NVIDIA GTC 2026 officially marked the change from AI as a model, to AI as an integral part of the economy. NVIDIA founder Jensen Huang's keynote highlighted the token as a fundamental component of contemporary computing, signaling a shift towards Agentic AI, which are autonomous systems that need levels of execution, scalability, and infrastructure that have never been seen before. 

With the increasing demand for AI and high-performance computing (HPC), the methods and locations of computing resources are strategic factors. Our team participated in the event to showcase the design of our renewable-powered data centers in Iceland and Finland, designed specifically for these applications. 

Building Data Centers, the Home for Vera Rubin and Feynman 

A significant advancement in vertical integration is indicated by NVIDIA's announcement of the Vera Rubin platform and the upcoming Feynman architecture. Now, AI factories need even higher power densities, and specialized cooling is needed for these systems in addition to conventional floor space.  

Our campuses in Finland and Iceland are already at the forefront of this development. Our facilities are built to support Direct Liquid Cooling (DLC) and high rack densities over 140kW. They are Nvidia-ready certified, designed in accordance with AI Factory standards. 

Lenni Leinonen, our Senior Sales & Partner Lead says: 

In the AI field, it’s no longer practical to plan 20 years into the future. Hardware evolves so fast that 'future-proofing' now means planning just 5 to 10 years ahead. At Borealis, we design our campuses modularly to meet NVIDIA’s future requirements by enabling increased power and supporting denser workloads, allowing us to scale and adapt alongside next-generation AI systems. 

Sovereignty in the Time of "AI Native" businesses 

The demand for GPUs is exceptionally high as mentioned in Huang’s keynote, driven by an influx of $150 billion into AI startups. AI becomes more important, and using the cloud can limit costs, performance, and ownership. That's why universities like Eindhoven University of Technology (TU/e) worked with us and MDCS.AI to create a sovereign AI stack. This stack gives them full control over data, models, and IP, and it is backed by a high-performance, integrated system that is designed for world-class research at scale.

Read the full TU/e sovereign AI deployment overview in the brochure ➔

But sovereignty isn't enough on its own; it needs to grow with the next generation of AI hardware. Here, location plays a critical role. 

"Combined with our Nordic climate and innovative cooling design—which already enables chiller-free free cooling for today's hardware—we are well positioned to deliver even greater efficiency gains as next-generation platforms supporting +45°C emerge," Leinonen says of our business practices.

In practice, the Nordic data center location advantage means less cooling overhead, less energy use, and more predictable OPEX, and as AI workloads become denser and more expensive, sovereign infrastructure stays both scalable and efficient. 

Sustainability at the Forefront  

Our facilities run on 100% renewable energy. Our sites are strategically located and prepared for hybrid cooling, harnessing naturally available free-air cooling where possible. This allows us to achieve a PUE as low as 1.15, enabling your AI agents to operate with a much smaller environmental footprint.

Whether you are scaling a pilot or deploying a hyperscale environment, Borealis provides the sovereign, sustainable foundation your models deserve. 

Join the leaders who are choosing control, performance, and 100% renewable energy. Explore Our AI & HPC Data Center Services or Contact Us to discuss how we can build your infrastructure together. 

FAQ  

What makes a data center "AI-ready"? 

A facility designed specifically for high-density computing and GPU clusters is known as an AI-ready data center. In contrast to conventional configurations, these "AI Factories" are built with robust power and cutting-edge cooling infrastructure to handle the massive needs of AI and HPC installations at any size, from early pilot projects to hyperscale environments. 

Why choose data center colocation in Iceland and Finland? 

Iceland and Finland offer a unique combination of renewable energy, stable power infrastructure, and naturally cool climates, making them ideal locations for high-performance and AI workloads. Both countries provide long-term energy predictability, strong connectivity to global markets, and secure operating environments. 

For businesses scaling AI and HPC infrastructure for example, this translates into lower total cost of ownership, high operational reliability, and a reduced environmental impact compared to more power-constrained regions.