Computational Analytics
Institute of Computer Science - Department I
We research and teach in both theoretical and practical aspects of Computational Analytics, covering both algorithmic data analysis and practical implementation on high performance systems.
One focus of the Computational Analytics department is the development, theoretical analysis, and practical evaluation of new algorithms and data structures for combinatorial optimization problems on data that can be modeled as graphs or networks. Our techniques are rooted in theoretical algorithmics as well as in application-oriented algorithm engineering and mathematical optimization.
Another focus of our department is on the hardware and software aspects that need to be considered when solving complex and large problems on supercomputers. These include co-design with users, the implementation of software stacks and the development of hardware prototypes. A particular focus here is on the Modular Supercomputing Architecture (MSA), which is being further developed in close cooperation with the Jülich Supercomputing Center (JSC).
Research Highlights
Algorithmic Data Analysis for Geodesy (AlgoForGe)
In our AlgoForGe research group, we study algorithmic issues relating to fundamental AI problems in geodesy.
HPC Cluster Marvin
"Marvin" is the university's high-performance computer. Above all, it enriches the research areas of High Performance Computing (HPC), Artificial Intelligence (AI) and Machine Learning (ML).
MSA: Modular Supercomputing Architecture
We investigate new hardware technologies and develop innovative concepts to integrate them into heterogeneous supercomputers.
Working Groups
Department I "Computational Analytics" consists of two working groups under the direction of Prof. Dr. Petra Mutzel and Prof. Dr. Estela Suarez.
The "Computational Analytics" working group, headed by Prof. Dr. Mutzel, is dedicated to algorithmic data analysis. This includes the development, theoretical analysis and practical evaluation of new algorithms and data structures for combinatorial optimization problems on data that can be modelled as graphs. These are used in various areas, e.g. cheminformatics, cartography, geodesy, bioinformatics, statistical physics or network design.
The "High Performance Computing" working group, headed by Prof. Dr. Suarez, focuses on researching and co-developing new HPC technologies at the Jülich Supercomputing Centre (JSC) of Forschungszentrum Jülich (FZ Jülich) with the aim of improving the overall performance, energy efficiency and user-friendliness of next-generation HPC systems. The work of our scientists includes hardware testing and prototyping, system software development, application porting, benchmarking and testing on these new platforms.
Working Group Leaders
Prof. Dr. Petra Mutzel
Computational Analytics
Room: 2.077
Phone: +49 228 73 69917
Research interests
- Computer-aided analytics
- Algorithm Engineering, esp. graph algorithms & data structures
- Algorithmic data analysis & graph mining
- Combinatorial optimization
- Network design & optimization
- Visualization of graphs & networks
- Analysis of chemical structures & biological networks
To the publications at Google Scholar
Prof. Dr. Estela Suarez
High Performance Computing
Room: 3.108
Phone: +49 228 73 69321
Research interests
- High Performance Computing (HPC)
- Heterogeneous HPC system architectures
- Modular Supercomputing Architecture (MSA)
- Hardware prototyping and evaluation
- Development of software environments
- Co-design and application optimization
To the publications at Google Scholar