Overview

• Both companies signed a letter of intent to deploy at least 10 gigawatts of Nvidia systems for OpenAI’s AI infrastructure, with final terms expected in the coming weeks.
• Nvidia’s funding will be deployed progressively as capacity comes online, and OpenAI will use the cash to buy Nvidia systems, with Nvidia designated a preferred supplier for chips and networking gear.
• The first phase is targeted for the second half of 2026 using Nvidia’s Vera Rubin platform, which Jensen Huang said equates to roughly 4 million to 5 million GPUs.
• Deal mechanics reported by Reuters include non‑voting, non‑controlling equity for Nvidia and an initial $10 billion tranche contingent on a definitive chip‑purchase agreement.
• The scale and vertical alignment could invite antitrust scrutiny, and the firms say the pact complements OpenAI’s collaborations with Microsoft, Oracle, SoftBank and the Stargate project.