SWI Compute launches with plans for NVIDIA AI factories in Europe

The first large-scale facility will be located in Cambridge, UK, starting with 30 MW of capacity, close to 7,000 Rubin GPUs, and 3,500 Vera CPUs. This initial phase is expected to come online by the end of 2027. The UK campus is planned to expand to 530 MW, serving as the first step toward a target of 2.3 GW of NVIDIA AI factory capacity by 2029. All sites will follow NVIDIA reference architectures, with future deployments using the full-stack NVIDIA DSX AI Factory Platform.

SWI Compute aims to serve European AI natives, labs, and enterprises that have struggled to secure capacity previously dominated by global hyperscalers. The company’s first customer is a US-based AI inference platform. SWI Group, formed from the merger of Icona and Stoneweg, allocates over 80% of its capital to digital infrastructure, positioning SWI Compute as the final component of its vertically integrated stack, which also includes AiOnX and SWI Digital.