Abstract

This paper examines the range of possible solution approaches within the areas of Operational Research (OR) and Management Science (MS) that could benefit from the application of High Performance Computing (HPC) utilizing a parallel processing paradigm implemented with Message Passing Interface (MPI). The results include an overview of the Infinite Horizon Discounted Markov Process (IHDMP) using Pre-Jacobi (PJ) Iteration with Porteus Bounds (PB) and a comparism of the results partially differentiated with respect to an isolated operational parameter, demonstrating how the variation of input affects the final solution set. This paper includes a graphical demonstration of the relationship between the monotonic and contraction properties of the IHDMP and the prime number system describing the Regular Rectilinear Markov Primes (RRMPS). Also described is a comparism of the computational performance of the serial and parallel algorithms.

D. LEE B.Sc., M.Res.