Parallelization of dissipative particle dynamics simulation

The dissipative particle dynamics simulation is usually used to study polymer in mesoscopic space. The traditional methods are resource intensive, especially when the scale of research is large. Therefore, improving computing efficiency is a key point in this research area. Two major issues are addressed in this paper. First, the DPD methods are analysed and the most time-consuming parts are identified: conservative force, dissipative force and random force. Second, we describe how to parallelize the existing serial application in the Windows Compute Cluster Server (WCCS) platform. The results show that the parallel algorithm not only effectively shortens the computing time, but also improves the resource utilization rate.