Two months ago, Nvidia CEO Jensen Huang and OpenAI CEO Sam Altman announced a historic agreement for Nvidia to invest up to $100 billion over several years, starting in 2026, to support OpenAI’s AI supercomputing infrastructure. While the announcement grabbed headlines, Nvidia reminded investors in its latest quarterly filing that “there is no assurance that any investment will be completed on expected terms, if at all.”

Nvidia has been actively deploying its cash, including a $5 billion investment in Intel and an agreement to invest up to $10 billion in Anthropic, alongside the OpenAI plan. Despite the uncertainties, Nvidia executives remain bullish. CFO Colette Kress highlighted OpenAI’s 800 million weekly users, 1 million enterprise customers, and healthy gross margins, while Huang emphasized that all of OpenAI’s operations currently run on Nvidia hardware.

However, OpenAI has also partnered with AMD, signing a formal agreement to deploy 6 gigawatts of AMD Instinct GPUs starting next year, complete with stock warrants and vesting milestones—something Nvidia’s deal currently lacks. The contrast highlights that while Nvidia’s OpenAI announcement signals massive potential, the actual execution and contractual certainty remain unproven, even as both companies continue deepening their AI collaboration.