Fine Grained Parallel Computing