Gordon Stitt is president and CEO of Extreme Networks Inc., of Santa Clara, Calif. Stitt co-founded Extreme in 1996 and led the company to a public offering in 1999. The company built its early success on Gigabit Ethernet and Layer 3 switching technologies and, although it was hit hard by the broad decline in the networking equipment market over the last year and a half, has attained nearly $500 million in annual sales. Prior to Extreme Networks, Stitt co-founded Network Peripherals Inc. and previously worked for Sun Microsystems Inc. eWeek Executive Editor Stan Gibson caught up with Stitt recently at the NetWorld+Interop show in Las Vegas to learn how Extreme can survive not just a down economy but a rival such as Cisco Systems Inc.
eWeek: Extreme hit a home run on Layer 3 switching and remained narrowly focused on Layer 3. Now, however, youre moving into Layer 4 through 7 switching. What is Extremes destiny, to be narrowly or broadly focused?
Stitt: The answer is a narrow focus. But if you look at what we do, its not so much about Layer 3. What we do is build world-class enterprise networks.
We have two goals. One is to be the clear alternative to Cisco—to build the best large- enterprise network. We do that through Layer 3, but we do believe Layer 4 to 7 is a critical part of that. Data centers have become more and more important as the model for applications has changed. Some of those data centers may be in the enterprise. Others may be on the Internet.
eWeek: Becoming the alternative to Cisco is an ambitious goal. ...
Stitt: In the large enterprise. Were not going to try to be what they are. Everyone who has tried that has failed. Nortel [Networks Ltd.] tried that; 3Com [Corp.] tried that; Cabletron [Systems Inc.] tried that.
Cisco doesnt make a lot of mistakes. But they spread their resources pretty thin across a broad range. That means they are not great at anything, but theyre good at putting it all together. Some people call that being a systems integrator. Our approach has been best of breed. Weve been very careful about what we have chosen to do. Were going to be the best at building large-enterprise networks.
The second goal that we have is to lead in the emerging metro Ethernet market. Our customer base is roughly 70 percent large enterprise and 30 percent metro Ethernet.
eWeek: 10 Gigabit Ethernet is criticized for being too expensive. Do you see prices coming down, and, if so, what is the sweet spot for large-scale adoption?
Stitt: Thats a tough question. The optics cost is whats driving the price.
This year is going to be a trialing year with small deployments. Some of the optics suppliers may choose to drive the price down to get the market going. Will it happen this year? I tend not to think so. But it will get pushed down. Right now its $60,000 to $100,000 per connection. That will come down, but its outside our control.
eWeek: What standards is Extreme participating in beyond 10 Gigabit Ethernet?
Stitt: Were very active in whatever speed comes next. Theres talk of 40 and 100 Gigabit Ethernet. Well maintain a very high level of participation in IEEE.
eWeek: How much longer do you see the current network equipment slowdown continuing?
Stitt: Two answers: If I knew, Id be rich, and if I knew, I wouldnt tell you. The truth is, I dont know. We get that question from Wall Street all the time. Weve publicly said theres been some stabilization in the United States. But I dont know if weve hit bottom or not. But to some extent, we dont care. Weve structured ourselves to succeed in a down market and to take share. The last quarter we grew 2 percent.
In June, it will have been an 18-month slowdown in networking. So, you look at corporate networks and ask if theres enough old stuff that people just have to replace.
eWeek: When will Extreme return to profitability on a GAAP [Generally Accepted Accounting Practices] basis?
Stitt: Were profitable now on a pro forma basis. Last quarter, we wrote off all the goodwill relating to our two acquisitions.
eWeek: So youll be profitable in the next quarter?
Stitt: Were not predicting that, but the two [GAAP and pro forma] statements will be very close.
eWeek: Youre committed to IP and Ethernet. Is there anything beyond that?
Stitt: Not as far as the eye can see.
Ethernet is going to continue to scale. It runs great over copper; it runs great over fiber; people can send Ethernet packets through the air. Its just such a simple, well-understood thing that makes for interoperability and keeps costs low. Is something going to replace IP? Were implementing IP V6 in some products today, and youll see more later this year.