|
Understanding Moore’s Law is a retrospective look at a common paradigm of the computer age—the exponential improvements in the performance of computers over time. Gordon Moore was the originator of this empirical rule. In 1965, he noted that the number of transistors in Intel microprocessors was doubling every year. That statement was based on just a few data points (5 points from 1959–1965). Remarkably, the trend continued over the next decade, and Moore’s name became attached to the phenomenon. In 1975, Moore revisited the data, and changed his prediction to a doubling rate of once every two years. That prediction also came true, and has held true more or less up to the present day (through 2005 data in the book). Different measurement criteria have been used over the years (e.g. size of lithographically produced circuit elements, or cost factors, rather than just the number of transistors), but the exponential relationship has remained. The book’s preface succinctly defines its purpose as answering a series of questions: “What is Moore’s law? Where did it come from? What is the underlying technology for making silicon chips? How has it changed, and who did the work? Who is Gordon Moore?” As you can tell, this relatively slender book seeks to serve as a brief history of the semiconductor industry, focused around those inventions that have markedly improved computer performance. The technology story is well told, and the role of chemistry is clear, from the initial need for high purity silicon to the current role of photoresists in nanofabrication technologies. Gordon Moore’s voice is clearly heard in this volume. His original 1965 paper, 1975 speech, and 2005 reflections on “Moore’s Law at Forty” are all reproduced here. The tone of the original writings is as much a part of the story as the factual details, especially with one individual reflecting on the industry repeatedly, over a long time period. It is a moderately technical overview of the industry over time, with a significant dose of human stories, showing both individual and group achievements. Overall, it is an interesting, quick read, on a fundamental trend in modern technology. Surprisingly, the one thing the book doesn’t do is give a concise, tight definition of Moore’s Law. The reader is left mostly to infer the law from the original article, and more recent updates, all of which discuss many different aspects of the industry and its productivity, rather than focusing on the fundamental definition of the law. Given the wide ranging uses this general concept has seen in the last 40 years, the lack of a tight focus on one specific formulation of the law is most appropriate!
|
