Yes, I guess I've always operated under the assumption that I've needed to parallelize dramatically. I usually operate on data from families of ~40 people with next-gen sequencing data, and the tools that I use generally finish within about an hour.