Further examination of Moore's law with data envelopment analysis

Although it has stood the test of time for over 30 years, Moore's law addresses but a single aspect of microprocessor design. As a proxy for technology, the number of transistors in an integrated circuit represents a limited perspective on the technology as a whole. In past work the authors proposed a set of metrics by which to measure a technology and a means to measure its progress over time utilizing data envelopment analysis (DEA). In this revised model, the assumption of state of the art (SOA) on product release is dropped, technical progress is measured iteratively over time, the effective time elapsed between the SOA and a no longer SOA has been refined to include a weighted average, and a means of utilizing proxy Decision Making Unit (DMUs), was implemented to maintain the dataset over time. (C) 2002 Elsevier Science Inc. All rights reserved.