Behavioral partitioning with exploiting function-level parallelism

Abstract

This paper proposes a method to efficiently generate hardware from a large behavioral description by behavioral synthesis. For a program with functions which are executable in parallel, this proposed method determines an optimal behavioral partitioning which fully exploits the function-level parallelism with simultaneously minimizing the area in the datapath and control path. This partitioning problem is formulated as an integer programming problem. Experimental results demonstrate the effectiveness of our proposed method.