"Large amounts of venture capital have lured researchers to leave large companies and universities to join startups," said panel moderator David Johns, a professor at the University of Toronto. "Are these startups doing innovative research or are they working on relatively short-term products to satisfy their investors?"

The answer, it seems, depends on the definition of innovation, panel members said. "Technology innovation is not product innovation," said Bryan Ackland, Vice-President of Circuits and Systems Research at Agere Systems, Allentown, PA. "Startups do product innovation based on technology invented somewhere else." Ackland was the most vocal advocate of the notion that real research that leads to real innovation can only be done in an environment without constant market pressures. "Innovation takes time, and pressure to quickly turn a profit to satisfy investors precludes time to think and innovate," he said.

"Innovation is more about bringing new ideas to market than the discovery of physical laws," responded Nav Sooch, president and chief executive officer of Silicon Laboratories (Austin, Texas), which he co-founded in 1996 after engineering stints at Crystal Semiconductor and AT&T Bell Labs. "Startups are more efficient at developing new ideas to market than any other enterprise."

Ackland countered: "We are eating our seed corn, we are busy exploiting research done in the '80s. Who is going to come up with the new great ideas?" He showed a diagram to illustrate how self-defeating startups could be: out one door of an established company exit researchers for a startup, where they collect their $2 million and reenter the established company through another door after the startup is bought up. Meanwhile, innovation at the established company suffers. "Successful startups invariably make money from ideas generated in some other setting," Ackland said.

Bob Hewes, vice president and director of DSP research and development at Texas Instruments Inc. (Dallas), took a middle-of-the-road stance. "Innovation is alive and well at both large companies and startups," he said. "Innovation is meaningful only when technology is successfully brought to market, and that involves invention, development, marketing, financing, manufacturing and many other activities."

But Hewes' remarks slightly favored established companies more than startups. In just four years, he said, TI was able to increase the number of die per wafer in its semiconductor lines from 310 in 1995 to 2,616 in 1999. "That type of innovation doesn't happen in a startup," he said.

The very essence of innovation seems to have changed over the past 20 years. Panelists agreed that basic research has suffered at the expense of applied research. "Electronic design is an applied science and the value of an invention depends completely on its merits to find application," said Klaas Bult of Broadcom Corp.'s design center in Bunnik, Netherlands, in a prepared statement. "The phrase that 'necessity is the mother of invention' says it all, and in the cutthroat environment of startup companies, necessity is highest."

Ron Rohrer, adjunct professor of electrical and computer engineering at Carnegie Mellon University (Pittsburgh), agreed and padded his answer with some harsh reality: "The quarter-to-quarter myopia of Wall Street discourages breakthrough research endeavors in non-monopolistic large companies that operate in competitive business environments." Rohrer contended that industrially-funded university research has wrongfully focused on highly improbable technology transfer, thus minimizing the probability of innovative breakthroughs.

Masao Hotta of Hitachi Central Research Labs (Gunma, Japan) offered a conciliatory tone to the debate. " While startups employ many outstanding engineers from either large companies or universities, thus slowing down fundamental innovative research, they contribute in bringing the results of innovative research to production," he said. "However, counterproductive venture companies pursuing only stock options waste good engineers."

Defending voices

About a third of the audience at the Tuesday (Feb. 6) panel debate was from startups and they were the most vocal during the Q&A period, when questions ended up sounding more like testimonials to the virtues of startups. One participant, who had joined a startup from a university teaching position and is now back working at a different university, said, "The experience has made me a better faculty member." Another participant gave his foreign perspective: "I don't understand this tendency to downgrade the rate of innovation in startups. From the Swedish perspective we always admired the innovation achieved in American startups. In fact, your example has spawned a few of our own, thereby enabling our engineers to work for more than the largest Swedish electronics firm."

Moderator Johns agreed. "Canada has taken the American example to heart — we have a lot of innovation coming from our own startups these days."