Implementation of a numerical solution of the multicomponent kinetic collection equation (MKCE) on parallel computers

Two different numerical solutions of the two-component kinetic collection e quation were implemented on parallel computers. The parallelization approac h included domain decomposition and MPI commands for communications. Four d ifferent parallel codes were tested. A dynamic decomposition based on an oc cupancy function provided the optimum balance between time performance and flexibility for ally number of processors. The occupancy function was defin ed according to the number of calculations required at each grid point in t he domain. Speed-up performance depended very much on the parallel code use d and in some cases very good results were obtained for up to 32 processors . (C) 2001 Academic Press.