September 23, 2025 – In a groundbreaking move, Nvidia and OpenAI have announced a strategic partnership that is set to reshape the landscape of artificial intelligence (AI) development. The two tech giants are currently finalizing the specifics of this collaboration, with details expected to be ironed out in the coming weeks.
Under the terms of the agreement, OpenAI will leverage Nvidia’s cutting-edge hardware to construct and deploy AI data centers with a combined capacity of at least 10 gigawatts (GW). This ambitious project will involve the use of millions of Nvidia’s graphics processing units (GPUs) to train and deploy OpenAI’s next-generation AI models, propelling the pursuit of artificial general intelligence (AGI). The first phase of this initiative is slated to go live in the second half of 2026, utilizing Nvidia’s Rubin platform.
To bolster this strategic alliance, Nvidia has pledged to invest up to $100 billion in OpenAI for each gigawatt of capacity deployed. This marks Nvidia’s largest investment commitment to date, underscoring the significance of the partnership.
Jensen Huang, founder and CEO of Nvidia, reflected on the longstanding relationship between the two companies, stating, “Over the past decade, from the first DGX supercomputer to the breakthrough of ChatGPT, Nvidia and OpenAI have consistently pushed each other forward. This investment and infrastructure collaboration represent a new leap forward – we will deploy 10 gigawatts of computing power to drive the next era of intelligence.”
Huang further elaborated that the 10 GW capacity is equivalent to 4 to 5 million GPUs, a figure that matches Nvidia’s total annual output for the current year and is double that of the previous year. He emphasized the scale of the project, calling it “massive.”
During a previous earnings call, Huang revealed that the cost of building a 1 GW data center capacity ranges from 50billionto60 billion, with approximately $35 billion allocated for the purchase of Nvidia chips and systems. Based on these figures, the total investment for the 10 GW project could reach into the hundreds of billions of dollars.
Sam Altman, co-founder and CEO of OpenAI, stressed the importance of computing power, saying, “Everything starts with computing. Computing infrastructure will be at the heart of the future economy. Through our collaboration with Nvidia, we will not only drive new AI breakthroughs but also empower individuals and businesses on a large scale.”
Greg Brockman, co-founder and president of OpenAI, added, “Since the inception of OpenAI, we have worked closely with Nvidia. We have utilized their platform to build AI systems that are used by hundreds of millions of people every day. We are thrilled to partner with Nvidia to deploy 10 gigawatts of computing power, pushing the boundaries of intelligence and expanding the inclusive value of this technology.”
As part of the collaboration, OpenAI has selected Nvidia as its strategic computing and networking partner. The two companies will jointly optimize their roadmaps, covering OpenAI’s models and infrastructure software, as well as Nvidia’s hardware and software.
This strategic partnership will complement OpenAI’s existing extensive collaborations with companies such as Microsoft, Oracle, SoftBank, and Stargate partners, all of which are working together to build the world’s most advanced AI infrastructure.
In a related development, Altman announced on Sunday that OpenAI plans to launch several new compute-intensive products in the coming weeks. Due to the high costs associated with these products, some features will initially be available only to Pro subscribers.
