An automated framework for power-critical code region detection and power peak optimization of embedded software

  • Authors:
  • Christian Bachmann;Andreas Genser;Christian Steger;Reinhold Weiß;Josef Haid

  • Affiliations:
  • Institute for Technical Informatics, Graz University of Technology, Austria;Institute for Technical Informatics, Graz University of Technology, Austria;Institute for Technical Informatics, Graz University of Technology, Austria;Institute for Technical Informatics, Graz University of Technology, Austria;Infineon Technologies Austria AG, Design Center Graz, Austria

  • Venue:
  • PATMOS'10 Proceedings of the 20th international conference on Integrated circuit and system design: power and timing modeling, optimization and simulation
  • Year:
  • 2010

Quantified Score

Hi-index 0.00

Visualization

Abstract

In power-constrained mobile systems such as RF-powered smart-cards, power consumption peaks can lead to supply voltage drops threatening the reliability of these systems. In this paper we focus on the automated detection and reduction of power consumption peaks caused by embedded software. We propose a complete framework for automatically profiling embedded software applications by means of the power emulation technique and for identifying the power-critical software source code regions causing power peaks. Depending on the power management features available on the given device, an optimization strategy is chosen and automatically applied to the source code. In comparison to the manual optimization of power peaks, the automatic approach decreases the execution time overhead while only slightly increasing the required code size.