Battle at the Bottom of the Sea

cover story

Battle at the Bottom of the Sea

Assemble a dozen telecommunications people and ask them the definition of "global network" and you'll likely get 12 simple, similar answers. But these simple answers will begin to get complicated as each respondent attempts to work out the details of what constitutes a "truly" global network.

Two global network companies—Global Crossing Ltd. and Project Oxygen Ltd. (both of Hamilton, Bermuda)—are working out such details now. At a time when bandwidth demand is being driven steadily upward by the Internet's popularity, worldwide deregulation, and a generally healthy telecommunications industry, carriers with international aspirations must decide now where to buy capacity to meet tomorrow's demands. Thus, how Global Crossing and Project Oxygen work out the details will determine which company will be known as the world's first—and preeminent—global optical capacity purveyor.

Details, details

The global network undertakings of both companies will be roughly two years in the making, so few observers doubt that pricing and availability adjustments—even compromises—will likely have to be made as each rolls out its infrastructure. Yet, distinctions between the competing approaches to bandwidth provision have already begun to appear.

"The main difference between Global Crossing and Project Oxygen probably centers around their tactics in selling capacity," says Stacey Yates, a senior analyst at KMI Corp., a Rhode Island-based fiber-optics market-research firm. "Right now, Global Crossing is selling as a carrier's carrier, while Project Oxygen is looking to be more flexible."

To be more specific, Global Crossing is selling capacity to carriers who, in turn, sell that capacity to end users or other carriers. Project Oxygen, on the other hand, shrugs off the carriers' carrier label, stating it will sell capacity to anyone who needs it and has the money to buy it. Splitting hairs? Not at all, says Tom Soja, an analyst who has followed Global Crossing since its inception.

"I asked Neil Tagare [founder and chairman of Project Oxygen] if he was planning to be a carriers' carrier, and when he said no, it was a shock to me," says Soja. "Selling to anybody who needs capacity, to me, is a very unfocused strategy. If you start selling capacity to end users and going around carriers who would be your biggest customer, why would those carriers want to support that kind of thing? It just strikes me that it would be as if a carriers' carrier was competing with its own customers."

That sounds a lot like the "pot calling the kettle black" when you consider Global Crossing's recent purchase of Frontier Corp. The acquisition gives the company a terrestrial link across North America—and appears to place Global Crossing in competition with its own customer base, including heavy-hitters like MCI WorldCom and Deutsche Telekom.

Another unique aspect of Project Oxygen's approach is its use of what it calls the "Internet model" for bandwidth. Like the Internet, customers will be able to jump on Project Oxygen's network at one point and connect with any location around the world. For one price, it doesn't matter where you get on or off the network, as long as you don't exceed the bandwidth purchased. The big selling point is flexibility, meaning you're not locked in to any one traffic route.

Origins

While these factors hint at some of the emerging differences between Global Crossing and Project Oxygen, a closer look at both network projects may be necessary to understand how two unique visions have spawned separate global networks.

Global Crossing's worldwide strategy wasn't formed overnight. It began with a transatlantic cable system called Atlantic Crossing (AC-1), which was announced in March 1997. Soja worked with Global Crossing on that project and says he doesn't know if the company truly had visions of building a global network when AC-1 was started. (Global Crossing did not respond to Fiber Exchange's requests for an interview and thus offered no comment on this issue.) At some point, however, an opportunity was recognized and Global Crossing's worldwide plan was created.

Having a plan and publicizing it turned out to be two different matters, however. "Until Global Crossing announced its South America Crossing, the public had no way of knowing where in South America it was going, whereas even though it's broken into several phases for South America, you at least had an immediate sense of where Project Oxygen was planning to go," says Yates. "Global Crossing has been a lot more guarded in its rollout."

Global Crossing chose a strategy of building the network, segment by segment. In addition to AC-1 and Frontier, the company has announced the undersea PC-1 between the United States and Japan, Mid-Atlantic Crossing, as well as the terrestrial Pan European Crossing network. The company also founded Global Access Ltd. last December to provide connectivity to major Japanese cities.

This formula has led critics to claim the Global Crossing system is merely a collection of connected point-to-point cables and not truly a global network. But by integrating the segments, adding terrestrial sections where it makes sense, and carrying traffic from the shore to international switching centers, it's tough to argue that global connectivity isn't taking place. With AC-1 already up and operating, the company enjoys revenues that can be applied to future segments.

As Yates says, Project Oxygen was clear in its global ambitions when it announced its network plans in June 1997. The initial scale of the project—275,000 km of cable connecting 262 landing points—raised doubts in the minds of many about the viability of the effort.

That initial vision has been scaled back into a phased rollout, whereby Phase 1 will consist of 168,000 km and 98 landing points. Subsequent phases for other segments of the network are still very fluid and could be significantly modified before actual construction begins.

Project Oxygen expects to begin laying cable for Phase 1A-1, a four-continent loop around the Atlantic, as soon as financing is closed. Plans include lighting up sections of the phase as they are completed. Capacity customers are not required to pay anything up front until service is actually provided, according to Dave Martin, vice president of business development for CTR Group, creator of the Oxygen concept and now a contractor of Project Oxygen Ltd. "There is no risk on the part of companies who sign up, because if we never get to them, they don't have to pay us anything," says Martin.

Last shall be first?

Global Crossing is addressing market demands by rolling out in a rapid fashion and cashing in on "targets of opportunity." The fact that it has already put cable in the water would appear to give it an initial advantage over its rival. But by building the network one segment at a time, will Global Crossing's initial technology be obsolete by the time the last portions are deployed? KMI's Yates says it may be an advantage for Project Oxygen to sit back and plan a network based on the latest in undersea optical technology. For instance, Global Crossing's AC-1 was deployed with a capacity of 40 Gbits/sec. Using dense wavelength-division multiplexing (DWDM) technology, plans are already underway to increase the cable to 80 Gbits/sec. Additionally, Global Crossing has announced an additional cable, AC-2, to be deployed alongside AC-1 for more capacity.

"We're about to put in cabling with a capacity of 1.28 Tbits," says Project Oxygen's Martin. "Sure, [Global Crossing] is putting more fiber in, but that's the second for them...and another build. The contracts we have with our suppliers say that we'll get their latest, most advanced technology and we'll always be installing the latest and greatest. If a higher-capacity cable is developed before we begin, such as 2.56-Tbit/sec, that's what will be deployed...and that scenario has actually happened twice since we began."

"It's true that the longer you wait, the better technology you'll have—along with less revenues and profit," counters Soja. "The point is that Global Crossing has the resources to invest in the same level of technology as any competitor in any particular market, globally or regionally."

Why not more networks?

With the anticipated success of projects like Global Crossing and Project Oxygen, aren't these combatants likely to face additional competitors? That's not likely, say observers.

"One major factor is the lack of experienced management team members with knowledge of how to put such a huge project together," says Yates. The same sentiment is echoed by Soja, who says a strong management team makes all the difference in the world between success and failure. Venture capitalists, he says, aren't just looking for another great idea, but a team capable of bringing a good one to fruition.

"As one venture capitalist put it, 'I'd much rather invest in a grade A team with a grade B idea than in a grade B team with a grade A idea,'" says Soja. "At this point in time, it would be very difficult to muster together the kind of team that could successfully put another global venture together."

Project Oxygen's Martin agrees but adds that worldwide connections are also essential for such a daunting task. "You've got to be able to waltz into the telecommunications companies in any country and convince them to be a landing party for you," says Martin. "There just isn't a huge community of people who have experience at doing that."

Yates points out other hurdles in creating global networks. For example, establishing backhaul facilities in countries dominated by monopolies is extremely difficult, although that situation is slowly changing as competition spreads. There is also the fact that global systems, because of their sheer size and the enormous amount of equipment involved, cost a ton of money. Resources are also a limiting factor. There is a question of having enough cable-making and cable-laying capability for deployment of another worldwide network. Last is the issue of timing. With companies like Project Oxygen and Global Crossing already years ahead, it would be extremely difficult to jump in with a new project.

Enough demand to go around?

Once Project Oxygen and Global Crossing have complete global networks in place and operational, their success relies on whether demand for capacity will be great enough to require the massive amount of bandwidth the two will provide. At least one of the two competitors is preparing for a price war, just in case.

"Global Crossing, about six months ago, was selling an STM-1 [155 Mbits/sec] across the Atlantic Ocean for something in the area of $9.5 million," reports Martin. "Once we are funded and running, we'll be selling an STM-1 anywhere in the world to anywhere else in the world for $2.5 million at our worst price. Don't you think Global Crossing will have to react by lowering their price? Conceivably, they would have to be way below us, because anyone buying from us isn't getting just transatlantic, they're getting trans-everywhere."

Currently, Global Crossing is competing with consortium systems and other carriers that are selling in the traditional indefeasible rights-of-usage (IRU) manner for point-to-point connectivity. Once another competitor comes along, pricing changes are inevitable for all players to remain competitive. Project Oxygen may be that competitor. If so, Global Crossing may not only have to change its prices, but its pricing strategy as well.

One thing is almost a certainty: Capacity availability and capacity demand will work hand-in-hand to determine the market parameters. "On a graph, capacity resembles a stair step, with sharp surges every time a new cable comes online," says Yates. "Demand, on the other hand, ramps upward on a smoother curve. This will result in short-term capacity gluts, but after the capacity spikes up again, prices drop, demand accelerates, and they begin to converge again."

There is no shortage of capacity users in sight, but those users want their bandwidth at the lowest price possible. When prices are down, the capacity is easily bought up. As demand increases, providers can charge more and the demand ebbs. Thus, competitive providers are key to lower prices in global capacity. Yates and KMI believe the market can definitely support two global providers such as Global Crossing and Project Oxygen.

Which to choose?

Which system will ultimately appear more attractive to the customer seeking global connectivity? That's dependent on a multitude of considerations. For example, we return to our original question: What is your definition of global connectivity? If you want several connections to multiple places around the world, with the option to change the connections at will, you would probably lean toward Project Oxygen.

If you consider global connectivity to mean crossing the Atlantic, for example, and the ability to backhaul your traffic anywhere in Europe through a single connection, you may want to consider Global Crossing via AC-1 or AC-2 and its pan-European network.

Global Crossing's status as strictly a carriers' carrier may push other users such as large video broadcasters or multinational corporations with huge bandwidth requirements toward providers like Project Oxygen that will sell bandwidth to virtually anyone who can pay for it. There's also the question of how well a global network can provide backhaul of traffic from its landing sites. Carriers might have more flexibility in this area, but end users might be more interested in a company that can provide a seamless solution.

"If they're a carriers' carrier who is coming online in, say 2003, the best pricing solution right now would be Project Oxygen if they could sign on today," says Yates. "But there are other variables to consider. There's the fact that Project Oxygen doesn't have anything in the water. So there's a certain amount of risk involved in signing on with a company that hasn't installed any cable yet and isn't generating any revenue. So if you're more interested in security issues for getting more capacity, Global Crossing would have great appeal."

According to Yates, technological advances are outpacing the advanced system costs. That means if prices hold steady, the percentage of capacity a provider needs to sell to break even drops dramatically. For example, if you install a 40-Gbit/sec cable, you might have to sell 50% of its capacity to break even. Yet, if a 160-Gbit/sec cable is installed, you only have to sell one quarter of the capacity to break even.

"Capacity plays a big role in your profitability," says Yates. "But with technology advances comes the assumption that prices will drop. So you are constantly up against the battle of putting technology in the water and having the price of capacity drop, requiring you to sell a higher amount of capacity to be profitable."

Getting revenue as early as possible may be an advantage in getting the business moving and, more importantly, showing that you are truly a player in the market. Still, says Yates, whoever has the latest and greatest cable in the water with the most capacity has the advantage because they need to sell less of their total capacity to break even.

As Project Oxygen prepares to begin installing cable for the first phase of its network, and while Global Crossing continues to add segments and tie its network together, there is still a common denominator for both projects...global capacity. Even if you dub one network a "bunch of connected point-to-points" or wonder what's taking the other one so long to get started, both are headed in a similar direction. Two worldwide systems capable of routing vast amounts of voice, data, and video traffic around the globe are sure to change the face of the future telecommunications industry.