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

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