Parallel Computational Modelling of Inelastic Neutron Scattering in Multi-node and Multi-core Architectures

Abstract

This paper examines the initial parallel implementation of SCATTER, a computationally intensive inelastic neutron scattering routine with polycrystalline averaging capability, for the General Utility Lattice Program (GULP). Of particular importance to structural investigation on the atomic scale, this work identifies the computational features of SCATTER relevant to a parallel implementation and presents initial results from performance tests on multi-core and multi-node environments. Our initial approach exhibits near-linear scalability up to 256 MPI processes for a significant model.

Towards Ad-Hoc GPU Acceleration Of Parallel Eigensystem Computations

“…We have developed an initial high performance parallel version of SCATTER (Roach et al, 2007), a new INS modelling tool, that has demonstrated linear scaling up to 1024 nodes on the Huygens prototype supercomputer in the SARA facilities in the Netherlands (Garba et al, 2010). Central to this parallel version are phonon mode calculations carried out with the support of EISPACK.…”