Nebius, the Amsterdam-based AI cloud infrastructure provider, has announced the launch of its first GPU cluster in Kansas City, Missouri, furthering its strategic expansion into the U.S. market. The deployment, scheduled to go live in Q1 2025, will house thousands of Nvidia GPUs, with plans to scale up to 40 megawatts of power and approximately 35,000 GPUs at full capacity. This move reflects Nebius’ ambition to establish a leading role in AI infrastructure, catering to the growing demand from U.S.-based AI developers and enterprises.

Landmark Transition: From Yandex to Nebius

Previously operating as Yandex N.V., often referred to as the "Google of Russia," Nebius emerged after a $5.4 billion divestment deal in July 2024. This deal saw Yandex split its Russian and international assets, with Nebius retaining the international business units. Rebranded and headquartered in Amsterdam, Nebius shifted its focus entirely to AI cloud services and infrastructure, unshackling itself from its Russian origins.

Arkady Volozh, the co-founder of Yandex, now leads Nebius, marking a new era for the company. His return to the helm follows his removal from European sanctions after publicly denouncing Russia's invasion of Ukraine.

Expanding U.S. Presence with Strategic Investments

Nebius is making significant strides to bolster its U.S. footprint. Beyond the Kansas City GPU cluster, the company has opened customer-facing hubs in San Francisco and Dallas, with a third office planned in New York by the end of 2024. This U.S. expansion aligns with Nebius' pledge to invest over $1 billion in AI infrastructure by mid-2025, ensuring the company can meet the increasing demand for AI-native cloud services globally.

CEO Arkady Volozh stated, “Our first GPU cluster and new offices represent a pivotal step in our U.S. expansion. Serving American customers from American facilities means lower latency and maximizes the advantages of our AI-native cloud.”

Technological Advancements Driving Nebius’ Growth

The Kansas City cluster will initially deploy Nvidia H200 Tensor Core GPUs, with Blackwell chips expected later in 2025. This cutting-edge infrastructure is purpose-built to support machine learning lifecycles, including data processing, training, fine-tuning, and inference. Nebius’ full-stack AI infrastructure provides a seamless experience for developers and enterprises, making it a preferred choice for AI innovation.

The launch of Nebius AI Studio, a user-friendly platform offering access to state-of-the-art open-source models, further strengthens its position. The platform boasts one of the lowest price-per-token rates in the market, enhancing its appeal to app builders and developers.

Navigating a Unique Path: From Delisted Entity to Public AI Powerhouse

Nebius’ journey is nothing short of extraordinary. Initially floated in 2011 as Yandex N.V., the company faced immense challenges following Russia's invasion of Ukraine in 2022. Nasdaq halted trading in Yandex shares due to sanctions, and a prolonged restructuring followed. The divestment of Russian assets and the subsequent rebranding as Nebius allowed the company to resume trading in 2024 with a renewed focus on AI infrastructure.

This unprecedented transformation—where a delisted company re-emerged with a new identity and business model—highlights Nebius’ resilience and strategic pivot to align with global opportunities.

Meeting the Soaring Demand for AI Infrastructure

The demand for AI infrastructure is exploding, with Nebius poised to meet this challenge through its robust U.S. expansion strategy. The company is already in advanced talks to establish a second, larger GPU cluster in the U.S., scheduled for deployment in 2025. These developments position Nebius as a key player in the competitive AI infrastructure market, rivaling established giants.

Nebius’ dedication to innovation is underscored by its investment in new data centers in France and the expansion of its Finnish site in Mäntsälä. These moves ensure that the company remains at the forefront of AI advancements, offering scalable and secure solutions for global customers.

Implications for AI Developers and Enterprises

The launch of Nebius’ Kansas City GPU cluster has far-reaching implications for AI developers and enterprises:

Enhanced Performance: The deployment of Nvidia H200 and Blackwell GPUs ensures cutting-edge computational capabilities for advanced AI applications.
Lower Latency: U.S.-based facilities minimize latency, improving efficiency for domestic users.
Scalability: The infrastructure’s design allows seamless scaling to meet future demands.
Cost Efficiency: Platforms like Nebius AI Studio offer developers affordable access to premium tools and models, fostering innovation.

The Future of Nebius: A Bold Vision for AI

Nebius’ strategic pivot from its Yandex roots to a standalone AI infrastructure provider reflects a forward-thinking approach. By leveraging advanced technology and expanding its global footprint, Nebius is well-positioned to capitalize on the booming AI industry.

CEO Arkady Volozh aptly summarized the company’s vision: “We will be building more fully owned GPU clusters across the U.S. to meet exploding demand for high-quality AI infrastructure. Our mission is to empower AI developers globally with unmatched resources and expertise.”

As Nebius continues its journey, its commitment to innovation and scalability ensures its place as a leading force in the AI infrastructure domain. The company’s ability to adapt and grow amidst challenges highlights its resilience and determination to redefine the future of AI.