DSL Deployments Hit the Throttle

The major local-exchange carriers are accelerating
digital-subscriber-line facilities deployments well beyond their current marketing
capabilities in hopes that by next year, they'll be in a position to address
potentially explosive demand more efficiently than they can today.

A tabulation of the latest round of increased deployment
commitments shows that the regional Bell operating companies and GTE Corp. combined plan
to have more than 40 million lines qualified as ready to deliver high-speed-data services
from installed DSL-access multiplexers (DSLAMs) by year's end (see chart). This is
more than twice the number of lines that were accounted for in announced plans at the
start of the year.

"With competitive local-exchange carriers and
cable-modem operators making strides in broadband-access markets, the ILECs [incumbent
LECs] couldn't afford to wait any longer to make their move," The Strategis
Group consultant Adrianne Brandt said.

"A lot of ILECs are deploying DSLAMs in COs [central
offices] on a large scale before moving to market the services aggressively because they
don't want to be in a position of telling the customer service isn't available
once they begin marketing," Brandt added.

By the second quarter of next year, self-installing
capabilities, together with the scale of deployments, will make it easier for ILECs to
proceed with more aggressive marketing efforts, Brandt said.

There has already been a significant increase in marketing
efforts, she added, buttressed in part by the fact that more advanced means of qualifying
local loop are now available and in part by self-installation capabilities introduced into
vendor modems over the past four months.

Many suppliers have introduced USB (universal serial bus)
modems that users can plug into their computers as they would any other modem.
"We've seen a tremendous improvement in the percentage of
self-installations," Brandt said, noting that at U S West, that figure is now at
about 92 percent.

Strategis estimated that DSL-modem shipments in North
America will hit 400,000 by year's end, representing a 900 percent gain over the
60,000 shipments registered by the firm at the start of the year. In contrast, the cable
industry anticipates that it will have more than 1 million users connected via cable
modems by year's end.

Some firms tracking high-speed-data deployments see cable
maintaining a strong lead for some time, although they have raised their estimations of
where DSL will be as the telcos have moved to deploy DSLAMs more aggressively this year.

Forrester Research Inc., for example, said the number of
DSL users will hit 1.8 million in 2001, compared with 9.2 million cable users, with cable
maintaining the lead through 2003 at 19.65 million versus 7.7 million.

But other observers suggested that projections based on
current telephone-industry marketing efforts miss the potential for rapid penetration
across a 40 million-line base of installed DSLAMs.

Retail stores are already part of the DSL-distribution
chain, Brandt said, since many suppliers are offering modems that can be directly plugged
into computers that conform to local DSL systems supported by ILECs.

The current approach to marketing and connectivity -- where
early adopters comprise the overwhelming majority of customers -- is no measure of what
the approach will look like a year from now, when self-installing technology will be
widely available, she added.

An important factor in the ability of the telephone
industry to aggressively market DSL services in the year ahead is the emergence of a new
generation of DSLAMs that allow telcos to reach customers served by digital-loop carriers,
TeleChoice Inc. DSL analyst Claudia Bacco said.

For example, she noted, the new Lucent Technologies DSLAM
"sets a new bar for DSLAM performance by doing everything a DSLAM should do in a
single, compact box."

Bacco added, "The problem for Lucent is that other
vendors that already have significant market share are producing a new generation of
systems that accomplish many of the same things."

For example, DSL-system supplier Copper Mountain Networks
Inc. -- which Dataquest Inc. said is the leader in DSLAM deployments for business use --
is introducing line cards designed to support the G.Lite DSL standard, employing
G.Lite-specific chip sets supplied by Centillium Communications Inc.

"We're able to support 24 ports per line card
using these chip sets, versus the four to eight ports that you can do on line cards that
are designed for rate-adaptive ADSL [asymmetric DSL]," Copper Mountain director of
product marketing Richard Sekar said.

Copper Mountain -- which, by year's end, will begin
shipping G.Lite line cards that can be inserted into its currently installed DSLAMs -- has
worked out an interoperability agreement with 3Com Corp. ensuring that all of 3Com's
G.Lite customer modems will work with Copper Mountain DSLAMs.

Even though G.Lite is now a standard, interoperability is
not guaranteed between DSLAMs of one vendor and customer-premises equipment of another,
due to the absence of standards for the applications interfaces below the physical
transport layer, Sekar noted.

This is a key area of concern for the telephone industry,
and it is being addressed in standards efforts and through interoperability testing
similar to what Copper Mountain and 3Com are engaged in, Brandt noted.

Telcos are finding that if different vendors'
equipment is used at the ends of the G.Lite connection, the specified transmission
distance of 18,000 feet for a 1.5 megabit-per-second data rate is often not attainable,
she said, adding, "The effective transmission distance is more like 12,000
feet."

While some telcos like BellSouth Corp. and Sprint
Communications Co. have voiced intentions to push G.Lite aggressively starting in early
2000, others are holding back while these problems are resolved.

Many companies are supplying microfilters that customers
can easily attach to their premises lines to support delivery of the non-G.Lite DSL
signals over those lines, which allows self-installation without requiring that the telco
offer G.Lite, Brandt noted.

"The technology exists to solve the G.Lite problems,
and there's a tremendous amount of focus by these big companies on doing that,"
she said. "I don't believe you can underestimate the power of the ILECs once
they've set their minds on accomplishing something."