I should clarify that the companies on my list of 2003 startups did not necessarily begin in 2003. They came to my attention in '03. Several are existing companies that offered their first EDA-related products last year.

There's much variety among the startups. Several are one-person, garage-shop-type operations started by unemployed engineers. Others involve sizable teams of EDA veterans and millions in venture capital funding. Many of the 2003 startups are engaged in electronic system-level design, and another hot area is design for manufacturability.

Several are focusing on complex FPGA design. Some are addressing verification, and there are several offering hardware-based, rapid-prototyping systems. Most of the startups are American, though a few European companies are on the list. Nearly all focus on chip design.

The good news behind all this activity is that it brings innovation and excitement to EDA. It also shows that there's a lot of venture capital money returning, and that it's not unusual for a startup to win a round of $10 million or so. This translates into more choices for users, and, hopefully, new technology that can address sub-100-nm problems.

But there are concerns. One is that the big EDA vendors seem to be relying more on startups for innovation. They're either acquiring startups and niche companies or partnering with them.

Another concern is marketplace confusion. The industry doesn't need 60 or 70 functional-verification vendors, but more keep appearing. To go to the Design Automation Conference and find 250 vendors, of which 230 are fighting over perhaps 5 percent of total EDA industry revenues, must be as befuddling for engineers as it is for reporters.

As for the 2004 startups, I say: Bring them on. But hopefully they will be well-thought-through efforts with some stability, funding and good technology. The industry doesn't need 46 new companies every year.

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