A multistage software router (MSSR) architecture is composed of several personal computers (PCs) to overcome single PC-based software router scalability issues. Although the architecture scales almost linearly with the number of internal elements, energy consumption could be a threat to its scalability features when building a core router with many internal components. Assuming a known 24 hour traffic load, this paper proposes energy efficient multistage software router design approaches, which enhance the performance of energy saving algorithms that minimize the energy consumption by tailoring the architecture to match the current input traffic demand. Simulation results show that a multistage software router architecture defined by the newly proposed design approaches saves roughly 10% of energy with respect to existing algorithms for similar cost. The energy saving may reach 20% depending on the initial admissible budget.