data centers today, all dedicated to a single entity’s goals.
The timeline is anchored by the year 2026. The “second half of 2026” is the target for the first phase of this new infrastructure to come online. This initial deployment will be built using Nvidia’s next-generation Vera Rubin platform, setting a concrete date for when the tangible results of this partnership will begin to emerge.
Together, these numbers paint a complete picture. The $100 billion provides the means, the 10 gigawatts defines the objective, and the 2026 target sets the pace. It’s a meticulously planned, numerically grounded strategy to build the world’s most advanced AI infrastructure and accelerate the journey toward super-intelligence.