AI Data Centre Firm Nscale Plans GPU Network Expansion
The race to build infrastructure for AI is intensifying as technology companies compete for access to graphics processing units (GPUs) and data centre capacity. The global shortage of these computing resources has created opportunities for new entrants in the market – particularly those able to combine data centre operations with GPU deployment and cloud services.
Against this backdrop, Nscale, a vertically-integrated AI infrastructure provider that develops and operates data centres and GPU clusters, has secured US$155m in Series A funding to expand its operations across Europe and North America. The funding round was led by Sandton Capital Partners, with participation from Kestrel, Bluesky Asset Management and Florence Capital.
Nscale’s Series A positions it for AI compute expansion
The company, which launched from stealth in May 2024 and owns its data centres and develops both the infrastructure for graphics processing units (GPUs) and the software stack provided to customers, will use the capital to accelerate the deployment of large-scale GPU clusters. These clusters are used for AI model training – the process of developing AI systems using large datasets – as well as fine-tuning and inferencing, where AI models process new data to generate outputs.
The funding will support Nscale’s expansion of its data centre pipeline, which has grown from 300 megawatts to 1.3 gigawatts since the company emerged from stealth in May 2024. Nscale plans to develop 120 megawatts of capacity in 2025 across sites in Ohio, Texas, the UK and Norway.
Full-stack approach drives AI infrastructure development
The company has developed a full-stack AI private cloud platform designed to support the entire generative AI lifecycle. This platform enables developers to run and train AI models through a combination of bare metal and virtualised GPU nodes. The system uses Kubernetes-native services (NKS) for container orchestration and management, alongside AI workload scheduling powered by SLURM, an open-source workload manager.
The company says it is preparing to launch a public cloud service in the first quarter of 2025, which will provide AI developers with access to training and inferencing capabilities in a flexible development environment. The platform will complement Nscale’s existing private cloud infrastructure, offering developers additional options for AI development and deployment.
Joshua Payne, CEO of Nscale and co-founder of the US$357m NASDAQ-listed battery metals SPAC Battery Future Acquisition Corp, brings significant entrepreneurial experience from capital markets, crypto mining, and renewable energy M&A to the venture. Joshua says: “Nscale manages every layer of infrastructure in the value chain to meet the intensive needs of large-scale AI customers. The largest risk to the market's ability to scale is the large contiguous tranches of electricity required to power these large GPU superclusters. Nscale has a 1.3GW pipeline of sites in our portfolio, which allows us to design from the ground up, the data centre, the supercluster and the cloud environment end-to-end for our customers.”
Sustainable infrastructure powers AI development
The company’s data centres are designed to support large supercluster deployments for global hyperscalers and large language model platforms. The facilities use closed-loop direct liquid cooling to manage the heat generated by intensive computing operations, minimising environmental impact while maximising performance and efficiency.
The company currently operates a 60-megawatt renewable energy-powered data centre in Norway, demonstrating its commitment to sustainable AI infrastructure development. This facility forms part of Nscale's strategy to build environmentally conscious computing capacity across its expanding network of data centres.
Rael Nurick, Co-Founder at Sandton Capital Partners, says: "With a notable founder track record, established industry partnerships, and a unique vertically integrated approach, Nscale is building the hyperscale AI platform to power enterprise AI at scale."
The Series A round follows Nscale’s US$30m seed funding in December 2023. The company has formed a partnership with Open Innovation AI, a GPU orchestration platform in the Middle East and North Africa region, to deploy 30,000 GPUs over three years. Nscale’s Svartisen Cluster, one of its computing systems, was included in the 2024 Top500 list of global super-computing systems.
Joshua says: “This means we can deploy bespoke GPU clusters tailored to customers' requirements, do this faster than competitors and at any scale, all while delivering superior unit economics.”
Explore the latest edition of Data Centre Magazine and be part of the conversation at our global conference series, Tech & AI LIVE and Data Centre LIVE.
Discover all our upcoming events and secure your tickets today.
Data Centre Magazine is a BizClik brand
