Josh Payne’s Nscale is moving fast enough to make even its own backers sweat. Since the start of 2026, the company has taken on a $1.4 billion GPU-backed loan and a $790 million financing facility in Norway. And in the only public accounts on record, covering the early stages from seven months to December 2024, Nscale showed a $24 million loss. That combination is the tension at the center of Europe’s AI infrastructure scramble: big capital commitments, big expectations, and early financial reality that does not magically keep up with hype.

To understand why that matters, start with what Nscale is trying to be. Spun out of Arkon Energy in 2024, the company now builds and operates computer centers, then rents processing power to firms that need to train and run large AI models. In Europe, where the U.S. had three-quarters of the world’s AI computing power as of May 2025 (with China around 15%) and the EU held less than 5%, “compute” is not a back-office detail. It is the difference between building frontier models locally and importing them at scale.

Nscale’s strategy is essentially “infrastructure, but with sovereignty and scale baked in.” Like other “neoclouds” (a label used for providers that build or finance AI-ready data centers and rent computing power), it pitches long-term access to cheap renewable energy. Nscale builds its own data centers and aims to own the hardware and software stack on top. But it also follows the industry’s capital-intensive pattern, piling on debt to fund the build-out. The bet is that cheaper power and faster deployment will outrun the cost of financing, even as Europe’s electrical grid becomes the bottleneck.

Europe’s grid constraints are not theoretical. Grid connection queues in the continent’s most sought-after markets stretch seven to 10 years on average, rising to 13 in the most congested cities. Ireland has imposed strict rules on new data center connections in Dublin. The Netherlands and Frankfurt have effectively banned new connections until at least 2030. Researchers have warned that without reform, Europe’s AI ambitions could end up as costly stranded assets. Nscale’s play is to respond by distributing compute, rather than piling everything into the familiar hubs like Frankfurt, Amsterdam, and Dublin. In Northern Norway, power is cheaper and fully renewable, and Nscale leans into that by building large AI campuses in places like Norway while also maintaining smaller national sites for sensitive or low-latency workloads.

This is where the company’s governance and ecosystem reach becomes more than résumé theater. Nscale’s board pitch matters because scaling AI infrastructure is part engineering, part politics, and part partnership with the chip supply chain. When Josh Payne recruited Sheryl Sandberg, it was a letter to his management team that convinced her. Sandberg tells Fortune she had only ever seen writing like that from Mark Zuckerberg: “so clear, so thorough, but so big in its vision.” Payne also brought in former UK Deputy Prime Minister Nick Clegg and former Yahoo president Susan Decker, alongside the strategic credibility needed to recruit customers and align investors with long time horizons.

But the biggest interdependency is Nvidia. Nscale has raised almost $4 billion from companies including Dell, Nokia, and Nvidia, and it has signed sweeping contracts with several tech giants. Nvidia has also been both investor and strategic partner. Nvidia CEO Jensen Huang has been explicit about the relationship in a recent podcast, saying, “Neoclouds wouldn’t exist” without Nvidia’s support. Company accounts show Nvidia agreed to guarantee up to $860.3 million of Nscale’s obligations under a lease for a Texas facility, in exchange for warrants to acquire a stake in the company, where Nscale is deploying Nvidia chips for Microsoft. For Nscale, that kind of backstop can accelerate scaling. For Europe’s AI builders, it underscores a second-order truth: even “sovereign” compute stacks are still tied to the leverage of the supply chain that makes the chips.

Nscale’s momentum also includes prominent deals like Stargate, OpenAI’s global infrastructure push unveiled in January 2025. Nscale was part of the UK’s Stargate plans, including 8,000 GPUs for OpenAI in an initial phase and potential expansion to 31,000 GPUs. In Norway, OpenAI said it had found the right conditions to bring Stargate to Europe through the Narvik project, positioning Nscale within a broader, half-trillion-dollar narrative of global AI capacity. That kind of customer anchoring helps explain the company’s headline valuation, but it also raises the stakes of what comes next: can demand and execution keep pace with the debt-fueled build?

Nscale’s early finance says “prove it.” Its early accounts show a $24 million loss through December 2024, and it is still fueling construction through large financing moves that do not care about valuations. Payne argues, in the context of negative commentators, that “every dollar is backed by real revenues.” Whether that is enough will likely come down to how quickly Europe’s compute needs translate into paid utilization, especially when energy and grid access are the limiting factors rather than AI ambition.

For executives and boards across the neocloud and infrastructure wave, Nscale is a live stress test. If Europe can translate renewable power advantages, federated workload deployment, and chipmaker partnership into durable economics, it strengthens the case for more capital into the continent’s AI capacity. If not, the same queues and regulations that shape this market could turn fast-growing plans into stranded assets. Nscale is betting it can outrun the clock. Its next milestone will determine whether that bet becomes a blueprint or a cautionary tale.