Thursday, April 07, 2005

The anniversary of Moore's Law--UPDATED

Few men have been as prescient or gotten richer than Gordon Moore.
April 7, 2005
This month is the 40th anniversary of one of the great intellectual feats of modern times: Moore’s Law. If you don’t understand Moore’s Law, you don’t understand the electronic revolution. If you run a business based on electronic technology you better grok it. In a paper published on April 19, 1965, Gordon Moore, one of the founders of Intel, predicted that the number of transistors that would fit per square inch in a microprocessor would double every year. He later revised that to every 18 months, and astonishingly, that's exactly what has happened. If you are old enough, you remember the first hand-held calculators out of Texas Instruments. They added, subtracted, multiplied and divided and cost several hundred dollars. We were dazzled. I wrote a book in the late 1970s on a Radio Shack TRS-80 (the infamous Trash Eighty) and blew away my friends with the capabilities of my 28 k memory. (It worked). I don’t remember the clock speed but let’s say “leisurely.” You can now buy calculator that does all that the old ones did, plus currency conversions and mathematical tricks like square roots for a few bucks. The chips inside are called “jelly bean chips” because that’s about what they are worth. I’m writing this on an Apple G-5 iMac with 1 gig memory and a clock speed of better then 1.6 gigahertz, several times more powerful and faster than the computers that were used on the moon expeditions in the 1970s. It has the entire Encyclopedia Britannica in storage and four days worth of iTunes files including most of Mahler’s symphonies, and it works effortlessly. In 1971, Intel put 2,300 transistors in its first microprocessor. This year they will have one that contains 1.7 billion! Moore, who incidentally is a very nice man (I interviewed him once), has been right on the money. "If the automobile industry moved this fast," he once said, "your car would move at a million miles per hour and it would get 50,000 miles a gallon." Most everyone, including Moore, knows that the law will eventually expire. It runs into the laws of physics in about 15 years, when the transistors get to the size of atoms and manufacturing the chips will prove impossible. Moore’s Law, nonetheless, has been a major factor in the modern economy, not only turning California’s Santa Clara Valley into Silicon Valley, but driving and challenging industry. The semiconductor business now has $213 billion in annual sales and powers the $1 trillion electronics business. If you have a business that relies on this technology, you have to factor in the law. The technology you use will change in 18 months, the business plan you devised will have to adjust every 18 months, and you better be right guessing how it will go. The world is littered with people who either guessed wrong or never thought of Gordon Moore. Bad mistake. And of course Gordon Moore got to be a very rich and very happy man. He earned it.

UPDATE: The estimable John Markoff points out in this morning's New York Times that while Moore gets the credit for the concept, he wasn't first to do so. The equally admirable Doug Engelbart was years ahead of him. That surprises no historian of Silicon Valley; Engelbart was ahead of most people most of the time and has probably received less publicity than anyone of the founders of this electronic revolution. He is best known as the inventor of the mouse, but he had his hands in many of the developments and inventions of the computer age, including the whole concept of aq desktop computer. Happy to set the record straight and knowing Gordon Moore, he probably is too.

1 comment:

Walter E. Wallis said...

Then there is Clarke's "law" about technology/magic. If you had handed a Pentium chip to a 1950s scientist he would have had no idea what he had nor any tools to find out. We do indeed stand on the shoulders of giants. Sometimes on the shoulders of midgets, too.