Nscale designs, builds, and operates AI data centers across the world, optimized for advanced model training and inference.
Nscale’s AI data centers combine multi-megawatt power, engineered cooling, and dense GPU server systems to meet the demands of advanced AI for enterprises, governments, and mission-critical workloads.
Scale AI sustainably with campuses that prioritize renewable energy, from Arctic hydro to coastal seawater cooling and circular heat reuse.
Add AI capacity faster without the lock-in, using pre-fabricated, chip agnostic infrastructure built for predictable performance.
Run large-scale training and inference designed for dense GPU superclusters with high-capacity power, low-latency connectivity, and advanced cooling.
Discover our global network of data centers engineered for performance, scale, and sustainable growth.
Our Glomfjord AI data center, located in Norway’s Arctic Circle, is powered by 100% renewable energy and uses the latest advancements in AI infrastructure to achieve best-in-class efficiency without sacrificing performance. The site currently delivers 30MW of operational compute capacity, expandable to 60MW, making Glomfjord an efficient, scalable hub for AI workloads.
In partnership with Aker and OpenAI, Narvik is the location of Stargate Norway, an AI gigafactory, powered entirely by renewable hydropower and optimized for low-cost, high-efficiency operation. The data center is designed to deliver 230MW of capacity with plans to expand by an additional 290 MW.
Located at the Start Campus data center in Sines on Portugal’s Atlantic coast, our hyperscale site is designed for GW-scale AI deployments. Its coastal location makes it a low-latency, high-bandwidth hub for European and trans-Atlantic AI traffic, and is powered entirely by renewable energy with seawater cooling.
Ward County is a roughly 240MW AI data center in West Texas, developed in partnership with Ionic Digital for rapid deployment and long-term hyperscale growth, with plans to expand its footprint to 1.2GW. Engineered specifically for AI workloads, Ward County is designed with closed-loop direct-liquid cooling and rear-door heat-exchangers to deliver industry-leading efficiency.
The location of our London-proximate AI campus, Loughton is designed for high-performance, high-density GPU deployments and excellent fibre connectivity to the City of London, with the ability to scale power allocation up to 90MW. The site’s architecture includes direct-liquid cooling and sustainability features such as reuse of waste heat and water.
One of Iceland’s largest liquid-cooled GPU installations, our Keflavik data center is set to deliver 15MW of capacity across Verne’s Icelandic campus. Designed for maximum efficiency and density, the configuration runs on 100% renewable energy from geothermal and hydropower.
