Intel’s (INTC) biggest show features a fireside chat with Intel’s biggest legend: Gordon Moore, the man who came up with Moore’s Law.
The interview has begun.
Gordon Moore is talking about how he ended up in Silicon Valley, and how he and others founded Fairchild Semiconductor.
Back then, engineers weren’t trained in semiconductors because they didn’t exist.
Silicon Valley got its name when a journalist in the mid-’60s coined the term, he recalls.
In the early days, he says, lots of people argued that integrated circuits didn’t make sense. One of the reasons people thought they wouldn’t make sense is that people thought the yields would be horrible — that many of those in each batch produced wouldn’t work, and the costs would be astronomical.
(It’s worth noting that the room here is packed full of engineers in awed silence.)
Fairchild got rid of two CEO within six months, and Bob Noyce was passed over for the job. When Noyce decided to leave, Moore decided to leave, too. (This is the reason Intel was born.) There were only about 10,000 computers being made at the time, so there wasn’t much of a market for what they wanted to do — so they decided to focus on making semiconductor memory.
They were really lucky, Moore said. They started out pursuing three different technologies: bipolar transistors (which others copied), multiple chip assembly that was too complicated. But they also tried silicon gate MOS. It was the Goldilocks strategy. It wasn’t too easy, wasn’t too hard. Other companies didn’t focus on it, and Intel had a monopoly on it for seven years. If it had been significantly more difficult, Intel would have run out of money, if it had been too easy, others would have copied it.
Do you ever regret that it got called Moore’s Law, Gunn asks? “I guess I don’t now,” he deadpans.
On the genesis of cubicles: Intel had just acquired a big building, Santa Clara 4. It was roughly square. They started looking at what would happen as they brought in a lot of engineers — offices down all the hallways would make it look like a windowless prison block. That doesn’t make sense, the leaders said. So they looked at cubicles, and decided to put everyone in them.
Moore said he still has the biggest cubicle at Intel, because he had a big, round table that wouldn’t fit anywhere else.
A question has come in: Is there an end to Moore’s Law?
There is, Moore says. Any physical thing growing exponentially comes to an end. Stephen Hawking said the speed of light and the atomic nature of matter are going to be limiting factors. We’re getting close to those limits, Moore says. But it’s amazing to him that the industry has continued to get around the limits. In another decade, decade and a half though, he expects that we’ll hit some kind of real limit.
He set a limit long ago that 72 is the maximum age for a board member. He said he now feels that it’s still a good choice, but fortunately there are other companies that don’t have that limit where he can serve.
Another question: Can other industries innovate as quickly? Moore says yes.
What would he advise people to study now? If he were going back today, he would probably look at something more in the biological end of things. Interfaces are key. The interfaces between computers and biology is a very complex area now. Moore went to Berkeley yesterday, and noted that in a biology lab, the scientists wouldn’t be able to do their work.
What would the world look like without the integrated circuit? Moore recounts a story that it was estimated that eventually a transistor would cost 68 cents. (Now they cost a tiny fraction of that.)
Steve Jobs recently said that the technology road is bumpy. What does Moore think about how hard it was for Intel to survive in the early days when Japanese competitors were copying Intel designs? He says he feels guilty for empowering the Japanese, who at the time didn’t have experience innovating in the area. They soon saw where things were going and the Japanese became formidable competitors on quality and price. Moore realized that the Japanese were doing a better job at manufacturing. Fortunately, the industry was able to improve manufacturing and catch up. But there was a period when it wasn’t clear that the U.S. semiconductor industry was going to survive.
What innovation has surprised him over the years that came out of Intel? A lot of them have, Moore says. The barriers that look impossible continue to melt as we get closer. Huge 300-millimeter wafers are one thing. “I never believed wafers were going to get that big,” he says. In the 70s he did a slide on wafer size (they were 3 inches at the time). He did one that said that in 2000 there would be 57-inch wafers, just to show how ridiculous it was to extrapolate that out. He allows that they didn’t grow that fast, but they’re still a lot bigger than he thought they would be.
If he had another career, what would it be? He really doesn’t know. He feels fortunate to have had the career he had, and to have gotten in early.
Taking time alone is a valuable thing to do, Moore says. He didn’t have a BlackBerry, went to places where there were no telephones. (It’s said he came up with Moore’s Law during a fishing trip.)
How has Intel affected him over the years? “Obviously it’s affected me financially rather significantly, and given me opportunities I otherwise wouldn’t have. But I’ve still got the same wife …” [applause.]
About the Moore Foundation for developing innovative technologies. It’s focusing on the environment, science education, and Moore’s wife is passionate about nursing. Part of the conservation focus arose because during fishing trips, he kept noticing that remote areas that were some of the best spots were being overrun by development.
How does he see humans and computers interacting in the future? One thing that has always seemed like it was going to happen was that we’ll get to the point where there’s good language recognition by a computer, and you can have computers do translation and hold a good conversation. When you can have a conversation with a computer like you talk to a colleague, that will bring a lot more people into the computing fold.
What does he still want to accomplish? “I’d like to get the paperwork in my office cleaned up,” he says. The foundation’s work. He’d like to see some of the things in development come to fruition. He’d love to come back 100 years from now and see what’s happened. “We’re living in just a fantastic time period.”
What advice does he have for new engineers? Remember the fundamentals that you learned in school and apply them regularly. The things that were most often needed in his work were the most basic things. Make them a part of what you do. He said he doesn’t know what a new engineer might be tackling these days.
What were the alternatives to the Intel name? We were originally incorporated as Moore-Noyce Electronics, just to get through the paperwork, he says. There were a list of 15-20 names that they tried to get cleared in California and New York. Even after they settled on Intel after having a few others rejected, they ended up having to buy the name from a motel company in the Midwest.
How do you make tradeoffs whether to build a certain technology or not? You have to make the decision eventually. The tough decisions are tough, usually, because there’s not much difference between the two choices, he says. The dangerous decisions are the ones where the choice seems easy.
The interview has ended, and Gunn gives Moore a refrigerator magnet that says, “If God didn’t make it, it’s technology.”
That ends the interview, and the crowd gives a standing ovation.