Research Groups


CSMM - Research Group Computational Science and Mathematical Methods

FiNE - Junior Research Group Fixed-Point Methods for Numerics at Exascale

With the explosive rise in hardware parallelism, running applications efficiently on modern high-performance computing systems requires a total redesign of the underlying numerical methods. This new paradigm must include implementations that focus on node-level parallelism, reduced global communication volume, and relaxed synchronization requirements. The Fixed-Point Methods for Numerics at Exascale (FiNE) research group aims to tackle these challenges head on by developing a new class of algorithms for extreme-scale computing.

Research Group Distributed and Parallel High Performance Systems

The DPHPS (Distributed & Parallel High Performance Systems) group is tightly integrated in SCC and thus allows a close interaction of research topics, system-related and operational aspects of IT research infrastructures and practice-oriented teaching. The focus is on three major subject areas: Data Management, Data Intensive Computing and Cloud Computing.

In Data Management, research & development is about methods and technologies for distributed, efficient and secure handling of research data and for data archiving for the entire data life cycle. These methods and technologies are used in practice in cooperation with national and international partners and in close cooperation with application sciences. A special focus is on the handling of large-scale research data.

In Data Intensive Computing, research & development focusses on efficient scheduling methods (e.g. to bring computing jobs and data together more efficiently) as well as methods and tools for data analysis. Furthermore, parallel simulations and data analysis applications are developed and scaled for use on state-of-the-art, heterogeneous parallel high performance systems and optimized for efficient use of thousands of processor cores.

In Cloud Computing, efficient methods for the management and use of distributed resources (e.g. through scheduling and auctions) are developed and their performance is evaluated. The above-mentioned topics are integrated in the research-oriented lectures "Parallel Computing and Parallel Programming" and "Distributed Computing" as well as in seminars and praktika. In (almost) all research topics, student projects and bachelor/master are possible. More details can be found here or directly contact Achim Streit.

Scheme of the complete Data Life Cycle supported by the DLC-Labs of SCC
Supporting the complete Data Life Cycle

MBS - Junior Research Group Multiscale Biomolecular Simulation

The tremendous growth of high-performance computational (HPC) resources in the last decades has enabled the use of computers to perform virtual experiments that study a wide range of questions in science. Concurrent efforts in theory, experiment, simulation, and data exploitation have emerged as a new paradigm to speed the discovery of scientific phenomena and new technological applications, in particular at the interface between life and physical sciences. The Research Group Multiscale Biomolecular Simulation focuses on method development and application for three main research directions: (a) quantitatively understanding the structural and dynamic mechanism of life at the molecular level, (b) simulating cancer development at the cellular and tissue level, and (c) driving the development of novel nano-scaled materials.

Overview Research

Name Tel. E-Mail
2 additional persons visible within KIT only.

RAI - Junior Research Group Robust and Efficient AI

UQ - Junior Research Group Uncertainty Quantification

Mathematical models are ubiquitous in applications across science and technology, where these models are used to describe complex processes. To systematically investigate and predict phenomena of interest using computer simulations based on these models, we need accurate, reliable, and efficient computational methods. 

Many complex models of practical relevance are affected by uncertainties, for example due to a lack of knowledge of material properties, intrinsic natural variability, or by incorporating noisy data. Such uncertainties inevitably impact the considered mathematical model and consequently the quantification of the reliability of simulation outcomes.

The junior research group Uncertainty Quantification (UQ) develops advanced mathematical and numerical techniques for the treatment and quantification of uncertainties in complex computational models. Our research focuses on theoretical and methodological aspects, as well as on interdisciplinary projects where theoretical sound methodologies are tailored to applications.