The first deal was flashy and ultimately hollow. The new one is quieter but considerably more real. Nvidia’s $30 billion equity investment in OpenAI — part of a $100 billion funding round that will value the ChatGPT maker at $730 billion — replaces a controversial $100 billion arrangement that fell apart after it emerged the commitment was never firm.
Unlike its predecessor, the new investment involves no chip purchase obligations. Nvidia will simply buy equity in OpenAI, joining a roster of major investors that includes Amazon, SoftBank, and Microsoft. It is a straightforward investment in one of the world’s most talked-about private companies, and it removes the conflict-of-interest concerns that made the previous deal so problematic.
When the original $100 billion deal was announced last September, Nvidia’s market cap briefly exceeded $5 trillion. Markets were excited, but the excitement masked a structural problem: the deal essentially had Nvidia funding its own chip sales. When reports surfaced that the arrangement was never binding and that OpenAI was already exploring chip alternatives, the deal collapsed and markets wobbled.
OpenAI has since confirmed partnerships with AMD and Broadcom, demonstrating that its hardware strategy has diversified significantly. That makes Nvidia’s ongoing commitment to OpenAI even more interesting — it is choosing to invest in a company that may no longer be its best customer, which suggests the investment is driven by strategic and financial calculus rather than supply chain logic.
OpenAI’s path forward is cluttered with challenges. Market share in the AI assistant space has dropped by more than 20 points. Enterprise customers are gravitating toward Anthropic. Advertising experiments have drawn competitive fire. And the $730 billion valuation demands a level of future profitability that the company has not yet come close to demonstrating. Nvidia is betting on a turnaround — a costly and high-profile one.