Reducing Leakage Power in Fixed Coefficient Arithmetic

Abstract

Most of the power consumption has in the past been related to the dynamic activities, in a CMOS circuit. However, the static power, i.e. leakage, is a major contribution to the total power consumption, in present nano-meter scale technologies. This paper discusses static power reduction methodologies on architectural and arithmetical level. Novel arithmetic techniques to reduce the static power consumption in digital applications for nano-scale CMOS technologies are addressed. An arithmetic reduction of the static power consumption down to 6% by using bit-serial arithmetic compared to bit-parallel is indicated.