Parallel implementation of alternate quadrant interlocking factorisation method on star topology

This paper discusses the parallel implementation of the solution of a set o f linear equations using the Alternative Quadrant Interlocking Factorisatio n Methods (AQIF), on a star topology. Both the AQIF and LU decomposition me thods are mapped onto star topology on an IBM SP2 system, with MPI as the i nternode communicator. Performance parameters such as speedup, efficiency h ave been obtained through experimental and theoretical means. The studies d emonstrate (i) a mismatch of 15% between the theoretical and experimental r esults, (ii) scalability of the AQIF algorithm, and (iii) faster executing AQIF algorithm.