An evolution programming approach on multiple behaviors for the design of application specific programmable processors

This paper proposes an Evolution Programming Approach for behavior-level area-efficient design of ASPPs (Application Specific Programmable Processors). This approach, based on a given behavioral-level kernel, randomly transforms each of the input behaviors, then the behavioral kernel is used in the evolution process to guide the survival of data flow graphs (DFGs). Finally, instead of the given DFGs, the surviving DFGs are used to synthesize a programmable architecture. This leads to an area-efficient design for all the input behaviors. Experimental results indicate this approach is encouraging.