A compute-ahead implementation of the fan-in sparse distributed factorization scheme

In this report, we consider a compute-ahead computational technique in the distributed factorization of large sparse matrices using the fan-in parallel scheme. Experimental results on an Intel iPSC/2 hypercube are provided to demonstrate the relevance and effectiveness of this technique. Fortran source code is also included in an appendix. 9 refs., 2 figs., 1 tab.