Code transformations for energy-efficient device management

Context-aware Battery Management for Mobile Phones

“…Prior research on dealing with the limited battery lifetime problem has concentrated on optimizing energy at different levels of the stack, starting with hardware all the way up to the application layer [5,3], including compiler-based energy optimizations [6]. There has been limited research on managing battery and treating energy as a firstclass operating system resource.…”

Inter-program optimizations for conserving disk energy

“…If the data streams were longer, CON + TIME would show a pattern of disk accesses starting close together and then drifting apart over time because the acceses are never synchronized. [5]. Related to this baseline is TIME, which assumes that data consumption rates for all applications are known.…”

“…Previous work has shown that applications which read data from disk in a streamed fashion (i.e., periodic access) can utilize large disk buffers to save energy [5]. These disk buffers are local to each application and serve to increase the idle period between disk accesses.…”

“…Our work extends the intra-program compilation framework as proposed by Heath et al [5]. In contrast to their approach, our compiler framework initiates disk power state transitions directly through appropriate system calls, i.e., the operating system is not involved in making decisions with respect to disk hibernation for the set of optimized applications.…”

“…This compilation strategy has been shown to work well in a single process environment [5,4,6], but may lead to poor overall results in a multiprogramming environment. In a multiprogramming setting, one program may finish accessing a resource and then direct the resource to hibernate over its idle period.…”