Nvidia has committed roughly $2 billion to AI cloud infrastructure builder Nebius through pre-funded warrants, taking a minority stake while helping finance an ambitious expansion of hyperscale AI capacity. The deal signals that the next phase of the AI boom is increasingly defined not just by better models and software, but by who can secure enough electricity, advanced chips, cooling systems, and physical data center capacity to keep large-scale AI workloads running.
Under the companies’ stated plans, Nebius aims to build more than 5 gigawatts of Nvidia-accelerated computing capacity by 2030. That is an enormous figure in data center terms and reflects the industrial scale now required to train frontier models, serve inference at global volume, and support the growing wave of agentic AI systems entering the market.
A strategic infrastructure partnership
Amsterdam-headquartered Nebius, which trades on Nasdaq under the ticker NBIS, is part of a newer class of AI-focused cloud providers often described as “neocloud” companies. Unlike legacy cloud platforms adapted for AI demand, Nebius has positioned itself as a provider built specifically for AI workloads from the outset. That distinction has become more important as customers seek infrastructure optimized for model training, inference, and large GPU fleet management rather than conventional enterprise computing.
The company’s roots go back to the restructuring of Yandex’s international operations. After the Dutch parent of the Russian search giant divested its Russian businesses in 2024, the remaining international assets were reorganized into Nebius. Since then, the company has expanded AI infrastructure operations across the United States, Europe, and Israel, moving aggressively to establish itself in a market where access to compute and power is increasingly decisive.
Nebius already runs large GPU clusters powered by Nvidia hardware, including H100 and H200 accelerators. As part of the new partnership, it plans to adopt future Nvidia technologies as well, including the Rubin architecture, Vera CPUs, and BlueField storage platforms. The collaboration is expected to span the full AI stack, from data center design and infrastructure buildout to inference deployment and GPU fleet operations.
For Nvidia, the investment is more than a financial bet. It also helps secure a major long-term ecosystem partner that is likely to consume substantial volumes of Nvidia chips over time. In practical terms, Nvidia is reinforcing demand for its hardware while also supporting the infrastructure layer needed to sustain continued AI growth.
Why scale now matters more than ever
Nvidia CEO Jensen Huang framed the deal as part of a new inflection point for the AI industry, with agentic AI driving fresh demand for compute and accelerating infrastructure expansion. Nebius CEO Arkady Volozh similarly emphasized that the company was built specifically for AI rather than adapted from legacy cloud services, arguing that this gives it an advantage as the market shifts toward specialized infrastructure.
The numbers attached to the expansion are substantial. Just days before the investment announcement, Nebius secured approval for a 1.2-gigawatt AI data center campus in Independence, Missouri. According to the company, the site ranks among the largest planned AI infrastructure projects in the United States. The development is expected to create around 1,200 construction jobs and generate an estimated $650 million in economic impact over two decades.
The Missouri campus is only one part of a much broader global strategy. Nebius has indicated that it could operate as many as 16 data center sites worldwide by the end of 2026, with contracted power capacity approaching 3 gigawatts within the next year. Those figures highlight the degree to which AI infrastructure is becoming an energy-intensive industrial buildout rather than a conventional software expansion story.
Spending surges ahead of revenue scale
The company’s financial profile reflects that reality. Nebius reported capital expenditures of about $2.1 billion in the fourth quarter of 2025 alone. For 2026, projected spending is expected to land between $16 billion and $20 billion as the company races to build ahead of demand. That level of investment is extraordinary for a company still in an expansion phase, underscoring how AI infrastructure providers are being forced to spend heavily before their revenue fully catches up.
By comparison, Nebius reported trailing twelve-month revenue of roughly $530 million. However, the company argues that its longer-term outlook is supported by contracted business and backlog. Among the most significant disclosed agreements is a multi-year deal with Microsoft tied to a data center project in Vineland, New Jersey, estimated at between $17 billion and $19.4 billion. Nebius has also disclosed a separate $3 billion arrangement with Meta Platforms.
These figures suggest that the market for AI infrastructure is increasingly being shaped by long-duration enterprise and hyperscale contracts rather than short-term cloud demand alone. As major technology companies lock in future compute capacity, providers like Nebius are trying to scale physical infrastructure fast enough to meet those commitments.
Investor reaction and the bigger message
Investors reacted positively to Nvidia’s move. Nebius shares rose roughly 13% to 16% after the announcement, pushing the company’s market capitalization above $24 billion in early trading. The market response reflects a familiar pattern often described as the “Nvidia halo effect,” where participation from the chipmaker is interpreted as a strong signal of strategic relevance.
Still, the more important takeaway is broader than one company’s share price. The Nebius investment highlights a structural shift in the AI economy: the bottleneck is no longer only algorithmic innovation. Increasingly, the limiting factors are access to electricity, cooling systems, advanced semiconductors, and the ability to construct and operate massive data center campuses at speed.
That dynamic helps explain why gigawatt-scale projects are attracting so much attention. Large language models, inference-heavy AI applications, and autonomous agent systems require persistent, large-scale compute. Meeting that demand means building infrastructure at a scale more commonly associated with utilities and industrial facilities than with traditional software businesses.
In that sense, Nvidia’s $2 billion commitment to Nebius is both a targeted investment and a broader statement about where the AI race is heading. The next winners may not be determined by software alone, but by who can assemble the capital, energy access, supply chains, and hardware ecosystems required to power AI at global scale.