For me, the opportunity is really about learning how to develop software for a multicore future, rather than using these for processing. There may not be too ma

For me, the opportunity is really about learning how to develop software for a multicore future, rather than using these for processing. There may not be too many embarrassingly parallel problems around, but that doesn't mean we should avoid developing algorithms for highly scalable and even tightly coupled problems.