Quenching his thirst

Wed, 02/28/2001 - 7:00pm
Leslie Ellis, Technology Analyst

CED: By joining Aurora Networks, you really put them on the map. Yet I'd bet there are still people who perhaps aren't all that sure they know the company. Let's start there.

Werner: Aurora is focused on one thing, and one thing only, and that's building optical network components and optical networking systems for existing and next-generation broadband networks. We'll have a wide array of components and products that are backward compatible with existing HFC networks, as well as a very unique fiber-to-the-home solution. So while we're focused on next-generation, we're extensible to fiber-to-the-home, and we're backward compatible with today's equipment.

CED: What's the head count at Aurora?

Werner: Forty. And of those, well over 30 do nothing but engineering. Digital engineers, a couple of mechanical engineers, and the rest on the optical engineering side. I think that's significant. People say, gee, how can you compete with the big guys? But if you look at the big guys . . .

CED: The big guys being Harmonic...

Werner: Harmonic, Motorola, ADC, Antec and S-A. So if you look at them, they have maybe 20 engineers that are focused on the same area that we are. The big guys have lots of engineers, but their focus is split. Digital set-tops, other products.

So as far as direct optical next-generation networking, using the best components and techniques, for the fastest and most cost-effective solution, I think we've got more of a dedicated focus than anyone out there.

CED: This was a fairly big decision for you. You went from AT&T, a company with tens of thousands of employees, to a company with 40 employees. I know that at the time, you were considering a few other offers. What was it that lit up this path?

Werner: A number of things. Yes, we had close to 40,000 employees, maybe more with MediaOne. A huge company. And AT&T was great; I'm glad I worked there. But I thought, all of a sudden I'm at the age where–if I'm ever going to do something entrepreneurial, if I'm ever going to do something where I've got control of it, this is the time. I looked at a number of very small startups, as well as some medium sized and some very large companies. Large companies are great, but I was … (pauses)

CED: Tired of bureaucracy?

Werner: Ready for a change. So I looked at everyone's business plan, the people, the products. I looked at where I believed the industry to be going. Ultimately, I decided that Aurora had the best product, business plan, group of people, and niche, for what I wanted to do.

CED: The people who are familiar with Aurora tend to correlate it with AT&T's Salt Lake City project, to drive fiber to 50- home or smaller nodes. But yet you've told me that's only part of the story.

Werner: That's something I'd like to address. We're working to dismiss that association as the only way to view Aurora. Not that we don't want to serve AT&T! (laughs) We'll serve them hopefully very well. But we're not solely an AT&T-centric company.

CED: You mentioned some beliefs in where the industry is headed?

Werner: Yes. I firmly believe that broadband to every home, certainly in major metropolitan markets, is on target. We will achieve that, as a country. It's just a matter of time.

I also believe that capacity requirements will get higher and higher as applications get richer, and as we get more and more broadband. Because of that, the trend toward fiber deeper will always be there.

CED: Sure, but how far, how fast?

Werner: I'd say there's consensus that the end game is fiber-to-the-home, and fiber-to-the-device. It's not if, it's when. I'm not going to say it's this year or next year. I'm not even saying it's necessarily this decade. But it certainly is this century. (laughs) Without a doubt. It's probably in the next couple of decades.

CED: Your plan, then, is to be there when it happens? All along the way?

Werner: Yes. We've looked at the latest optical components that are out there. We're very close to several of the companies we work with, sort of under the radar screen, that have the latest and greatest in optics. We think we have the leading technology in EDFAs. We've got some neat stuff in dense wave division multiplexing. And we've got a really good group of digital engineers that are doing all sorts of digital techniques, like digitizing the entire upstream spectrum.

CED: All the way from the home? Or from the node back?<

Werner: We can do it from the home. But our initial implementation is at the node. We digitize the 5 to 40 MHz, and then we do what's essentially a digital daisy chain, to bring it back. It allows us to take very inexpensive optics and get DFB (distributed feedback) performance, out of say, an F-P (Fabry-Perot), or some of the other unique lasers we work with.

CED: So the idea is to keep the cost at the F-P level, with functionality of a DFB?

Werner: Exactly. And, we're mindful of space restrictions at the headend. One of our pieces serves 256 optical nodes off three rack units. So if headend space is a problem–which, trust me, it is (laughs)–we are certainly aiming to help that.

CED: What's the common reference point for nodes served versus headend space requirements?

Werner: Typically, it's around 12 nodes per three rack units. In addition to this, if you have standard HFC, there are additional benefits. Like, we have 16

100-base-Ts that we can run over the optical node, that come out of this digital stream. It's a sort of free byproduct.

It's useful in HFC situations where you may say, I'm going to drop this node in a business park, or in a downtown area. Then you can sell Ethernet, at native protocols, right off of our interface. For very good economics.

CED: Interesting. Because the commercial side has always been sort of like, yeah, it's important, and it's probably big money, but we'll get to it later. This makes it potentially sooner, rather than later?

Werner: Especially for small and medium sized businesses. You can drop this node in right behind it to serve it. We're also working on a GR-303 interface. So you could do phone as well as fast Ethernet, yet it's the same transport that can carry all the 5 to 40 MHz and the legacy equipment people have. DOCSIS, set-tops. Then, on top of all that, it's totally dynamic. You can change it via software from the headend. Another useful application is bulk data to MDUs–where the PCs in each apartment only need an Ethernet card, and you supply the high-speed Internet.

CED: In your opinion, is cable's HFC network future-proof?

Werner: Nothing in this world is future-proof. But I think the HFC architecture is probably the best network to provide voice, video and data services today. Over time, with extraordinary success, modifications will be needed to address bandwidth consumption, but it's still the network best suited for the challenges. That goes for today and tomorrow.

CED: Let's talk about fiber counts. And node sizes. For the past couple of years, the trend has been 2,000-home nodes down to today's rule of thumb, around 500 homes per node. Does the 500-home node continue, or does it keep slimming down?

Werner: It depends. I think there's an awful lot of capacity in a 500-home node. You can do an awful lot of high-speed data, and other services over 500 homes. The question is, what makes you drive fiber deeper?

Well, over time, we know that the Internet will get richer. And we know that we've never overbuilt bandwidth in this industry. It's always been, yeah, 400 (MHz), 550 (MHz), we've got what we need, 600 (MHz), 750 (MHz), 860 (MHz). There's never been too much bandwidth, and I doubt there ever will be too much bandwidth.

But, 500 or 600 homes at 860 MHz is a lot of bandwidth.

CED: Downstream, yes. But upstream can get scary.

Werner: Yes. Peer to peer computing trends start to put pressure on upstream, as do other things. So we're putting together tools and products that take the economics, and take the knee in the cost curve, which had traditionally been 500 homes, and pushing the knee closer to 100 homes, or 50 homes. To the point where you can go passive from the node to the termination, and have a nice evolution if you ever do want to go to fiber-to-the-home, let's say in a really dense area. Or in an apartment building. This product is there, and it's extensible. So we're bringing the tools to push the knee in the curve further out.

CED: When does that start to happen? AT&T has been the most aggressive in the really little nodes, but I've not seen much momentum from the other MSOs.

Werner: Actually, there is. I think there is. And here's where the economics work pretty clearly: If you're in a green field situation, so let's say it's plant extensions, it's cost parity today.

CED: To go to . . . ?

Werner: Passive networks. Depending on the density, you'll get between 50, 100, 125 homes off a single node. And you can do that at cost parity for plant extensions, today. With our technology. Not with everybody's. Remember that. (laughs.)

CED: Geez, Tony, you sound like a vendor. (laughs.)

Werner: If you go in an upgrade situation, it depends on how much aerial, how much underground, and how much coax you have to replace. If you have to replace a lot of coax, or if it's primarily aerial, we can still provide a solution that's pretty cost comparative. Even if all the coax is good, and you have to replace very little, and you've got, say, 40 to 50 percent underground, then it's very hard to compete. Because we'd be requiring them to dig up some of the underground, and that's where the cost comes in.

But places where you have a lot of coax to replace, and you've got more aerial than underground–say 75/25–and in new areas, there's no doubt that this is where the knee in the curve is.

CED: Let me just be the devil's advocate for a moment. If you look at construction spending trends, you're entering the game pretty late. These days, everybody's saying, well, we're almost done building. You're coming into this with a series of products, at the tail end of the big spend.

Werner: You're the add 'em up and divide it person, so you know this better than I. (laughs.) If you go back and look, there have been a series of construction waves in this infrastructure. And I think there will continue to be a series of waves. There's still a lot out there.

Even if you just take plant extensions, there are tons. That will continue. There are still a lot of networks that haven't been upgraded, slated for this year. There are a lot of them that fell to the end of the last (spending) cycle, that need all their coaxial cable replaced. So that's good for us. There are the international markets.

There are pretty good market opportunities out there. And, we're coming into a market where there are no other new IPOs. There's not as much new thinking. The industry has four or five very solid manufacturers with good vertical momentum–Motorola, S-A, and others–that have digital set-tops, cable modems, and all the other stuff they make, in addition to opto-electronics. I believe there's an opportunity to come in here and be the leader in just the opto-electronic segment.

To your point, if we had done this three years ago, we surely would've had three great big years. But I think there's still a couple of awfully good years left, even in this round. And as people start splitting nodes–this evolution that will happen as services get better–we will have some natural products that lay into that network.

CED: Say I'm a cable operator that maybe isn't 75 percent, and I'm really just to the point where penetrations are starting to saturate my bandwidth capacity. Then what would you say to me?

Werner: We've got some great tools. And, I'd point out that we're building modules that are compatible with other companies.

CED: I wouldn't have to switch everything out and go to you. Even though you'd probably prefer that.

Werner: (laughs.) Right. That would be the best thing.

We'll not build product for everyone that's out there. But we have started to work with one of the manufacturers that's got a very high market share of optical nodes and other equipment, and is a good partner. We're working with them to use their node platform, and put our product into it, for instances where a customer wants their node platform. Or if a customer's already got their node platform, our product will buckle into it.

CED: Let's move up to the business level. You're a CEO. That means you have to talk about things like your top-three strategic goals?

Werner: The strategic goals come in waves. The first is to sew up as much intellectual property as we can. Patents. So that anyone else coming in has a very hard time penetrating and getting past our level of innovation. We want to be the most innovative, sew that up carefully, as best we can.

Second is to get good market understanding, good market acceptance, good relationships with our customers so they know what we are, and don't misunderstand what we are. A lot of people think we're just one very small niche, and we're not.

The third piece is more tactical, but it's to hit the cash break-even point, which we're forecasting to be first quarter 2002. Once we get there, then we've got an awful lot of leverage and options.

CED: Speaking of which, how is Aurora funded? And, if the stock market ever claws its way out of the dungeon, are you thinking IPO?

Werner: We've got enough cash, even without exercising credit, to work through this year. We expect to start shipping product in the second and third quarter. We will be raising some additional funding here in the first quarter, more to bring in some people we consider strategic partners. And to make a few small acquisitions. We won't be an acquisition binge company at all, but there may be a couple of small acquisitions that will tie into some of our intellectual property stuff. Plus, then we don't have any worries as far as ramping production. But the money we've got is sufficient to build the inventory we need and to take on the additional employees for this year.

CED: So if you were going to go the IPO route, that'd be 2002 or after.

Werner: An IPO is certainly a nice place to get to, in some respects. But if you talk to a lot of the people that work in public companies these days, they'd rather have stayed private for some time.

CED: A lot of scrutiny.

Werner: Tons of scrutiny, and you get forced to do things that aren't always necessarily in the best interest of your long term business. So our idea is, the IPO may be a byproduct, but it's not an objective. Our objective is to be a company with good fundamentals, that over time, becomes a cash positive company. Then, build value and culture on that. Now, from that we will likely do an IPO, because eventually the venture people will want their liquidity. I wouldn't mind some liquidity, and some of the founders may want liquidity. But we'll do it for that reason. For liquidity. Not for operations.

CED: There's a lot of advancement, or at least talk of advancement, around optical switches. The advantage being the ability to send light as light for long distances, without having to take it down to RF and retransmit it. Is there a point in time when that becomes applicable to a cable network?

Werner: I think it does. All of these networks are converging to be a lot more than cable networks, anymore. Optical switching starts to make sense.

CED: I've always understood that the benefit is that you can go much farther. So I wasn't sure if that applied, for cable.

Werner: There are tons of benefits. One is what you said, distance. Another is you don't add in a lot of latency, from processing. You simplify sites. As the circuits get larger and larger, optical switching makes sense.

It used to be, in the telecom world, you'd get a big customer here, a big customer there. They'd want a circuit. You'd nail up a circuit. The circuit would last for 20 years, and the telecom provider, whoever it was, AT&T or whoever, would be really happy.

Today, that doesn't happen. You have a big company here, they set up a branch here. They shut down that branch, and move it over there. It's always changing. So the ability to set up and tear down circuits quickly and efficiently will drive it, too. I don't know if cable necessarily will have a huge appetite for it this year or next year, but within five or six years, you'll see a good proliferation of it.



Share This Story

You may login with either your assigned username or your e-mail address.
The password field is case sensitive.