A group within the Global Semiconductor Alliance (GSA) is trying to figure out how to prevent chip startups from dwindling towards extinction.

OK, maybe it's not that dramatic. But the point is that chip startups are now rare. The concern is that without chip startups to assert new ideas, the industry's development could slow down, at least in the U.S.

The problem is that semiconductor development is unavoidably expensive and any revenue payoff is years in the future. A mobile-app startup can happen with two programmers and a long weekend. Chip companies need tens of millions of dollars and years of work. Which one do you suppose venture capitalists gravitate towards?

The Capital Lite Working Group, started by the GSA in mid-2011, wants to develop new models for nurturing chip startups. The group hasn't been a secret -- it got some shout-outs during the GSA's awards banquet in December -- but it's getting its formal launch Monday, when the GSA plans to divulge details about its mission and tactics.

The group is gunning for quick results, says Ralph Schmitt, CEO of chip vendor PLX Technology Inc. (Nasdaq: PLXT) and chair of the Capital Lite executive committee.

"We want to get something funded in this model by the second quarter of 2012," Schmitt told Light Reading in November. "I think we have a really good chance of making that happen."

To that end, Capital Lite has recruited venture-capital firms to create a semiconductor-only fund. But Capital Lite also wants to change the fundamentals behind starting up a chip company.

The cost to produce the first prototype of a chip can be US$35 million, the group estimates. Schmitt and others think the aggressive use of cores -- reusable chip designs -- can bring that to $15 million or $20 million. Cores would let engineers skip the mundane parts of a chip and focus on the innovative aspects of a design.

Where to get those cores? Maybe from bigger companies -- maybe in exchange for future royalties or even for equity in the startup. That's one proposition explained in a paper by Capital Lite member Amer Haider, also vice president of business development at Cavium Inc. (Nasdaq: CAVM).

There might also be room to change the way chip companies pay for the software tools they use. Electronic design automation -- software for designing chips -- is really expensive and sold by a small clique of vendors that includes Cadence Design Systems Inc. and Synopsys Inc. (Nasdaq: SNPS).

Of course, there's a trade-off. If it becomes cheaper and faster to start a chip company, it might also mean lower profits for investors. The kinds of companies funded would more likely be acquisition targets than IPO candidates. But at least they'd have a better chance of getting started.

About 15 semiconductor startups get Series A funding annually, and the industry has been churning through 50 mergers or IPOs per year, Schmitt figures. "So we're about to hit a wall. There's not going to be anything coming through the pipe any more that people can acquire," he says.

— Craig Matsumoto, West Coast Editor, Light Reading