One golden moment at the Intel Developer Forum this week in San Francisco was Tuesday's sighting of company cofounder and living techno-legend, Gordon Moore. He sat, smiling and fully engaged, business casual with open-collared shirt and blue blazer, for a live interview with NPR's "Tech Nation" host Moira Gunn. Still sharp as a tack, self-deprecatingly funny, and wise beyond measure, he recalled the early days at Fairchild and Intel, spoke of the development and runaway notoriety--and predicted duration---of his eponymous law, and waxed scientifically and philosophically on a number of topics. All in all, it's a great piece of Web-enabled oral history.

Here are a few samples. A questioner from the Internet wondered, "Is there an end-state to Moore's Law?" Gordon replied that "any physical quantity that grows exponentially predicts disaster, an ending. You can't go beyond certain limits." He noted that when Stephen Hawking was asked about the fundamental limitations of microelectronics, he answered 'the speed of light and the atomic nature of matter.'" When does Gordon see the law finally hitting the wall, once and for all? "A decade, decade and a half."

The size of wafers, once again a controversial topic in the chipmaking world, is an innovation that Moore admitted to being amazed by. "We're at 300 mm. I never believed wafers were going to get that big." He talked about a wafer-size calculation he did in the early '70s, looking forward from the 3-inch substrates in those days to the year 2000. "I extrapolated a size of 57 inches, which just shows how foolish extrapolating exponentials can be." (I wonder what he thinks about his company's push for 450-mm wafers!)

Wired's Epicenter blog and Semiconductor International have posted stories about Gordon's appearance, but if you want to see the whole charming "fireside chat," click here to go to the archived two-part video webcast.