Nscale is expanding its AI infrastructure agreement with Microsoft, adding more than 30,000 NVIDIA Rubin GPUs to its Narvik data center campus in Norway. The deployment, scheduled for 2027, will significantly scale compute capacity at the 230 MW site, positioning it among the largest onshore AI infrastructure projects in the region.
The Narvik expansion builds on a broader March 2026 announcement in which Nscale outlined plans to deploy more than 100,000 GPUs across Europe based on the NVIDIA Vera Rubin platform, including systems integrating Rubin and Grace Blackwell architectures. Those deployments, spanning sites in the UK, Norway, and other locations, position Nscale as one of the first providers globally—and the first outside of Microsoft—to roll out the Vera Rubin NVL72 platform at scale for enterprise and hyperscale customers.
Located in northern Norway, the Narvik campus reflects a growing trend toward siting AI infrastructure in regions with access to renewable energy and favorable cooling conditions. The scale of the GPU deployment underscores accelerating demand for AI compute, particularly as hyperscalers and enterprise customers expand deployments of generative AI and advanced machine learning models, supported by tightly integrated compute, networking, and software stacks designed to operate as unified AI supercomputing platforms.
- Adds more than 30,000 NVIDIA Rubin GPUs at the Narvik campus
- Total site capacity reaches 230 MW, among the largest AI infrastructure builds in Norway
- Deployment scheduled for 2027 as part of expanded Microsoft agreement
- Part of a broader rollout exceeding 100,000 GPUs across Europe based on Vera Rubin platform
- Includes Vera Rubin NVL72 systems integrating Rubin and Grace Blackwell architectures
- Nscale assumes full control following Aker Nscale joint venture consolidation
- Targets enterprise and hyperscale AI training, fine-tuning, and inference workloads across multiple geographies
“Customer demand for advanced AI infrastructure continues to accelerate across markets, and our focus is on bringing the latest technology online in the right locations and at real scale. Deployments in Narvik reflect strong demand from customers and Nscale continues to respond with pace and relentless execution,” said Josh Payne, Founder and CEO of Nscale.
🌐 Analysis: The Narvik expansion should be viewed as one component of a much larger European AI infrastructure buildout centered on the Vera Rubin platform. With plans exceeding 100,000 GPUs, Nscale is positioning itself as a regional capacity provider aligned closely with Microsoft and NVIDIA, at a time when hyperscalers are diversifying supply chains and securing long-term access to next-generation AI systems.
🌐 Analysis: The emphasis on NVL72-scale systems and integrated software stacks reflects a broader industry shift toward treating AI clusters as tightly coupled supercomputers rather than loosely connected GPU farms. Comparable strategies are emerging across the ecosystem, including vertically integrated approaches from hyperscalers and infrastructure providers deploying Blackwell-class systems and beyond.
🌐 We’re tracking the latest developments in AI infrastructure and data centers. Follow our ongoing coverage at: https://convergedigest.com/category/data-center/
