A month after Optical Solutions expanded the reach of its equipment to 52 miles in February, Flexlight Networks--at the moment perhaps Optical Solutions’ only challenger in the burgeoning gigabit passive optical networking (GPON) market--announced a trial by British Telecom using dense wavelength-division multiplexing to extend GPON to a distance of 135 kilometers (almost 84 miles). The company named a new chief executive officer in January, and its chief technology officer, Oren Marmur, sat down with Telephony’s Ed Gubbins at the OFC/NFOEC show earlier this month to talk about going the distance in the GPON market.

On the BT trial: It’s something pretty unique to BT and their 21st century initiative. While it relies on some unique features of our GPON system, it also dictates a completely different architecture all the way up to the core network. As far as we’re aware, most of the [fiber-to-the-premises] projects here in the U.S. focus on access networks; there are no additional changes in metro or core networks planned. The main benefit is to further reduce capex and opex. The more sites you have, the more maintenance and operational costs you have. If you can deploy GPON in hundreds of sites instead of thousands and serve the same number of customers, that could lower the cost of operations and equipment.

On customers: By now, I think we have at least one large customer on each major continent. We’ve deployed a few networks with China Netcom, the second-largest operator in China. We have some networks deployed with a few small cable operators in Korea. We’re working with a new customer that will hopefully be finalized soon in Japan, which is important for us to show that GPON can really penetrate Japan as well, which so far has been much more EPON. We’ve also sold systems to NTT through Bell Labs. We have a fairly large number of networks deployed here in the U.S., with one of the largest MSOs. That’s been going on for over a year now, but they’re sensitive about approving the press release. Hopefully we can release that soon. The MSOs are very active in the space and much quicker to move than the RBOCs. We have quite a lot of activity in Europe as well, with Eastern Europe being a bit quicker to move. We’ve deployed a few networks in the Ukraine and some of these countries. And we’re going to do a lot of trials with most of the major PTTs in Europe--BT, FT, DT, Telecom Italia and so on.

On MSO traction: Telcos are just as much a focus. One of the reasons MSOs have become pretty active with us is our focus initially had been more fiber-to-the-business. That’s where we started. We’re moving into fiber-to-the-home and the residential space these days with some new products we’re launching. The MSOs have the residential space covered with their cable modem and access technology, but they’re trying to compete more with telcos in the business space. And GPON is a very compelling technology.

On the joint Bell RFP of 2003: We did not have a BPON offering; we submitted our GPON just to sort of make sure we’re on the radar screen. It was a marketing activity to make sure we were remembered. I think they even had a short clause in the RFP asking vendors to discuss their future migration plans toward GPON. They started to think about GPON even back then. Part of the evidence is the fact that, of the three RBOCs, only one, Verizon, actually deployed BPON. I think Bellsouth and SBC are waiting on GPON. Rumors in the industry point to the fact that sometime this year--could be toward the end of this year--another [Bell] RFP could come out for GPON. We’re working with a few partners right now. We want to make sure we have the right tier-one partner to deal with the RBOCs.

On the GPON standard: One of the problems is that there are too many options. For bit rate, the official standard has eight options. There’s 1.25 Gb/s and 2.5 Gb/s downstream. Then there’s four options for the upstream: 155 Mb/s, 622 Mb/s, 1.25 Gb/s and 2.5 Gb/s. Eight options is a bit too much. My guess is we’ll eventually see only two options: the 2.5 Gb/s down, 1.25 Gb/s up--which is what we’re doing today--and possibly the 1.25 Gb/s symmetrical. In FSAN, the operators have circulated a questionnaire among the vendors asking them what will be the price of those different options. All the different speeds and bit rates at decent volumes would be pretty much the same cost. There’s fierce competition going on with the optical vendors. [Price is] basically only a matter of volume. If you’re going to pick the high-speed version, that would drive volumes, and eventually you’d get to the same cost point as you would with the lower speed options.

On optical network terminals: [Outdoor ONTs] are very unique to the North American market. The other markets we’re playing in, Asia and Europe, often require an internal ONT. In [a multi-dwelling unit] environment, it’s less of an issue because you usually put the ONT in the basement of the building. Obviously the RBOCs want it outdoors because it’s easier for the technicians to come over and fix or check on the ONT without having to schedule an appointment with the customer. That makes the [potential] integration with a home gateway much more difficult. The home gateway option is something we’ve definitely heard from several customers. Still, I think the market requirement for such a product is still in its infancy. Some are also a bit wary of this because all these home gateway boxes are extremely cheap, and customers tend to replace them much more frequently than you’d replace an ONT. There’s the concept of having the ONT as the service provider equipment that will be installed and probably live there for at least 10 years. Then you have your Linksys or whatever low-cost router at the home that you replace every year or two because there’s a newer unit with new features. It’s a complex tradeoff between tighter integration and possibly lower cost and, on the other hand, flexibility. You probably have a router at home. You can go in and open ports and change the configuration. No service provider would want the customer to access the ONT portion. That’s another obstacle. How do you allow the customer to access the unit but only portions of it and not other portions that could affect the service itself?

On ONT economics: We started in the business market. Price is always the number-one factor. Even in the business portion of the access market, margins are tight. Still, the price target is not as harsh or as aggressive as it is on the residential side. So we’re producing our own business ONTs, and there are nice margins on those ONTs, even in the relatively low volumes we’re selling to right now. The residential markets are much more challenging. We’re launching some of our residential products later this year, and we definitely see pricing being a great challenge there. That’s why we’re trying to find partners that could help us initially, at least with production. Later on, maybe we can even license part of the technology and have them mass produce the ONTs and sell those ONTs to other customers as well. There’s no real example of this type of vendor in the marketplace yet.