Resource Allocation and Binding Approach for Low Leakage Power

  • Authors:
  • Chandramouli Gopalakrishnan;Srinivas Katkoori

  • Affiliations:
  • -;-

  • Venue:
  • VLSID '03 Proceedings of the 16th International Conference on VLSI Design
  • Year:
  • 2003

Quantified Score

Hi-index 0.00

Visualization

Abstract

We propose a leakage power minimization approachbased on Multi-threshold CMOS (MTCMOS) technology.A clique partitioning-based resource allocation and bindingalgorithm is presented, which maximizes the idle periodsof modules in the data-path. Modules with significantidle times are selectively bound to MTCMOS instances. Wedeveloped a parameterizable MTCMOS component library,characterized with respect to sleep transistor width. Usingthis characterization, the leakage power-delay trade-offis analyzed and optimal sleep transistor widths are identified.For three well known HLS benchmarks, we obtainan average leakage power reduction of 22.44%. Themain disadvantage of MTCMOS technology is performancedegradation. We present a performance recovery techniquebased on multi-cycling and introduction of slack. Withthis technique, the performance penalty reduces to as lowas 14.28%. We obtain an average leakage power reductionof 17.46% after performance recovery. The averagearea overhead incurred due to the introduction of MTC-MOSmodules is 10.21%. Results are presented for 0.18µmCMOS technology.