Research Group: High Efficiency, Low Power Systems (HELPS)






HAPPI: A Homogeneous Architecture for Power Policy Integration

Technical Report: Implementation Guides for a Homogeneous Architecture for Power Policy Integration

A significant volume of research has concentrated on operating-system directed power management (OSPM). The primary focus of previous research has been the development of OSPM policies. Under different conditions, one policy may outperform another and vice versa. Hence, it is difficult, or even impossible, to design the ``best'' policy for all computers. In this paper, we explain how to select the best policies at run-time without user or administrator intervention. We present a software framework called the Homogeneous Architecture for Power Policy Integration (HAPPI). This architecture is portable across different platforms running Linux. HAPPI specifies common requirements for policies and provides an interface to simplify the implementation of policies in a commodity OS. We describe how policies may be improved using HAPPI. Our experiments indicate that many policies published in the literature require tuning to achieve significant energy savings. Our approach allows these policies to be compared simultaneously and improved iteratively without rebooting the system. By using HAPPI, we are able to improve the energy savings of several policies by up to 41 percent. HAPPI can select the best policy among a set of distinct policies at run-time. This mechanism provides a critical step towards the acceptance of sophisticated power management policies in commodity OSs.

We have posted two sample policies for those interested in how policies may be implemented in HAPPI.

Cross-Layer Energy Management by Architectures, Operating Systems, and Application Program

Energy conservation has become a major challenge in system design. Most existing techniques focus on individual layers: devices, circuits, architecture, or software. The proposed research conserves energy through a collaborative and coherent approach across different layers. In the hardware layer, special registers report cumulative energy consumption of different hardware modules. These registers are called energy counters, similar to performance counters in modern processors. Operating systems use the information from energy counters and follow a set of accounting rules to determine the amount of energy consumed by each process. These processes can adjust their behavior to use energy efficiently. For portable systems powered by batteries, some processes may off-load computation to a remote server that is grid-powered. A programming environment will be developed to assist programmers in improving energy efficiency. This proposal has broader impacts on improving the energy efficiency of electronic systems to provide services for longer periods of time for mobile users.


These papers are protected by copyright laws.


Any opinions, findings and conclusions or recomendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the National Science Foundation (NSF).