Happy 40th birthday, Moore’s Law

Launching a little company called Intel
Moore, a chemist by background, had joined Fairchild after leaving a firm set up by William Shockley, one of the transistor’s inventors in 1947. Robert Noyce, a pioneer in integrated circuits, also was part of the “Traitorous Eight” defectors.

After a series of management changes at Fairchild, Noyce and Moore bolted again — this time forming Intel — short for Integrated Electronics — to make advanced circuits and find new uses for them.

“It was a very courageous thing to do — to believe this doubling could continue on for a significant number of generations,” Williams said. “But he understood it and placed some pretty significant bets — and boy did it ever pay off.”

Last year, Intel reported $34.2 billion in sales. Now retired, the 76-year-old Moore still ranks among the world’s wealthiest people with an estimated fortune worth $4.3 billion. He remains an active philanthropist.

Success industrywide
Success, however, was not limited to Intel. The global semiconductor industry reported sales last year of $213 billion. The consumer electronics industry, which relies largely on semiconductors, pulled in $1 trillion.

On a fundamental level, Moore’s Law gives chip makers, their equipment suppliers and their customers headlights for the future.

But cramming more and more transistors onto a chip is only half the story: Companies are left to figure out how to use them.

Recently, Intel and others have moved toward integrating two processor cores on a single chip, a move that could be planned years in advance thanks to guidance from Moore’s Law, said Fred Weber, chief technology officer at Advanced Micro Devices Inc., an Intel rival.

“The reason we had the confidence five years ago to (announce) our multicore design was because we saw where the number of transistors would move, and we knew what computer architecture would make possible,” he said.

Moore's Corollary: Nothing lasts forever
But, as Moore himself is quick to point out, nothing can grow at an exponential rate forever. Within the next 15 years, it’s expected that engineers will reach the point where any further shrinkage would require atoms to be split.

“Any exponential extrapolated far enough predicts a disaster,” Moore said.

Then again, predicting the end of Moore’s Law has become a favorite pastime in the Silicon Valley. Its demise has been a few years away for the last 20 years, though each seemingly insurmountable barrier has been bypassed.

“There’s been every reason given why it shouldn’t work anymore, every reason why it’s dead, every reason why it’s about to die,” said Craig Barrett, Intel’s chief executive.

Despite Moore’s influence, there’s not a whiff of arrogance about him. As he walked outside the other day, he offered to hold the door for a delivery man who was carrying a handheld computer but didn’t recognize the man who 40 years ago had predicted such technology could exist.

For years, Moore was uncomfortable uttering the words “Moore’s Law,” and he said he is surprised at how his prediction has been adapted — and sometimes mangled — by others. “I’m going to be as well known as Murphy one of these days,” he said.

MORE FROM GADGETS

Gadgets Section Front

Apple, AT&T need to upgrade customer communications Apple blames batteries for overheated iPods The PC gets more personalized Forget HD: Toshiba focuses on plain old DVD Mobile services mushroom for locating friends iPhone dropping calls? Blame the Infineon chip Intelligent GPS devices tap away traffic Tech that's too clever for its own good Intel PCs to wake up for phone calls   Jobs confirms iPhone ‘kill switch’ Gadgets Section Front   Add Gadgets headlines to your news reader: