Comparative Analysis of Power Consumption in the Implementation of Arithmetic Algorithms

Historically, energy management in computer science has been predominantly treated as an activity of hardware optimization. A great deal of the effort in this area is concentrated on component activation, deactivation, and resource scheduling in order to provide a reduction of total power consumption. This work focuses on the study of power consumption from the developer´s point of view, using a reliable power measurement framework to validate the literature´s premise that programming options, such as multiplication operations, are high consumers of power energy. Besides some elementary operations and authors´ suggestions about alternatives for power consumption reduction on the programming stage, we also compare and evaluate two well-known and widely applied algorithms for large number multiplication: Karatsuba and Toom-Cook. The obtained results provide guidelines to the developer in the programming phase to choose, in some cases, the best technique to reduce power consumption, speed up the software, and establish a maximum power limit for the completed software.