1 Introduction
As a result of technology trends, leakage (static) power dissipation has emerged as a first-class design consideration in high-performance processor design. Historically, architectural innovations for improving performance relied on exploiting ever larger numbers of transistors operating at higher frequencies. To keep the resulting switching power dissipation at bay, successive technology generations have relied on reducing the supply voltage. In order to maintain performance, however, this has required a corresponding reduction in the transistor oxide thickness to provide sufficient current drive at the reduced supply voltages. At the technology node, CMOS processes will have oxide thicknesses of to [1]. Since the Metal Oxide Semiconductor Field Effect Transistor (MOSFET) gate tunneling leakage current increases exponentially with a reduced oxide thickness [15], gate tunneling leakage power dissipation will grow to be a significant fraction of overall chip power dissipation in modern, deep-submicron processes [12] [18]. As the gate oxide thickness gets thinner, gate tunneling leakage could surpass weak inversion and drain-induced-barrier-lowering leakage as the dominant leakage mechanism in future technologies [8] [13] [10].