Energy optimization has become an important issue in hardware and software design. In particular, optimizing the energy consume by the memory system has been found to be very important. This paper looks at the influence of state-of-the-art compiler optimizations on the memory system energy considering the overall impact on instruction and data accesses. Our results indicate that while these optimizations provide significant savings in energy consumed by data accesses, they increase the energy consumed by instruction accesses. This, in turn, results in an increase in overall energy in most cases. Consequently, compiler writers for energy-aware systems need to consider both instruction and data locality in a unified optimization framework.
|Original language||English (US)|
|Number of pages||10|
|Journal||IEEE Workshop on Signal Processing Systems, SiPS: Design and Implementation|
|State||Published - Dec 1 2000|
|Event||2000 IEEE Workshop on Signal Processing Systems (SIPS 2000) - Lafayette, LA, USA|
Duration: Oct 11 2000 → Oct 13 2000
All Science Journal Classification (ASJC) codes