SANTA ANA PUEBLO, N.M. Network equipment and communication IC vendors attending this week's Integrated Broadband Networks Summit took little comfort in the words of executives from local and long-haul service providers, who said their only capital investments in the next few years will be for systems and software that can reduce month-to-month operating expenses.
"Everything else is out of the picture. There is no money being spent," said Kay Burin, director of advanced technology engineering at AT&T Local Network Services.
Venture capitalists at a session on investment had to think long and hard before finding bright spots where new research could lead to payoffs. Greg Rossman of Pequot Ventures and Michael Golden of Neveric Merchant Bank said they saw some possible realms to exploit in semiconductor products in wireless networks and Internet Protocol traffic management. Infonetics Research president Michael Howard countered that he saw few investment areas in telecom beyond middleware for making carriers' operations support systems (OSS) work better with IP networks.
"The hardware plays seem to have hit a brick wall for the time being," Howard said. "There are some tactical opportunities in software for vertical OSS functions, but money must be invested very cautiously."
Panel sessions at the conference were replete with talk of falling off a precipice, and of a "nuclear winter" in telecom. Dan Moffat, chief executive and founder of New Edge Networks, warned that "even the biggest players are in danger of going into a death spiral," while TeleChoice Inc. chief executive Danny Briere predicted that more than 95 percent of telecom startups formed in 1999 and 2000 could go out of business.
The chief technical officers of long-haul carriers such as Williams Communications, Sprint and Level 3 Communications all talked of difficult pruning of national broadband networks. While debt burdens are less severe at incumbent local-exchange carriers, the ILECs bore the brunt of criticism at the conference for their perceived slowness to deploy digital subscriber line (DSL) services, once competitive local-exchange carriers (CLECs) had gone out of business. Joseph Lardieri, director of emerging technologies at BellSouth, quipped that he felt he had to carry the flag for proving there were still ILECs interested in providing broadband services.
Briere said in his keynote Monday (March 25) that industry investors still live with perceptions that all transport functions in the network have been commoditized, that the broadband residential market has become saturated at the 7 percent penetration level, and that cable TV multisystem operators represent the only competitors left for ILECs. In reality, Briere said, there are still competitive LEC models that can work, and profitability can be realized through fairly simple functions above the transport layer.
Moffat's New Edge Networks is an example of a CLEC that survived by concentrating on second- and third-tier cities, and by slicing the size of its anticipated footprint from 1,200 to 600 cities.
O. Matthew Beal, chief network architect at Williams, said that interexchange carriers could take a lesson from the surviving CLECs at challenging ILECs directly for services. Beal said that "we can't run away from businesses just because they might compete directly with an ILEC."
Briere predicted that new broadband-access pricing models should take into account the fact that many consumers care more about the always-on "presence" created by DSL and cable-modem service, than about the specific upstream and downstream speeds offered.
The tough choice for most carriers, said Sprint vice president of research, architecture and design Fred Harris, is to know when to transition to the simplest form of IP network possible for all traffic, while retaining support for frame-relay and asynchronous transfer mode (ATM) networks that continue to show strength.
"I direct virtually all my research today to IP and optics," Harris said. "We will protect our frame and ATM customers, but we will ride the horse that will win, and that is IP."
Burin of AT&T went a step further, saying the bulk of network traffic will be based on IP on Layer 3, and Ethernet frames on Layer 2. She said that the bulk of multiservice provisioning platforms offered by equipment suppliers in the past few years simply could not aggregate Sonet/time-division multiplexed, ATM and IP traffic as advertised. At the end of the day, Burin said, metropolitan Ethernet transport offered savings of five to 10 times per bit over ATM.
"I'm an ATM bigot, but Ethernet is where it's going to be," she said.
Nobel laureate Arno Penzias took a slightly more nuanced view in an evening keynote. Penzias, now a venture partner at New Enterprise Associates, said that the extreme IP scenario assumes that multiprotocol label switching works as advertised, allows IP flows to act just like circuits, and IP takes over all traffic soon. In the other scenario, an all-optical network allows for switched private networks to be created anywhere in a global mesh, and Internet-based service becomes one higher-layer service among many. Trends may favor the former scenario, Penzias said, though the ability of optical wavelength-switched nets to handle switched virtual circuits may keep support for residual non-IP services alive.
Carriers are becoming aware at all scales that any packet transmission business is subject to commoditization, speakers agreed. The problem lies in being able to shape bandwidth and add higher-layer services on top of the frame and packet layers, while spending as little money as possible.
At the conclusion of his keynote Briere said, "Baby steps rule!" In an impoverished 2002-2003, he said, those that can offer simple virus protection, e-mail filtering, voice-over-IP virtual private networks and IP PBX features may rule the day. More complex bandwidth provisioning will be demanded sooner or later, he said, but carriers need to live through the current dry spell first.