SAN JOSE, Calif.--The rules of the game are changing in venture capital funding for semiconductors, in which the fabless model has become less appealing for VCs in the marketplace, according to a former chip executive and now venture capitalist.
"There is a real question about the model of the fabless company," said Yoav Nissan-Cohen, the former co-chief executive with Israeli silicon foundry provider Tower Semiconductor Ltd., in a recent interview with Silicon Strategies.
"The whole fabless model has become less appealing for the VC," said Nissan-Cohen, now a venture partner with Pitango Venture Capital, one of the leading venture capital firms in Israel. The Herzliya-based firm has helped start over 80 companies, managing funds in excess of $700 million in committed capital.
This is not to say that venture capital funds have dried up. For example, so far in 2004, up to and including Wednesday April 14, Silicon Strategies has reported on 58 startup chip and related companies that have raised $833.5 million from more than 80 venture capital firms and other investors (see April 28 story).
In fact, the VC market is heating up to some degree after a lull-- especially for semiconductor, wireless, and networking technologies--but the risks and funding requirements are much different than before, according to Nissan-Cohen.
In the past, for example, a semiconductor startup could receive from $10-to-$15 million in funding and expect to break even over time, he said. "Now, it's two times that amount," he said. "Today, a startup may require $40-to-$50 million just to break even," he said.
VCs are also taking a harder look at chip startups. "Today, the NRE costs are much higher to bring a product to market," he said. "It also seems like many product markets move so fast. They become commodities before the startup can make money."
One example is Wi-Fi or wireless local area (WLAN) networking. In recent times, a plethora of Wi-Fi chip startups formed to capitalize on the booming market, but few have actually made any money in what is now an overcrowded and commodity business.
"Me-too" products have become less interesting for VCs, he said. "We're looking at disruptive technologies," he said. "You are also looking for a group of people who can deliver."
Among those "disruptive" technologies include 3G, video streaming, and video for mobile applications. "Wireless continues to be interesting, especially as it applies to the home," he said. "Many of the new applications are headed for the digital home and TV."
It is also unlikely the market will see a plethora of new foundry startups, he said. In the mid- to late-1990s, a collection of foundry startups emerged in order to take advantage of the booming chip-outsourcing trends.
Gone are the days of the "mom and pop" foundry startup. "Nobody can start a foundry without substantial government support," he said. "Countries want big technology clusters."
As many have predicted, Nissan-Cohen believes consolidation is the future for the foundry business. "To start a foundry industry, you have to do it large scale or come up with a unique technology," he said. "I'm not sure if the foundry industry has the room for the many players we see today."