Few phrases have rung louder in the history of innovation than “Moore’s Law,” the amazingly potent guess, published 50 years ago on April 19, that the capacity of electronic circuits would double every couple of years.
Since the 1968 founding of Intel, the world’s leading semiconductor-maker, that prediction has formed the basic doctrine of the enterprise. By 2006, the calculation foretold, the number of ultra-tiny transistors produced each year by Intel and its competitors would be a million million million (1018). That’s about as many as the grains of rice that humanity grows each year, and 10 times E.O. Wilson of Harvard’s figure for the number of the world’s ants. As usual, the big anniversary in 2015 for Intel co-founder Gordon Moore’s light-hearted prophecy induces people to ask how much longer the prediction can last. As usual, the rough answer is maybe another 10 years.
For most people the concept is an abstraction—a line on a semilog plot on graph paper that hardly suggests the underlying physics, chemistry, and manufacturing challenges involved in making the prediction come true. Yet Moore’s words in now-defunct Electronics magazine have had a tremendous impact. They’ve applied a sustained blowtorch to the electronic revolution. By holding out a sense of the possible to generations of clever people, Moore’s prophecy has helped shrink computers from the size of a big room down to a device in the palm of your hand or, soon, on your wrist.
Innovation needs driving ideas. So people shouldn’t think that Moore’s calculation is the only sustained technological driver in the 250-year history of the global Factory Revolution. That revolution began with several near-simultaneous developments in England. One brought machines together in textile mills starting in 1771, slashing drastically the price of goods like cotton thread and cloth. Another, around 1776, produced a more efficient steam engine that was quickly adapted to driving machinery. A third, in 1780, involved “puddling” smelted iron so its chemical makeup could be controlled more precisely, allowing metal to displace wood in factory machinery. The result was an inexorable push for ever-greater precision in making things. To cash in on the new capabilities, you couldn’t measure in fractions of an inch any more. Machine tools began evolving toward millionth-of-an-inch accuracy.
Another example of lickety-split, sustained development comes to mind: It was just over half a century from the discovery of cheap petroleum in western Pennsylvania in 1859 to Henry Ford’s mass production in 1913 of automobiles, a revolution in daily life for millions only a decade after radio, moving pictures, and the Wright brothers’ first test of a heavier-than-air flying machine. And don’t forget that those Model T motors (made of vanadium steel) also made nice powerhouses for farmyards and farm fields.
I mention manufacturing—often using very big machines to make something very small—not only because it is at the heart of Moore’s prediction. Manufacturing also is going to be central to holding down prices of the rising tide of new biotechnology medications. At Intel and its competitors, fabricating semiconductor memories and microprocessors calls for a fantastic set of steps, involving ultrapure materials like 12-inch slices of silicon, arcane lithography at the scale of nanometers, special chemical coatings and then chemicals to etch them, and visits to ultrahot electric furnaces, followed by elaborate mounting of the resulting “chips” in stable structures linked to many others.
All this made a big impression on me when I started reporting for the New York Times about Moore and his Intel colleagues Robert Noyce and Andrew Grove in the mid-1970s. To reach the targets of Moore’s Law they had already adopted two benchmarks for Intel: 10 percent of sales spent on R&D, and 10 percent profits after tax (both figures twice the industry average).
Moore’s law has created a very big tent. Many of the systems which keep Moore’s law in force hadn’t even been thought of when he first made his graph of how many more transistors you could fit on a fingernail-sized square of silicon.
All of this innovation is subject to immense pressures to use as little energy as possible. It’s not just that battery life is king. You have to minimize the operating temperature of, say, a laptop or smart watch. Also required are billions of dollars, committed years in advance, to build immense new ”wafer fabs.” These factories are designed to produce devices of specified dimensions, but the manufacturer’s ability to predict a chip’s functions and market years in advance is limited. Getting it done calls for unrelenting human discipline, and a fanatic drive for uniformity, not just from a few genius designers but from thousands of industrial workers. As Moore sighed to a recent interviewer, “There aren’t many easy businesses, and this certainly isn’t one of them.”
A lot of people say that the endless rollercoaster ride of modernity is driving us nuts. Yeah, maybe, but we don’t get off the train. Not even Gordon Moore as he relaxes at age 86 on a beach on Hawaii’s Big Island.
[Editor’s Note: This is the sixth of a series of notes about major anniversaries in innovation and what they teach us. You’re invited to suggest other milestones of innovation for in the Xconomy Forum. Example: This year will mark the 75th anniversary of Vannevar Bush obtaining President Franklin Roosevelt’s OK for mobilizing U.S. scientists in World War II.]
Further Reading:
Moore, Gordon (2006). “Chapter 7: Moore’s law at 40”. In Brock, David. Understanding Moore’s Law: Four Decades of Innovation. Chemical Heritage Foundation. pp. 67–84. ISBN 0-941901-41-6.
Chris Mack, IEEE Spectrum, Mar. 30, 2015
Gordon Moore, Rachel Courtland interview, IEEE Spectrum, Mar. 30, 2015
