Browse by year:
October - 2000 - issue > Cover Feature
All-Optical Network Is Only A Dream
Friday, November 21, 2008

Q: Kids in grade school learn about Thomas Edison, Alexander Graham Bell, and so on. How do you think the work that you and others in the optical networking industry are doing right now stacks up in importance to those kinds of innovations — the light bulb, the telephone?

A: I think it’s important, even though the magnitudes are very different. In that part of the industrial revolution, the innovator, entrepreneur, and businessman were all one person — and the whole industry grew at a certain rate. Today, the network itself creates a very new environment, because in the connected world, people can create value wherever they are, and in whatever way they want — and the business cycle moves very rapidly. The credit that the entrepreneurs deserve is the rate at which they can take all these pieces together and put them to good use.

Q: What needs to happen before the dream of the all-optical network is realized?

A: Well, I’m not sure that the all-optical dream is a valid dream — not in my mind. It’s crazy, the way Wall Street works. People like optical networking and, therefore, if something is “all optical” it has to be much better than just “optical!” That’s the theme that a lot of people have been playing up. All-optical means that you convert information into an optical format and you leave it in the optical format all the way until it gets out at the other end — which is not feasible. It’s not possible to keep the information in the light format for that long. You have to convert it back into the electrical format because signals get corrupted and so on. So what you do is you create a lot of these optical islands and you connect them. What happens as time goes on is that those islands get bigger and bigger — maybe someday you’ll have all optical.

Q: Won’t we be able to back an optical pipe straight up to a computer?

A: No, no, no, that’s very far away. For example, just connecting you and me for a data connection, optically, is light-years away. There are just a bunch of pieces that need to get connected before we can talk. And all those connections will not happen in a pure optical format — let alone my computer talking to your computer directly and optically. Even just getting a connection where my information gets converted into an optical format and leading it, in the form of light, until it reaches you is a long, long way off.

But that’s not the real issue. The real issue is, conceptually, you can shoot these laser beams and get these rapid connections going — and somewhere in the network, along the way, you may have to convert them back into electrical format or whatever format you need. But the excitement is the fact that you have the ability, now, to get these big fat connections in a very quick fashion, and change the economics of that whole connection.

So you were saying, “What does it take to get there?” I think the challenge is not so much technical — it’s true that the technology will improve and the product will get better. But it’s just the market dynamics that have to change, because the people who own the right of way in the bandwidth game are the telephone companies, and the telephone companies like to sell bandwidth as multiples of phone calls. That’s why the connections are very expensive. For example, a high-speed connection from Boston to California could be half a million dollars a month, because the way they figure it out is they take the pipe, see how many phone calls they can fit into it, and then they’ll give you a discount. Whereas for data, technically it will be possible to drop it way down. But then the reality is that the whole market has to get restructured, and the best way to do that is to break the monopolies. In the US, that seems to be happening a little faster than in the rest of the world, because of the fact that things are moving a lot faster here due to competition. If you look at India, there’s DoT (Department of Telecommunications) and just one company. So what’s the motivation for them to adopt all this new technology and drop the price? So what every country has to do is to restructure the telephone business and create more innovation.

Q: So, all optical or not, how would you envision tomorrow’s network?

A: All optical or mostly optical, the Internet has to get restructured. The Internet is not a workable model as it’s set up right now. For example, when you get a connection to the Internet, you pay 30 dollars a month, and you’re going to use it more and more every month. But, at the same time, that thirty-dollar price is going to come down. So it’s an equation that’s doesn’t work — you’re going to pay less and less, but you’re going to use it more and more. So the ISP has to pay more money to get more bandwidth — that doesn’t work. So what needs to happen is that you’ll probably get a connection to the Internet for free, but then you’ll get onto the Internet and conduct your e-commerce. Let’s say you go and buy a movie for two dollars, and the movie gets downloaded in one second, maybe twenty cents of that two dollars should go toward paying for that transport, and that’s possible with the kind of technology that Sycamore is developing, where you can get bandwidth on demand. So, what has to happen is you need the ability to set up and tear down these connections in a very rapid fashion and somehow relate the cost of the transport to the cost of the e-commerce.

Q: How much of this has been invented?

A: I think architecturally we have technology in the labs that can make a lot of these things happen. Two things have to happen: One, the carriers have to take this technology and put it into business practice, and secondly the technology has to go through its own learning curve and improvements so that it becomes more and more affordable and therefore viable in more and more applications. But the basic architecture, I think, is fairly well understood.

Q: What about software applications?

A: That comes on top of our software. In other words, “what are all the applications you can develop if, in fact, you have the ability to open these connections on the fly?” For example, a couple of years from now I’m sure we’ll be having this conversation on some video conferencing Web site — you go click, I go click, and bang, we’re connected. We can’t do that today because there’s no way to get a high-speed connection on the fly.

Q: Sycamore is in a hot position right now — the envy of a lot of people in the industry. What’s your strategy to stay on top?

A: I think in this business, the fact that we are ahead of the game is yesterday’s news. It’s so competitive that you have to keep fighting every day. We’re very close to the market and the customers — that really helps us define what needs to get done. And then you get a lot of talented people and just empower them to go make it happen.

Q: Is this business a kind of arms race?

A: This business is good for another twenty years. And so you need two things. You need innovation, because the technology is changing very rapidly. At the same time, to bring that innovation to the marketplace, you need a company that has stability and the ability to survive on its own. And usually those two don’t go hand in hand. And that’s what unique about Sycamore. Sycamore has the agility of a startup, and it’s able to create a lot of products but at the same time it has something that most of the startups don’t have — a management team with a lot of connections in the carrier networks. Because, after all, at Cascade we built most of the networks that are data networks in the world today.

Q: But let’s talk about those startups and emerging companies. What about the “best of breed” networks that those companies are creating versus “end to end” solutions? Where do you see that balance?

A: It’s a question of when you say “best of breed,” where do you cut that line? Do you take it all the way down to the best screw, best board and best nut?

I break the network up into three pieces. There’s a piece where you convert the information into IP packets — that’s one piece of the business, and I think somebody could do the best job and survive that marketplace. The second piece of the market is switching the IP packets — and that’s what Juniper and Cisco and these guys do. They switch the packets and get the packets ready to be transported. The third piece of the business is creating these pipes. These pipes have to be created across the city, across the nation, across the globe. And that’s a business by itself again.

We are in the business of creating pipes. In order to survive in that business of creating pipes, you need four technologies: You need access technology, you need switching technology, you need transport technology, and then you need the network management. You need those four pieces, and then you need a company that has a distribution channel, customer service and professional services. Once you have all those pieces, then I think you have a shot at creating a sustainable business. I don’t think you need to have all three pieces [of the overall network], but if you want to have a business within that “creating the pipes” business, you do need all of those four pieces — switching technology, network management et cetera. You can look at the startups and say “will this startup survive?” and it’s a question of “can they grow into all four pieces?”

Q: Is your goal to grow into a company of the size and scope of, say, Cisco?

A: Absolutely — the opportunity that we have building this infrastructure for the world is a 30, 40, 50 billion dollar a year business, and it’s growing rapidly. And so the opportunity is there for us to build a substantial company. And Sycamore was founded with that goal in mind.

Q: How do you maintain nimble and aggressive innovation as you grow into a large public company?

A: Number one is the mission. The mission has to be very clear, and exciting enough that it unites the people. I think it’s the fact that we’re doing these exciting things and changing the world that continues to attract a lot of very good people. Once you have a lot of good people, it’s a question of empowering them so that they feel like it’s their game. We’re up to about 650 people now, and each one of them feels like they’re a founder of the company — and that’s something that’s very unique.

Q: Obviously, that engineering talent is life and death to you. So how do you go out and find these people and attract that intellectual capital?

A: That’s the most important thing that the company can be doing. Constantly, we’re in touch with a lot of people — the thought leaders in the industry. A lot of times, surprisingly enough, we actually get approached by people, and then it’s a network — it’s very synergistic, because the excitement is very infectious. So a lot of it is just networking, and truly believing and acting and putting your money where your mouth is. That’s really helping us build.

In some ways, I think it’s easier to build these companies on the East Coast than on the West Coast. On the West Coast, it’s good that people are entrepreneurial, and that they’re “hustlers.” But there’s just a little too much hustle there. So I’m not sure I’d want to be building a company on the West Coast at this point.

Q: You mean it’s easier for you to keep people in house on the East Coast?

A: People [on the East Coast] are very focused. They really believe in doing something for the long term. It’s really the true excitement about building something as opposed to everybody just talking about stock options and stock prices. I’m sure, for the people trying to build a business on the West Coast, it must be a nightmare.

Q: Do you have any plans to start more companies in the future?

A: No. I’m going to make sure that I give all my energy to building Sycamore.

Share on LinkedIn