The TSC Streetside Chat: Desh Deshpande of Sycamore
By TSC Staff
Originally posted at 5:17 PM ET 12/9/00 on RealMoney.com
Sycamore Networks (SCMR:Nasdaq - news) co-founder and chairman Gururaj "Desh" Deshpande saw where networking technology was headed some time ago. So he set up shop far out on the optical Internet frontier, seeking to capture that business. Now, he's putting the pieces together and adding one or two customers every quarter.
TheStreet.com's Scott Moritz had a chat with Desh recently in New York, in which the executive dispelled the notion that Sycamore needs to pair with a router maker like Cisco (CSCO:Nasdaq - news) or Juniper (JNPR:Nasdaq - news) to fulfill its strategy. Deshpande also explained the economic imperatives of gigabit Ethernet, his take on why the equipment spending slowdown won't hurt his business, and why his critics are wrong about the strength of his software.
--------------------------------------------------------------------------------
TSC: An assertion I hear a lot is that some companies would fit well together with other companies because the technologies match well. For example, optical and Internet protocol, the thinking's been that there needs to be a big marriage of the two technologies. Would this be viewing it the wrong way, or is there a way you can migrate to IP on your own, without the need of a big player in IP already?
Deshpande: I think we can split up the business into converting information to IP packets, switching IP packets, and the creation of the pipes to connect those packets. Why break it up that way? Not only because of convenience for a business model, but also because I think that's roughly the way the world gets built, as well: The people who have right-of-way tend to lay the pipes. And then the ISPs use the routers and convert the traffic into packets in the network.
The Ciscos, Nortels (NT:NYSE - news) and Lucents (LU:NYSE - news) like to be in all three segments. They like to do end-to-end. We are focused on just creating the pipes for the business. The big advantage is the fact that we understand what these pipes get filled up with, the ATM [asynchronous transfer mode], frame relay, IP and all that stuff. But at the same time, the innovation in creating the pipes is just so rapid and also the guys that buy the equipment are just so different -- I think just the fact that it's a global business, a huge business, so we're going to stay focused on our product group. And then, when we change that business substantially, at that point, once we've conquered this part of it, what we should do is open up on the other sides. But right now we're going to stay focused on that part of business.
TSC: A lot's been made of gigabit Ethernet. It seems to be gaining momentum in the network as a trusty and simple method of carrying the IP traffic. How does Gig-E figure in with your plan?
Deshpande: Actually, it's part of it. I'm really bullish on gig-E's because the way you rent the pipe has a lot to do with the way you fill the pipe. For example, if you've got a 10-gigabit pipe today, it's maybe a $250,000 a month, 10-year contract, so that's $25 million to $30 million. Now, the way you fill it is by getting OC192 cards on the routers. Each interface card on the router is about $250,000, but then it's totally irrelevant because you're paying $250,000 a month to get that pipe.
So who cares, right? Let's say you can get this pipe, not for a 10-year contract, but just for three weeks for $10,000 to $15,000. So you haven't quite changed how expensive the connection is, but you're allowed to get it for a shorter period of time, then obviously paying a quarter of a million dollars to get this connection is an issue. And the way you fill a pipe, let's say, within this building, is very different from the way you fill the pipe on a wider network, because within a building, once you put in the wire and the capital, the connections are free.
So the Ethernet has driven away the cost of the volume. You can get a gigabit connection for $800 and a 10-gigabit will be probably $5,000, and it will be going down to $1,000 in another two years. So, in the local networking context, we can fill a pipe more cheaply, a lot more cheaply than you can on the wider network. And so, as the economics of the pipes change, I think people will start filling up those pipes more and more with gigabit interfaces. In fact, we have a very large network we're rolling out in Sweden -- it's called UTFORS.
TSC: How's that?
Deshpande: U-T-F-O-R-S. Stan Nardell, the chief technology officer there, he came up with this idea about two years ago, and that's when we met for the first time. Together we designed this network that provides gigabit connections all over the place in Sweden. And so we were the first guys to ship those gigabit connections on the optical network. He provides 1-meg, 10-meg and 100-meg connections to people in Sweden. They can get a 10-meg connection for $5,000 a month. And on that 10-meg connection, he has a way to carry your phone calls, your [virtual private networks], all kinds of stuff. It's an amazing network; it's like one big Ethernet for all of Sweden.
TSC: And you guys are making the gigabit Ethernet cards to go into your optical switches?
Deshpande: Yes, we've been shipping them for over a year now.
TSC: And your prices on those cards aren't as high as what you were just mentioning?
Deshpande: We're changing the economics. The fact is that the pipe is getting cheaper and cheaper and what connects to it costs less, so it's all getting cheaper.
TSC: Is UTFORS like the Telseon and Yipes kind of model of delivering gigabit Ethernet connections?
Deshpande: It's a little bit like that, except he's got a more complete solution. He's more a full-fledged carrier. He carries everything, so as an end user, you get a complete service. It competes with Telia, the big telco in Sweden. So he'll be the number two carrier who provides all of this, but he provides all of that with the new architecture.
TSC: Is that how you see things heading?
Deshpande: Oh yes. You know, in Sweden particularly, phone calls and everything else are very expensive, but with this architecture, you know, you pick up a phone, make a call wherever you want, anywhere in the Nordic countries, free. It's pretty amazing.
TSC: One of the things you're probably asked all the time is how's Sycamore's progress with the AT&T (T:NYSE - news) contract? I've talked to a lot of people on this AT&T contract, and it just seems to be either on hold or possibly still on track but smaller than first anticipated. AT&T has said it may scale back the project. How does this affect you?
Deshpande: These are all large accounts, and they all have their own sell cycles. The world is going to transition at its own pace. In fact, this tightening in the marketplace may hasten that transition. Nothing ever goes from 0 to 100% over night. Because of this tightening, people may smarten up and start transitioning from the old to the new architecture a lot faster.
TSC: Which would still require spending. The telcos large and small have much less cash to spend next year. How can that help you at all?
Deshpande: It's more making use of the capital you have in a better way, but that requires a little bit of risk taking, because you're doing something different than you've done in the past. But that may be the impetus to take that risk, because this is some of the classical disruptive technology thing, because every time there's a new thing that comes along, even if the new thing is a lot better than the old thing, there's always a certain resistance to go to the new thing because there's a risk, right? And maybe this tightening will provide that impetus for people to look at that new thing more seriously and put out the pilot projects and get more aggressive about that and act fast.
TSC: In that, can I read your strategy on how you might crack the Bells, in that they might have to move to your technology faster than you'd have to move to their standard for certification?
Deshpande: Yes, absolutely. There's a maturing of the technology that has to happen, which will happen over the course of time. So they have to change, we have to change, and then, the things meet at a certain time. And all we can hope for is that that transition will happen faster and faster and faster.
TSC: That's a good point. As things change, you say they'll eventually come around to your technology.
Deshpande: And they won't yet. You know, these guys are all smart guys. But it's just that there's a right time for them to jump in. Because these are large networks, they cannot jump in too early. So they can't trust large networks on a technology that's not mature enough, even though it has the right feel to it. The technology has to mature and their business plans have to change as well.
TSC: Great. Now, changing gears a bit, you've heard this a lot. Your competitors like to dish out the FUD, the fear?
Deshpande: Fear, uncertainty and doubt?
TSC: There you go. It's been said that your switch software is inferior and that you will always be at a disadvantage until those bugs are ironed out. How is the software these days?
Deshpande: We'd love for all of the competition to fight us on the software. That would be ideal for us. Because if the game is really in software, and really in applications, and understanding the IP and how to route, re-route, and everything else, I think we have the game beat.
Most of the guys right now at Ciena (CIEN:Nasdaq - news), Tellium and Nortel have reached a point where they've gotten their hardware to work, and it's basically an electronic patch panel. But the way it's being done is fundamentally very different than our approach. And that difference doesn't show up until your network starts scaling. And so, if people want to fight the battle on scaling the network, on software, on setting up and tearing down a lot of these connections, on building a network and chopping it up into a hundred different pieces, and letting each of the pieces survive on its own merit and all those kinds of things, that's right up our alley.
TSC: One last question: Your stock price has certainly taken a major hit lately. Is this on news of anything, or do you think it's just market volatility?
Deshpande: You are more of an expert on that than I am. We only build companies.
thestreet.com
Jack |
