Title :
Dynamic compilation for energy adaptation
Author :
Unnikrishnan, P. ; Chen, G. ; Kandemir, M. ; Mudgett, D.R.
Author_Institution :
Dept. of Comput. Sci. & Eng., Pennsylvania State Univ., University Park, PA, USA
Abstract :
While previous compiler research indicates that significant improvements in energy efficiency may be possible if properly optimized code is used, the energy constraints under which a given application code should be optimized may not always be available at compile-time. More importantly, these constraints may change dynamically during the course of execution. In this work, we present a dynamic recompilation/linking framework using which the energy behavior of a given application can be optimized while the application is being executed. Our preliminary experiments indicate that large energy gains are possible through dynamic code recompilation/linking at the expense of a relatively small increase in execution time.
Keywords :
constraint handling; hardware-software codesign; integrated circuit design; optimisation; program compilers; application energy behavior; compile-time; compiler energy efficiency; dynamic code recompilation/linking framework; dynamic compilation; dynamically changing constraints; energy adaptation; energy gains; execution time; optimized application code; Batteries; Computer science; Constraint optimization; Design optimization; Dynamic compiler; Energy efficiency; Feedback; Optimizing compilers; Power engineering and energy; Runtime;
Conference_Titel :
Computer Aided Design, 2002. ICCAD 2002. IEEE/ACM International Conference on
Print_ISBN :
0-7803-7607-2
DOI :
10.1109/ICCAD.2002.1167528