Parallel operations

Ok.? Thanks for the feedback.? What about using bigmemory and bigmatrix as shown here: http://files.meetup.com/1781511/bigmemoryRandLinearAlgebra_BryanLewis.pdf?? I can't figure out if vam is a proprietary package from Paradigm4, or an R package (didn't find when I searched CRAN).? Is there a way utilize the foreach approach illustrated using distributed R packages without incurring the I/O penalty?