Nvidia, OpenAI appear stalled on their mega deal. But the AI giants still need each other

Nvidia's CEO, Jensen Huang, and OpenAI's CEO, Sam Altman, made headlines in September when they announced plans for a monumental $100 billion deal aimed at revolutionizing the artificial intelligence (AI) landscape. However, five months later, no contracts have been finalized, and the anticipated financial transactions remain unrealized. Recent reports from The Wall Street Journal indicate that negotiations have stalled, raising concerns among investors. Some Nvidia executives have expressed skepticism regarding OpenAI's business model, leading to speculation about the future of their partnership. This has been a significant point of discussion in AI circles, especially since Nvidia warned in its quarterly risk assessments that definitive agreements with OpenAI were not guaranteed. Despite the existing tensions, both companies continue to rely on one another. Altman has indicated that OpenAI's ambitious revenue targets depend heavily on Nvidia's AI chips, while Huang emphasized the importance of OpenAI as a customer in showcasing innovative services that drive Nvidia's sales. Nvidia's stock soared to over $5 trillion at its peak in October, although it has since dipped approximately 15%, bringing its market capitalization down to $4.4 trillion. OpenAI, valued at $500 billion in the private market last year, is reportedly seeking a valuation exceeding $800 billion in its next funding round. Huang expressed optimism regarding OpenAI's growth, stating, "We are looking forward to Sam closing it and he's doing terrifically. There is no question about that." Nvidia's first investment in OpenAI occurred in October 2024 as part of a $6.6 billion funding initiative. Huang has reassured stakeholders that there is no "drama" in their relationship—an assertion echoed by Altman in a recent post on X, where he affirmed OpenAI's commitment to being a significant customer of Nvidia. Yet, progress on the ambitious infrastructure deal announced in September, which involves a power requirement of 10 gigawatts, appears to be lacking. According to the initial agreement, Nvidia's $10 billion investment would be activated upon the completion of the first gigawatt, anticipated to come online in late 2026. Nvidia's participation in OpenAI's current funding round is separate from last year's arrangement. Huang has indicated that Nvidia is eager to explore additional investments in OpenAI and is also interested in participating in the AI lab's IPO. On the ground, Nvidia's shares have experienced a slight decline, contributing to a broader downturn in tech stocks. Over the last decade, Nvidia and OpenAI have been closely intertwined, with OpenAI being one of the first organizations to utilize Nvidia’s early AI systems back in 2016. Following the release of ChatGPT, Nvidia reported a significant revenue increase, from $6 billion to $57 billion within a year. Analysts note that Nvidia holds over 90% of the market for graphics processing units (GPUs), while ChatGPT continues to lead the chatbot space with 800 million weekly users. OpenAI forecasts annual sales could reach $20 billion but is not expected to become profitable until 2030. The underlying tension in their relationship, which both companies publicly downplay, stems from their respective strategies to diversify by collaborating with competitors. Nvidia has strategically invested in other partners as its customer base becomes more concentrated, while OpenAI has been forging partnerships with other semiconductor firms, including AMD and Broadcom, to meet its substantial computing needs. In a recent statement, OpenAI's infrastructure executive Sachin Katti reiterated the fundamental nature of their relationship with Nvidia, emphasizing that "Our entire compute fleet runs on Nvidia GPUs. The demand curve is unmistakable. The world needs orders of magnitude more compute."