NVIDIA announced a $2 billion strategic investment in AI infrastructure company Nebius aimed at accelerating the deployment of large-scale GPU cloud capacity designed for AI training and inference workloads. The partnership reflects the rapid growth in demand for accelerated computing infrastructure as enterprises and AI developers seek access to large clusters of GPUs optimized for model development.
Nebius said it plans to expand its global data center footprint with infrastructure designed specifically for AI workloads, including large GPU clusters connected through high-performance networking fabrics. The systems will be built on NVIDIA’s accelerated computing platform, including upcoming Rubin GPUs, Vera CPUs, and BlueField data processing units designed to support large-scale AI workloads and data movement between compute nodes.
The collaboration is intended to expand access to AI compute resources while enabling Nebius to deploy large GPU clusters across multiple geographic regions. NVIDIA said the investment supports its broader strategy of working with cloud providers and infrastructure companies to accelerate the deployment of AI computing capacity globally.
Key Points
• NVIDIA makes a $2 billion investment in AI infrastructure company Nebius
• Infrastructure expected to deploy Rubin GPUs, Vera CPUs, and BlueField DPUs
• Nebius expanding global GPU cloud infrastructure for AI workloads
• Partnership aimed at accelerating deployment of large AI compute clusters
• Infrastructure designed for model training and large-scale inference workloads
“AI is transforming every industry, and that transformation requires a new class of computing infrastructure,” said Jensen Huang, founder and CEO of NVIDIA. “By working with partners like Nebius, we are helping expand access to accelerated computing platforms that allow developers and enterprises to build and deploy the next generation of AI applications.”
🌐 Analysis
The announcement reflects NVIDIA’s continued effort to expand the global ecosystem of GPU cloud providers capable of deploying large-scale AI infrastructure. While hyperscale cloud providers remain the dominant operators of AI infrastructure, a growing number of independent cloud platforms are emerging to provide access to specialized GPU clusters designed specifically for AI workloads.
Nebius Group is an AI-focused cloud infrastructure company headquartered in Amsterdam, Netherlands, formed from the 2024 restructuring of the international assets of Yandex following the divestiture of its Russian operations. Led by founder and CEO Arkady Volozh, the company is building a hyperscale cloud platform optimized for artificial intelligence workloads, providing high-performance GPU clusters, cloud services, and developer tools for training and deploying large AI models. Nebius’ platform integrates advanced AI infrastructure—including NVIDIA GPU-based compute clusters, high-speed networking, and scalable storage—targeted at AI startups, research institutions, and enterprise developers. The company operates major data-center infrastructure in Europe and is expanding globally to support large-scale AI training workloads. Nebius is publicly traded on Nasdaq and has attracted strategic partnerships across the AI ecosystem, positioning itself as a European-based alternative to U.S. hyperscale cloud providers.
