CLEC Getting Started

DSL Prime News Weekly: The Inside Source  August 7, 2001

Dave Burstein
DSL Prime

Quote: "I told them where they could stick their modem and cable equipment and proceeded to cancel my cable TV and modem accounts," an Adelphia customer shut down for running a server.

Verizon's Sarah Deutsch knows better, "asking for subscriber termination is a very drastic remedy that infringes on people's rights and speech."

Worldwide, DSL is adding ten million subs this year and beating cable, great news for the industry. In the U.S., Verizon and SBC last quarter added fewer subscribers than Korea Telecom has been known to install in one week (KT hit 160K at peak, Verizon netted 120K and SBC 83K in 13 weeks, adjusting for shut-offs).

SBC had originally targeted 350K, Verizon not far behind, and have lowered estimates to avoid further embarassment.

The CLECs are too crippled to make a difference. Rhythms has virtually a fiduciary duty to file a massive antitrust suit against the telcos, even if Rhythms is filing bankruptcy. The Cartesian calculation is plain: any plausible chance of victory, with the rewards literally in the billions, justifies the relatively small expenditure in legal fees. The logic in bringing the suit is clear, but please don't read this item as a DSL Prime opinion on the merits.

Slowing deployment holding back US subscribers
Error in our previous numbers

A smart observer just proved to us our availability numbers for the telcos are significantly distorted. We have been working from telco figures like this one from Verizon:

"48M lines are passed by DSL (i.e., come out of C.O.s that have been equipped), out of a total of a little less than 63M lines. Of the 48M, about 62 percent are loop qualified, which means about 29.8M Verizon lines can get DSL from Verizon."

We calculated that meant 47 percent of Verizon customers could get DSL. SBC's similar number is 55 percent. We now realize we were mistaken, because DSL deployments are concentrated in the larger cities, where businesses (which have many lines) are typically closer and more likely to be served. This means the actual percentage of subscribers eligible is actually lower, and we hope the telcos will help us get more accurate figures. (Bell Canada is over 70 percent, BellSouth 60 percent, Qwest far behind even after the new DLC rollout).

Either number is far less than planned, despite Whitacre and Salerno both asserting the spending cutbacks were not affecting "growth initiatives." SBC was supposed to be over 80 percent with Pronto in 17 months (universal soon after), while GTE/Verizon West expected to be ahead of them. Our guess is that 20-40 percent of the shortfall in DSL subscribers is due to folks who should have been already reached.

The problem must be solved. DSLAMs are now so cheap that there is a fast return on investment for putting one in every CO that's already fiber connected, and holding back is simply sacrificing profits for this year's earnings per share or bad accounting.

Proper line testing, British Telecom has demonstrated, can qualify 5 to 15 percent more lines at a cost that's more than repaid in reduced service problems. Reach is getting better, with more improvements on the way. Repeaters are on the market from Charles, Symmetricom, GoDigital, and 2Wire, with Adtran on the way. The cost will come down in telco order sizes.

Dan Reingold of CSFB sees a transition happening in telecom, with new services (data and wireless) taking up the slack as traditional voice fades. Wall Street in general, however, is pushing so hard for short term profit, telcos are under pressure to compromise their future by underinvesting. US telcos right now are benefiting from a quick $B profit rise from reduced reciprocal compensation, and will get a volume rise with long distance; that's not enough for the future, we believe.

$15/port for VDSL basement DSLAMs
Infineon 8 port chip drops the bar

Infineon has upped the density and lowered the list price on its very interesting 10 meg Ethernet over VDSL chips, an interesting choice for basement and shorthaul networks. At the prices being offered, we do not understand why MTU units would use anything slower, whether ADSL or G.SHDSL. Even if the capacity is not needed today, the upgrade path has to be worth the cost in most deployments, when the chip hits the mark later this year.

VDSL is inherently cheaper than ADSL at the chip level, requiring fewer transistors and less power. Volume has been so small, however, that the effective price has been higher. Infineon is pricing down the learning curve, making it possible for manufacturers to jump on now.

Similarly, other fast components (backplanes, ATM switches) add little to the delivered cost, and dramatically extend product life.

Copyright 2001 Dave Burstein.
The DSL Prime Newsletter is reprinted with permission.

DSL Prime News
DSL Prime Examines Verizon
DSL and the US Government
DSL Prime People
DSL Prime News Briefs

Email this article to a colleague