We recently brought you news of Silicon Navigator (see Nov. 15, page 4), which plans to build a platform around the OpenAccess database that will let startups, research groups and universities build EDA applications quickly. With no need to build a database infrastructure, the developers can devise algorithms and simply write engines that plug into an existing infrastructure.

"When EDA companies start up, they spend 20 percent of their time on a new idea and 80 percent of their time rebuilding what other guys have," said Silicon Navigator CEO George Janac. "If the industry is going to continue to innovate, we have to lower this barrier."

Another provider taking the software component model approach is Verific Design Automation, which this week will roll out a SystemVerilog parser (see story, page 46). By providing VHDL, Verilog and now SystemVerilog parsers to small EDA vendors and startups, Verific is allowing those companies to field HDL-based applications faster. Similarly, Concept Engineering provides schematic viewing software that's used by many EDA vendors.

What all these approaches do is lower the barriers to entry for startups, internal CAD groups, university researchers, small EDA vendors and engineers who wake up one morning with a brilliant idea. It turns EDA into more of a "cottage industry," accessible to a wider range of contributors. That could be good news for users, but what about commercial EDA?

The short answer is that the big vendors will face more competition. If an internal CAD group or startup devises a better synthesis or signal-integrity program, for example, it could replace a big-vendor tool.

But the big EDA vendors can take another approach. What Mike Fister was able to accomplish at Intel, and hopes to replicate as CEO of Cadence Design Systems, may be the answer. The idea is to forge partnerships with large customers that involve the right combination of tools, services and intellectual property to get a product out the door.

That's not something a small startup can do, and it's a "wheel" that will be hard to reinvent.

Richard Goering is managing editor of Design Automation for EE Times.