scientific article; zbMATH DE number 17454
A parallel algorithm for the modification of the lower and upper triangular decomposition factors of a banded matrix is presented. The algorithm represents an extension of the method proposed by \textit{J. M. Bennett} [Numer. Math. 7, 217--221 (1965; Zbl 0132.36204)]. The parallel implementation is discussed for the SIMD-type machine ICL DAP and the MIMD-type T-RACK transputer system. A description of architectural features of both systems is incurred. The algorithmic version for the DAP processor is expressed by means of DAPFORTRAN primitives. Despite timings of concrete implementations are not provided, the paper illustrates on a linear algebra example contrasts of the SIMD versus MIMD parallelization principles. [For the entire collection see Zbl 0728.00018.]
This page was built for publication:
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q3974774)