Abstract
We carry out a performance study on a single processing node of the HITACHI SR8000. Each processing node of the SR8000 is a shared memory parallel computer which is composed of eight scalar processors with a pseudo-vector processing facility. In this study, we implement highly optimized codes for basic linear operations including matrixmatrix product, matrix-vector product and vector inner-product. As a practical application of matrix-vector product, we examine the performance of two iterative methods for linear systems: the conjugate gradient (CG) method and the conjugate residual (CR) method.KeywordsConjugate GradientOuter LoopProcessing NodeVector OperationBlock DistributionThese keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have