I'm coming at this from more of a physics viewpoint than from CPSC. To me, parallel programming is all about dividing a big problem into many little ones. The

I'm coming at this from more of a physics viewpoint than from CPSC. To me, parallel programming is all about dividing a big problem into many little ones. The meat therefore lies in boundary conditions. I don't have time to read this pdf right now, but it contains just two instances of the word "boundary". That sets off alarm bells for me.