Computational design alternatives with microprocessor-based systems

Abstract

This paper examines and characterizes four elemental hardware computational design alternatives (CDA's) and presents a structured approach to computational section design which incorporates a rigorous, theoretic foundation. The DIRECT CDA incorporates a single microprocessor (μP) and memory. The AU CDA contains a μP, memory, and arithmetic unit. A μP, memory, and calculator chip comprise the CALC CDA. Finally, several μP's and memories in a Master/Slave arrangement implement the multiple-μP mμP CDA. A common set of attributes — precision, speed and cost — facilitates comparison. Using these attributes, Multiattribute Utility Theory assesses a numeric quantity, the utility, to represent each CDA's relative usefulness. Thus, design involves selecting the CDA with the greatest utility.