Speaking at the Fabless Semiconductor Association (FSA)'s Distinguished Speaker Series and Expert Roundtable here, Sutardja said that VCs need to realize that the traditional exit strategy — an initial public offering (IPO) — is no longer a realistic goal for the vast majority of fabless companies.

Citing statistics compiled by the FSA, Sutardja said that in the past six years the total number of fabless companies has doubled, yet there have been only 16 fabless company IPOs since 2000.

He blamed the dearth of fabless IPOs on two things. For one, he said, most fabless companies have limited human resources — according to the FSA, 67 percent of fabless companies have fewer than 100 employees.

Secondly, Sutardja said, most fabless companies misuse their capital, spending far too little on R&D. He said the typical fabless startup splits its capital expenditures equally among four categories: R&D, management/overhead, fixing mistakes and sales/marketing. Even within the 25 percent spent on R&D, he said, not enough is spent on developing core intellectual property (IP) — too much money is spent on expensive EDA tools, licensing unproven third-party IP and other items that are not key to product differentiation.

Sutardja said VCs, using Marvell's success as a yardstick, want fabless companies to develop multiple products. But, he argued, it took Marvell years to get established as a multi-product company, and with the changes that have taken place in the industry over the past 10 years, it could take today's fabless companies even longer. VCs should be pushing fabless companies to concentrate on markets where they can provide a truly differentiated product, he said.

"One could say that the VCs are part of the problem," Sutardja said. "They don't realize that things have changed. A lot of things that [Marvell] did have gotten a lot more complex."