CoreWeave Expands Its AI Infrastructure in Europe

CoreWeave, a U.S.-based AI cloud provider, is expanding its computing capacity with a new location in Denmark while also investing in a global high-performance network. The infrastructure is designed to connect data centers across four continents and make it easier for companies to access powerful AI resources.

CoreWeave, a cloud provider specializing in artificial intelligence, is driving forward the international expansion of its infrastructure. The Nasdaq-listed company has brought additional computing capacity online in Denmark and plans to build a global network specifically designed to meet the demands of large-scale AI applications. In addition to North America and Europe, locations in the Middle East and the Asia-Pacific region are also set to be integrated in the future.

With this expansion, CoreWeave is responding to the growing demand for computing power for the training and productive use of AI models. Companies increasingly require high-performance GPU infrastructures as well as fast and secure connections between different data center regions.

CoreWeave is therefore pursuing an integrated approach in which computing capacity and international data connections are expanded in tandem. This will enable customers to transfer large volumes of data between locations and operate AI applications across multiple regions.

New Location in Denmark Expands European Capacity

A key component of this expansion is the DEN01 data center operated by infrastructure provider atNorth in Copenhagen. CoreWeave has already brought the majority of the planned 22.5 megawatts of IT capacity online there. This is the company’s first location in Denmark.

The facility uses GPU systems based on the NVIDIA Blackwell architecture, which are interconnected via NVIDIA Quantum-X800 InfiniBand. This technology is specifically designed for computationally intensive applications such as training large AI models.

With the new facility, CoreWeave’s European infrastructure expands to nine data centers across five countries or regions. In addition to Denmark, the company already has a presence in the United Kingdom, Spain, Sweden, and Norway.

Energy efficiency also plays a role at this location. The data center has an agreement to utilize excess waste heat, which will be fed into the local district heating network. This allows some of the heat generated during operations to be used to supply households.

Global AI Network to Connect Four Continents

In parallel with the expansion of its computing capacity, CoreWeave is developing its own international network designed to transmit large volumes of data for AI applications.

Starting in the fourth quarter of 2026, additional network nodes are scheduled to be established in the United Arab Emirates, India, Indonesia, Singapore, Australia, South Korea, and Japan. By mid-2027, the company plans to have a continuous network structure in place connecting four continents.

According to the company, the network is expected to enable data transfers in the petabit range in the long term. At each connection point, multiple lines, each with a capacity of 400 gigabits per second, are planned. Hardware-based encryption and redundant connections are designed to ensure security and reliability.

“Our customers need more than just access to computing capacity. They need an infrastructure that functions reliably even on a large scale.”

Sachin Jain, Chief Operating Officer of CoreWeave

These investments underscore the growing importance of specialized infrastructure in the competition for AI cloud services. In addition to the sheer availability of high-performance graphics processing units, international data connections and the geographic distribution of computing capacity are becoming increasingly important.

As of June 30, 2026, CoreWeave operated 51 data centers worldwide with an active power capacity of more than 1.5 gigawatts. In August, the company estimated its contractually secured capacity at approximately 4.2 gigawatts. CoreWeave was founded in 2017 and has been listed on the Nasdaq since March 2025. By expanding its infrastructure, the company is positioning itself as a specialized cloud provider for AI developers, startups, and large enterprise customers.