Photo by anna-m. w. on Pexels

OpenAI and Nvidia are exploring a landmark $100 billion collaboration to bolster AI infrastructure development, sources report. A preliminary agreement outlines Nvidia’s intention to provide at least 10 gigawatts of its advanced hardware to support OpenAI’s ambitious next-generation AI models geared towards achieving superintelligence.

Nvidia’s investment in OpenAI could reach $100 billion as the system, leveraging Nvidia’s upcoming Vera Rubin platform, is deployed, potentially commencing in the latter half of 2026. This signifies a move towards tighter integration between AI chip manufacturers and AI software developers.

Under the proposed deal, Nvidia would acquire a financial interest in OpenAI, a key client, while OpenAI would gain guaranteed access to Nvidia’s essential processors and crucial funding. This agreement, however, could raise antitrust concerns for some competitors regarding Nvidia’s substantial control in chip manufacturing and OpenAI’s leadership in AI software.

Sources suggest Nvidia would obtain non-voting shares in OpenAI, with OpenAI using these funds to then purchase Nvidia chips. OpenAI CEO Sam Altman has frequently stressed the importance of robust compute infrastructure as the cornerstone of future economic expansion and groundbreaking AI innovations.

While OpenAI explores potential custom chip designs, its partnership with Nvidia will not impact its other technology ventures, including those with Microsoft. The deal’s purpose is to increase OpenAI’s ability to achieve artificial general intelligence (AGI) and could influence various other industry partnerships. Given its magnitude, however, it is susceptible to regulatory review.

Analysts have noted the possibility of ‘circular’ investment concerns, where Nvidia’s investment may indirectly return to the company via OpenAI’s chip purchases. Nevertheless, the potential for significant advancements in AI capabilities enabled by this collaboration remains substantial.