Exploring module selection space for architectural synthesis of low power designs

Architectural synthesis for low power design is a complex optimization problem due to the interdependence of power, delay and area. In order to obtain the optimal low power architecture where both the delay and area are efficient, full design space of module selection must be explored. In this paper we formulate the module selection as a multi-objective optimization problem and propose a branch and bound approach to explore the large design space of module selection. Experiments show that this approach can produce far better results than traditional architectural synthesizers where power, delay and area are all globally optimized simultaneously.