Unsung innovators: Lynn Conway and Carver Mead

Computerworld - There is an analogy Lynn Conway brings up when trying to explain what is now known as the "Mead & Conway Revolution" in chip design history.

Before the Web, the Internet had been chugging along for years. But it took the World Wide Web, and its systems and standards, to help the Internet burst into our collective consciousness. "What we had took off in that modern sort of way," says Conway today.

Before Lynn Conway and Carver Mead's work on chip design, the field was progressing, albeit slowly. "By the mid-1970s, digital system designers eager to create higher-performance devices were frustrated by having to use off-the-shelf large-scale-integration logic," according to Electronic Design magazine, which inducted Mead and Conway into its Hall of Fame in 2002.

The situation at the time stymied designers' efforts "to make chips sufficiently compact or cost-effective to turn their very large-scale visions into timely realities." But then Conway and Mead introduced their very large-scale integration (VLSI) methods for combining tens of thousands of transistor circuits on a single chip.

And after Mead and Conway's 1980 textbook Introduction to VLSI Design -- and its subsequent storm through the nation's universities -- engineers outside the big semiconductor companies could pump out bigger and better digital chip designs, and do it faster than ever.

Their textbook became the bible of chip designers for decades, selling over 70,000 copies.

"We distilled simple methods that opened up a creative area of electronic design that hadn't been accessible to computer engineers outside of a handful of companies," says Conway, who is now professor emerita at the University of Michigan.

Conway's background included a stint at Xerox Palo Alto Research Corp. (PARC), and she had experience in supercomputer design at IBM. For his part, Mead's Caltech background included work in device physics and integrated circuit design.

Their design revolution was all about simplicity. "We figured out how to remove tons of unnecessary design rules and optimizations, so that it all came clear," says Conway.

This was in contrast to the way chip designers at the time worked -- with different teams working on different aspects of chip design. The end result was usually far more complicated than necessary.

"We visualized a more streamlined way to do it, and we could see it was dynamite," she says. Their ideas also opened the door for innovative VLSI computer-aided design (CAD) tools that ran efficiently on personal computers. The small computers were then in just early stages of development. Conway and Mead saw that PCs would become ever more powerful as chip densities inexorably increased.