How small and still effective a CMOS-SoC could ever be?

Abstract

In this paper, effects of expected physical limits of CMOS technology on the performance of small-scale System-on-Chips (SoCs) are described. The exponential progress of CMOS technology has entered to the saturation phase. This could be called, if we like, a third phase of the Moore's law. In this third phase of development, the peak-performance of SoCs is not any more in the main concern. Instead of that, we could see an explosion of creative small-size applications: bionic appendages, smartphones with smart sensors, networks of tiny sensors, and a host of other applications we have yet to imagine. The International Technology Roadmap for Semiconductors (ITRS) lists expected performance parameters for CMOS technology up to the year 2028. In this paper, it is shown that starting from the simple RTL architecture parameters, e.g., gate count, and using performance figures taken from ITRS, we are able to figure out the (space-time-energy) performance limits of CMOS implementation of any logic architecture. As a practical example, the study of performance limits of a novel digital receiver suitable to be used in different capillary networks of Internet-of-Things applications is described, here. Design space exploration technique described in this paper can be used to find out performance limits of wide range of smart object applications.