PowerQuest: Trace Driven Data Mining for Power Optimization

Abstract

The paper introduced a general framework, called PowerQuest, with the primary goal of extracting "interesting" dynamic invariants from a given simulation-trace database, and applying it to the power-reduction problem through detection of gating conditions. PowerQuest adopts machine-learning techniques for data mining. The advantages of PowerQuest in comparison with other state-of-the-art dynamic power management (DPM) techniques are: 1) quality of ODC conditions for gating; 2) minimization of extra logic added for gating. The validity of the approach was demonstrated in reducing power through experimental results using ITC99 benchmarks and real-life microprocessor test cases. Power reduction of up to 22.7 % was presented, in comparison with other DPM techniques