A heterogeneous mixed-mode execution model for massively parallel systems

In this paper, we consider a massively parallel system that is composed of heterogeneous processors, that is, processors with different processing pow er, and that combines the advantages of the SIMD and MIMD architectures. Th e heterogeneous mixed-mode (HeMM) execution model is composed of two main c omponents, which operate in the well-known SIMD and MIMD paradigms. The mai n computing power comes from a component that is composed of a massive numb er of processors and operates in a data parallel manner. The other componen t is composed of a few (or even one) fast processors which operate in the M IMD paradigm. The operation of a small number of processors in an MIMD para digm has been well demonstrated through actual systems. The processors in t his component add flexibility to the execution of the parallel programs suc h that it adjusts to-the changing parallelism of the program to enhance the performance. Based on this execution model we analyze the gains in perform ance that is obtainable by this new system. We show that substantial perfor mance gains can be obtained by using the HeMM system. (C) 1999 Academic Pre ss.