Nvidia's latest big move in the artificial intelligence arena involves to put $2.000 billion on the table to strengthen Nebius' AI cloudWith this move, the chip manufacturer consolidates its role not only as a hardware supplier, but also as a financial driver of the infrastructure that will support the next generation of AI services.
The operation places Nebius, Amsterdam-based, Nasdaq-listed cloud AI infrastructure companyAt the heart of the global advanced computing landscape, the company is committed to deploying more than five gigawatts of data center capacity by 2030, a figure that gives an idea of the scale leap it is preparing to make to power increasingly powerful models.
Nvidia gains 8,3% and secures future computing demand
According to the documentation submitted to the US regulator, Nvidia will buy $2.000 billion worth of Nebius sharesat a price of $94,94 per share. This capital injection gives it a stake of around 8,3% in the Dutch group, without aspiring to control it, but with a notable capacity to influence its technological roadmap.
In parallel to the announcement, Nebius shares reacted with strong gains on the NasdaqThe company's shares rose by as much as 12% to 14% during the session. Its value has multiplied several times over the past year, and its market capitalization now comfortably exceeds $25.000 billion, reflecting the market's appetite for those with significant financial resources.
The agreement states that Nebius will deploy more than five gigawatts of data center capacity by the end of 2030based on successive generations of Nvidia's accelerated computing platform. In terms of the electrical system, that power is roughly equivalent to the consumption of more than four million US homes concentrated in infrastructure dedicated to AI.
For Nvidia, the interest is not limited to revenue from chip sales. The company It uses these investments to secure long-term demand and anchor its partners in its technology ecosystem, from the GPUs H100 and H200 current to future architectures like GB200. It's a way of closing the circle: it finances operators who, in turn, become some of its most important clients.
Nebius, the European “neoncloud” that scales at a rate of gigawatts
Nebius has positioned itself as one of the new players specializing in high-density AI cloudThey are moving away from the general-purpose cloud model. Based in Amsterdam and operating in Europe, the United States, and the Middle East, their offering focuses on providing GPU clusters optimized for training and deploying next-generation models, with an architecture specifically designed for AI developers.
The company fits into the category of “neocloud“or “neonube”: operators that prioritize technology clients, large AI labs, and platforms that need continuous, intensive computing, rather than a multipurpose cloud for any type of workload. This approach has allowed it to close long-term contracts with giants like Microsoft and Meta, including a deal valued at around $17.000 billion with Azure and another worth several billion with Facebook's parent company.
Their strategy relies on data centers designed from the ground up for AI, with higher GPU density, advanced cooling, and low-latency networkingIn contrast to the approach of traditional clouds forced to reconcile very diverse services, Nebius seeks to maximize performance per watt and per euro invested in accelerated computing, a detail that is key for large training sessions.
The business has attracted the attention of investors: since its IPO, Nebius has experienced revaluations of several hundred percent and very rapid revenue growth linked to AI infrastructure. Nvidia's entry as a strategic shareholder is understood as a further step in consolidating its profile as a leading operator in the pure AI cloud.
An AI cloud for the era of autonomous agents
The public discourse of both companies highlights a new phase in the development of AI. Jensen Huang, founder and CEO of Nvidia, maintains that “AI is at another turning point: agentic AI”, referring to systems that act as autonomous agents, capable of making decisions and coordinating complex tasks without constant supervision.
According to Huang, Nebius is building an AI cloud designed for this new stage“fully integrated from silicon to software” and supported by Nvidia’s next-generation accelerated computing. The shared goal is to scale the cloud to meet the growing global demand for computing power for more complex models and more intensive applications.
For his part, Arkady Volozh, CEO of Nebius, emphasizes that Their platform was designed from the ground up for AIand not as a later adaptation of a general-purpose cloud. The alliance with Nvidia, he explains, allows them to extend that specialization “to the entire stack”: from large, gigawatt-scale AI factories to inference services and the software used by developers.
Both companies will collaborate on the design and support of these AI factoriesThis will be achieved through enhanced access to design materials, shared review processes, and the early integration of successive generations of Nvidia hardware. The result, if the deadlines are met, will be a network of extremely powerful data centers aligned with the chipmaker's technology roadmap.
Gigawatt-scale AI factories: from marketing to the real economy
The term "AI factoryThe term "data center" has become commonplace to describe large campuses of data centers dedicated almost exclusively to artificial intelligence. In practice, such a facility is a complex housing tens of thousands of state-of-the-art GPUs connected by high-capacity networks, with reinforced power supplies and advanced cooling systems.
Discuss more than five gigawatts of contracted capacity until 2030 This involves moving from projects of hundreds of megawatts to facilities that rival the power of large electricity generating plants. This leap forces governments and regulators to treat AI data centers as critical infrastructure, with a direct impact on energy planning, water availability, and industrial land use.
Nebius is already working on plans for campuses of several hundred megawatts in the United States and EuropeAnd it has made progress in deployments in countries like Finland, France, and Iceland, which offer a good combination of climate, electrical grid, and, in some cases, abundant renewable energy. The alliance with Nvidia accelerates this agenda, incorporating the requirements of the dominant AI accelerator provider from the design stage.
From an industrial perspective, these AI factories become a new type of basic infrastructure: They contribute investment, skilled employment, and indirect economic activity.However, they can also concentrate significant energy consumption in certain regions. The decision to host these types of projects is already part of national and regional competitiveness strategies.
Nvidia's strategy: to finance its own customers
The deal with Nebius fits into a broader pattern. In recent years, Nvidia has multiplied its investments in companies linked to the AI ecosystem, from model developers to data center operators and suppliers of key components such as advanced optics for interconnection.
In the last week before this announcement alone, the company allocated $4.000 billion to Lumentum Holdings and CoherentTwo companies specializing in optical systems for high-capacity data centers. In parallel, it has closed financing or equity agreements with players such as CoreWeave, another AI cloud provider, and has committed significant resources to software and semiconductor companies.
From September 2025, Nvidia's commitments to investments and agreements related to AI They are approaching 82.000 billion dollarsAccording to figures from analysts, this package includes significant investments in fundamental modeling laboratories such as OpenAI, Anthropic or xAIas well as stakes in European and North American startups dedicated to generative AI, infrastructure and developer tools.
The financial muscle to sustain this strategy comes from a Free cash flow close to $97.000 billion in the last fiscal yearThis growth is supported by a revenue increase of around 65%, reaching almost 216.000 billion. This position allows Nvidia to go beyond simply selling hardware and build a network of partners that it finances, equips, and supports in their expansion.
Industry critics describe this model as a form of “circular finance”, in which The same actor that supplies the critical component enters the capital of those who buy it.This ensures future demand. Its proponents see it as an efficient integration of the value chain. In any case, the result is an ecosystem where Nvidia's influence on strategic AI cloud decision-making is constantly growing.
Europe and Spain facing the new geography of computing
The fact that a key Nvidia ally is headquartered in Amsterdam is not a minor detail. Europe has been seeking to strengthen its digital sovereignty for years.However, in the field of advanced computing, the reality is complex: much of the core technology and supply chain remains concentrated outside the continent, even though the parent companies of some operators are European. This situation is strained by measures such as... AI chip crackdown driven from the United States.
Nebius offers the region a window to attract high-value AI infrastructureCombining European localization with deep integration into Nvidia's technology stack, its data center projects in Nordic and Central countries, supported by renewable energy and cold climates, align with EU priorities for energy efficiency and emissions reduction, but also reinforce reliance on a single acceleration provider.
For Spain, the situation presents both an opportunity and a challenge. The country It competes to attract data centers due to its potential in renewable energy, land availability, and new interconnections. with France and North Africa. A gigawatt-scale AI campus would involve billions in investment, direct and indirect jobs, and a boost to the local digital ecosystem, from service providers to universities and research centers.
At the same time, projects of this size can strain on the electricity grid, water planning, and the debate on energy useThe fundamental question is whether the computing capacity installed in Spanish territory will translate into local added value —in the form of companies, talent and advanced services— or will function mainly as an exportable computing infrastructure for third countries.
European institutions, through regulations such as the AI Act and climate and energy regulations, will need to define to what extent these mega data centers are considered strategic infrastructure And what requirements for efficiency, transparency, and contribution to the local economy will be demanded of them? The line between industrial opportunity and technological dependence narrows as the scale of the projects grows.
Risks of concentration and energy bill
Beyond the headlines about record investments, the alliance between Nvidia and Nebius raises two structural risks that also affect Europe and Spain. The first is the concentration of power around a single AI hardware providerwhich also participates in the capital and in the design of the infrastructure where that hardware is consumed.
In such a scenario, access to advanced computing can depend on bilateral agreements, purchase volume and preferential relationshipsThis makes it difficult for startups, technology SMEs, universities or research centers that do not have the scale or budget of the large hyperscalers.
The second major risk is the energy bill associated with these deploymentsFive gigawatts of AI capacity entails massive electricity and water consumption for cooling, even with highly efficient technologies. Without coordinated planning that includes new generation plants and grid upgrades, these types of projects can put pressure on prices and complicate national and European climate goals.
Some regulators are beginning to raise specific limits, requirements for the use of renewables, or energy storage mechanisms for large data centers. Formulas are also being discussed to ensure that the regions hosting these infrastructures obtain clear returns in the form of employment, innovation, and preferential access to advanced computing services for their own productive sectors.
In this context, Nvidia's $2.000 billion investment in Nebius is not just another financial deal: It marks a further step towards a model in which AI computing becomes a global strategic infrastructure, heavily influenced by a small number of technological actors and with direct implications for the European economy and industrial policy.
The alliance between Nvidia and Nebius is thus understood as a key piece of the future architecture of the artificial intelligence cloud: a strong capital commitment to ensure gigawatts of capacity, strengthen a network of partners that already includes hyperscalers like Microsoft and, at the same time, pose to Europe and countries like Spain the dilemma of how to take advantage of the wave of investment in AI data centers without losing decision-making power over the technology, energy and the benefits they will generate.
