Huang Renxun reveals that NVIDIA may only invest $30 billion in OpenAI, with the $100 billion plan essentially put on hold

On March 5th, NVIDIA CEO Jensen Huang stated that the company’s previously rumored plan to invest $100 billion in OpenAI is currently “not realistic,” and the project has been significantly scaled back. Huang revealed at the Morgan Stanley conference in San Francisco that NVIDIA and OpenAI have essentially finalized a cooperation arrangement, with an expected investment of about $30 billion.

Huang pointed out that this investment could be one of the most strategically significant AI projects NVIDIA has participated in. He noted that OpenAI has a very prominent technological influence in generative AI, and NVIDIA’s core task is to provide the large-scale computing power needed to train and run AI models. As the demand for computing power continues to grow, AI chips and data center infrastructure are becoming key competitive areas in the global tech industry.

As early as 2025, NVIDIA proposed a larger-scale cooperation plan, aiming to invest up to $100 billion in AI infrastructure projects related to OpenAI over the next few years. According to that plan, the two sides would jointly build at least 10 gigawatt-capacity AI data centers and use NVIDIA’s GPU systems to support large language model training. However, the latest statements suggest that this massive investment framework may not be fully implemented.

Huang also mentioned that NVIDIA’s previous nearly $10 billion investment in AI company Anthropic is likely another important capital deployment in this field. Anthropic is the developer of the Claude AI system, and the market generally expects it to have the potential to go public in the future.

Regarding rumors that there are disagreements within NVIDIA about investing in OpenAI, Huang has previously publicly denied this. In a recent interview, he expressed strong support for OpenAI and its CEO Sam Altman, believing that the company plays an important role in advancing AI technology.

Currently, OpenAI remains a private AI company, but its position in generative AI and large language models has made it a focal point for global capital and tech companies. As the demand for AI training compute power continues to rise, investments around data centers, power supply, and high-performance GPUs are rapidly expanding.

Noted investor Michael Burry also previously pointed out that the emergence of ChatGPT is driving a massive AI infrastructure race. Many tech companies and capital institutions are continuously increasing their investments in AI compute and data center construction, a trend that could profoundly impact the global tech industry landscape in the coming years.