When the cost of chip design is going out of sight, why is there no corresponding EDA innovation to solve the design problems and reduce the costs? The EDA business is not delivering this crucial value through innovation, even as it complains that it does not receive value for what it does deliver. How did EDA get into this situation?

Let's start by debunking the myth that spiraling mask costs are to blame. In fact, as Gary Smith of Gartner Dataquest pointed out at the 2005 Design Automation Conference, mask cost amortized over gate count is falling. But Smith steered toward what's really going on when he noted that roughly 27 percent of engineers are using proprietary tools developed in-house. That's about twice the historical level--and a reversal of the dynamic that created the EDA business.

Bernard Meyerson, chief technologist of IBM's Systems and Technology Group, has posited that scaling "broke" at 130 nanometers because "atoms don't scale." That scaling discontinuity presents a growth opportunity for EDA, but the response of the leading EDA suppliers has been inadequate.

The economic consequences are clear. According to a Gartner/Dataquest report from December 2005, industry revenue grew from slightly over $3.7 billion in 2001 to somewhat more than $3.9 billion in 2004--a compound annual rate of less than 2 percent.

The leading EDA suppliers' failure to deliver the requisite sub-100-nm tools could stifle innovation in a chip industry dependent on design technology to leverage the latest processing technology. And any inability of the chip industry to service its high-growth markets will stifle EDA's own growth.

Over the past 10 years, the dominant EDA companies have accessed innovation by acquiring smaller companies--after the smaller players' technologies have achieved market success. That "place your bets on the horse that won" strategy was effective before 130 nm, but it is not effective at delivering the pace of innovation necessary for sub-100-nm success. Worse, the dominant EDA players' "all you can eat" bundled pricing strategies are commoditizing the business. Price competition depresses revenues. Bundling increases the barriers to entry. And the dominant companies consume customer EDA budgets to the point of crowding out the innovators.

EDA innovators must race to support new process nodes. They must understand the customers' problems and the fabrication process world, and they must take responsibility for outcomes. Today's giants got to the top by meeting changing needs in timely fashion. It's time for history to repeat itself.

By Lucio Lanza (lucio@luciolanza.com), managing director of Lanza TechnVentures (Palo Alto, Calf). an early-stage venture capital and investment firm.