A systolic array architecture for linear and inverse matrix systems (Q1208524)

The authors use the LU decomposition method for systolic arrays given by \textit{W. M. Gentleman} and \textit{H. T. Kung} [Matrix triangularisation by systolic arrays, SPIE in Real-time Signal Processing 298, 19-26 (1981).] to develop a systolic array architecture for linear and inverse matrix systems (SALIMS). The authors first give a brief treatment of methods for solving linear systems and matrix inversion. Then they describe the systolic array in detail, the proposed new algorithm and possible applications. This is followed by performance analysis. The Occam program for the systolic architecture is given in Appendix 1 and numerical examples in Appendix 2.