Nebius has launched its first UK-based deployment of NVIDIA Blackwell Ultra AI infrastructure, marking one of the country’s earliest large-scale rollouts of next-generation GPU technology. The move brings hyperscale AI compute capabilities – previously the domain of global cloud giants – directly to British enterprises, researchers, and public institutions seeking to build and deploy advanced artificial intelligence models domestically.
The new installation, powered by NVIDIA Blackwell Ultra GPUs and NVIDIA Quantum-X800 InfiniBand networking, delivers cutting-edge performance for training and scaling generative AI systems and large foundation models. Located in the London area, the site integrates energy-efficient cooling and resilient power systems, emphasizing both sustainability and compute density. Nebius said the facility’s infrastructure will support “the next wave of AI innovation across healthcare, life sciences, financial services, robotics, and public sector applications.”
The launch expands the UK’s sovereign compute capacity at a time when national and regional governments are increasingly concerned about AI infrastructure independence and data security. It also aligns with the UK Government’s AI Opportunities Action Plan, which calls for broader domestic access to advanced computing power as a foundation for AI-driven innovation.
NVIDIA’s Anthony Hills, Director of Enterprise Business for the UK and Ireland, said the collaboration represents a major boost to the country’s AI readiness. “The UK’s industrial future is being reshaped by AI,” Hills said. “With Nebius deploying advanced AI infrastructure built on the NVIDIA accelerated computing platform, we’re helping unlock the potential of British data to drive innovation and economic growth.”
Nebius’s deployment gives UK organizations direct access to enterprise-grade, AI-optimized cloud computing that competes with global hyperscalers, but with local data control and regulatory compliance. The Nebius AI Cloud platform – now in its third iteration, known as Aether – includes SOC2 Type II, HIPAA, GDPR, and CCPA compliance, making it suitable for healthcare and other regulated sectors handling sensitive data.
Aether’s integration with the NVIDIA AI Enterprise software suite enables end-to-end model training and inference at scale. It also supports advanced orchestration and automation features for developers building AI applications using large language models (LLMs), multimodal systems, and scientific simulations.
The company said its Reference Platform NVIDIA Cloud Partner (NCP) status ensures performance parity with NVIDIA’s reference architecture and close technical collaboration for performance validation and workload optimization. Nebius was among the first partners to achieve “Exemplar Cloud” designation for achieving training efficiency within 95% of NVIDIA’s benchmarks on Hopper GPUs.
Nebius’s early UK customers highlight the platform’s potential to accelerate innovation in data-intensive fields. Basecamp Research, which applies AI to molecular biology and programmable medicine, said the infrastructure allows it to handle massive proprietary datasets without compromising speed or data sovereignty. “Having this performance here in the UK means we keep greater control over our data and can compete at the cutting edge of AI-driven life sciences,” said Basecamp co-founder Oliver Vince.
Similarly, Prima Mente, a neuroscience startup developing AI models to understand and treat brain diseases such as Alzheimer’s and Parkinson’s, has adopted Nebius’s infrastructure for large-scale model training. “Training multi-billion parameter foundation models on biological data demands infrastructure that can scale rapidly without breaking our workflow,” said Ravi Solanki, Prima Mente’s CEO. “Having this capacity locally means we can accelerate diagnostics for patients across the UK.”
Industry analysts note that Nebius’s move strengthens the UK’s positioning as an emerging AI infrastructure hub, alongside major data center expansions from hyperscalers like AWS, Microsoft, and Oracle. However, Nebius differentiates itself by offering AI-native cloud services optimized for high-performance training workloads rather than general-purpose compute.
For Nebius, the UK launch marks another milestone in its global AI infrastructure rollout, following similar deployments in Europe, the US, and Israel. The company’s strategy centers on building regional AI supercomputing clusters that combine hyperscaler simplicity with localized sovereignty and compliance controls – an increasingly attractive model as AI regulation tightens worldwide.
As enterprises shift from experimentation to full-scale deployment of generative and agentic AI systems, compute demand continues to rise exponentially. With NVIDIA’s latest architecture at its core, Nebius aims to deliver the performance, sustainability, and control needed for organizations to build and run AI safely and at scale within their own jurisdictions.
