Talk about kamikaze business plans: Fujitsu, NEC,
Mitsubishi and Toshiba are diving into the digital-subscriber-line chip-set market, with
an eye toward supplying core components into a market that remains, well, up in the air.
The Japanese companies' collective entry in the past
few months puts them in competition with about one-dozen U.S. and European companies, such
as Lucent, Alcatel and Texas Instruments, which are also rushing into the DSL-chip market.
What are they thinking? Market share? Some kind of fin de
siécle technological dominance?
The slow-to-develop DSL market hardly needs a price war.
Yet the onslaught of suppliers will drive the price of DSL set-top chips down from
today's $15 range into the $5 range, according to some semiconductor analysts. That
could make DSL equipment attractively cheap -- and it could also drive marginal
competitors out of the market.
The entire DSL world appears on shaky turf right now -- not
unlike its competitor, the cable-modem business. A new study from TeleChoice Inc. foresees
sluggish growth for both categories during the next two years, although cable-modem
shipments could run at more than triple the pace of DSL units.
According to TeleChoice, cable modems will go into about
1.5 million homes during 2000 and nearly 2.8 million homes in 2001. Compare that with
DSL's installed base of 400,000 homes by the end of 2000 and 750,000 households one
Just to keep things interesting -- and to totally confuse
anyone trying to make comparative estimates -- another forecaster, In-Stat Inc. (a sister
company to Multichannel News), foresees 14 million DSL chip sets reaching the
market by 2001.
Of course, that's a worldwide figure, which still
confuses the issue, since DSL appears to be lagging behind in appeal overseas even more
than its yawn-inducing impact in the United States. The In-Stat prediction includes chips
for the low-end version, called "G.Lite," as well as for the ADSL (asymmetrical
DSL) and VDSL (very high-speed DSL) formats.
Fabricating and distributing the chips is just part of the
problem: The telephone companies that are ostensibly promoting DSL are finding it hard to
meet their own objectives.
Bell Atlantic, for example, which extolled its DSL agenda
in limited markets in October, is apparently having trouble rolling out the service, even
in the targeted areas.
Although Bell Atlantic said it would introduce DSL
throughout specific central offices, it turns out that prospective customers are now being
checked to make sure that their households are within the requisite 12,000 to 29,000 feet
from a Bell switch. That's the reliable reach of the DSL feed, depending on the
architecture of the local loop.
Pacific Bell customers have reported similar problems in
Distance from the central office also affects the bandwidth
-- a variation on the shared-user-capacity problem facing cable-modem customers.
Meanwhile, the Bells' avowed efforts to deploy DSL are
quickly being overshadowed by interexchange carriers and competitive local-exchange
carriers. Both groups are concentrating initially on offering DSL to their business
customers, with CLECs in particular pushing DSL in neighborhoods where they have their own
That effort to maximize facilities utilization for business
customers may reduce the competitive impact on cable-modem providers, which are expected
to concentrate on residential customers. But if CLECs capture business customers now, it
lessens the opportunity for cable-modem providers to pursue those commercial users in
For their part, Sprint and, more significantly, MCI
WorldCom are also beginning to chase DSL business, seeking to leverage their relationships
into the high-speed realm. If they can achieve early, metrowide entry into some markets,
their presence may stymie later offerings by cable modem or even Bell DSL providers.
All of this would-be competition takes place in an unreal
environment of early-adopter pricing, where we see wacky cable-modem fees of up to $70 per
month on systems that require users to maintain a phone line (or install a separate line)
for hybrid return paths. That's almost as unrealistic as the $79 to $109 monthly fees
for some Bell DSL services.
Those are suicidal prices, given that one-half of
residential online users remain complacent about their 28.8-kilobit-per-second services,
and they are satisfied with the $20 to $25 monthly fee for that speed.
The influx of DSL chip-makers and component suppliers may
induce telcos and competitive carriers to offer more attractive deals -- and more
competitive services. Most researchers continue to dismiss 1999 (as they did with 1998) as
a showcase year for high-speed access. This is just another year of getting ready, of
getting all systems into place.
The race really begins in 2000, although it's hard to
say right now how many survivors there will be.
I-Way Patrol columnist Gary Arlen covets Compaq's
"triple-play" PC, which is equipped to handle all three access formats: DSL,
cable and satellite.