Nvidia announced it will invest up to $100 billion in OpenAI , boosting the construction and deployment of data centers based on the company’s AI processors. OpenAI plans to build infrastructure worth hundreds of billions of dollars , requiring about 10 gigawatts of power — a metric used to describe the largest clusters of artificial intelligence chips.
10 gigawatts equals 5 million GPUs
Nvidia CEO Jensen Huangexplained that 10 gigawatts of power is equivalent to 4 to 5 million GPUs, double the amount produced last year. Huang described the project as “giant” and highlighted the importance of working closely with OpenAI, with Sam Altman and Greg Brockman in attendance.
See also: Nvidia – Wayve: They discuss investing in autonomous driving
First phase: 10 billion investment
The initial $10 billion investment will be triggered by the completion of the first gigawatt of infrastructure. The investments will be made in phases, based on the company's current valuations, according to a source familiar with the details who spoke on condition of anonymity.
The announcement sent Nvidia shares up 4%, boosting the company's market capitalization by nearly $170 billion to nearly $4.5 trillion. The momentum reflects investors' expectations for future GPU demand due to the rapid growth of AI.
OpenAI faces three critical challenges
Sam Altman emphasized that OpenAI must focus on three axes: leading research in artificial intelligence, creating products that users want to use, and managing the unprecedented infrastructure challenge. Nvidia's investment mainly covers the third axis, securing the necessary chips and computing power for the next generation of AI models.
Nvidia as a preferred strategic partner
Nvidia will provide not only GPUs but also networking equipment, strengthening its position as a key supplier for OpenAI. Huang called the partnership “monumental” and stressed that the investment is in addition to what has already been agreed upon, underscoring the company’s strategic importance to AI development.
Competition in the AI market
Despite Nvidia's dominance, the company faces increased competition from Advanced Micro Devices (AMD) and cloud providers developing their own AI systems. Building a gigawatt of data center infrastructure costs $50 billion to $60 billion, of which about $35 billion goes to Nvidia chips and equipment, Huang said.
See also: OpenAI signs deal with Oracle for Project Stargate
Collaborations with Microsoft, Oracle and SoftBank
Nvidia's investment complements OpenAI's existing strategic partnerships with Microsoft, Oracle, SoftBank and the Stargate project. Microsoft, an early investor in OpenAI, is integrating its AI models into the Azure platform and Microsoft Office, while other investors include Thrive Capital.
Nvidia invests in startups and competitors
In addition to OpenAI, Nvidia is strengthening its market presence through investments in Intel ($5 billion stake), Nscale ($700 million), and Enfabrica ($900 million). These moves reinforce the company's strategy to control critical AI technologies and infrastructure globally.
The future of AI requires massive infrastructure
The Nvidia-OpenAI partnership highlights that the development of the next generation of artificial intelligence is not only based on algorithms but also on high-powered infrastructure. The ability to build and maintain giant-scale data centers will determine the speed and scope of innovations in the field of AI.
See also: Nvidia invests $5 billion in Intel for chip development
Nvidia's investment in OpenAI marks a new era for artificial intelligence, where scale and processing power become the defining factors. This partnership is not just about technology, but also about strategic hegemony in the rapidly evolving world of AI.
