A parallel optimization method for numerical solving the system of polaron equations using the partitioning algorithm

The previously developed method for the numerical simulation of the formation of polaron states in condensed media is modified using the partitioning algorithm, which provides a significant speedup in the parallel computations on multiprocessor systems. The software implementation is based on the MPI technology. Numerical results obtained on the multiprocessor cluster installed at the Laboratory of Information Technologies (Joint Institute for Nuclear Research, Dubna) with various numbers of processors and with various computational parameters show that the proposed approach is efficient for the numerical solution of the system of nonlinear differential equations describing the polaron dynamical model.