COLLISION PROBABILITY CALCULATION AND MULTIGROUP FLUX SOLVERS USING PVM

Collision probability evaluation and flux computation are the most tim e-consuming aspects of applications based on the linearized time-indep endent transport equation. Parallelization for collision probability c alculation and multigroup flux computation are investigated. Particula r techniques pertinent to the two-step energy/space iterative process of solving a multigroup transport equation are described. The parallel performance is studied in cases where the cyclic tracking technique i s used to integrate collision probability. Parallelization is achieved by distributing either different energy groups or different regions o n a set of processors. These algorithms were tested on a four-processo r IBM SP2 and an eight-processor SPARC 1000 as well as on networks of workstations using the public domain PVM library. Typical run times ar e provided for unit cell calculations.