From Letter of Intent to Global Infrastructure: The Timeline of the $100B Build-out

This week’s stunning announcement of a $100 billion Nvidia-OpenAI partnership began with a simple “letter of intent,” but it kicks off one of the most ambitious infrastructure projects in modern history. The timeline laid out provides a clear, albeit challenging, roadmap from this September 2025 agreement to a fully operational, globe-spanning AI supercomputer.
The first major milestone is the deployment of the initial gigawatt of computing power. This is a critical trigger point, as it unlocks the first $10 billion tranche of Nvidia’s investment. The successful execution of this first phase will be a proof-of-concept for the entire 10-gigawatt vision and a major focus for engineering teams over the next year.
The target date for this initial phase to come online is the second half of 2026. This gives the partners roughly a year to finalize plans, secure locations, and begin building the new data centers that will house Nvidia’s next-generation Vera Rubin platform. This aggressive timeline underscores the urgency both companies feel in the competitive AI landscape.
Beyond 2026, the project will scale up, with subsequent phases adding more gigawatts of power until the 10 GW goal is reached. This staged build-out allows for a more manageable construction process and enables the partners to incorporate technological advancements and learnings from earlier phases into later ones.
From the signing of a letter in late 2025 to the launch of a revolutionary computing platform in 2026 and beyond, this timeline is a testament to the speed and scale of the AI revolution. It’s a sprint to build the foundation for the next era of intelligence, and the clock is now officially ticking.