Design Automation of Low Power Circuits in Nano-Scale CMOS and Beyond-CMOS Technologies.

Abstract

Today’s integrated system on chips (SoCs) usually consist of billions of transistors accounting for both digital and analog blocks. Integrating such massive blocks on a single chip involves several challenges, especially when transferring analog blocks from an older technology to newer ones. Furthermore, the exponential growth for IoT devices necessitates small and low power circuits. Hence, new devices and architectures must be investigated to meet the power and area constraints for wireless sensor networks (WSNs). In such cases, design automation becomes an essential tool to reduce the time to market of the circuits. This dissertation focuses on automating the design process of analog designs in advanced CMOS technology nodes, as well as reciprocal quantum logic (RQL) superconducting circuits. For CMOS analog circuits, our design automation technique employs digital automatic placement and routing tools to synthesize and lay out analog blocks along with digital blocks in a cell-based design approach. This technique was demonstrated in the design of a digital-to-analog converter. In the domain of RQL circuits, the automated design of several functional units of a commercial Processor is presented. These automation techniques enable the design of VLSI-scale circuits in this technology. In addition to the investigation of new technologies, several new baseband signal processor architectures are presented in this dissertation. These architectures are suitable for low-power mm3-scale WSNs and enable high frequency transceivers to operate within the power constraints of standalone IoT nodes.