With the presentation of the certificate by Dean Prof. Dr.-Ing. Höffer the junior professorship of Jun. Prof. Dr. Vogel was extended on the 17th of September. He was also assigned the status of a civil servant with limited tenure.

Jun. Prof. Dr. Vogel spoke with the CompEng Support Team about his research as well as his role in the Computational Engineering Master's programme.

What is the central topic of your research program?

My research is centered around high-performance computing. I develop algorithms and suitable implementations on modern computer architectures which make best use of given hardware opportunities. The overall aim is always to numerically address scientific and engineering research questions. For this purpose, the employment of state-of-the-art computers is one of the key methods to gain insight via predictive simulations, data mining, machine learning and similar techniques. However, a massive computing power is commonly required to address many of these questions by computational methods, and we therefore employ computer clusters or hardware accelerator cards to speed up the solution process. For such hardware architectures, we then investigate which algorithms are best suited, how they have to be adapted, and in which way we have to implement them for the scientific and engineering practice.

What makes your field of research particularly important for the future?

Scientific computing has become one of the main pillars for scientific discovery and serves as an underpinning tool for many engineering tasks. We see a trend of more and more powerful hardware installations which allow the successful employment of modern numerical methods supporting engineers and scientist in all kind of fields not only for predictive simulations based on partial differential equations but also in the emerging area of machine learning. High-performance computing enables many of these techniques for large-scale applications and is therefore considered as an important building block for future developments.

Which subjects do you teach for CompEng students?

I currently offer two courses on high-performance computing. One course focuses on multi- and manycore architectures as they are already found in today’s commodity hardware. We teach the students the general concepts and deepen their understanding via accompanying exercises in the computer lab. The second course is devoted to cluster computing. We discuss the required mathematical properties of algorithms such that a good scaling on modern supercomputers can be achieved. In addition, the technical programming aspects are covered and, finally, implemented by the students in a parallel computation code written from scratch. As applications, we focus on classical simulation techniques, such as the finite element method, and on methods from the field of artificial intelligence, such as deep learning.

In your opinion, what makes CompEng unique?
CompEng combines in a well-balanced way the modeling aspects of engineering sciences with modern numerical methods for the efficient treatment of complex problems and offers students an international and inspiring atmosphere for their studies.

We are pleased that Mr. Vogel continues to be a member of the faculty and thus CompEng and wish him every success in the implementation of the research projects already begun!