Intel co-founder Gordon Moore has predicted that his famous processor 'law' is running into a "fundamental block" and will no longer apply within 20 years.
Forty years ago today he wrote an article in Electronics Magazine predicting that the number of components on a processor would double every year and that costs would fall commensurately.
Moore revised the law in 1975 to processor complexity doubling every two years, but speaking to vnunet.com from Hawaii he admitted that the law had only 20 years to run.
"Something like this cannot continue forever," he said. "The dimensions are small enough now that we're approaching the size of atoms and that's a fundamental block. I think the law has another 10-20 years before fundamental limits are reached."
When Moore wrote the original article commercial chips contained about 30 components, and he claimed that he had one with 60 in the Intel laboratory.
Looking back to the start of transistor production, he predicted that the doubling of integrated circuit complexity was a valid projection for the next 10 years.
Rereading the original article a few years ago, Moore also found he had predicted that chips would be used to build home computers. However, he said he does not want to claim credit for this.
"Of course I had no idea what it would look like," he said. "The first time I discussed a home computer at Intel the only reason we could think of to use one was for a housewife to store recipes on."
In 1975 Moore realised that most of the spare space on chips had been taken up with processors, so he revised the law. Intel employee David House then came up with the notion of computer power doubling every 18 months.
But even after Moore's Law ends the Intel chief was optimistic about the future of the industry. Moore remains as chairman emeritus of the chip giant.
"Even then we'll be able to make bigger chips and there'll be a host of other innovations that will boost performance," he claimed.
"Over the next 40 years, society has a lot of problems to overcome but the technology that's possible is mind blowing."
Moore said that he would not be making any more predictions, as he is no longer close enough to an advanced field. Instead he preferred to "rest on my laurels".
He also praised Intel's new president, Paul Otellini, who is the first person without a PhD to lead the company.
"I often quipped with [chairman] Andy Grove that when he got into management he'd finally got over his PhD," Moore joked.
"Otellini has been steeping in processor technology, more so than I or Andy ever were, and you can't expect education to be a constant. Although I found out recently that I've been pronouncing his name wrong for 15 years."
Do you agree?
Have your say on this article