Green Mining: Keeping Software Sustainable by Engineering for Power Consumption

点击翻译按钮获取中文摘要

Contemporary computation is dominated by power consumption issues. For example, battery life is key to mobile devices and power usage and resource conservation is important within server farms. Any savings in power consumption provide direct savings in power costs. Power consumption was once the domain of computer and electrical engineers, but mobile and cloud computing shifted some of the responsibility for power consumption to software engineers. Since applications can consume power in many different ways (disk, network, radio, video, etc.), practitioners seek methods of engineering software to reduce power consumption. I propose Green Mining, a research agenda dedicated to helping software engineers reduce the power consumption of their own software by estimating the impact on power consumption that their modifications will have. Green Mining is all about how software maintenance and evolution impact a system's power usage. It leverages the huge corpus of publicly available software to help learn power consumption patterns in software. Questions that this research program will address include: · What kinds of software designs affect power consumption? · What kinds of changes to software affect power consumption? · What kinds of software development processes affect power consumption? Our long term goal is to be able to estimate the impact that a software change has on a system's power consumption, using a model based on historical, static and dynamic analysis. We want to help programmers produce more energy efficient applications. This research is very relevant to companies involved in mobile and server based computing, such as RIM, Microsoft, Apple, HTC, and Intel, who would hire HQPs trained by this proposal. This research has positive resource and environmental impacts.