Digital Demand Meets Limited Grid Capacity
IT energy demand is growing rapidly—driven by AI processing, cloud services, and the increasing adoption of edge computing. For Europe, building significantly larger computing capacities of its own is not an option but a strategic necessity. This is the only way to secure digital sovereignty and gradually reduce dependence on non-European technology conglomerates. But there are hurdles along this path: In metropolitan areas such as Frankfurt or Eindhoven, electricity capacity is barely sufficient to supply new data centers. Elsewhere, such as in northern Germany, more electricity is generated than can be consumed. Due to a lack of transmission capacity, some of this electricity is simply curtailed instead of reaching the places where it is needed. Europe thus faces a dual challenge: growing digital demand on the one hand, and a limited, unevenly distributed power supply on the other. This is where the collaboration between T-Systems and Fraunhofer comes in. In the project “DEM4DCF – Dynamic Energy Management for Data Center Flexibility” —funded by the Federal Ministry for Economic Affairs and Energy and the European Union under the IPCEI CIS (“Important Project of Common European Interest on Next Generation Cloud Infrastructure and Services”) funding framework—both partners are pooling their expertise to establish data centers as active, flexible participants in the European electricity market, thereby strengthening both Europe’s energy efficiency and its digital sovereignty. In the predecessor project “IPCEI CIS Sustainability-Focused Orchestration in the Edge Cloud Continuum,” T-Systems International and the Fraunhofer IUK Technology Group had already collaborated on optimizing IT infrastructures to improve energy efficiency across the entire technology environment of the Edge Cloud Continuum. The Edge Cloud Continuum describes the flexible shifting of data processing between decentralized on-site devices on the one hand and centralized cloud data centers on the other—depending on which option makes more sense at any given time for efficiency reasons.
Shift Workloads, Reduce Network Load
DEM4DCF takes an innovative approach: workloads—such as those for AI training or cloud services—are shifted intelligently and dynamically so that they can be deployed in a way that benefits the network and is cost-effective. This so-called “load shifting” directly links workload management with energy supply—thereby overcoming a fundamental weakness of existing systems. Until now, energy supply and IT infrastructure have largely operated independently of one another, without any awareness of the other side. The principle is familiar from everyday life: Someone with a solar panel system on their roof will run the washing machine during the midday hours or charge their electric car when the sun is shining—and then tap into its battery again when needed. DEM4DCF applies a similar logic—across time and space and beyond European national borders—to the workload and electricity consumption of data centers—on a scale that can help stabilize entire power grids. Boris Otto, Chairman of the Fraunhofer IUK Technology Group, summarizes the benefits of this solution: “Intelligent load management creates genuine win-win situations. Grid operators can cushion peak loads, users benefit from lower electricity costs, and data center operators can tap into entirely new business models. Above all, however, by combining two research domains within the Fraunhofer IUK Group—cloud computing and energy systems—we are making a concrete contribution to Europe’s technological sovereignty. Distributed computing capacities are coordinated across the entire continent and operated independently of non-European players.” Christine Knackfuss-Nikolic, Executive Board Member and Chief Sovereignty Officer at T-Systems International, adds: “Data centers are not just energy consumers; they can actively contribute to stabilizing the European energy system. Together with Fraunhofer, we are developing approaches to intelligently link AI and cloud workloads with energy availability. Building on our successful collaboration within the IPCEI CIS framework, we aim to put these innovations into practice. In this way, we are laying the foundation for the sustainable scaling of digital and AI-based applications while simultaneously strengthening Europe’s digital sovereignty.” Conceptually, the solution relies on an orchestrator—software that allocates resources, compensates for outages, and scales workloads without the need for manual human intervention. Within the control triangle of computing capacity, energy, and communication, the energy requirements of data centers are continuously forecasted—both for the IT load and for cooling. Based on these forecasts, an optimization algorithm determines when and to what extent tasks should be executed, postponed, or marketed as flexibility. Via a control interface, the system translates these decisions directly into specific operating parameters and continuously monitors their implementation.
How the Market Benefits
Internal Fraunhofer project management is handled by the Fraunhofer Institute for Energy Economics and Energy System Technology IEE, whose key development areas include the digitalization and resilience of energy systems.
Institute Director Martin Braun: “Data center operators have so far hardly considered their facilities as energy market players—yet that is precisely where significant economic potential lies. Those who strategically market their load flexibility via spot markets—and, with the appropriate prequalification, via balancing energy markets—can tap into new revenue streams while simultaneously reducing their energy costs. The best part is that these benefits can be achieved using existing facilities without requiring significant new investments.”
As part of the project, the Fraunhofer Institute for Industrial and Economic Mathematics (ITWM) is establishing the so-called FlexLab—a real-world laboratory where industry partners can test concrete use cases for dynamic energy management.
“What sets this project apart from many others is the ability not only to model ideas but also to test them under real-world conditions. “In the FlexLab at Fraunhofer ITWM, we operate real computing infrastructure—from CPU to GPU systems—and can test live how computing loads and energy management can be intelligently integrated. Operators and aggregators can develop and validate specific usage scenarios here before rolling them out on a large scale,” said Anita Schöbel, Director of Fraunhofer ITWM.
The Fraunhofer Institute for Software and Systems Engineering ISST, the Fraunhofer Institute for Applied Information Technology FIT, and the Fraunhofer Institute for Optronics, System Technology, and Image Analysis IOSB are also participating in the “Dynamic Energy Management for Data Center Flexibility” project in various roles.
The white paper being developed as part of the project will analyze the technical and regulatory prerequisites for flexible data center operations. Furthermore, using scenarios for 2030 and 2035, it will highlight the flexibility potential of European data centers in the electricity market and derive concrete business models for marketing this flexibility. The project will run until the end of 2026.
Journalists who would like to receive the white paper upon its publication can sign up now by emailing dem4dcf-whitepaper@iuk.fraunhofer.de.
Fraunhofer Institute for Software and Systems Engineering