- Led by Norwegian industrial group Aker, the funding values Nscale at about $3 billion and includes Nvidia’s pledge to supply up to 300,000 AI chips.
- Nscale aims to position itself as a “sovereign” AI infrastructure provider with local data-centre capacity, but some critics highlight its Australian ownership and question how “British” it truly is.
What happened: Record-breaking raise, global backers — but questions over Nscale’s British identity
Nscale, a UK-based AI infrastructure firm, has closed a $1.1 billion Series B round — the largest such round in British and European history. The round was led by Aker ASA, which invested approximately $285 million in a mix of cash and land, acquiring a 9.3 per cent stake in the company. Among the other participants are Nvidia, Dell, Fidelity, Nokia, and G Squared.
Nvidia has committed up to $500 million as part of this broader backing, and in return anticipates supplying as many as 300,000 GPU chips to Nscale over time. Part of the financing is tied to infrastructure commitments such as a joint-venture with Aker in Narvik, Norway, which already involves a $6.2 billion contract with Microsoft for access to data-centre capacity.
Nscale also aims to back projects in the UK under the “Stargate UK” framework (in collaboration with OpenAI and Nvidia) and to build large scale AI campuses, including a supercomputer in Loughton. Its technology stack spans bare-metal and managed clusters, serverless inference, and AI fine-tuning services.
However, some commentators note that although Nscale is headquartered in the UK, it was spun out from Australia’s Arkon Energy — which according to public filings remains its sole shareholder. This raises questions about how “British” the company truly is in terms of ownership.
Also read: Meta widens Llama AI access to US allies
Also read: Ciena acquires Nubis to power AI data centres
Why it’s important
The AI boom has accelerated demand for massive compute and GPU capacity, and Nscale’s raise underscores a shift: infrastructure is now central to the AI race, not just models and algorithms. By securing this capital, Nscale is better positioned to deploy data centres closer to enterprise and governmental clients, helping manage latency, regulatory and data sovereignty constraints.
From a strategic viewpoint, backing from Nvidia aligns the chipmaker not only as a vendor but as a patron of its own downstream ecosystem; by supporting infrastructure providers, Nvidia helps cultivate future demand for its hardware.
Yet the narrative of “sovereign AI infrastructure” carries implicit risks. Building and operating AI-scale data centres is capital-intensive, energy-hungry and complex. Meeting promises of energy efficiency, adherence to local regulation, and secure operations is materially harder than pitching them.
And on the question of national allegiance: should government contracts or incentives flow to firms labelled “British” if their ownership, control or origin lies elsewhere? The tension between branding and substance may become a live issue as the UK (and other nations) seeks to anchor AI development domestically.
