Scheduling and optimization for multiprocessor systems

Studies the problem of scheduling arbitrarily nested loops on a tightly-coupled multiprocessor system. First, based on the concept of concurrent iterations, some interesting properties of arbitrarily nested loops are explored. Then an approach is presented for scheduling these nested loops in a way that their completion times can be even and diminished. In addition, two optimization techniques are utilized to schedule more loops on idle processors. Therefore, the system and program performance can be enhanced