Optimizing modulo scheduling to achieve reuse and concurrency for stream processors

Both reuse and concurrency are performance-critical for stream processors. When applying loop unrolling and software pipelining separately to stream-level loops, either reuse or concurrency or both...