Optimized parallel simulations of analytic bond-order potentials on hybrid shared/distributed memory with MPI and OpenMP

Analytic bond-order potentials (BOPs) allow to obtain a highly accurate description of interatomic interactions at a reasonablecomputational cost. However, for simulations with very large systems, the high memory demands require theuse of a parallel implementation, which at the same time also optimizes the use of computational resources. The calculationsof analytic BOPs are performed for a restricted volume around every atom and therefore have shown to be wellsuited for a message passing interface (MPI)-based parallelization based on a domain decomposition scheme, in whichone process manages one big domain using the entire memory of a compute node. On the basis of this approach, thepresent work focuses on the analysis and enhancement of its performance on shared memory by using OpenMP threadson each MPI process, in order to use many cores per node to speed up computations and minimize memory bottlenecks.Different algorithms are described and their corresponding performance results are presented, showing significantperformance gains for highly parallel systems with hybrid MPI/OpenMP simulations up to several thousands of threads