High-Performance Computing in Materials Science (HPC)
HPC group photo, December 2016.
A most efficient implementation of computer simulations methods is a key element in theoretical research on materials. The development and implementation of simulation methods for parallel computers has become an indispensable way to tackle nowadays materials science problems in order to map the requirements for size and complexity. The research group High Performance Computing in Materials Science at ICAMS is working on the development of parallel methods and algorithms for the efficient simulation of materials science applications on different levels of approximation. Apart from original research, the group has strong links to the other ICAMS departments and supports the development of simulation codes and efficient
parallelization of programs, developed at ICAMS. Currently the work of the group is focused on various topics, represented by the three departments at ICAMS, i.e. Atomistic Modelling and Simulation (AMS), Scalebridging Thermodynamic and Kinetic Simulation (STKS) and Micromecanical and Macroscopic Modelling (MMM). A focus is given on highly efficient implementations of particle based simulation methods, e.g. molecular dynamics, the combination of Monte Carlo and molecular dynamics (MD) in a parallel environment, parallel phase field simulations, and the problem of load balancing for different methods.
|Comparison of a parallel simulation with OpenPhase using a domain decomposition without (top) and with (bottom) load balancing, based on a block decomposition. Left: distribution of processors; Right: color coding for work distribution.|
ContactDr. Godehard Sutmann
Institute for Advanced Simulation
Jülich Supercomputing Centre (JSC)
D – 52425 Jülich
Tel. +49 (0)2461 61-6746
C. Teijeiro, G. Sutmann, G.L. Taboada and J. Tourino, Parallel simulation of Brownian dynamics on shared memory systems with OpenMP and unified parallel C, The Journal of Supercomputing (in press)
C. Teijeiro, G. Sutmann, G.L. Taboada and J. Tourino, parallel Brownian dynamics simulations with message-passing and PGAS programming models, Computer Physics Communications (in press)
C. Begau, J. Hua, A. Hartmaier, A novel approach to study dislocation density tensors and lattice rotation patterns in atomistic simulations, Journal of the Mechanics and Physics of Solids, 60, 711-722, (2012)