Parallel schedules for MOC sweep in the neutron transport code PANDAS-MOC

Abstract

Method of characteristics (MOC) is a commonly applied technique for solving the Boltzmann form of the neutron transport equation. In the PANDAS-MOC neutron transport code, MOC is used to determine the 2D radial solution. However, in the whole-code OpenMP threading hybrid model (WCP) of PANDAS-MOC, it is found that when using the classic parallelism, the MOC sweeping performance is restricted by the overhead incurred by the unbalanced workload and omp atomic clause. This article describes three parallel algorithms for the MOC sweep in the WCP model: the long-track schedule, equal-segment schedule, and no-atomic schedule. All algorithms are accomplished by updating the partition approach and rearranging the sweeping order of the characteristic rays, and their parallel performances are examined by the C5G7 3D core. The results illustrate that the no-atomic schedule can reach 0.686 parallel efficiency when using 36 threads, which is larger than the parallel efficiency obtained in the MPI-only parallelization model.