High-performance computing (HPC) has become an important method in many scientific fields, such as drug and climate research. However, the data centres with energy-intensive supercomputers release quite a bit of CO2. “Energy-optimised supercomputer networks through the use of wind energy” (ESN4NW): Behind this title is a new, nationwide joint project led by the SICP — Software Innovation Campus Paderborn at the University of Paderborn. In cooperation with the SICP member company WestfalenWIND IT and their brand windCORES, the idea was born to develop a new HPC infrastructure with a sustainable concept. The Federal Ministry of Education and Research (BMBF) is now funding the project for the next three years with around 2.5 million euros. The consortium also includes the University of Passau, the Fraunhofer Institute for Reliability and Microintegration IZM and the companies AixpertSoft GmbH, Rittal, Atos Deutschland and Zattoo.
Together, the partners are researching and demonstrating the potential of so-called “windCORES”. These are wind turbines that house data centres in their towers and supply them with the wind power generated on site in an almost climate-neutral manner. The aim of the project is to develop the infrastructure and operational management of an HPC cluster within several wind turbines (WTs). The direct, locally available renewable energy is to be incorporated into the operational management in order to make maximum use of it. In addition, the waste heat generated is also to be considered as a limiting factor.
For demonstration purposes, an alternative — beyond the state of the art — architectural approach will be pursued and implemented. The project partners are developing an infrastructure that continuously adapts the computing power of the decentralised HPC cluster through intensive data utilisation and artificial intelligence (AI) methods via an operating platform. The cluster is distributed over several WTs and must optimally exploit the fluctuating energy availability. The project also explores the impact of different architectures and operating strategies on the sustainability assessment of the target system. AI-based operation management requires behavioural and predictive models of the infrastructure and the local energy grid, which are not yet available. “The scientific and technical challenges of the project lie, among other things, in the intensive data development and use (energy grid & infrastructure), in the integration of artificial intelligence into control loops, as well as in the systemic merging of all models with the help of a digital twin,” explains Dr. Gunnar Schomaker, manager at SICP. “The focus is on predictions of weather-dependent local energy availability, as well as precise behavioural models of all interacting trades with regard to their power consumption and heat output,” Schomaker adds.
A special feature and innovation of the project is that the tower of a wind turbine is integrated as a potential heat sink if no waste heat receiver is available or if it does not need the waste heat at the moment. The tower can then absorb the waste heat from the IT systems and cool it down. This function, together with energy availability, is to be used as a guideline for the operation of the HPC cluster. In concrete terms, this means that the system will only run if renewable energy is available and waste heat can be dissipated in an almost CO2-neutral manner. Initially, the project will focus on the technical and economic feasibility, the functional and didactic demonstration of the operation and the sustainability assessment of the system in the extended system life cycle. However, the intended results should be transferable not only to wind turbines, but also to other HPC systems with waste heat generators and waste heat users.
“The project particularly addresses the challenges of the energy transition and digitalisation. We want to show that the increasing energy demand of digitalisation is not a dead end for more sustainability and that these growth needs can also be covered flexibly in terms of time and space by regenerative energies,” says Dr Fiete Dubberke, Managing Director of WestfalenWIND IT, describing the project goals. “In doing so, we are asking ourselves the following questions: How can we deal with waste heat from the most diverse industrial sectors, i.e. use it? Which systems of generators and users are to be connected in a meaningful way and what tasks can artificial intelligence methods take over?”, Schomaker elaborates. The “windCORES” concept, which was awarded the “German Data Centre Prize”, has already shown that merging a wind turbine with a data centre can also be implemented successfully in economic terms.