Nvidia plans to invest $100 billion in OpenAI, giving the ChatGPT maker a huge war chest to build data centers using the chipmaker's chips.
OpenAI will have access to millions of Nvidia chips in the coming years.
“It all starts with computing,” Sam Altman, chief executive of OpenAI, said in a joint press release. “Computing infrastructure will be the foundation of the future economy, and we will use what we build with Nvidia to both create new breakthroughs in artificial intelligence and empower people and businesses.”
The $100 billion investment in OpenAI will be phased in as each gigawatt of the 10 planned gigawatts of capacity needed for data centers is used. The first gigawatt data center is planned to be put into operation in the second half of 2026.
“This investment and infrastructure partnership marks the next step forward—10 gigawatts of power to power a new era of intelligence,” Jensen Huang, founder and CEO of Nvidia, said in a press release.
Companies around the world are expected to spend $375 billion on artificial intelligence infrastructure in 2025. Nvidia GPUs power most of today's artificial intelligence data centers. That makes its chips a critical commodity in the artificial intelligence boom and has catapulted it to the forefront of global geopolitics.
Nvidia has become the puppet master of the artificial intelligence ecosystem. The company has been at the center of several major deals in the past year, including a $250 million investment in CoreWeave, a cloud services provider that leases Nvidia GPUs. Shortly after the investment, CoreWeave signed a $6.3 billion deal with the chipmaker, ensuring that all unsold cloud computing capacity would be acquired by Nvidia.
Last week Nvidia acquired 5 billion dollars shares in its weak rival Intel in developing data center and PC products. In the same week, Nscale, another London-based cloud provider using Nvidia chips, received 500 US dollars–million investment from Nvidia, followed by US$2.7–billion commitment to developing artificial intelligence infrastructure in the UK
Critics are concerned that this could be an example of “circular finance,” in which a company invests in its largest customers, who then use the money to buy the investor's products.
The concern is that this could lead to inflated demand for Nvidia chips. Nvidia supporters see the $100 billion investment in OpenAI as another sign of huge pent-up demand for AI as companies and countries rush to roll out computing infrastructure.
“We believe Wall Street is underestimating demand in the Fourth Industrial Revolution,” said Dan Ives, senior equity analyst at Wedbush Securities. “We believe there is a multiplier effect because Nvidia is investing in the future of large tech companies, and for every $1 Jensen invests in Nvidia, they will generate $8 to $10 in returns in the coming years.”
In 2016, Huang donated one of his first eight-GPU supercomputers to OpenAI, which will launch ChatGPT in 2022, ushering in the generative AI boom.
OpenAI is currently valued at $500 billion and has 700 million lightly active users.
In January, OpenAI, along with Japanese investor Softbank and database provider Oracle, committed $100 billion to build artificial intelligence infrastructure in the United States. As part of the Stargate project, companies began to build 4.5 gigawatts data center in Texas. They plan to invest $500 billion over the next four years.
OpenAI signed deal with in May to build a one-gigawatt data center complex in the United Arab Emirates, a deal that was brokered by the Trump administration. Oracle, Nvidia, Softbank, Cisco and G42, an Emirati artificial intelligence company backed by the royal family, are committed to supporting the project.
