Nvidia has deepened its infrastructure-first AI strategy with a roughly $2 billion investment in Nebius Group, a fast-growing AI cloud provider that is positioning itself as a purpose-built platform for the next wave of model training, inference, and agentic AI workloads. The investment, structured through pre-funded warrants, gives Nvidia a minority stake while helping finance Nebius’s ambitious expansion plans.
At the center of the deal is scale. The two companies said the broader buildout is expected to exceed 5 gigawatts of Nvidia-accelerated computing capacity by 2030, a figure that underscores just how far the AI race has shifted away from software alone and toward the physical backbone of the industry: power, cooling, real estate, and semiconductor supply.
A Massive Bet on AI Factories
Nebius and Nvidia are not talking about ordinary cloud expansion. Their collaboration is aimed at building large AI “factories” capable of training frontier models, running inference at industrial scale, and supporting the growing wave of agentic AI systems entering the market. In practical terms, gigawatt-scale data center capacity points to infrastructure on the level of major utilities and regional economic projects rather than incremental server deployments.
Nvidia founder and CEO Jensen Huang described the current market as another turning point for artificial intelligence, saying that agentic AI is driving a new surge in compute demand and accelerating infrastructure deployment. Nebius, he said, is building an AI cloud designed specifically for that era.
For Nvidia, the rationale is clear. Backing Nebius means supporting a long-term ecosystem partner that is likely to consume enormous volumes of Nvidia hardware over time. For Nebius, the agreement provides both capital and strategic alignment with the dominant supplier of AI accelerators.
Who Is Nebius?
Headquartered in Amsterdam and listed on Nasdaq under the ticker NBIS, Nebius has emerged as one of the more closely watched names in the so-called “neocloud” segment. These companies are being defined by a simple idea: instead of adapting legacy enterprise cloud platforms for AI, they are building cloud infrastructure around AI workloads from the ground up.
Its origins trace back to the restructuring of the Dutch parent linked to Russian search giant Yandex. After divesting its Russian businesses for about $5.4 billion in 2024, the company reorganized its remaining international assets into Nebius. Since then, Nebius has developed AI infrastructure operations across the United States, Europe, and Israel.
The company already operates large GPU clusters using Nvidia hardware, including the H100 and H200 accelerators. Under the new partnership, it also plans to adopt future Nvidia technologies such as the Rubin architecture, Vera CPUs, and BlueField storage platforms. According to the announcement, the collaboration spans the full AI stack, from data center design to inference deployment and GPU fleet management.
Expansion Plans Are Already Underway
The scale of Nebius’s ambitions is not merely theoretical. Just days before the Nvidia investment was announced, the company secured approval for a 1.2-gigawatt AI data center campus in Independence, Missouri. The site is one of the largest planned AI infrastructure developments in the United States and highlights the company’s strategy of moving quickly to lock in power and capacity.
The Missouri project is expected to generate around 1,200 construction jobs and produce an estimated $650 million in economic impact over two decades. For local and regional policymakers, such projects are increasingly being framed not just as data centers, but as industrial anchors tied to employment, grid planning, and long-term capital formation.
Missouri is only one part of a much broader rollout. Nebius said its larger expansion strategy could see it operating up to 16 global data center sites by the end of 2026, while contracted power capacity could approach 3 gigawatts within the next year. That pace reflects an industry-wide reality: companies believe demand for AI computing may outrun available supply unless infrastructure is built aggressively in advance.
Spending Big, Revenue Still Catching Up
Like many infrastructure-heavy AI companies, Nebius is spending ahead of revenue. The company reported approximately $2.1 billion in capital expenditures in the fourth quarter of 2025 alone. For 2026, projected spending is expected to fall in a range of $16 billion to $20 billion, an enormous sum that illustrates how expensive it has become to compete in AI cloud infrastructure.
Revenue, by contrast, remains comparatively modest for now. Nebius reported roughly $530 million in trailing twelve-month revenue. Still, the company says its long-term contracts and backlog support expectations for rapid scaling as new sites come online and customer demand deepens.
Among the major agreements already cited by the company is a multi-year deal with Microsoft connected to a data center project in Vineland, New Jersey, estimated at roughly $17 billion to $19.4 billion. Nebius has also disclosed a separate $3 billion arrangement with Meta Platforms. These partnerships suggest that hyperscale and enterprise demand for dedicated AI infrastructure remains robust despite the capital intensity involved.
Wall Street Responds Quickly
Investors reacted positively to the Nvidia announcement. Nebius shares rose between 13% and 16% after the news, pushing the company’s market capitalization above $24 billion in early trading. The move reinforced what market observers often call the “Nvidia halo effect,” where backing from the chipmaker is interpreted as a strong signal of strategic relevance and future demand.
That response also reflects a broader market belief: in AI, the chokepoint is increasingly not ideas, but infrastructure. Advanced models require more than software talent. They depend on enormous electrical capacity, cooling systems, physical campuses, supply-chain access to advanced accelerators, and the ability to deploy and operate these systems at scale.
The Bigger Message for the AI Industry
The Nvidia-Nebius deal captures a defining shift in the AI economy. The industry may be powered by algorithms, but its near-term competitive edge is increasingly determined by who can secure land, energy, chips, and financing fast enough to build the backbone of the next computing era.
By investing directly in Nebius, Nvidia is doing more than supporting a customer. It is helping shape a wider ecosystem of AI infrastructure designed to absorb future generations of accelerators and support the rapid expansion of training and inference demand. For Nebius, the partnership provides credibility, capital, and technical alignment at a moment when purpose-built AI clouds are racing to establish themselves before the market consolidates.
If the companies execute on their plans, the result could be one of the most significant examples yet of how AI is evolving from a software trend into a capital-intensive industrial buildout. And as that transition continues, deals like this may become the clearest signal of where the next phase of AI competition will be won: not only in model performance, but in the scale of the infrastructure underneath it.