Overview

• The planned investment would finance roughly 10 gigawatts of new AI data-center capacity for OpenAI using Nvidia hardware.
• Funding is reported to be phased, with an initial $10 billion released after the first 1 GW comes online, and Nvidia would receive an equity stake.
• The companies issued a joint statement confirming a letter of intent and offered limited specifics, with Nvidia representatives declining further comment.
• The scale reflects OpenAI’s heavy compute needs, including ChatGPT’s roughly 700 million weekly users and recent hints of compute‑intensive product launches.
• Some reports say the first phase could begin operating in the second half of 2026 using Nvidia’s Vera Rubin superchip, though details remain to be confirmed.