7. September 2022

The HPC CLAIX Projects Annual Report 2021 is online!

Lots of exciting projects, an insight into research, technology and High-performance Computing: The HPC CLAIX Projects Annual Report 2021 offers interested people plenty of reading material. High-performance computers such as CLAIX help to understand global issues and solve problems. Interesting projects ran on CLAIX in 2021 and helped to address these issues.
For example, the project Numerical analysis of wind-induced pressure fluctuations on open volumetric receivers in solar power tower plants (SPTP) explores the field of solar tower power plants, where the technology of the open volumetric receiver (OVR) has proven to be very robust and efficient at experimental scales as well as at the pilot power plant in Jülich (1.5 MWel). The next step in research and development of the technology is the scale-up to- wards market-relevant sizes. This project investigates a reference power plant with a tower height of around 200m and a thermal receiver placed in three separate cavities facing south, north-east and north-west with a combined thermal power of around 350 MWth. The aim of this study is to simulate the wind flow around the reference plant with transient cfd simulations utilizing DES turbulence modelling.

Also the topic of data security is becoming increasingly important. The security of complex systems and networks often depends on data aggregated from several sources such as enduser devices, or industry control systems. These devices can create a vast amount of alerts, that make it hard for analysts to discern attacks from benign behavior. There exist several approaches for Intrusion Detection Systems (IDS) and Intrusion Prevention Sys- tems (IPS) to handle these alerts, including machine learning approaches, to filter relevant output. However, such approaches raise questions on privacy, as they often include interactions with enduser systems and can be used to create user profiles or monitor user activity. Furthermore, the question arises how threats can be addressed preemptively, e.g., by leveraging public information to identify and secure vulnerable systems. The goal of the project Privacy-Preserving Machine Learning for Intrusion Detection is to research privacy-preserving machine learning approaches to several known problem domains.

The figure presented here is an example of the project Molecular Dynamics study of mutant IDH1 for in silico drug design against neurological diseases.

You can find more information about these projects and also other interesting projects in this CLAIX Report 2021.