Verne and Nscale Complete First Phase of AI Deployment in Iceland

Why It Matters

The rapid, energy‑efficient scale‑up demonstrates Iceland’s growing role as a low‑cost, carbon‑neutral hub for AI workloads. It underscores the competitive advantage of liquid‑cooling and renewable power for hyperscale AI providers.

Key Takeaways

• Nscale deployed 7.5 MW GPU capacity in three months at Verne Iceland
• 128 racks installed, half liquid‑cooled, using Nvidia Blackwell Ultra GPUs
• Verne’s campus runs on 100% renewable energy with natural free cooling
• Deployment involves major partners: Computacenter, Lenovo, Vertiv, Nvidia
• First liquid‑cooled AI deployment in Iceland, part of 140 MW campus

Pulse Analysis

The Icelandic data‑center market has become a magnet for AI‑intensive workloads because its abundant geothermal and hydro power delivers near‑zero‑carbon electricity at a fraction of the cost of traditional grids. Verne’s 40‑acre campus, authorized by Nvidia DGX, offers more than 140 MW of IT capacity and leverages the island’s sub‑zero climate for natural free cooling. By powering the facility entirely with renewable energy, operators can achieve dramatically lower PUE figures, making Iceland an increasingly attractive location for hyperscale AI cloud providers. The low ambient temperature also reduces the need for mechanical chillers, further cutting energy consumption.

Nscale’s first‑phase rollout demonstrates how quickly AI infrastructure can be scaled when liquid‑cooling and modular rack designs are combined. Within three months the company installed 128 racks—64 of them liquid‑cooled—housing Nvidia’s latest Blackwell Ultra GPUs and delivering 7.5 MW of compute power. The deployment, which meets the original plan of 85 % liquid‑cooled density, was executed alongside Computacenter, Lenovo, Vertiv and Nvidia, underscoring the collaborative ecosystem required to bring cutting‑edge AI hardware to market at speed. Such density enables AI providers to serve latency‑sensitive workloads while keeping footprint and cost in check.

The successful launch in Iceland is a blueprint for Nscale’s broader Nordic expansion, which already includes projects in Finland and Norway totaling several hundred megawatts. As AI models grow in size and inference demand spikes, providers that can combine renewable energy, high‑density liquid cooling and rapid deployment will gain a competitive edge. Investors and enterprises alike are watching these developments, expecting lower operating costs, reduced carbon footprints, and the ability to meet the escalating compute needs of generative AI services. The model demonstrates that sustainability can be a core differentiator rather than a compliance checkbox.