Silicon Chips in the Northwest

Lawrence Snyder
November 1995

In the summer of 1979 the "micro-chip" revolution came to the Pacific Northwest. The story is one of cooperation between the University of Washington and the region's high tech companies.

Invented in the 1950s, semiconductor technology had advanced by the late 1970s to the point where integrated circuits were in wide use in computers. Chips containing hundreds or even a few thousand transistors had been created. At Caltech, Prof. Carver Mead headed a demonstration project to place an entire computer on a single silicon chip. In the course of this work, Mead realized two important facts. First, the number of transistors per chip would rise explosively, heralding the era of Very Large Scale Integration (VLSI): switches, wires and memory would no longer be dear, as they had been throughout the history of computing, but would become almost free. Second, the design and fabrication of chips could be separated: no longer did the chip designer have to be familiar with the semiconductor fabrication process in order to produce a working chip; "anyone" willing to follow a few rules could become a chip designer. With his colleague Lynn Conway, Mead developed the abstractions necessary to support this separation of design from fabrication, and the software necessary to support these abstractions. Mead then embarked on a mission to bring his message to the world.

By summer of 1979 the Mead Crusade had visited MIT, Carnegie Mellon, Stanford, and Berkeley. Faculty and students had learned chip design from the master, and were teaching others. When asked by Boeing to teach his vision to their engineers, Mead stated he was only interested in teaching to those who would teach, and asked his long-time colleague, UW Computer Science Professor Ted Kehl, to put together a joint UW/Boeing summer school on VLSI design with Mead as the instructor. Class members were CS students and faculty, Boeing engineers, and the original members of the DECwest Engineering group which had just located in Seattle. During the four week class, Mead presented his vision of the "tall, thin man," one who becomes accomplished in all aspects of chip design, from algorithm creation to layout, from concept to chip. Within weeks, faculty, students and engineers alike were showing off their then-state-of-the-art 6m nMOS chips to anyone with a magnifying glass. The class generated considerable electricity (and launched DEC's MicroVAX I, the first microprocessor VAX), and the Mead crusuade permanently changed electronic design in the Northwest.

But more aggressive electronic design was not the only effect of the class. UW and Boeing had cooperated to bring Mead to the Northwest, and the class' spectacular success was evidence to the organizers that cooperation could be a powerful tool for strengthening the region's high technology base. Education and industry each would win. From that reasoning came a model collaborative research agreement, the University of Washington Northwest VLSI Consortium.

Computer Science chair Robert Ritchie and Kehl formulated the concept of an industry/university/government consortium to support research and education in high technology, specifically VLSI. Five companies were recruited to be the industrial partners: Boeing, John Fluke, Honeywell Marine Systems, Microtel Pacific Research (a Vancouver BC firm), and Tektronix. After extensive technical and legal negotiations among the companies, the University and the Defense Advanced Research Projects Agency (a government funding agency), the UW/NW VLSI Consortium opened in January 1982.

During the 1980s the organization funded education and research at the University, where students learned microelectronics using state-of-the-art equipment and technology. The Consortium taught courses like Mead's to industrial engineers from the Northwest, the US and Canada. And UW CS produced a package of computer aided design tools to help engineers design VLSI chips. These tools were released to companies, labs and schools across North America.

By the late 1980s the revolution was complete, and the technology, though continuing to advance exponentially, had become standard. The Consortium closed its doors, and donated its equipment and technology to the Department of Computer Science & Engineering's Laboratory for Integrated Systems, which continues cutting-edge research and education in VLSI under the leadership of Professors Gaetano Borriello, Steve Burns, Carl Ebeling, Ted Kehl, and Lawrence Snyder.