A new asset class for Danish RE investment firm: AI-ready, sustainable data centers
Copenhagen-based real estate firm Thylander is set to make history with Denmark’s first domestically owned hyperscale data center, marking a step toward data sovereignty
As production-scale deployments of artificial intelligence have exploded, enterprises are rethinking where and how to run these resource-intensive workloads. While cost is certainly a big part of the equation, the decision calculus also includes considerations such as data sovereignty, sustainability, and protection of intellectual property.
Thylander, a Copenhagen-based private equity real estate firm, recognized the emerging opportunity. The 40-year-old company has plans to develop a new 100-megawatt data center (with potential to expand to 200 MW) in Denmark and lease rack space, networking, power, and cooling to businesses that require bleeding-edge graphic processing units (GPUs) and other hardware for AI workloads.
In 2026, Thylander will open the doors to the Nordic country’s first Danish-owned and -operated hyperscale data center. It’s a step toward achieving data sovereignty in a market currently dominated by international operators. Denmark has become an attractive location for many of the established hyperscalers, thanks to abundant green energy, low-latency fiber networks, and one of Europe’s most reliable power grids, supported by underground high voltage transmission lines. Stable political, economic, and weather climates add to Denmark’s appeal. Now, a local company is seeking to capitalize on these resources.
From local support to global ambitions
Beginning in 2024, the company recognized an untapped opportunity for the firm to develop and operate Danish co-location data centers. The reasoning is clear: Some businesses require their data to be stored and processed locally, and many want to raise the standards for energy optimization in data centers.
“Looking at data sovereignty and thinking about who actually owns data centers was the starting point for us to say that we want to do something Danish for Danish companies,” says Anders Mathiesen, chief executive of Thylander Data Centers, “but also for externals who think the Danish markets are valuable.”1
The domestically owned and operated data center is the latest indicator of the global race to build out AI infrastructure, something Mathiesen expects to continue ramping up. “The United States invests 15 times as much into high-risk and venture opportunities as we do in Europe,” he says. “We need to start having the kind of risk tolerance and venture capital available to invest in these opportunities.”
Thylander aims to be a step ahead of that demand. While it would take an international company starting from scratch four to five years to build AI-ready data centers, according to Mathiesen, the real estate firm believes it can achieve this within 12 months with available land, green energy resources, local government support, and fiber-optic connectivity.
Sustainability at the core
Thylander’s website says the data center is a “new asset class at the intersection of real estate and low-carbon energy,” highlighting the firm’s commitment to sustainable investments.
Globally, most data centers rely on existing power grids, which may run on coal power or natural gas, and require massive water-based cooling systems that can deplete or contaminate local water sources, according to Mathiesen. Thylander is aiming for more sustainable operations, using low-carbon electricity and air-cooling, which is possible thanks to Denmark’s temperate climate.
Thylander’s data center will leverage pre-existing resources by repurposing the seawater cooling infrastructure from the Esbjerg power plant. This approach helps reduce construction time and costs while aligning with the company’s commitment to sustainability. What’s more, the data center will provide heating to local communities by recycling waste heat. “The Nordics are quite a good place to build data centers on an environmental level,” Mathiesen says.
Thylander’s infrastructure will be optimized for GPU clusters, which power most of today’s AI workloads, and high-density computing. The company’s scalable offerings will range from single racks to full halls. For Mathiesen, having these options close to home will enable Danish organizations to take a hybrid approach, putting some hyperscale workloads in the public cloud and others in their private cloud. “They’ll be able to say, ‘I need to have some control over this,’” says Mathiesen.
Editorial consultant: Stephanie Overby
Design consultant: Heidi Morrow
Cover image by: Meena Sonar