power aware 31 - The Case for Higher-Level Power...

Info iconThis preview shows pages 1–2. Sign up to view the full content.

View Full Document Right Arrow Icon
Abstract Reducing the energy consumed in the use of computing devices is becoming a major design challenge. While the problem obviously must be addressed with improved low-level technology, we claim there is potential value in a higher-level perspective, as well. In our approach, the needs of applications serve as the driving force for the development of power-management functions in the operating system and of a power-based API that allows a partnership between applications and the system in setting energy policy. The development of a PalmPilot application is used as an illustration. We advocate that reducing energy consumption should be raised to first-class status among performance goals when software is being designed. In support of this objective, new programming models, measurement tools, and system support mechanisms must be developed. These needs motivate our Milly Watt Project. 1. Introduction Mobile computing is dramatically changing our day-to-day lives, especially with the popularity of small devices such as personal digital assistants (PDAs) and with the embedding of substantial processing capabilities in devices such as telephones and cameras. Reducing the energy consumed in using these devices, thereby extending the lifetime of the batteries that power them (and reducing the weight of carrying spares), is one of the major challenges in designing such systems. Power consumption is an issue that extends well beyond the realm of battery-powered mobile devices to any computing platform in which heat or fan noise production may be factors (e.g. medical applications). Finally, energy efficiency of computers is desirable in its own right from both the economic and environmental points of view. This problem can be addressed at various levels: by improving battery technology, by engineering more efficient electronics and components, and by designing computer architectures and software with power as a primary measure of performance. Power is a critical, limited, and shared system resource. Traditionally, the operating system plays a major role in managing such resources. We also believe that the applications hold an important key, in the application-specific constraints and opportunities for saving energy that can be known only at that level. Therefore, our approach is to have the needs of the applications be the driving force for the development of power management functions in the operating system and of a power-based API that allows a partnership between applications and the system in setting energy use policy. Our higher-level perspective on the energy use problem grew out of firsthand experience with the design of a palmtop application, which we discuss in the next section. In that exercise, we were frustrated in our attempts to tailor the energy use to the specific needs of the application by several factors: a lack of feedback on what effect various design decisions might have on energy use, an inadequate programming model of power consumption in the target platform, a gap between our
Background image of page 1

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Image of page 2
This is the end of the preview. Sign up to access the rest of the document.

Page1 / 6

power aware 31 - The Case for Higher-Level Power...

This preview shows document pages 1 - 2. Sign up to view the full document.

View Full Document Right Arrow Icon
Ask a homework question - tutors are online