Nvidia to invest $2b in Nebius for hyperscale AI cloud

Nvidia will invest US$2 billion in Nebius to jointly develop and deploy a hyperscale AI cloud for enterprises and AI-first builders.

Nebius is an Amsterdam-based AI cloud firm listed on Nasdaq that has been deploying Nvidia infrastructure, including multiple gigawatt-scale AI factories in the US.

The firms said the partnership covers AI factory design and support and the creation of an inference and agentic AI software stack for developers and enterprises.

They also said it includes fleet management, including GPU health monitoring and software recommendations, and the early adoption of Nvidia computing architectures including the Rubin platform, Vera CPUs, and BlueField storage systems.

The companies said the partnership and Nvidia’s support will enable Nebius to deploy more than five gigawatts of capacity by the end of 2030.

🔗 Source: Nvidia

🧠 Food for thought

Implications, context, and why it matters.

Nvidia’s investment helps Nebius serve existing hyperscaler clients

The deal supports Nebius as it works through large contracts with AI buyers, including a five-year $3 billion deal with Meta and a $17.4 billion agreement with Microsoft 1.
Nebius is building out the infrastructure for that demand, including a planned 1.2-gigawatt AI data center campus in Missouri 2.
Growth has been fast, Q4 revenue rose 547% year over year to $227.7 million 2.
The funding builds on earlier ties, NVIDIA previously co-led a $700 million investment in Nebius 3.

Nvidia is building a new generation of cloud providers to secure its dominance

The backing fits NVIDIA’s plan to finance specialized neocloud operators, cloud providers built mainly for AI workloads, that take on traditional cloud players 3.
It follows a similar $2 billion stake in Nebius rival CoreWeave, shaping cloud firms that run on NVIDIA’s technology stack 3.
The approach helps NVIDIA push back as Google and Amazon develop their own AI accelerator chips 3.
Funding operators that can lock in gigawatts of power targets a choke point for AI rollout, grid interconnection and steady energy supply 2.