NVIDIA Pumps $2 Bn in Nebius to Scale Hyperscale AI Cloud Infra

NVIDIA has partnered with AI cloud provider Nebius, which will see a $2 billion investment to build next-generation hyperscale infrastructure for artificial intelligence workloads.

The collaboration will focus on developing large-scale AI cloud platforms designed to support the rapidly growing demand for AI training, inference, and agentic AI applications.

As part of the partnership, Nebius plans to deploy more than 5 GW of AI computing capacity by 2030, powered by NVIDIA’s accelerated computing technologies.

The companies will work together across the entire AI stack—from data center architecture and infrastructure deployment to inference software and system optimisation. 

The partnership also includes early access to NVIDIA’s latest computing platforms. These include the upcoming NVIDIA Rubin platform, NVIDIA Vera CPU processors, and NVIDIA BlueField data processing units, which will help Nebius scale its AI cloud infrastructure globally.

The collaboration will also involve joint work on AI factory design, GPU fleet management, and optimised inference stacks for developers and enterprise customers building AI-driven applications.

Jensen Huang, Founder and CEO of NVIDIA, said the industry is entering a new phase of AI development driven by autonomous software systems. “AI is at another inflexion point—agentic AI is driving incredible compute demand and accelerating infrastructure buildout,” Huang said in a statement.

Nebius CEO Arkady Volozh said the company’s cloud platform was built specifically for AI workloads rather than adapted from traditional cloud infrastructure.

“Now with NVIDIA, we are extending that throughout the stack—from gigawatt-scale AI factories to inference and software,” he added.

Nebius currently operates a rapidly expanding global infrastructure footprint, including multiple AI factory deployments in the United States.