Navigation and service

The Future of Scientific Computing: Predictive Hierarchical Simulation (PHS)

With Moore’s law coming to an inevitable end, the next generation in scientific computing cannot only rely on improved hardware, but has to follow an integrated approach based on the simulation tripod consisting of the three cornerstones: models, methods, and data. The full dynamic and hierarchical interaction between these cornerstones defines “The Future of Scientific Computing: Predictive Hierarchical Simulation (PHS)”, which results in the development of the ultimate tailored simulation process. This cluster determines the scientific core of the new Center for Simulation and Data Sciences established by Forschungszentrum Jülich and RWTH Aachen University. Through its Jülich Supercomputing Centre, Forschungszentrum Jülich is strongly involved in the PHS research units Methods and Data as well as in the PHS Hub, a cross-cutting working area with a particular focus on high-performance computing. For these units, hierarchical methods, software engineering on heterogeneous supercomputer and innovative parallelization techniques play a major role. The Helmholtz Excellence Network proposed here targets these topics by proposing four activities in the fields of parallel-in-time integration techniques, parallel I/O, HPC tools and flexible software infrastructure. Along these activities, the PIs will make first key steps towards a successful participation of Jülich in the PHS cluster, strengthening the role of Forschungszentrum Jülich in the excellence initiative and the Center for Simulation and Data Sciences.


Forschungszentrum Jülich GmbH, Germany
The contact person is Robert Speck

RWTH Aachen University, Germany

PHS is funded by the Helmholtz association under Grant Agreement EXNet-0017-Phase 2-3

The grant period is October 2017 until December 2019.