Convergence /

2002: End of the tech wreck?

By Thomas Nolle
Network World, 12/03/01

Last year about this time, I predicted that - rosy forecasts notwithstanding - 2001 would be the worst year in the history of our industry. Unfortunately, I was right. This year, despite uniformly gloomy prognostications about service provider and enterprise capital expenditures, I predict the coming year will be a lot better - for some of us.

For service providers, success in the coming year will depend on the cost of borrowing money. Service providers have to obtain capital to build or modernize their networks. The cheaper they can borrow, the higher their profits on a given sale. Big incumbents borrow pretty much at the prime rate; small emerging players borrow at junk-bond rates. Regulations and technology don't factor in when the financial disparity is 10 percentage points wide.

Big incumbent carriers are big and incumbent because they have large customer bases, presumably buying lots of services. What services? Frame relay, ATM, leased lines, toll-free numbers, long-distance voice and the rest of the dull, pedestrian service suite of the past. Faced with overwhelming success, do these guys toss aside those services to embrace IP VPNs? Not unless they have a death wish. They extend and promote what they're already winning with.

Frame relay will expand big-time in 2002. So will ATM and transparent LAN services. Even leased-line and SONET services will expand. This growth will shift infrastructure investment toward the creation of "legacy" services, away from the IP-driven past. You can already see that occurring in the positioning of carriers and equipment vendors. Competition between interexchange carriers and regional Bell operating companies as the latter enter the long-distance market will only magnify this shift, as carriers try to win the loyalty of the current national data service customers.

But IP isn't dead; it's just not being promoted with mindless enthusiasm. No one believes the consumer will embrace frame relay, so eventually (meaning beyond 2005) it's going to turn into an IP world. This means infrastructure deployed today will be installed based on expansion of legacy service and sustained over time by IP revenue. Multi-protocol Label Switching (MPLS) offers the best bridge between these service extremes, so MPLS will be a much bigger player in the coming year.

The problem is that the common conception of MPLS is that it's a protocol that runs inside the Internet. In 2002, we'll see the first clear indications that MPLS is really two related concepts - a technology to optimize IP and a technology to generalize connection-based services. In its latter form, MPLS makes a great core technology for ATM and frame relay networks, which is exactly what the carriers will want in 2002.

There will be changes at the edge, too. Carriers will look to consolidate the service edge functions of their networks. They'll want to support DSL, Gigabit Ethernet, IP, frame relay and every permutation thereof, out of a single platform. Many will say that this is a "service switch" application, but most service switches are just glorified IP access routers. Expect start-up and incumbent vendors to scramble to make the right products.

Expect a lot to fail. Incumbent players will survive based on their support of the legacy services wave, even if they miss some of the MPLS and edge opportunities. Most start-ups don't have the flexibility or cash reserves to reposition. Those that started before the big Nasdaq crash probably have been focused on the wrong product set and will face serious challenges. Those starting up now have a unique opportunity to create a product set that meets real near-term requirements. In between, it's a crapshoot.

That's why a "better" 2002 won't feel all that good to many of us. Thousands of the best and brightest in the industry have bet their futures on "emerging" technologies and carriers that are emerging no longer. Those start-ups created the dynamism and excitement of the industry. It's a shame they couldn't have created real economic value, too. But inventiveness will eventually be rewarded again, and our industry will reabsorb the talent that large-scale company layoffs and failures have idled. It starts in 2002, folks. Happy New Year.

Nolle is president of CIMI, a technology assessment firm. He can be reached at (856) 753-0004 or tnolle@cimicorp.com.