Daily Archives: March 19, 2010

VCs Rue 10GbE Chip Bets

The Wall Street Journal ran a piece yesterday on the bath venture capitalists have taken on their investments in startup companies developing chips for 10-gigabit Ethernet (10GbE) networks.

Citing numbers from Venture Source, Scott Denne writes that VCs have put more than $800 million into 10GbE semiconductor vendors since 2001.

How’s that working out for them? Not so good, apparently. Venture-funded startups in the space, such as Neterion Inc. and NetXen Inc., have been sold in the past year for a fraction of the amount that had been pumped into them.

It’s easy to understand why 10GbE semiconductor firms can be had for garage-sale prices. To this point, the market signals haven’t been bullish. Consider, for example, that vendors of 10GbE networking gear — from Cisco to Juniper to 10GbE specialists such as Arista Networks — spent less than $130 million on the networking chips last year, according to Dell’Oro Group. The previous year, Dell’Oro estimates, less than $100 million had been spent on 10GbE semiconductors.

Yes, the market is experiencing year-on-year growth, but do the math: If more than $800 million has been funneled into investments in 10GbE chip startups and those companies have generated about $230 million in aggregate sales during their two best revenue years, we don’t require the services of a PhD in statistics to calculate that the valuations of those startups are under intense downward pressure.

As for why 10GbE chip startups have underperformed, the WSJ cites technological factors, such as challenges associating with suppressing power consumption, and market competition from semiconductor-industry stalwarts such as Intel and Broadcom. Those are factors, I’m sure.

Not mentioned in the article is, well, the economy. I think it’s the most salient factor in the retarded development of the 10GbE market. If you look at the value chain, it’s obvious that revenue realization must begin with customers willing to buy what’s being sold.

Now let’s follow this particular chain in reverse: the 10GbE chip purveyors depend on a robust demand from 10GbE networking vendors, who depend on strong demand from service providers and large corporate datacenters, who in turn depend on demand from their customers, who might rely on the healthy pocketbooks of consumers. As spending contracts, and less money enters the value chain, everybody suffers, including the VCs who funded the 10GbE chip vendors.

As nearly every shred of available evidence shows, the high-flying Internet economy of 2000 never regained its former glory after the bubble burst later that year. What’s more, for most of the past two years, we’ve had a very dysfunctional value chain in the 10GbE space, which is why this upgrade cycle for the latest and greatest iteration of Ethernet has not been as vigorous as its predecessors.

James Carville could explain what’s happened to the 10GbE chip startups: It’s the economy, stupid.