NVIDIA-Backed Nscale Raises $2 Billion — Europe’s Largest Series C Ever

A UK-based AI infrastructure startup just closed the biggest Series C in European history. Nscale, the company behind Stargate Norway and a massive Microsoft GPU deal, pulled in $2 billion at a $14.6 billion valuation. The investor list reads like a tech industry wishlist: NVIDIA, Dell, Lenovo, Nokia, Citadel, Jane Street, and Point72. And to top it off, former Meta COO Sheryl Sandberg and former UK Deputy Prime Minister Nick Clegg are joining the board.

This is not a typical AI startup story. Nscale is not building chatbots or fine-tuning language models. It is building the physical backbone that makes all of that possible — GPU-dense data centers designed specifically for AI training and inference at scale.

From Crypto Infrastructure to AI Powerhouse

Nscale’s origin story is unusual. Founder and CEO Josh Payne, an Australian entrepreneur and former coal miner, started in cryptocurrency mining infrastructure through a company called Arkon Energy. Around 2024, the company pivoted to AI, acquired Kontena (a high-performance computing data center firm founded in 2018), and relaunched as Nscale.

The pivot proved prescient. Since emerging from stealth in May 2024, demand for Nscale’s AI infrastructure has been, by all accounts, overwhelming. The company raised a $155 million Series A, then quickly moved to larger rounds as hyperscalers and AI companies scrambled for GPU capacity.

The $2 billion Series C, led by Norwegian conglomerate Aker ASA and 8090 Industries, technically includes a $433 million pre-Series C SAFE from October backed by Blue Owl, Dell, NVIDIA, and Nokia. Even accounting for that, the numbers are staggering for a company barely two years old.

What Nscale Actually Builds

Nscale develops vertically integrated AI infrastructure — combining GPU computing, networking, data services, and orchestration software into modular, high-density data center packages. The company currently operates facilities in the UK, Norway, Portugal, Iceland, and the United States.

The key differentiator is density and specialization. Unlike traditional cloud providers that bolt AI capacity onto general-purpose infrastructure, Nscale designs everything from the ground up for large-scale AI workloads. Alex Sharp, president of Nscale’s Data Center Group and a chartered engineer with 30+ years of experience, has led the design and construction of 54 data centers, including some of Europe’s largest.

Three major deals highlight the scale of operations:

Stargate Norway: A joint project with OpenAI and Aker to build a 100,000 NVIDIA GPU “AI Gigafactory” in Kvandal, near Narvik in Northern Norway. The facility runs entirely on renewable hydropower. Both parties committed roughly $1 billion to the initial 20MW phase, with a target of full deployment by end of 2026. This is OpenAI’s first data center initiative in Europe.
Microsoft partnership: Nscale signed an expanded deal to deliver approximately 200,000 NVIDIA GB300 GPUs across three European data centers and one US facility, in collaboration with Dell.
Aker consolidation: The Aker-Nscale joint venture announced in July 2025 is being rolled fully into Nscale, consolidating delivery and governance under one entity. Aker CEO Øyvind Eriksen continues on the Nscale board.

Why Sandberg, Clegg, and Decker Joined the Board

The board additions tell a story about where Nscale is headed. Sheryl Sandberg brings operational scaling expertise from her years as Meta’s COO. Nick Clegg, who served as Meta’s President of Global Affairs and previously as UK Deputy Prime Minister, brings policy and regulatory navigation skills — critical for a company building infrastructure across multiple jurisdictions. Susan Decker, former Yahoo President, adds governance and media industry experience.

These are not advisory roles. For a company that needs to navigate European data sovereignty regulations, secure government partnerships, and manage relationships with the world’s largest tech companies simultaneously, this board composition signals serious ambitions beyond just building data centers.

How Nscale Stacks Up Against the Competition

The “neocloud” market — specialized GPU cloud providers built for AI workloads — has gotten crowded. Here is how Nscale compares to the major players:

CoreWeave is the most direct US competitor, developing a $6 billion data center campus in Pennsylvania and expanding into the UK with a $3.4 billion investment. CoreWeave went public and has deep ties to NVIDIA. But its focus remains primarily on GPU-as-a-Service rather than vertically integrated infrastructure.

Lambda Labs targets ML researchers specifically, offering pre-configured GPU instances with PyTorch, TensorFlow, and CUDA ready to go. It is a strong developer tool but operates at a smaller infrastructure scale than Nscale.

Nebius (spun out from Yandex) and Crusoe Cloud (with sites in Iceland and the Nordic region) compete on the European and sustainability fronts, but neither has matched Nscale’s funding or its direct partnerships with OpenAI and Microsoft.

Traditional hyperscalers (AWS, Azure, GCP) have massive GPU capacity but lack the specialization and density optimization that Nscale offers. Ironically, Microsoft is both a hyperscaler and an Nscale customer — a dynamic that shows how demand for AI compute still outstrips even the biggest providers’ capacity.

Nscale’s edge is the combination of European sovereignty positioning, renewable energy access, direct NVIDIA backing, and partnerships with both OpenAI and Microsoft. No other neocloud has that particular constellation of advantages.

What This Means for Europe’s AI Ambitions

The $2 billion raise is significant beyond Nscale itself. Europe has long worried about falling behind the US and China in AI infrastructure. Most major AI training runs happen on American soil, using American cloud providers. Nscale represents a credible European alternative — one that can offer data sovereignty, renewable energy, and regulatory alignment that US-based providers cannot easily match.

The Stargate Norway project is particularly symbolic. Powered entirely by Norwegian hydropower, it directly addresses concerns about AI’s environmental footprint while keeping compute capacity within European borders. For European companies and governments with data residency requirements, this matters.

The funding also signals that investors see the AI infrastructure shortage lasting well into the late 2020s. Despite massive buildouts by AWS, Google, and Microsoft, GPU demand from AI training and inference continues to outpace supply. Companies like Nscale are betting that this gap creates room for specialized providers to capture significant market share.

FAQ

How much funding has Nscale raised in total?
Nscale raised $2 billion in its Series C round (which includes a $433 million pre-Series C SAFE from October). This follows a $155 million Series A. The Series C values the company at $14.6 billion, making it Europe’s most valuable AI infrastructure startup.

Who are Nscale’s main investors?
The Series C was led by Aker ASA and 8090 Industries, with participation from NVIDIA, Dell, Lenovo, Nokia, Astra Capital Management, Citadel, Jane Street, Linden Advisors, Point72, and Blue Owl.

What is Stargate Norway?
Stargate Norway is a joint project between Nscale, Aker, and OpenAI to build a 100,000 NVIDIA GPU data center in Kvandal, Northern Norway. Powered by renewable hydropower, it is OpenAI’s first European data center initiative, with initial deployment targeting end of 2026.

How does Nscale differ from CoreWeave?
Both are GPU-focused neocloud providers, but Nscale takes a more vertically integrated approach — designing data centers from the ground up for AI workloads rather than primarily offering GPU-as-a-Service. Nscale also has a stronger European presence and direct partnerships with both OpenAI and Microsoft.

Is Nscale publicly traded?
No. Nscale is a private company. With a $14.6 billion valuation and high-profile board members, an IPO is a possibility in the future, but the company has not announced any plans to go public.

Top AI Product