This paper presents a parallel adaptive version of the block-based Gauss-Jordan algorithm used in numerical analysis to invert matrices. This version includes a characterization of the workload of processors and a mechanism of its adaptive folding/unfolding. The application is implemented and experimented with MARS1 in dedicated and non-dedicated environments. The results show that an absolute efficiency of 92 is possible on a cluster of DEC/ALPHA processors interconnected by a Gigaswitch network and an absolute efficiency of 67 can be obtained on an Ethernet network of SUN-Sparc4 workstations. Moreover, the adaptability of the algorithm is experimeted on a non-dedicated meta-system including both the two parks of machines.
Nordine Melab, El-Ghazali Talbi, Serge G. Petiton