Nvidia Pivots to $30B Equity Stake in OpenAI, Scrapping $100B Compute Pact

The shift from a $100 billion multi-year commitment to a $30 billion direct equity investment marks a significant recalibration in the relationship between the world’s most valuable chipmaker and its most prominent AI customer. Originally conceived in September 2025 as a decade-long strategic alliance, the $100 billion "compute-for-equity" pact was designed to secure OpenAI’s hardware pipeline while providing Nvidia with a massive, guaranteed revenue stream. However, the collapse of that preliminary agreement in favor of a more traditional equity injection reflects a cooling of the hyper-aggressive infrastructure spending models that dominated 2024 and 2025.

The original deal’s structure—ten installments of $10 billion—was inherently complex, tied to the deployment of 10 gigawatts of new computing capacity and the purchase of millions of Nvidia’s Blackwell and Rubin-class processors. By January 2026, investor caution began to mount as US technology stocks fell 17 percent year-to-date. This market volatility likely forced both parties to simplify the arrangement. A $30 billion direct equity stake provides OpenAI with immediate, flexible liquidity to fund its "Stargate" supercomputer projects and other infrastructure needs, while allowing Nvidia to maintain its strategic influence without the rigid long-term delivery obligations of the prior $100 billion framework.

“

The shift from a $100 billion multi-year commitment to a $30 billion direct equity investment marks a significant recalibration in the relationship between the world’s most valuable chipmaker and its most prominent AI customer.

This investment is part of a broader, record-breaking $100 billion funding round for OpenAI. With SoftBank reportedly in final negotiations for its own $30 billion stake and Amazon potentially joining the cap table, OpenAI is positioning itself as a sovereign-scale entity in the AI ecosystem. At a pre-money valuation of $730 billion, OpenAI is now rivaling the market caps of established tech giants like Meta and Alphabet. This "circular economy" of AI—where chipmakers like Nvidia invest billions into their own customers, who then use that capital to buy the chipmaker's hardware—has become a defining characteristic of the current market cycle. It ensures that Nvidia remains the primary beneficiary of OpenAI’s growth, effectively locking out competitors like AMD and Cerebras from OpenAI's core infrastructure.

However, the sheer scale of this $30 billion investment will almost certainly draw intense regulatory scrutiny. Antitrust authorities in the US and EU have already expressed concern over "vertical integration" in the AI sector, where a dominant hardware provider holds significant equity in the dominant software provider. By pivoting to a direct equity stake, Nvidia and OpenAI may be attempting to create a cleaner legal structure than the previous compute-sharing agreement, though the competitive implications remain the same.

Looking forward, the success of this deal hinges on OpenAI’s ability to translate this massive capital infusion into tangible revenue growth through its ChatGPT Enterprise and Sora video-generation products. For Nvidia, the $30 billion investment is a defensive masterstroke, ensuring that its largest customer remains tethered to its ecosystem even as rivals like AMD’s Instinct MI450 and Groq’s LPU architecture attempt to gain a foothold. The deal, which could close as early as this weekend, signals that while the "blank check" era of AI infrastructure may be evolving, the scale of capital required to stay at the frontier of machine learning is only increasing.