Parallel processing of topological operations by using a hybrid MPI/OpenMP approach

The topological analysis of spatial objects is computationally very expensive, and therefore the applicability of existing codes is still limited to small datasets. Parallel computation provides an opportunity to reduce run times. This paper discusses a hybrid MPI/OpenMP approach to exploit two levels of parallelisms in software and hardware to reduce computing time on a PC cluster. The overall task is divided based on record size and then assigned to individual cluster nodes, where the calculation of spatial relations is parallelized using OpenMP. The parallel code is specifically tested with two case studies: Within and Overlap. Significant performance increases are seen in all applications, demonstrating the advantage of the present parallel scheme.