Digital technologies have become an integral part of science and industry – but data centers are among the biggest energy consumers in the IT sector. High-performance computers in particular are essential for complex calculations in research, but at the same time cause high electricity costs and CO2 emissions. To meet these challenges, RWTH Aachen University is focusing on innovative solutions for energy efficiency and sustainable resource utilisation.
Efficient Technologies for Environmentally Friendly HPC
A central component of our sustainability strategy is CLAIX-2023, RWTH’s latest high-performance computer. It combines maximum computing power with state-of-the-art cooling technology: a hot water cooling system channels the waste heat generated via a circulation system to the roof of the building, where it is cooled by outside air – without the need for energy-intensive chillers. This not only saves electricity, but also significantly reduces CO2 emissions. In addition, CLAIX-2023 is powered exclusively by renewable energies.
In addition to energy-efficient hardware, we are also working on intelligent software solutions to further optimise energy consumption. The BMBF project IT-Zauber is developing an integrated monitoring environment that combines detailed operating data from computer systems and supply infrastructure. This simultaneous approach enables us to recognise local optimisation opportunities and identify cross-system savings potential.
One particularly innovative approach is the modelling of a digital twin – a virtual replica of our HPC system including the infrastructure. This simulation enables more precise predictions to be made about energy consumption, cooling requirements and resource utilisation. Based on this data, intelligent control strategies can be developed to optimise operation in real time.
AI-Supported Energy Saving in IT
Artificial intelligence also plays a key role in our sustainability efforts. In the BMBF project ENSIMA, we are researching how AI-supported methods can reduce the energy consumption of simulation and production processes. These methods make it possible to perform complex computing tasks with fewer resources, significantly reducing both computing time and energy consumption.
The aim is to develop resource-saving HPC applications that can be used not only at RWTH Aachen University, but also in industry and other research institutions. The close integration of AI and HPC creates new opportunities to make energy-intensive calculations more sustainable.
Vision for the Future: CO2-Neutral Data Centers
The knowledge gained flows directly into the planning of new data centres. One particular project is the CO2-neutral data centre in the Campus West university expansion area. Here we are focusing on an even closer integration of data centre, power grid and heat recovery. The idea is that the waste heat from the HPC cluster should not go to waste but be fed into the local heating network. This could be used to heat buildings on the campus in an environmentally friendly way.
We are also working on concepts for intelligent load distribution: HPC workloads could be specifically shifted to times when a lot of renewable energy is available in the grid. This would make it easier to balance out fluctuations in the power supply and optimise grid utilisation.
Sustainability as a Central Task for IT
The IT sector faces the challenge of combining increasing computing requirements with ecological responsibility. At RWTH Aachen University, we show that high-performance computing and sustainability do not have to be a contradiction in terms. Through energy-efficient hardware, intelligent software solutions and innovative research projects, we make an active contribution to reducing the CO2 footprint.
With these initiatives, we ensure that the IT of tomorrow is not only powerful, but also sustainable.
Responsible for the content of this article are Malak Mostafa and Christian Terboven.
Leave a Reply