System-on-chip design: impact on education and research

Abstract

Deep-submicron technology is rapidly leading to exceedingly complex, billion-transistor chips. Within a decade, these chips will deliver enormous computing power as well as RF and analog interfaces for information and communication technology. At the same time, portability and the need for inexpensive packaging will limit power consumption to a few watts or less. These systems-on-chip (SOCs), designed at the processor-memory level, will fuel the future information society. Such designs depend, however, on SOC architects who can bridge the gap between software-centric system specifications and their implementation in novel, energy-efficient silicon architectures. Designing such hardware-software platforms will require a global-system approach from concept to implementation, which may well require a rethinking of present engineering schools. Chips will no longer be stand-alone components but complete silicon boards encapsulating complex system knowledge. These boards will be specified far above the hardware-description-language level and implemented in new, heterogeneous architectures designed at the processor-memory level. Today we know very little about the nature of these future architectures, let alone the design methodology. We will instead need the design research center concept. The research center's goal will be to perform cross-disciplinary system design research to create new methodologies, tools, libraries, and courses-distributed via the Internet-to produce enough SOC architects worldwide.