Power consumption and portability issues are becoming increasingly significant in embedded system architectures. Therefore, it is important that chip architects and integrated circuit designers focus on power and portability, they must also be conscious of choosing the right set of parameters for an embedded processor. An evolutionary approach to configure the parameters of an embedded processor can be used to address these issues. From a given set of parameters the design space is explored using a simple genetic algorithm. This paper investigates the sensitivity of these parameters as they change during the process of the genetic algorithm.
James Northern III, Michael A. Shanblatt