Statistical leakage power optimization of asynchronous circuits considering process variations

  • Authors:
  • Mohsen Raji;Alireza Tajary;Behnam Ghavami;Hossein Pedram;Hamid R. Zarandi

  • Affiliations:
  • Department of Computer Engineering and Information Technology, Amirkabir University of Technology, Tehran Polytechnic, Tehran, I. R. Iran;Department of Computer Engineering and Information Technology, Amirkabir University of Technology, Tehran Polytechnic, Tehran, I. R. Iran;Department of Computer Engineering and Information Technology, Amirkabir University of Technology, Tehran Polytechnic, Tehran, I. R. Iran;Department of Computer Engineering and Information Technology, Amirkabir University of Technology, Tehran Polytechnic, Tehran, I. R. Iran;Department of Computer Engineering and Information Technology, Amirkabir University of Technology, Tehran Polytechnic, Tehran, I. R. Iran

  • 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

Increasing levels of process variability in deep sub micron era has become a critical concern for performance and power constraint designs. This paper introduces a framework for the statistical leakage power minimization of template-based asynchronous circuits considering process variation. We propose a statistical Dual-Vt assignment of asynchronous circuits that considers both the variability in performance and leakage power consumption of a circuit. The utilized circuit model is an extended Timed Petri-Net named Variant-Timed Petri-Net which captures the dynamic behavior of the circuit with statistical delay and leakage power values. We applied a genetic algorithm that uses a 2-dimensional graph to calculate the fitness to each threshold voltage assignment. Experimental results show that using this statistically aware optimization, leakage power can be reduced by 40.5% and 54.4% for the mean and the variance values.