Loop Pipelining in Hardware-Software Partitioning

This paper presents a hardware-software partitioning algorithm that exploits a loop pipelining technique. The partitioning algorithm is based on iterative improvement. The algorithm tries to minimize hardware cost through hardware sharing and hardware implementation selection without violating given performance constraint. The proposed loop pipelining technique, which is an adaptation of a compiler optimization technique for instruction level parallelism, increases parallelism within a loop by transforming the structure of an input system description. By combining this technique with our partitioning algorithm, we can further reduce the hardware cost and/or improve the performance of the partitioned system. Experiments show about 19% performance improvement and 44% reduced hardware for a JPEG encoder design, compared to the results without loop pipelining.