1 hitherto practical conventional parallelization strategies could achieve a complexity of o np with respect to the inter processor communication p being the number of processors