DevBlacksmith

Tech blog and developer tools

← Back to posts

Nscale Raises $2 Billion in Europe's Largest-Ever Series C to Build AI Infrastructure

Nscale Raises $2 Billion in Europe's Largest-Ever Series C to Build AI Infrastructure

What Happened

On March 9, 2026, UK-based AI infrastructure company Nscale announced a $2 billion Series C funding round, led by Aker ASA and 8090 Industries. The round values Nscale at $14.6 billion, making it Europe's most valuable AI infrastructure startup.

This is the largest Series C in European history.

The investor list reads like a who's who of tech and finance: Nvidia, Dell, Nokia, Lenovo, Citadel, Jane Street, Point72, and Astra Capital Management all participated.

Why This Matters

The AI Compute Bottleneck

The entire AI industry runs on a single constraint: compute. Training and running large language models requires enormous amounts of GPU capacity, and that capacity is concentrated in a handful of hyperscalers — primarily in the United States.

Nscale is building the alternative. The company operates vertically integrated AI infrastructure — from GPU compute and networking to data services and orchestration software — designed to provide AI compute capacity without depending on the major U.S. cloud providers.

European Sovereignty

The geopolitical dimension is impossible to ignore. European governments and enterprises increasingly want AI infrastructure that isn't controlled by American hyperscalers or subject to U.S. export controls and data sovereignty concerns.

Nscale is positioning itself as the answer: European-headquartered, globally distributed AI compute that gives European organizations a non-U.S. option for training and running AI models.

The board appointments reinforce this narrative:

  • Sheryl Sandberg — former Meta COO, brings deep experience in scaling technology operations globally
  • Nick Clegg — former UK Deputy Prime Minister and Meta's President of Global Affairs, brings political and regulatory connections across Europe
  • Susan Decker — former Yahoo President and Berkshire Hathaway board member, brings financial and strategic expertise

The Numbers

Metric Detail
Round size $2 billion
Valuation $14.6 billion
Lead investors Aker ASA, 8090 Industries
Strategic investors Nvidia, Dell, Nokia, Lenovo
Financial investors Citadel, Jane Street, Point72, Astra Capital, Linden Advisors
Target capacity 5+ gigawatts of data center capacity by 2030
Geographic scope Europe, North America, Asia

Five gigawatts of data center capacity is a massive number. For context, a single modern AI training cluster might use 50-100 megawatts. Five gigawatts could support dozens of frontier model training runs simultaneously.

What Nscale Actually Builds

Nscale isn't just renting out GPU servers. The company's value proposition is vertical integration across the AI infrastructure stack:

  1. GPU compute — large-scale clusters of Nvidia GPUs optimized for AI workloads
  2. Networking — high-bandwidth interconnects designed for distributed training across multiple data centers
  3. Data services — storage and data pipeline infrastructure for feeding training data to GPU clusters
  4. Orchestration software — the control plane for managing workloads, scheduling jobs, and optimizing resource utilization

This full-stack approach is significant because AI workloads are different from traditional cloud computing. Training a large model isn't just about having GPUs — it's about having GPUs that can communicate efficiently, storage that can feed data fast enough, and orchestration that keeps everything running at peak utilization.

The Aker Connection

Alongside the funding round, Nscale announced that the Aker Nscale joint venture — originally announced in July 2025 — will be fully rolled into Nscale. Aker ASA is a Norwegian industrial investment company with deep expertise in energy and infrastructure, two areas that are critical for large-scale data center operations.

The Aker connection gives Nscale access to energy procurement expertise at a time when securing reliable, affordable power for data centers is one of the biggest challenges in the industry. Norway's abundant hydroelectric power is a strategic advantage for AI infrastructure — it's renewable, reliable, and relatively cheap.

What This Means for Developers

More Options for AI Compute

If you're training models or running inference at scale, the GPU market has been dominated by the big three cloud providers (AWS, Azure, GCP). Nscale and other AI infrastructure startups are creating alternatives with competitive pricing and, in some cases, better availability.

European Data Residency

For teams building AI applications that need to comply with European data protection regulations, having European-headquartered AI infrastructure simplifies compliance. Running training and inference on EU-based infrastructure eliminates some of the cross-border data transfer concerns that come with U.S.-based cloud providers.

The Infrastructure Race

The $2 billion raise is part of a broader trend: massive capital flowing into AI infrastructure. In the same week, other AI infrastructure companies announced significant raises, and Nvidia invested $2 billion in Amsterdam-based Nebius. The bet across the industry is that demand for AI compute will continue to outstrip supply for years to come.

The Skeptic's View

$14.6 billion is a staggering valuation for an infrastructure company that's still building capacity. The AI infrastructure market is capital-intensive, with long payback periods and significant execution risk. Building and operating data centers at gigawatt scale requires not just capital but also land, power agreements, construction expertise, and the ability to hire and retain specialized engineering talent.

The competitive landscape is also formidable. AWS, Azure, and GCP have decades of experience operating global infrastructure at scale. Oracle, CoreWeave, and Lambda are also building AI-specific compute capacity. Nscale's differentiation — European sovereignty and vertical integration — is real, but it remains to be seen whether it's enough to carve out a durable market position.

That said, if the demand forecasts for AI compute are even directionally correct, there's room for multiple winners. The question isn't whether the market is big enough — it's whether Nscale can execute at the scale and speed its valuation implies.


Sources