Chipped away

Venture capitalists have posited all sorts of strategies for surviving the tough economic climate, from throttling back investments to investing more to ensure a portfolio company can survive longer with no additional infusions. It's becoming clear the latter does not apply to chip startups.

Semiconductors, which provide the foundation for just about every information technology product and an ever-increasing number of consumer products, are themselves facing a shortage in the investment needed to spur innovation. After weeks of anecdotal signs that venture investments in chip companies were falling sharply, the Global Semiconductor Alliance this month reported that third-quarter venture investments in the sector had plummeted 57% from the year-ago quarter and about 44% from the second quarter.

Mike Kwatinetz of Azure Capital Partners in San Francisco is one of many VCs who admit they'd be hard-pressed to fund a chip company, given the large amounts of capital -- generally between $25 million and $50 million -- it takes to get one off the ground. For some, the prospect of investing in semiconductors is so far-fetched it's a laughing matter.

"I wouldn't invest in [a chip startup]," says Brad Farkas, managing partner at I-Hatch Ventures in New York, who paused and had a chuckle before answering the question. Farkas says that while there is an undeniable need for innovative technologies, the economics of starting a chip company are daunting.

Unlike so many other businesses that can initially develop a product on the cheap before investing aggressively to ramp up mass production and marketing, a startup working on a single new chip design generally will have to spend at least $25 million on development before it can show a new product that could compete against those made by Intel Corp. or other established chip companies. It's a big gamble in any climate, and in the current economy, many view it as one not worth taking.

The challenges of raising the money to support even those more mature startups with established products was underscored recently when Fluidigm Corp., a South San Francisco, Calif., company that had raised about $150 million in venture capital, canceled a planned $80 million initial public offering. The company has about $32 million in cash remaining.

Another promising chip startup that did manage to go public three years ago was MathStar Inc. of Hillsboro, Ore. But the company, which made customizable chips with a proven performance, found it was still hard to take on more established chipmakers because customers were fearful of lengthy business disruptions that could result from switching to a new supplier. Earlier this year, MathStar hired Core Capital Group LLC and said it was looking to sell off its assets.

Mark Heesen, president of the National Venture Capital Association, argues that the recent flow of money away from chip startups and into other industries such as cleantech is part of the ongoing cycle in which industries become commoditized and gradually cease to attract investor interest.

"Semiconductors have been in a rough patch for a long period with no new whiz-bang changes in technology," he says. "I don't know of any VCs who see any dramatic change in this space."

Heesen says that although there remains some innovation around the edges of chip technology, it's no longer the sort of transformative work that once occurred. "You have to look at the history of information technology. This is the same thing that happened with computer hardware years ago. Chip technology may still be a good investment but not a good VC investment," he says, referring to the exceptionally large returns that venture capitalists look for.

Some, however, insist there is a great need for more efficient and higher-performance chips to make smartphones smarter and faster and add functionality to a host of household items.

"Innovation is going to always be important to our industry," says John Greenagel, spokesman for the Semiconductor Industry Association. "Every year, someone comes up with something that, five years earlier, we all would have marveled at. But it's a very challenging industry because the costs of bringing a new product to market are extremely high. The costs are the same whether you are Intel or XYZ startup. If you are a small company, you need to be prepared to get a very high valuation right away."

Clearly, some companies remain hopeful. About 35 chipmakers did manage to raise $511.7 million in the most recent quarter.