NVIDIA Corp. announced a $2 billion investment in Nebius Group NV, a sizable expansion of its strategy to bankroll the infrastructure providers that buy its chips.
The deal, announced Wednesday, sent Nebius shares surging 10% as the two companies enter a strategic partnership to build hyperscale artificial intelligence (AI) factories.
The collaboration is to develop a full-stack AI cloud platform, moving beyond simple hardware sales to include integrated software and architecture design. Under the agreement, Nebius plans to deploy 5 gigawatts of capacity by 2030 — an energy footprint comparable to powering nearly four million homes.
As the race to build the physical backbone of the AI economy intensifies, NVIDIA’s latest move suggests the company is no longer content just being a supplier: It is now a primary architect and financier of the cloud itself.
Nebius occupies a unique position in the neocloud sector, a new breed of providers that eschew traditional general-purpose computing to build data centers exclusively for AI. The company emerged from the remnants of the Russian internet giant Yandex; following a 2024 restructuring triggered by the invasion of Ukraine, the firm rebranded and pivoted entirely to Western-based AI infrastructure.
The $2 billion commitment from NVIDIA is a major escalation from its previous $33 million stake. It grants Nebius early access to NVIDIA’s next-generation Rubin platform, Vera CPUs, and BlueField storage systems.
“AI is at another inflection point — agentic AI,” said NVIDIA CEO Jensen Huang, referring to autonomous AI systems that require massive compute power. “Nebius is building an AI cloud designed for the agentic era.”
Although the deal cements NVIDIA’s dominance, it also invites fresh scrutiny over its circular investment model. By providing capital to its own customers — a pattern seen in recent billion-dollar deals with CoreWeave, OpenAI, and Anthropic — NVIDIA ensures a steady pipeline for its high-margin GPUs.
Critics argue this practice could potentially inflate market demand and valuations. However, for Nebius CEO Arkady Volozh, the partnership is a validation of the company’s AI-first architecture. “We are building one of the first and largest clouds for all AI builders everywhere,” Volozh said.
The partnership will focus heavily on inference, the process of running AI models for end-users, which is expected to be the next major frontier of commercial demand. Beyond hardware, the companies will collaborate on specialized blueprints for gigawatt-scale facilities, real-time GPU health monitoring and software optimization, and optimizing NVIDIA’s latest libraries for Nebius’s global platform.
SHARE THIS STORY
Originally published by Techstrong.IT. Republished with attribution.
