MPI backend for an automatic parallelizing compiler

Many naive parallel processing schemes were not successful as many researchers thought, because of the heavy cost of communication and synchronization resulting from parallelization. In this paper, we will identify the reasons for the poor performance and the compiler requirements for performance improvement. We realized that the decisions for parallelizing should be derived by the overhead information. We added this idea to the automatic parallelizing compiler, SUIF. We substitute the original backend of SUIF with Our backend using MPI, and gave it the capability of validating of parallelization decisions based on overhead parameters. This backend converts shared-memory based parallel program into distributed-memory based parallel program with MPI function calls without excessive parallelization that causes performance degradation.