How are OpenAI and Nvidia Partnering on AI Infrastructure?
OpenAI and Nvidia have confirmed a strategic collaboration to roll out 10GW of GPU-powered systems for AI data centres.
The agreement centres on building the compute infrastructure required to support OpenAI’s future models and accelerate the development of superintelligence.
The project involves a phased investment of up to US$100bn by Nvidia, linked to each gigawatt of capacity delivered.
The companies plan to activate the first phase of the infrastructure in the second half of 2026, launching on Nvidia’s Vera Rubin platform.
Infrastructure for scaled compute
Jensen Huang, Nvidia CEO, says the agreement is the next step in a long-running collaboration between the two companies.
“Nvidia and OpenAI have pushed each other for a decade, from the first DGX supercomputer to the breakthrough of ChatGPT,” he says.
“This investment and infrastructure partnership mark the next leap forward – deploying 10GW to power the next era of intelligence.”
Speaking to CNBC, Jensen also called the effort “the biggest AI infrastructure project in history,” describing the deal as being “about building an AI infrastructure that enables AI to go from the labs into the world.”
OpenAI Cofounder and CEO Sam Altman links the expansion directly to the company’s model performance and commercial strategy.
“There’s no partner but Nvidia that can do this at this kind of scale, at this kind of speed,” he told CNBC. “Building this infrastructure is critical to everything we want to do. This is the fuel that we need to drive improvement, drive better models, drive revenue, drive everything.”
He also underlined the strategic relevance of compute systems, stating: “Everything starts with compute. Compute infrastructure will be the basis for the economy of the future, and we will utilise what we’re building with Nvidia to both create new AI breakthroughs and empower people and businesses with them at scale.”
Greg Brockman, Cofounder and President of OpenAI, hightlighted how collaboration and partnership has bonded the two companies for several years.
“We’ve been working closely with Nvidia since the early days of OpenAI,” he says.
“We’ve utilised their platform to create AI systems that hundreds of millions of people use every day. We’re excited to deploy 10GW of compute with Nvidia to push back the frontier of intelligence and scale the benefits of this technology to everyone.”
Aligning model and platform development
The scale of this deployment sets a new benchmark in AI infrastructure planning.
The 10GW target involves constructing data centres equipped with millions of GPUs (graphics processing units), specialised hardware used for high-performance compute tasks such as training AI models.
These systems are designed to support OpenAI’s next generation of AI-driven products and services.
As part of the deal, Nvidia will become OpenAI’s preferred partner for strategic compute and networking. The companies will coordinate their development plans, aligning OpenAI’s model and infrastructure software with Nvidia’s hardware and software platforms.
The 2026 Vera Rubin deployment represents the starting point, with each subsequent gigawatt-phase adding to OpenAI’s compute footprint.
Taking this kind of staged approach is intended to enable OpenAI to meet rising global demand for AI at scale.
Supporting an expanding AI ecosystem
The announcement builds on multiple collaborations OpenAI and Nvidia maintain across the AI ecosystem.
The companies continue to work alongside Microsoft, Oracle, SoftBank and Stargate partners to deliver advanced infrastructure for AI operations at hyperscale.
These alliances aim to offer the compute and networking backbone required for global AI rollouts, enabling enterprises, developers and public sector organisations to integrate advanced AI capabilities into their workflows.
OpenAI now records over 700 million weekly active users, with strong adoption from businesses and technical communities. The expanded infrastructure is designed to sustain this growth and provide capacity for wider deployment of AI models.
Jensen views the scale of the initiative in terms of long-term impact. “We’re literally going to connect intelligence to every application, to every use case, to every device – and we’re just at the beginning,” he says. “This is the first 10GW, I assure you of that.”
Both companies state that formal agreement details will be finalised in the coming weeks, with further updates expected as the first deployment approaches in 2026.
