In 1965, a relatively young Gordon Moore penned his now legendary paper with the scholarly title, “Cramming More Components onto Integrated Circuits.”
That paper, by the future Intel co-founder, is widely celebrated as the original inspiration for Moore’s Law, which states roughly that the number of transistors that can be installed on an integrated circuit doubles every two years. His breakthrough observation became a sort of guiding principle for the still very young computer industry. During the next 30 years, it was often difficult to tell if chip and computer manufacturers were proving Moore’s theory, or following it like some kind of law.
However, a closer look at the formation document reveals something interesting: Moore never, at least not in that paper, said what his law has come to mean. Yes, the 1965 paper written for Electronics magazine focuses heavily on how many components can fit on a circuit, and which factors will impact that growing number; but it wasn’t until 1975 that the idea that transistors would double every year or two emerged.
The 1965 paper, which celebrates its 50th anniversary on Sunday, can therefore be viewed as something that eventually gave rise to one of the computer industry’s most iconic intellectual underpinnings. However, even without the now legendary law, the groundbreaking paper marked the start of a new movement and some incredible intelligence and foresight on Moore’s part.
Moore’s paper starts off with a bang, and is as close as he gets to the Moore’s Law concept:
With unit cost falling as the number of components per circuit rises, by 1975, economics may dictate squeezing as many as 65,000 components on a single silicon chip.
In a section that deals with the cost of building these circuits, Moore explains that the cost of squeezing in additional components falls until the yield levels — the number of final circuit boards that actually work— start to fall. At that point, the costs tend to increase dramatically.
As components are added, decreased yields more than compensate for the increased complexity, tending to raise… The cost per component. Thus there is a minimum cost at any given time in the evolution of the technology. At present, it is reached when 50 components are used per circuit. But the minimum is rising rapidly while the entire cost curve is falling (see graph).
Moore noted, however, that the minimum cost for these circuits would fall dramatically in five years. “In 1970, the manufacturing cost per component can be expected to be only a tenth of the present cost,” he said.
Looking at this section, it’s also clear the “law” was already in Moore’s head. He states that they had, in 1965, roughly five components per circuit. In 1970, he expected “about 1,000. It’s not exactly a yearly doubling, but it’s pretty close.
The paper never discusses the number of transistors on a chip. In fact, he only mentions the term “transistors” twice in the entire thing.
By 1975, Moore formalized the concept when he revised his original notion, and said the number of integrated circuit (IC) components would double every two years.
None of this makes Moore’s Law any less valid. It’s just worth noting that what we’re celebrating was, 50 years ago, a somewhat more vague notion than most people currently believe.
Have something to add to this story? Share it in the comments.