Microsoft’s Massive GPU Land Grab
Nebius Group (NASDAQ: NBIS) has secured a multi-billion-dollar AI infrastructure contract with Microsoft, fundamentally validating the “neocloud” model of specialized artificial intelligence compute. The five-year agreement guarantees Nebius a minimum of $17.4 billion, scaling up to $19.4 billion if Microsoft exercises additional capacity options.
The deal provides Microsoft with exclusive access to over 100,000 Nvidia GPUs at Nebius’s upcoming data center in Vineland, New Jersey. This deployment underscores the massive capital expenditure tech giants require to train large language models (LLMs) and run advanced generative AI applications.
“The deal’s economics are compelling, but more importantly, it accelerates the expansion of our AI cloud business,” Nebius CEO Arkady Volozh stated, hinting that Microsoft is just the first in a pipeline of planned hyperscaler partnerships.
Markets reacted decisively. Following the September announcement, Nebius shares surged over 50% and have maintained momentum, trading near their 52-week high of $141.10.
From Geopolitical Turmoil to Nasdaq Heavyweight
Nebius’s current trajectory follows a complex structural evolution. The company originally operated under Yandex N.V., often called the “Google of Russia.” After Russia invaded Ukraine in 2022, Nasdaq suspended Yandex trading, and Volozh faced European Union sanctions.
A $5.4 billion divestiture of Yandex’s Russian assets in July 2024 allowed Nebius Group to emerge as an independent entity. Retaining international assets, the newly formed company focused on AI cloud infrastructure, autonomous driving technology (Avride), and edtech (TripleTen).
The EU lifted sanctions against Volozh in March 2024. By October, Nebius resumed trading on the Nasdaq under the ticker NBIS, successfully navigating early volatility to establish a strong upward trend.
Capitalizing on the Global GPU Squeeze
Nebius’s aggressive infrastructure pivot aligns perfectly with the ongoing global shortage of high-performance computing resources. In early 2025, Nvidia allocated nearly 60% of its chip production to enterprise AI clients. Supply chain bottlenecks, exacerbated by earthquake damage to TSMC wafers, have kept specialized GPUs trading significantly above retail pricing.
Nebius reported staggering financial growth amid this squeeze. Second-quarter 2025 revenue hit $105.1 million—a 625% year-over-year increase—driven primarily by its AI cloud infrastructure segment growing more than ninefold. The company is racing to bring 220 megawatts of power online by late 2025, targeting over 1 gigawatt by 2026.
Financing the Neocloud Model
Rather than draining its balance sheet, Nebius is leveraging the Microsoft contract to secure favorable debt financing, utilizing Microsoft’s AAA credit rating. This follows a $700 million funding round in December 2024 featuring Accel Partners and Nvidia, which acquired a 0.5% stake.
The company’s core business achieved positive EBITDA ahead of projections in Q2 2025—a rare feat for capital-heavy AI infrastructure startups. Full-year 2025 revenue guidance is set between $500 million and $700 million.
Navigating a Hyper-Competitive Landscape
While legacy providers like AWS, Microsoft Azure, and Google Cloud dominate general enterprise compute, specialized “neoclouds” like Nebius and CoreWeave are capturing market share by offering immediate GPU availability, tailored pricing, and vertically integrated hardware-software stacks.
Nebius is rapidly expanding its geographic footprint:
- Vineland, New Jersey: A 300MW flagship facility custom-built for the Microsoft contract.
- Mäntsälä, Finland: A primary European data hub.
- Paris, France: A GPU cluster within an Equinix facility.
- Kansas City, Missouri: A dedicated GPU cluster currently under construction.
Execution Risks vs. Industry Upside
Despite the bullish outlook, analysts caution about execution risks. Building a 300MW facility and deploying 100,000 GPUs on schedule requires flawless logistics. Furthermore, as semiconductor supply chains eventually normalize, the current premium pricing for AI compute capacity may compress.
Yet Wall Street remains optimistic, with analysts raising price targets to $125 and fair value estimates reaching $153. By securing a foundational contract with Microsoft, Nebius has cemented its status as a critical supplier in a global AI infrastructure market that Nvidia projects will hit $3 trillion by 2030.
Related Technology Intelligence & Hardware Benchmarks
- NVIDIA Blackwell Enterprise Servers Claim 45x AI Performance Gains for On-Premises Infrastructure
- SingularityNET’s 3M Decentralized Supercomputer: Breaking the AGI Hardware Bottleneck
- TruthUpfront Enterprise Economics & Labor Realignment Intelligence
Technical Standards & Empirical Frameworks
- NIST Cybersecurity & AI Risk Management Framework (AI RMF): https://www.nist.gov/itl/ai-risk-management-framework
- IEEE Computer Society & ACM Digital Computing Standards: https://www.ieee.org/















