Nvidia Commits $2B to Nebius, Targets 5GW AI Cloud by 2030
Nvidia has agreed to invest $2 billion in Nebius Group to co-develop hyperscale AI cloud infrastructure and expand capacity to over 5 gigawatts by 2030. The partnership covers AI infrastructure design, deployment of Rubin GPUs, Vera CPUs and BlueField storage, and joint AI inference and agentic AI development.
1. Strategic Partnership and Investment
Nvidia and Nebius Group have entered a strategic agreement under which Nvidia will invest $2 billion to jointly develop next-generation AI cloud platforms. The deal positions Nebius to leverage Nvidia’s capital and expertise to accelerate hyperscale infrastructure build-out for AI workloads.
2. Infrastructure Expansion Plans
Nebius plans to scale its global AI cloud capacity to more than 5 gigawatts by the end of 2030, building multiple gigawatt-scale data centers across the United States. The expansion aims to meet rising demand for agentic and large-scale AI model training and inference.
3. Technology Deployment and Collaboration
The collaboration will integrate Nvidia’s future accelerated computing architectures, including the Rubin GPU platform, Vera CPUs and BlueField storage systems, into Nebius’s facilities. Joint efforts will also focus on AI factory design, system software integration, AI inference capabilities and fleet management tools to optimize performance and reliability.