Parallel Block Iterative Solvers For Heterogeneous Computing Environments

. We study the parallel implementations of a block iterative method in heterogeneous computing environments for solving linear systems of equations. The method is a generalization of the row-projection Cimmino method where blocks are obtained by partitioning the original linear system of equations. The method is referred to as the Block Cimmino method. In addition, we accelerate the Block Cimmino convergence rate by the Block Conjugate Gradient method (Block-CG). Firstly, we present three different implementations of the Block-CG to compare different parallelization strategies of the method. Secondly, we present a scheduler to balance the computational load of the parallel distributed implementation of the Block Cimmino iteration. The computational resources used in these experiments are a BBN TC2000, and a network of Sun Sparc 10 and IBM RS-6000 workstations. We use PVM 3.3 to handle the interprocessor heterogeneous communication. KEYWORDS. Block iterative methods, Cimmino method, co..