Software energy profiling: comparing releases of a software product

Abstract

In the quest for energy efficiency of Information and Communication Technology, so far research has mostly focused on the role of hardware. However, as hardware technology becomes more sophisticated, the role of software becomes crucial. Recently, the impact of software on energy consumption has been acknowledged as significant by researchers in software engineering. In spite of that, measuring the energy consumption of software has proven to be a challenge, due to the large number of variables that need to be controlled to obtain reliable measurements. Due to cost and time constraints, many software product organizations are unable to effectively measure the energy consumption of software. This prevents them to be in control over the energy efficiency of their products. In this paper, we propose a software energy profiling method to reliably compare the energy consumed by a software product across different releases, from the perspective of a software organization. Our method allows to attribute differences in energy consumption to changes in the software. We validate our profiling method through an empirical experiment on two consecutive releases of a commercial software product. We demonstrate how the method can be applied by organizations and provide an analysis of the software related changes in energy consumption. Our results show that, despite a lack of precise measurements, energy consumption differences between releases of a software product can be quantified down to the level of individual processes. Additionally, the results provide insights on how specific software changes might affect energy consumption.