DSL Prime, DOCSIS Report, Fiber News, A Wireless Cloud

August 31

• Steve Wozniak: Use a Verizon MiFi as Backup for your iPhone http://bit.ly/cscnGY
   
• NZ: Competition in Mobile, Govt. Fiber in Fixed http://bit.ly/bcFUQl
   
• Dane Bringing High Speed, Low Price to California http://bit.ly/bhNz95
   
• Verizon Turns on ADSL2+ to 1/3 http://bit.ly/cVEGuY
   
• Lantiq's Latest: 16 Channel VDSL, HD-speed 802.11, 2 Giga NP http://bit.ly/cz4RAq
   
• Apparently Inappropriate RUS $32M Grant http://bit.ly/adBFPt
   
• Julius Genachowski Gnoblog by NotJulius http://bit.ly/dmm5Co
   
• Briefs on: GPON realistic speeds, Bouygues, Covad, China Unicom jailbreaks iPhone, Comcast-Barry Nolan, component shortage easing?

Reply "subscribe" to be added, "un" to be dropped

"I stand by what I wrote." A prominent journalist who is trying to bluff his way out of an error in a broadband story. Pointing out a mistake is the biggest favor you can do for me.

*** Lantiq's Vinax is Industry’s Lowest Power, Smallest Footprint VDSL2 chipset. http://bit.ly/cef0qI (ad)

Apple founder Steve Wozniak loves his iPhone but recommends carrying "a second Verizon phone for backup” or to “carry a MiFi and rely on Skype on your iPhone.” (below)

     Dane Jaspers at Sonic.net is bringing "high speed, low price, high volume" to California, the model Masayoshi Son established in Japan a decade ago. Finally. 
 
    Australia's next Prime Minister may be decided by support for the national broadband network. The election ended 73-73 with three rural independents the key. They tend to support the conservatives, but want Labour's NBN for their territories. Paul Budde expects a compromise that preserves some but not all of the NBN plans. 

     From Kenya to Hong Kong, 4G finally is deploying widely. LTE is about to roll across most of the U.S., with Metro PCS opening Las Vegas and Dallas in September and Verizon a third of the country before yearend.  Jennie's doing the video for Carl Ford's 4G Wireless Evolution Conference in Los Angeles October 4. Say hello.

*** VINAX V3 chipset has industry-leading power consumption of just 0.9W per channel and meets the EU Code of Conduct (CoC) on Energy Consumption of Broadband Equipment. http://bit.ly/cef0qI (ad)

Steve Wozniak: Use a Verizon MiFi as Backup for your iPhone http://bit.ly/cscnGY

 

Apple founder Steve Wozniak loves his iPhone but recommends carrying "a second Verizon phone for backup” or to “carry a MiFi and rely on Skype on your iPhone,” he told dealer Henk van Ess when ordering a Mifi 2352 to use in Europe. Woz was angry after getting a $7,000 bill from AT&T for a half day's use of his iPhone, especially because he was supposed to be on an international unlimited plan.

     Woz wrote "If you can afford it, carry a second Verizon phone for backup. Another option is to carry a Verizon MiFi and rely on Skype on your iPhone. I have used this MiFi technique to rescue my own, and others’, iPhones on occasion. If you buy a Verizon Palm Pre, you get free MiFi on it so that is possible the best ‘compromise’ solution, to carry a Verizon Palm Pre along with your AT&T iPhone 4. ...

     I was in Germany 1 year ago for the World Segway Polo championships and after half a day got notified that I had a $7000 bill for data. I called AT&T’s customer support and told them that I always had an international plan as I travel a lot and it must have accidentally been removed when I bought my iPhone (3 GS?)."

   They keep trying to take my unlimited plans back when they think I’m not looking. ... I’ve heard of that sort of problem countless times, from friends and in the press. I had done everything a person could have done to be safe about this, but they still tried to take my plan away behind my back, figuring I wouldn’t notice it for months." http://bit.ly/d3IrRM

 

*** ASSIA -- the fastest way to dramatically reduce your DSL network operating costs.  Fewer truck rolls, fewer customer calls, better diagnostics.http://www.assia-inc.com/

 

(ad)

Lantiq's Latest: 16 Channel VDSL, HD-speed 802.11, 2 Gigabit NP  http://bit.ly/cz4RAq
Imran Hajimusa of Lantiq demo'd four TV streams going wirelessly across a large room to four HDTV's several months ago. They are now shipping the WAVE 300 in volume. Ulrich Huewels is confident it can support HD video. Lantiq tells me a large U.S. carrier will soon surprise by moving to wireless home networks. They've also doubled the density of their Vinax VDSL linecard chips and are sampling the two gigabit GRX gateway processor.

    Replacing wires has long been the grail for in-home networking. Vendors have been making promises for years, but field tests were not up to carrier grade. Carriers can't accept networks that only work for 95% of homes because the truck rolls to the other homes can be brutally expensive. They need close to 100% real world performance. I'm going to be skeptical until carriers prove the promises in the field, but the buzz for the new beam-forming chips is good.

    The new VINAX V3 supports 16 channels for the 50 meg VDSL2 Profile 17a and 8 channels for the 100/100 meg 30a. Power is 0.9W per channel. Bonding is supported and VINAX "is ready to support full System Vectoring, a VDSL2 enhancement that will reduce crosstalk." At last year BBWF in Paris, Lantiq had an impressive demo of vectoring nearly doubling speeds running on FPGAs. I'll be following the progress closely.

    The GRX288 has a 32-bit MIPS CPU as well as a protocol-processing engine. Lantiq says GRX288 "achieves routing throughput up to 2 Gbps." VOIP support, built-in gigE switching, IPV6, TR-069 and more are also supported in GRX network processors. These high speeds are coming just in time for all the new gigabit fiber offerings being promised from Vermont to Australia. http://bit.ly/cz4RAq

*** 4G Wireless Evolution October 4-6, 2010 - Los Angeles Convention Center
Unmatched networking opportunities and a robust conference program representing the wireless ecosystem. This year's conference features top professionals from AT&T, Sprint,T-Mobile, and Verizon who will be leading dynamic sessions
http://bit.ly/d0GftE

 

(ad) Programmed by Carl Ford, who programmed the great VON conferences, so you know the program will be strong.

Dane Bringing High Speed, Low Price to California http://bit.ly/bhNz95
Xavier Niel's 30 euro unlimited triple play took 5M customers from France Telecom, transformed the European Internet, and made him a billionaire. Dane Jasper's Sonic.net is (finally) bringing the same "low price, maximum speed, high volume" model to California. Sonic.net is offering 100's of thousands of Californians "up to 20 megabits" + unlimited national phone service for $56, about the same price as Verizon is charging for the 10-15 megabit DSL service alone.

    Verizon charges about $75 for similar 10 meg + voice and AT&T probably $84, about 50% more. Unless you live far from the exchange, Sonic.net offers a better deal than any large U.S. carrier. Dane has some interesting ideas about TV to implement as soon as practical, although I think he's going to be very busy just filling orders for a while.

     The word came first from Paris. Benoit Felten, Europe's most interesting fiber analyst, wrote I should read his interview with Dane. Now that I've reported from the states, look for reporters to check this out and create a storm. I told Dane - who's been asking me for years whether the Free.fr model would work in the U.S. - that the low price, high volume model has proven itself time and again. He hasn't quite brought U.S. prices down to French levels, but this is the biggest move in that direction since Mike Powell's rules killed the last big CLEC in 2003-4.

     With today's low bandwidth prices, it's an unnatural act to charge one price for a one meg service and much more for ten meg. The four major British carriers don't even seem to have an offering below ten megabits; the four in France not below 8 megabits. Most are at "up to 20." 50-100 meg cable is $25-45 as part of a package. With very high volume prices for bandwidth now below $2/megabit in major markets, the difference between one meg and ten meg service is dimes/month. Where competition is strong, almost everyone get high speeds at low prices.

     Dane is likely to do extremely well because he will now make substantially more on each customer. Including voice, even cheaply, ups the gross margin. Offering a great deal eliminates most of the expensive "customer acquisition."  Once he has the network in place, he just pays for the bare line (UNE-L), covers his voice and bandwidth costs, and has a generous margin on every new customer. He's been in the business since 1994, runs a very efficient operation. Dane is a good operator and can look to ten times as many customers as his current 35K. Sonic.net has a remarkable customer rating of 4.64 at DSL Reports, a full point above telco DSL and well above Verizon FiOS, generally America's favorite service.  Dane has ADTRAN DSLAMs in 50 Bay Area exchanges and enough fiber connectivity to avoid any problems. SSonic’s other smart moves include 2.5 meg upstream at no additional cost, bonding available throughout (double speed, double price for now) and no contract required, a big deal.

      I've been waiting for this since 2002 when Masayoshi Son in Japan proved low prices yield their own reward. This is exciting.

**** Vinax has the smallest package size and highest channel count (per chipset) in the industry allows double VDSL2 linecard density compared to all on-market alternatives. Each VINAX V3 chipset occupies an area of 29 mm x 29 mm for 16 channels in VDSL2 Profile 17a. http://bit.ly/cef0qI (ad)

NZ: Competition in Mobile, Govt. Fiber in Fixed http://bit.ly/bcFUQl

 

2degrees has quickly taken 5% of the New Zealand mobile market by charging (USD) 15 to 31 cents/minute. They've now introduced 3G data at prices between (USD) 11 & 14/cents per gigabyte. Telecom New Zealand and Vodafone were doing what duopolists prefer, ripping off customers for (USD) 62 cents/minute.

     In France, the planned entry of Iliad/Free as the fourth mobile in 2012 is already bringing down prices, Merrill Lynch believes. New entrants are forcing the big 3 to drop mobile prices in Canada, long far more expensive than the U.S. for the same service. All of this is just what you'd expect where competition is weak. Bringing in more competitors is a natural solution, which is what the U.S. broadband plan is hoping will happen around 2015 because of added spectrum.

    New Zealand recognizes more competitors are unlikely for landlines in less than a decade, if then. They consider the result unacceptable, because weak competition is not enough to persuade TNZ or anyone else to upgrade to a world-class fiber network. So they've decided to spend $1.5B to bring fiber to 3/4ths of the country, with a dozen bidders including Telecom applying. Telecom has agreed to fully split wholesale and retail in order to qualify.

*** ASSIA -- the world leader in dynamic spectrum management, is now supporting over 35 million lines of DSL.  Our performance and reliability are unmatched. http://www.assia-inc.com/ (ad)

Verizon Turns on ADSL2+ to 1/3 http://bit.ly/cVEGuY

 

Verizon has been installing ADSL2+ in most new equipment for years but not offering speeds above 6 megabits because only a small fraction of their 30M lines were covered. They now are offering 10-15 megabit down to about 4M customers for about $55 to about $70/month. Given that the majority of customers are beyond the 7,000 foot cutoffs for the 10-15 meg service, that means they have ADSL2+ to about a third of the network. Most of the rest are behind remote terminals or connected to DSLAMS 5-12 years old, neither of which are scheduled for volume upgrades.

      Experience from Britain, France and the UK has been that almost no one gets the 20 & 24 megabits possible with ADSL2+ and only a minority can even get 10 megabits down. DSM has improved things some, especially for the lines that were marginal for any given speed, and minor improvements in the chips keep improving performance. Current state of the art VDSL is that 15-25 megabits down is typical about half a mile from the DSLAM, less than 10 megabits from around a mile and a half. These are realistic averages, but there's an enormous difference from home to home and office to office. Your speed may well be different than the suggested averages. For example, I heard today of a customer at 5400 feet getting 21 megabits.

     Rob Pegoraro at the Washington Post tested the addresses of 13 friends and co-workers across the District and Alexandria at Verizon's site, and none came up as eligible for its fastest DSL. More at http://bit.ly/cVEGuY

 

*** Feature integration on the VINAX V3 chipset reduces the external Bill of Materials (BOM) cost by 50 percent. http://bit.ly/cef0qI (ad)

corrections

• John Killian is CFO of Verizon, not AT&T, of course. Typo.
• Fawn Johnson had switched from telecom to covering the SEC at the Wall Street Journal before she left for a job at the National Journal, a D.C. reporter writes.
• Reporting on Kapriel Karagozyan at Celano, I remembered that I was very optimistic about replacing DSL chips with software when he was at Smartlink years ago. That never found a market.
• Apologies to Rob Pegoraro of the Washington Post for a mis-spelling I've already corrected on the website.

Email

• Australia's 1 gig to 32 homes would oversubscribe 2.4 meg GPON 16:1-20:1 with a possible congestion problem, Steve Ross pointed out. 50-1 oversubscription worked fine in 2002 but many networks are cutting to 10-20:1 and Steve knows some going for 8:1. I'm guessing we won't see the problem for years because those with the highest speeds use far less of their bandwidth. We won't know for sure for several years when networks are fully loaded at high speeds, however.  Steve edits Broadband Properties, which is much more than a magazine for those connecting apartments and developments. BP probably does the best reporting about fiber in North America. Definitely worth asking for a free digital subscription at https://www.cambeywest.com/subscribe/?p=bbp&f=new
   
   
• An Australian asked “How do you pronounce Bouygues?” “Bweg” was my closest guess for France's third wireless carrier, but Wikipedia suggests “Bwig,” probably phonetically more accurate. France is the most competitive broadband market in the West so it's astonishing a newcomer jumped to a leading position in less than two years. They did that with an attractive price on quadplay, a strategy many are likely to copy.
• A veteran of rural telephony writes about the stimulus, “I believe it was a mistake that the gov’t sought broadband funding applications rather than do the mapping first and request RFPs for the top un-/under- served areas. Isn’t the first step in most projects a “needs analysis”?” With hindsight, we knows that's right and I hope RUS thinks that way going forward.

Briefs

• Jailbreak your iPhone? China Unicom shops offer that as a service despite Unicom being the official iPhone distributor in China. http://bit.ly/aas1vD
   
• Southwest Air, leading U.S. budget airline, is looking hard at satellite broadband for in-flight connections. (Via Satellite)
• Bravo to USTA for changing their speaker policy for their next event. “Research companies will not pay for their presentation slot. There will be the opportunity for research companies to sponsor events such as receptions and lunches. However, willingness to sponsor events is not a factor when assessing the worthiness of your presentation.” One reason USTA events have been so boring is past “pay to play” requirements. One CEO a while back was told if he wanted a speaking slot he would have to buy a $12,500 sponsorship.
• Covad, once the proudest broadband provider in the world, is no more. They were folded in to former Covad dealer, Megapath. So was Speakeasy, once one of the best ISPs in the U.S.

Craig Young of Megapath now is Chairman and CEO. They deserve a better sendoff and I'd be happy to print your recollections. press

• Terry Ann Knopf   (Columbia Journalism Review) reports Brian Roberts of Comcast fired Emmy-winning reporter Barry Nolan because Bill O'Reilly complained. O'Rielly suggested Comcast-Fox News business relationships might suffer.  Nolan had politely criticized O'Reilly of Fox News winning a journalism award. A comment suggested “the Comcast and Fox merger must be suspended as a matter of First Amendment enforcement.” Brian would be well-advised to settle with Nolan and publicly apologize. FCC Chair Genachowski speaks often of his journalism at Columbia.
• Gilad Rozen, Sharon Peleg, and Ronen Peleg, once at Actelis, are now at Celano and are convinced their 802.11 chips are robust enough for HD TV. Fiberhome, supplying Shanghai Telecom, agrees. Kapriel Karagozyan, their VP of sales, was at Smartlink when they developed software to replace DSL modem chips by running on the processor. See corrections, above.

wall street

• Amitabh Passi of UBS sees the current component shortage easing soon. “Recent sales/inventory trends at Tech OEMs e.g. CSCO, HP, Dell, NTAP, and LCD cos, point to mixed Enterprise trends and generally weaker data points in the PC and non-mobile consumer electronics (CE) markets.” He sees sales leveling off later this year outside of mobile. That's little consolation to telcos having problems getting gear today  like AT&T. http://bit.ly/cd28IB
   
   

policy

I've created a gnoblog for Julius Genachowski by notjulius. Satire, or perhaps what he really wants to say. http://bit.ly/dmm5Co

 

Julius just killed the free 768K service in the 2155-2180 MHz AWS-3 band from M2Z. He needs to look in the mirror and ask "Which side am I on?"

*** CITI Columbia October 15, 2010
The State of Telecom - 2010 Aligning Supply and Demand for Next-Generation Broadband. Featuring a Trans-Atlantic Dialog Organized in conjunction with IDATE  http://bit.ly/9wKHbN (psa) They haven't announced the speakers yet, but several will be worth making a trip to New York for. See you there.

Apparently Inappropriate RUS $32M Grant http://bit.ly/adBFPt
Montana Opticom was approved for a $32M RUS grant to build fiber to 7K+ homes in Gallatin County that were said to be "underserved." I have data from Opticom's web site, local governments for a majority of the population, local news reporting and their competitors that the area is well served. If so, the grant should be immediately frozen.
  According to the map posted at Opticom's website the majority of homes they intend to cover can already get 10 megabit cable modem service from Bresnan. I checked this with Bresnan and with local officials for Belgrade, Montana; Manhattan, Montana and Four Corners, Montana. Those three contain a majority of the homes to be covered. They already have two broadband choices, at least one at 8-10 megabits. In addition, most of the additional area on the Opticom map has DSL service as well as a local WISP.
  In addition, Opticom currently offers broadband to 300 homes at some of the highest prices in the world. Unless they firmly committed to offer honestly "affordable" prices, Julius Genachowski at the FCC should personally step in, review the situation with RUS, and make sure that this project will deliver his repeated promise of "affordable broadband." Opticom is charging $79.95 for a 2 meg download service and $149.95 for ten meg, similar to what Bresnan delivers to most homes in the territory for a third the price. Even 1 meg is $45.95. While past performance is no guarantee of future choices, I'd look very hard before giving $10K/home served to an outfit that charges prices like this.
  If in fact the full $32M grant & $32M loan are spent building the network, as promised, I hope RUS has a guarantee from a rock solid outside party for the loan. There's more, including improbable estimates of how many businesses are in the territory, a price tag that appears too high and claims for an almost impossible number of jobs to be created.
   Something is almost certainly rotten in the state of Montana.
-----------------
Jonathan has a choice: look more closely at the proposal or just let it go through even if a mistake was made. The odds are good that if he looked closely he would find grounds to cancel this one. It requires a great deal of courage for any politician to say "I made a mistake." I believe Jonathan has that courage. Lots more at http://bit.ly/adBFPt

Announcing:  Julius Genachowski Gnoblog by notJulius http://bit.ly/dmm5Co
A gnoblog is not the blog of the person named. Gnoblogs are fiction. The purpose is satire, humor, commentary or … Julius didn't write this, although perhaps he should have. Take a look and give me advice on how to make it funny. Recent postings:

• Shutting Up Telcos August 23, 2010 http://bit.ly/aH59im
   
• AT&T has made a major mess of a $B U.S. Treasury contract, Amy reports in the WSJ. What caught my eye was her comment “Criticism of the White House budget-cutters’ efforts has been muted, however, since few tech companies want to publicly criticize a major customer.” Next time Cicconi and friends criticize us, maybe I should ask Treasury or OMB to have a word with them.
• The Buck Stops Here on Broadband Failure August 21, 2010 http://bit.ly/bY48ld
   
• Q2 was the worst quarter for broadband in this century.  2009 was almost certainly the worst year for new broadband deployments in a decade but it’s August and I don’t even have an updated figure. It’s 18 months since Barack was elected.  I have to take responsibility. My policies of incentives aren’t working, nor is the $7B broadband stimulus. So I’ve asked my staff for ideas that won’t take years to have an effect.
• I have not been “neutered” August 21, 2010 http://bit.ly/ad6We1
   
• Heading “Genachowski, Man Up!” is a low blow, and I’m disappointed to see GigaOm, one of the most read tech blogs worldwide, shooting me down like that.  Stacey Higginbottom’s GigaOm article was perhaps right I’m not a “quintessential politician,” but we’ve only been in charge for 18 months. Change in Washington takes longer than that.  The FCC has not “abdicated.”Stacey simply doesn’t understand D.C. I have balls, it’s just that Jim Cicconi of AT&T is holding tight and squeezing them. He turned out 70 Democrats and all the Republicans against Net Neutrality, enough to vote down anything I proposed.  After that, it took courage even to keep the issue on the table by proposing the Third Way.  I’m proud of what I’ve accomplished. In a year, I’ve taken an agency that was totally demoralized and revitalized it with dozens of very smart people. Eddie Lazarus, my Chief of Staff, clerked for the Supreme Court. Sharon Gillette is from MIT. A dozen more I’ve appointed have extraordinary experience.  Ask anyone who’s worked at the FCC recently how the atmosphere has changed.
• Randall Stephenson’s Truth August 18, 2010 http://bit.ly/ca42nB
   
• Randall earned my respect by calling and saying “I made a mistake,” something I don’t hear often from folks with 200,000 employees.  He had said that Title II “would undermine investment in broadband networks.” When he got back to AT&T headquarters in Dallas, his people told him that was wrong.
• “AT&T has nowhere left to cut,” he told me. On wireless, they are a year behind Verizon on the 4G buildout and everyone knows their iPhone capacity problems. They are paying a high price for cutting capex $2B two years ago. On wireline, T went negative on net adds for the first time because they slowed U-Verse buildouts by a third last year. “Net neutrality” will have minimal effect on investment, friends on Wall Street tell me. It’s amazing how different the world looks when I get away from the lobbyists in D.C.
• For Broadband, Bring Down the Rural Internet Cost August 18, 2010 http://www.julesgenachowski.com/?p=26
   
• One way to get more broadband was supported by just about everyone. Immediately bring down the sometimes exorbitant cost of rural backhaul.  When I went to the OPASTCO meeting of rural telcos, their leader, John Rose, told me this was a crucial issue for them and that “OPASTCO would be at the FCC every day to get this done.”  I got the message. Last September, the broadband plan presentation to the commissioners pointed out these excessive backhaul costs were 25%-40% of the problem.  I made a bad mistake not moving on this immediately because I didn’t understand the issue.  If I had, hundreds of thousands of homes would be connected by now and prices across rural America would be falling. And it wouldn’t require a penny of public money.
• Stupid.  So I’ve asked the wireline bureau to use special access where wholesale Internet connections cost more than 3x the national average. We have an open proceeding with enough of a record that we can come to a decision any time. Even next week.

Reply "subscribe" to be added, "un" to be dropped
Volume 11, #14 August 31 2010   

Perhaps because of the election, Mike Quigley announced the GPON NBN would run at 1 gigabit rather than 100 megabits  with no increase in cost.  Nobody outside the industry believes 2.4 gig down GPON can deliver a gig to any customer, but so few applications today can even support 100 megabits I'd expect to reach the gig 95% of the time. Verizon demo'ed a gig in GPON, but don't expect it anytime soon for consumers.

     A true gig of active Ethernet to every home is coming to Springfield, Vermont's VTEL and 5+ megabit LTE to 100% of the state, with the help of an RUS grant. Michael Guite of VTEL asked me “Dave, can you think of any deployment, anywhere in the world, better than what we are doing here?” LTE is the best you can get in today's wireless and 1 gig active Ethernet is as good as anything I know being deployed. I'm proud to help on the technology strategy on this one.  Any friends with depth on LTE and wireless contracts? I need some advice.

     The “separate peace” between Verizon and Google has been signed, about which the less said the better. I was perhaps the first reporter on this, which I picked up from Dave Kaut of Stiefel. Maybe my colleagues in D.C. will investigate the demand included by V & G that the FCC move to do away with intercarrier compensation within the year. http://bit.ly/bBOLOJ
 
Julius hasn't announced a press conference in months so I can't ask him how much the Bells will net from his ICC/USF proposals. It's literally $billions. Just “follow the money.”
   
*** Lantiq: 20 years' track record of leading industry innovation. Our semiconductor solutions are deployed by major carriers and found in home networks in every region of the world. http://www.lantiq.com
 
(ad)

Editorial: Gigabit on All New Networks
Fiber is expensive to deploy, whether in ditches or on poles. Electronics and bandwidth to run faster are (relatively) cheap.
A small increase in investment (~5-10%) takes you to a full gigabit.
… Therefore all new fiber networks should be a gigabit.

Tasmania to New York: We're Faster http://bit.ly/aGZa1i
 
Larissa Bartlett startled New Yorkers by telling us the Internet runs faster in Tasmania. If Labour wins the election next week, Mike Quigley's National Broadband Network will bring 100 megabit and even a gigabit to 93% of the country; the opposition promises to kill off the NBN as too expensive. Prime Minister Julie Gillard flew to Hobart for the official opening of the NBN -- and some politicking. ...

      The impact of broadband is overstated in Australia as it is in other countries. I don't go as far as Grahame Lynch, who writes in the Australian "NBN is welfare for tech-heads." I believe a great Internet is worth even a high cost. Grahame is right that broadband alone is unlikely to live up to political promises. Mr. Larissa Bartlett, the Tasmanian Premier David Bartlett, released details of their trip. Every politician should do similar. http://bit.ly/aGZa1i

 
Australia's Gigabit: Cheapest Upgrade in History http://bit.ly/9l7ENp
 
Australian Minister Stephen Conroy announced the National Broadband Network would offer speeds of 1 gigabit without spending a penny more of capex. Sounds like the usual politician's promise but actually is 95% true.  The opposition wants to kill the $43B project as too expensive; the government warns that a vote against them will condemn Australians to a second rate Internet for a decade or more. Both are right.

    The shared 2.4 gig down, 1.2 gig up GPON Mike Quigley has chosen can in fact deliver a gigabit to the home - as along as your neighbors aren't doing much on the Internet. If three of the 32-64 users on a node want a gigabit, it can't deliver. Today, so few web services deliver even 100 megabits I'd guess you could get the 800+ megabits gigabit 95% or even 98% of the time. Even in five years, likely traffic patterns would allow actual speeds of 400 megabits or more most of the time. Charging the customers who want the gigabit $5-10 more per month would easily cover the increased operations costs.
 
   We've learned that in practice even dramatically faster speeds produce surprisingly modest increases in total demand.  Hong Kong Broadband Network is the only carrier I know doing customer experiments with a gig over GPON and hasn't discussed the results yet. The efficiency of sharing at high speeds over GPON is unproven. I'd very much like to hear, probably off the record, from any well-informed engineer.

   10GPON, four times faster, was demonstrated a few months ago by Huawei and Verizon. Commercial units are years away however. Active Ethernet inexpensively provides a full gigabit. It's being deployed at that speed in Sweden, Singapore, and the Vermont Tel network I consult with. It requires a strand to every home and more lasers, but is a simpler network to manage at high speeds.

  For any fiber network built in 2011 or later, the natural speed is a gigabit. http://bit.ly/9l7ENp

 
*** CommsDay Australasia Australia and New Zealand's telecommunications daily since 1994. COMMSDAY International The weekly capacity newsletter. Free trials at http://www.commsday.com/commsday/?page_id=833
 
(ad)

Worst U.S. Broadband Quarter Ever ~350,000 http://bit.ly/9VNpVl
 
The numbers are clear. AT&T actually went negative. They are getting killed by cable were they haven't done their DSL upgrades (U-Verse), The 1/3rd cut in U-Verse spending in January, 2009 by now leaves them 3M homes behind their earlier plans. Cable did somewhat better, but was significantly down. The numbers are from Leichtman and do not include the smaller carriers.

    Comcast was up 116K, Time Warner 96K, and cable overall +343K. That's far better than the telcos but down from the prior year and quarter. AT&T was down 92K, which I believe is the first time T has actually fallen. Verizon after slowing FiOS was up only 28K, and Leichtman's telco count is -7K. With the smaller telcos included, the telco totals would probably be slightly positive and the overall total somewhat about 350K. Full chart http://bit.ly/9VNpVl

 
*** ASSIA -- the world leader in dynamic spectrum management.  Our performance and reliability are unmatched. http://www.assia-inc.com/
 
(ad)

Bouygues Quadplay Sweeping France http://bit.ly/dmXfzj
 

25-35% of French broadband net adds are going to Bouygues/Ideo, ahead of stalwarts Free.fr and France Telecom. Last year Bouygues launched a 45 euro ($60) Quadplay and it's become the hottest product in Europe. Bouygues, France's third wireless carrier, has come from nowhere to among the landline leaders with an offer of about a 25% discount. The land of the 30 euro triple play has become the land of the 40-60 euro quadplay, as SFR and France Telecom have now had to offer quad. 45 euro at Bouygues gets unlimited landline calls to 100 countries, "up to 20 meg" ADSL, 90 (not so popular) TV channels and 120 minutes of mobile calls.

     Frederic Jeanmaire of Merrill Lynch has put a buy on Bouygues. He acknowledges that mobile prices will be driven down by the entry of Xavier Niel and Free.fr in 2012, perhaps 10-25% more. He also believes the government program to bring down mobile and sms termination will continue working. Jeanmaire thinks those factors are already in the stock price after a large fall.

   Viviane Reding was on target when she called for eliminating termination charges across Europe. The cuts - fiercely resisted by most countries - have resulted in huge consumer savings across Europe. Prices set in the retail market for the services are more competitive than the charges of a terminating monopoly.  That's a strong argument for eliminating ICC in the U.S. as well, so long as it's not structured as a multi-billion giveaway to the bells.

*** ASSIA -- competitive advantages:  DSLAM vendor independence, field experience, high-speed automatic line re-profiling, and unmatched scalability. http://www.assia-inc.com/
 
(ad)

Free, France Telecom enterrent la hache de guerre http://bit.ly/bjNvc8
 
"Burying the hatchet" between Free.fr and FT became easier when "good cop" Stephane Richard became CEO, replacing "bad cop" Didier Lombard. Xavier Niel may be a bad boy, but with 5M customers and strong profits he's not going away. Fiber in France has been delayed nearly two years by FT causing disputes at the regulator, and after the peace agreement should be able to grow rapidly.

     Niel is dropping his competition suit against FT while FT is dropping their libel suit against Iliad/Free. There's probably a payment or unannounced side deal. The evidence that France Telecom abused competition law was strong enough to justify hundreds of millions in fines, so Free.fr had a likely winning case. Niel certainly said many things about France Telecom that would be libelous if false, but I believe them all substantially true.

    Figaro still calls Xavi "adepte de la provocation" but he's also a successful businessman who will need to cooperate with the other telcos on the mobile and fiber networks. France Telecom is spending most of their expansion capital and attention outside of the country, including a current offer to buy into Moroccan mobile. They also benefit from toning down the war on the home front.

    "Burying the hatchet" in the U.S. is illustrated with a picture of a native American, so I was surprised to find a similar phrase in France. However, no one in North America likely matched the fervor of the Celts and Gauls who invaded Italy and sacked Rome around 387 BCE, Roman accounts have the Celts as naked painted warriors whose ferocity caused trained Roman troops to break ranks and flee in terror.

    Did Gauls and Celts carry hatchets?

China Q2: 5.81M, Not 5.5M http://bit.ly/ajQpDr
 
The 5.5M estimate in the last DSL Prime was based on April and May. Lee Ratliff of iSuppli writes the June numbers are in and a strong month raised the actual Chinese total to 5.81M.
      Lee raised his 2014 estimate to 223M because " 2009 and the first half of 2010 have been white hot. There's a new trajectory on the curve." He remains concerned about the potential of a melt-down in China's economy, but notes that kind of thing is almost impossible to forecast.
       Both he and I would very much appreciate data about the rapidly expanding cable industry in China. It's moving to digital, starting to add modems, and I expect it will be a powerful incentive for all companies to invest more and grow.


Dado Chops 20% as ADSL Falters at Ikanos    http://bit.ly/bZG89A
 
Ikanos wrote down $12.9M of ADSL chips. Sales Q3 will be down 22% or more from $55.6M in Q2. 20% of employees are on the way out. Dado Banatao, legendary chip entrepreneur now in control of the company, has brought in new management from SiRF, another Tallwood VC portfolio company.
  Dado is now a VC looking for a way out of a troubled investment. He pointed out on the call that "VDSL is a growth segment." He sees the future of the company as highly differentiated next generation products. He expects revenue from bonded DSL in 2011, with both the U.S. and Asia interested. Dado claimed "Our vectoring is beautiful compared to our competition," but like then-CEO Mike Gulett he did not provide enough details to confirm that. Dado doesn't expect vectoring out of the lab until 2012.
   New CFO Dennis Bencala is blunt.” ADSL is a declining market. We have decided not to participate in RFPs that do not meet our requirement for margins." That's possibly true on the CO side, but most of the 5.8M new broadband customers in China needed ADSL modems. China's adding fiber, but they will require millions of DSL chips every quarter for a long time.
    John Quigley, who joined in June, will take over as CEO and Dado will remain in charge for Tallwood Capital. Michael Kelly and Jim Murphy are also coming from SiRF.

** Lantiq: Providing innovative solutions for over 20 years. Lantiq offers a broad portfolio of highly integrated, end-to-end solutions targeting Next Generation Networks and the Digital Home. These advanced SoC (system-on-chip) solutions address a wide variety of technologies, including VDSL, ADSL2+, SHDSL and VoIP. Our products include: Broadband Customer Premises Equipment, Home Networking, Digital Cordless, Broadband Access, Wireless Backhaul, Voice Access and CPE, Enterprise & ISDN. (ad)

$4B Cut in Verizon, AT&T Fiber+ DSL Spending http://bit.ly/dvkgyc
 
Verizon's wireline capital spending in the first six months of 2010 was 3.35B down nearly $1B from last year. For the full year that is nearly a $2B drop, which corresponds to their plan to cut the FiOS build in 2010 by 2/3rds. Since they've also cut the post 2010 FiOS build by 2-4M homes, this is probably a permanent drop. The numbers at AT&T are similar but not broken out. AT&T cut U-Verse by 1/3rd last year, one reason they went 92K negative on broadband this quarter.
     Spread over 4 years, the total cut in Verizon wireline/FiOS spending would be about $7B, about the same as the total government money spent on the broadband stimulus. It also corresponds to the Verizon's likely share of the broadband tax. (I am writing separately that the broadband tax will go mostly to the company's bottom line, not expanding broadband.)
    These multi-billion dollar cuts came after the U.S. enacted a stimulus program and now is talking about huge subsidies supposedly for broadband. Ivan is a smart guy who told investors that he thinks the government will pay up if he doesn't invest. The stimulus, as Tom Hazlett predicted, resulted in fewer new broadband connections as company after company reduced spending hoping the government will pay instead. Verizon is claiming 20-30% of their phone lines require a subsidy and asking for billions or they might discontinue voice service. Ironically, V & T just reassured wall street their wireline margins are staying up.
Brett Feldman if Deutsche Bank just put a buy on T, citing among other factors the stability in wireline.
      CFO Killian projects the total AT&T capex for the year about flat at $17B. Most of this year's wireline/FiOS cut will invested in wireless. John Hodulik projects they will cut back even further this year, perhaps another half-billion. Verizon's capital spending is significantly down as a % of sales and per customer, as Alltel has made them a much bigger company while the gross spending has been about flat.  Verizon's spending does remain above depreciation and by most measures Verizon is investing more than any other large North American telco.
     AT&T dropped capex in 2009 an amazing $3B, which is the actual cause of the iPhone network problems. They paid a high price and have to restore some of that this year. CFO Lindner pointed to cable clobbering them in areas without U-verse as the primary cause of the decline. "We have seen competitors become more aggressive with promotions targeted specifically at our non-U-verse areas. And we will respond."  Meanwhile they are both upselling and raising prices. "Our overall consumer broadband ARPU is up year-over-year over 5%. And our consumer revenue per household is up about 7.5%. ... it gives you the opportunity to maintain and strengthen margins."

Correction:  The 5.5M estimate I published last issue for China Q2 is updated with an actual figure above of 5.8M.

email

    * A rural telco guy writes "Beyond USF, the FCC wants ICC rates to go to zero. That will put RLECs out of business unless there's compensation, or at the minimum, put all capital spending on hold." OPASTCO and NTCA, the trade associations for the little telcos, warn of bankruptcies and major defaults on RUS and CoBank loans.

briefs

    * China Tietong will achieve 10M broadband customers by the end of the year and pass Verizon according to a Huawei press release.  Now affiliated with China Mobile, the #3 Chinese broadband carrier may add more customers this year than all the U.S. carriers combined.
    * Sycamore announced IQstream, which will be very useful for mobile backhaul if it works as promised. Using a node at the cell site and a master at the base, Sycamore claims they can dramatically increase the throughput for 3G whether the backhaul is microwave, copper, or fiber. This would be enormously useful if backhaul is congested, a problem that has called for major upgrades. (AT&T says they have fixed it, but won't release enough data to convince people.) Since fiber can take months to run, IQstream would be a good interim improvement. It can also bridge a modest amount of time between when you should upgrade and when you have the budget. In some cases, it might eliminate the need to upgrade. But Sycamore refuses to supply enough information about how it works, actual performance, or cost. They are a reputable company, so I'll keep a watch when they ship them. That's by the end of the year, they hope. If you have a major need for interim backhaul improvements, it's probably worth a look.

press

    * Karl Bode at DSLR “asked the FCC what the likelihood was that they'd enact EU-style rules that require carriers provide consumers with tools to cap monthly mobile data spending but they refused comment. The agency also wouldn't comment to Broadband Reports on ...” It is now months since Julius announced a press conference, and that was scheduled for only 15 minutes. Shall we start calling him “Julius the Scared?”
    * Darren Pauli at Computerworld Australia headlines Budde barney over NBN blowout. Malcolm McKenzie speculated the broadband network could cost $89B instead of $34B. Analyst Paul Budde thinks that's “idiotic.” Budde's almost definitely right. The Australian cost estimates are substantially higher than similar networks at Verizon and AT&T so should be realistic. Australia has made the smart choice of using 5 megabit satellite for the last 1% who can be brutally expensive to reach. The U.S. plan calculated that ½ of 1% cost into the tens of thousands per home. Using satellite for those last few halves the cost. Reaching 100% landline would cost $20-30B; 99% landline and 1% satellite is $10-15B. That's the right choice but the plan couldn't say so directly. Instead, they suggested that “money should first be spent where the bang for the buck is biggest.” That could be enforced by “placing low priority on any homes requiring subsidies of $5,000,” with the unspoken inference they'd never be wired.
    * “Wicked smart” Fawn Johnson moves from covering telecom at Dow Jones in D.C. to the National Journal. Her new editor-in-chief Ron Fournier believes she knows “how to dig beyond the talking points and spin.” If more D.C. policy reporters actually did dig beyond the talking points, web reporters outside D.C. wouldn't be constantly out-reporting those inside the beltway.
    * Jennifer Loven, top AP White House correspondent, joined the Glover Park Group. Reporters, like government officials, like the million dollar salaries common in the D.C. flack and lobby business. 

people

    * Phil Anschutz, the guy in charge of Qwest when everything went bad, has given up his planned grand luxe American Railway Explorer. (Business Insider) “21 Streamliner-era rail cars graciously adorned with rich leathers, dark wood accents and rare fabrics” were set to travel through Crater Lake,the Columbia River Gorge, Yellowstone, Yosemite, the Grand Canyon, and across the country. Joe Nacchio used to complain about having to clear things with Anschutz. So Phil's statements to the Congressional investigating committee about not being involved rang hollow. Billionaires rarely go to jail, of course. Henry and Henry of Broadcom beat the rap on a technicality despite overwhelming evidence.
    * Larry Irving, longtime D.C. advocate now keeping a low profile, will visit Tasmania for Hewlett-Packard after a visit from the Tasmanian Premier claiming Tasmania will have the best Internet service in the world in 2014. (Slightly exaggerated.)

wall street

    * Richard Greenfield of BTIG echoed the common opinion that neutrality and related FCC fights will have little impact on earnings/investments for years, if then. In which case, the decline last week was overdone and it's a buying opportunity if cable stocks are where you want to put your money.


policy

I've set up a new site, DCjester.com, to pay all due respect by laughing out loud at lobbyists' and politicians' appropriate remarks. Help me by passing on LOL quality absurdities from all sides of the policy discussion.  Rick Whitt of Google is the first honoured for saying “Google has a close business relationship with Verizon, but ultimately this proposal has nothing to do with Android.” Rick, I have a bridge to sell you in downtown Manhattan.
     For some related D.C. gossip, check http://bit.ly/bBOLOJ
 
and http://julesgenachowski.com
 


NTIA's Smart Stimulus Review  http://bit.ly/bRrmDQ
 
Tamia True's proposal to review the results of the NTIA Broadband stimulus is thoughtful and right on target. If delivered to specification, it will be one of the most carefully done studies of any broadband program in the world. The contractor must produce "a report of sufficient methodological, theoretical, and pragmatic rigor to withstand critical scrutiny by technically proficient practitioners."
    The proposal is clear that that BTOP's #1 priority is "To provide access to broadband service to consumers residing in unserved areas of the country." The total number will almost certainly be so small that the program has to be judged a failure at its primary goal.
      Broadband policy to date has been shaped by a series of "studies" that fall apart when anyone competent takes a look at them. Nearly all are funded by the companies wanting government money or favors, especially Verizon. They are garbage in, garbage out quality work.  
     There are two huge problems with this review. The first is that the primary results will not be available until 2014, long after the money has been spent. The officials responsible will probably have moved on. It's also likely that the terms of the contract will eliminate most, if not all, qualified bidders. No one can bid unless they are qualified under the GSA Mission Oriented Business Integrated Services (MOBIS) Schedule. It takes months, at least, to become MOBIS qualified - long after the contract is awarded.
    There are only a few dozen people with the experience to do this job well and as far as I can determine none of them works for a MOBIS contractor. So the odds are the contract will go to one of the usual beltway bandits who have fancy paper credentials and a knowledge of the system but not the subject. They make their living off government work and nearly always deliver a glowing report that praises the agency.
    More, including a pointer to some fraud, http://bit.ly/bR

 
Reply "subscribe" to be added, "un" to be dropped
Volume 11, #13 August 18 2010   

"Profits for members of the New York Stock Exchange totaled $61.4 billion in 2009, the most ever." NYT

Paul Kagame, Rwanda's President, co-chairs the U.N. ITU Broadband Commission. Under his regime, Rwanda has dramatically increased both international connectivity and local computer use. But is he the right choice if "Rwanda Opposition Figure Found Dead" is the newspaper headline?  The story is below for you to judge.
     In Australia on the winter side of the world, Julia Gillard hopes the 100 meg national broadband network will help help re-elect her Prime Minister. "Polling shows that the NBN is regarded as almost a universal positive for the government, especially now that initial sites are deploying," Grahame Lynch reports. Gillard warns voters that if they vote for the other party their Internet will be slow for the next decade.
    On the summer side of the world, Jennie and I are heading to Washington looking for honorable politicians. Everyone talks about broadband but no one is doing anything about it. 2Q will be the worst quarter for U.S. broadband "since the service became available," predicts John Hodulik of UBS. He expects cable to take three out of four of the half million or fewer new subscriptions.   
    D.C. just confirmed one failure. $4B for NTIA stimulus is not bringing the promised broadband to the unserved. According to NTIA's incomplete spreadsheet, they are spending $1,759,530.79 for each home newly offered service. (Policy, as always at end.)
   This one's for Bradley Manning and Julian Assange. Truth to power, whoever is in power. Say hello to the round fellow with a beard and the irrepressible Jennie Bourne at MMTC in D.C.

Pakistan "Taking Off"
"Broadband has taken off, with its subscriber base up 1.5x to 0.82mn, where PTCL leads with 0.48mn subscribers." Muhammad Saqib Sajjad of KASB & Merrill writes. "We expect broadband subscribers to rise to 2.8mn by FY12E. Unmet demand from underserved/unserved areas is the key driver, we believe, as operators increase their outreach, backed by USF subsidy."
 Pakistan has only 3M phone lines for 100M people, so DSL will soon be dwarfed by wireless. For now, DSL is 55% of that 0.82M. Sajjad includes in "broadband" WiMAX and EVDO, which will be the natural way to count in Pakistan, India, Indonesia, and most of Africa. Wireless data will be dominant despite the relatively modest speeds.

*** ASSIA Dynamic Spectrum Management. The crucial first step to better DSL performance. (ad)

Broadband's Great Middle: France, Germany, UK and (yes) U.S. http://bit.ly/aeO8ep
France has 30.4 lines per 100 people, Germany 30.3, Canada 29.6 and Britain 29.5. The penetration is effectively identical. While the U.S. is lower at 26.4, much of the difference is larger family sizes and more extreme poverty. Scandinavia, Switzerland and Korea are far ahead, but most of the larger countries are surprisingly close.
    France has the 30 euro triple play but as Scott Wallsten mentions is not much more deeply penetrated. Japan in the past had very attractive prices. Although that edge has eroded, it's surprising to find them at 24.8. I believe that's because many in Japan are happy reading their email etc. on their cellphones. Canada had been far ahead of the U.S. for years, but CRTC has allowed Bell and the cablecos to effectively create a cartel and raise prices.
 Those unhappy with U.S. policy or wanting government charity have been screaming "the U.S. is #15." That's true, but the gap with most countries is modest. My opinion is being in the great middle is unacceptable and that the U.S., the first in large deployments, should be among the leaders. The FCC obviously doesn't agree, because they've just allowed another series of demand killing price increases.
      The data, which may surprise you, at http://bit.ly/aeO8ep

*** Kevin Werbach's Supernova Forum 2010: Perestroika is a superior gathering of articulate Internet folk. July 29-30, 2010 at the Wharton School in Philadelphia, PA. An ideal chance to share ideas with some of the best, including key startup entrepreneur John Borthwick (Twitter, among others), Jeff Jarvis (most interesting thinker in new journalism), danah boyd (Microsoft), Beth Noveck (White House), and several dozen others. (psa) Code sn75 to attend for just $75

Free U.S. Wireless Ain't Dead - M2Z is Baaack! http://bit.ly/cbDdHh
John Doerr and Milo Medin visited the FCC 13 July with a crazy plan. They would provide 768K wireless free across 95% of the U.S.. Even with a good deal on the spectrum, it will require investing $5-7B before breakeven. Or maybe it's not so crazy.  Doerr is a legend who was the money behind Google, Amazon, Compaq and a dozen others. Milo built most of the cable networks in the U.S. as CTO of @Home.  The audience included the Chairman and six chiefs. Julius is taking this seriously, and Doerr is confident they have the cash needed.  There's essentially no downside to Julius saying yes except the predictably violent opposition of the incumbents. The telcos are screwing him so many ways he might just go ahead.
    Doerr's partners include Al Gore as well as Republicans Colin Powell and Tom Perkins. The Senate is pressing Julius hard because he's not doing enough on affordability. This wouldn't have much impact until 2015 or so, but is a partial answer at the very low end.
    M2Z wants the 2155-2175 AWS-3 spectrum. That 20 MHz is as much as Verizon is devoting to a worldclass LTE network, although at the higher frequencies far more towers would be needed to match VZ. The business plan begins with selling at market rates the higher speeds and probably voice. Doerr is on the board of Google, so they've presumably have thought through the advertising. Google put $3.5B on the line for spectrum in the hope of breaking the wireless cartel and has advanced $billions to win ad contracts from AOL and others. With over $20B on Google's balance sheet, they are a natural source of financing.
    They promise to reach 50% of the country by 2016 and 95% by 2021. The technology will presumably be LTE, for which inexpensive gear should be widely available in 2013-2015. They would require radios on 3,000-10,000 towers in 2016 and can plan the build for where towers and backhaul are available at competitive prices. They also will have a natural partner if Harbinger makes it. Some think that a long shot, but Phil Falcone just filed for a $400M debt offering for Harbinger.
    Especially because Doerr is one of Obama's most important supporters and active in the White House.

*** ASSIA -- the world leader in dynamic spectrum management.  Our performance and reliability are unmatched. http://www.assia-inc.com/ (ad)

Chunghwa: Double Speed, Same Price to Find Subscribers http://bit.ly/dfmjvi
Taiwan's achieved 80% household penetration - higher than the U.S. - by keeping prices down. The market is so close to saturation that modest problems led to Chunghwa actually losing broadband customers in 2009. To counter that, and with government encouragement, they've doubled speeds at the low end. ~$8/month DSL will go from 256K to 512K down. The ~$22/month offering goes from 1M/64K to 2M/128K. That's after an 8% price drop last year. Low income customers get 50% off the line rate and disabled customers also get discounts.
 Much of the island can get 100 megabit VDSL and IPTV. Any building promising to sign up 20 apartments will get 100 megabits within a few months. They've just ordered 200K IPTV set tops for $90 each in an open tender. The suppliers - DinYen and HwaCom - surely would love to bring their low priced units to other markets.
 Chunghwa often pays less for equipment than giant companies such as Telstra and AT&T because open tenders work.
The U.S. needs to apply them for our $2B schools and libraries program. The information is invaluable. Alcatel once demanded I retract a story Telstra's costs of doing FTTN were out of line because they were twice as high as similar builds elsewhere. I wrote that implied either Telstra was incompetently organizing the build, fudging the numbers, or paying their supplier Alcatel twice as much as others paid Alcatel for similar gear. I pulled out Alcatel's bid to Chunghwa, which was far below what they were charging Telstra. They promptly shut up and never did apologize.
 Shyue-Ching Lu of Chunghwa will be in Paris in December for the Broadband World Forum. It's going to be a good show.

*** ASSIA DSL Expresse:The choice of the most demanding carriers in the world http://www.assia-inc.com/ (ad)

AT&T's Bonding for the Millions; No Speed Increase, Just More TV http://bit.ly/cUFCiF
Why would anyone want more than 24 megabits?" AT&T now-CEO Randall Stephenson asked me years ago. The remarkable sales of 50 and 100 megabit DOCSIS 3.0 around the world have answered that question. Users want speed unless the price is ridiculous, with cablecos in Holland, France, and England dramatically increasing their growth rate. In Holland, Mike Fries of Liberty Global/UPC tells me DOCSIS 3 is actually outselling fiber where they compete head to head.
   However, T will only use bonding to extend the range of their IPTV from about 3,000 feet to close to the original specification of 5.000 feet. They will continue to hold speeds to 18-24 down, 3 up. 70% or so of U-Verse homes - including all of them now served - could get 50 meg down and 5 up, but overpriced cable competition in the U.S. leads T to believe they wouldn't find a market.
   Karl Bode at DSL Reports reports AT&T originally told him bonding would be used in 2007, so they are nearly 4 years late. DOCSIS 3 prices in the U.S. are literally twice the price in England and France, although it's less expensive to deliver 50 meg DOCSIS 3 today than 6 meg service a few years ago. If U.S. cablecos priced 50 meg at the $30-50 European prices instead of $99, there would be a mass exodus from the telcos. There's obviously a "gentlemen's agreement" preventing that,

Email
John Egan writes about Ed Eckert, who is moving on from the standards job at Ikanos. “I have known Ed for going on 10 years and have a ton of respect for the man as a person and expert in the field of telecom. His leadership and strategic abilities are evident whenever he becomes involved in a subject and he will be sorely missed. He has had my respect from when we first met, even though we were on different sides of the VDSL line code battle. I believe the success of the other side was in no small part due to his contributions. Since that time he has made his mark at ITU, Broadband Forum, and elsewhere. As I said before, he’ll be sorely missed. I look forward to crossing paths with him in the future and tossing back a few.” The technical work of the standards groups more than a decade ago is while we don't have brutal interference problems in today's DSL networks. They remain active with local retransmission and other improvements on the way.
Press
FT's Andrew Parker & Paul Taylor revived the rumor of a merger between Sprint and T-Mobile, the U.S. #3 & #4 without any fear of a rejection by competition authorities. Reducing competition certainly allows the companies to increase prices, which is why the U.S. regulator is basing policy on having more competition. Andrew & Paul: Are you implicitly assuming the regulator is lying, stupid or impotent? Example after example prove that fewer but larger companies do not make for more effective competition. If Julius isn't lying about wanting more competition he will oppose the deal if he isn't stupid. Or perhaps the FT guys think he has no power. President Obama is building close ties to Brian Roberts of Comcast and Ivan Seidenberg of Verizon, part of a campaign to show he's not anti-business. Does the FCC still have the political clout to prevent anti-competitive mergers?
Bob Fernandez at the Philadelphia Inquirer has been doing exceptional reporting about Comcast so I'm very glad they've settled their contract without a threatened lockout. The reporters are paying a high price, about a 15% cut. If broadband didn't exist, the papers would still have been very profitable without needing the wage cut. Broadband is great for many, bad for some.
Ian Scales at TelecomTV asks "UK Digital Economy Act. ...‘Why?’ Why, in the glow of what was clearly the fading embers of the Labour government’s 13 year stint, was it deemed necessary to rush through The Digital Economy Bill. Let’s just say that we will be watching carefully to see which politicians end up within the entertainment industry... rub [bing] shoulders with movers and shakers in music and entertainment." My uncle the movie producer has some very funny stories of "very rich guys hanging around Hollywood looking to meet actresses." Actually, Lord Minister Stephen Carter, once a slick ad guy, has moved on to be hypemeister for unglamorous Alcatel. Hiring Carter is not a good sign for Alcatel's future; his previous experience includes leading NTL cable into bankruptcy and strategy advice for Gordon Brown on his way to an historic defeat. Perhaps Lord Peter Mandelson, will get a job with one of his friends like David Geffen or Paul Allan. After dining with Geffen at the Rothschild's, Mandelson made the bill even more industry dominated. For that achievement, Lord Peter was recently named UK Internet Villain of the Year.
people
Randall the Silent CEO is coming out.  Randall Stephenson as AT&T CEO has been publicly almost invisible, letting Donovan, Stankey, and de la Vega do the public speaking and even boycotting his own trade association. I'm glad to see that change, with several public events scheduled over the summer. As CFO, he was "a good finance guy" but made frequent faux pas about the Internet and other parts of the company, a good reason to hide. Since he's taken over, exceptional results have shown he can learn and delegate. He's impressed me and I'd love to hear more of his thinking these days. Maybe one day he'll start taking reporter's questions again.
Marc Zionts was CEO at Westell a decade ago when they had hopes to be the DSL leader worldwide. His current company, Ortiva Wireless, was featured in an article at Fierce Wireless http://bit.ly/aCBUz6 on Mobile video traffic: Alleviating the capacity crunch. The article reports Matthew Oommen of Sprint and Colson Hillier of Verizon are looking at ways to reduce the traffic load as more people watch video.  It appears that optimized streaming to mobile can have a major impact on network capacity.
Wall Street
Nikos Theodosopoulus of UBS put a buy on Calix because of their prospects of selling gear to the stimulus winners. They announced $100M in stimulus related sales and the stock is up 20% after earlier disappointments.
Policy International
Paul Kagame was introduced to me by a U.S. State Department official.who told me Kagame was doing remarkable things bringing the Internet to his country. However, his government has recently been criticized by human rights advocates and reporters.  Numerous readers of DSL Prime are involved in UN and ITU - I'd appreciate perspective from those who know his work in broadband.

U.N. Broadband Chief's Opponent Dead http://bit.ly/chnKGc
Rwanda President Paul Kagame co-chairs the U.N. ITU Broadband Commission. Under his regime, Rwanda has dramatically increased both international connectivity and local computer use. But is he the right choice if "Rwanda Opposition Figure Found Dead" is the newspaper headline? Josh Kron (NYT) reports“the nearly beheaded body of party official Andre Kagwa Rwisereka was found by the banks of a river." Reporters without Borders called on the European Union to suspend their assistance to the Rwandan government "following a series of grave press freedom violations," including the murder and arrests of journalists. He has also been involved in the war in Congo, where millions have died.
    On the other hand, Kagame played an important role in ending the massacre of hundreds of thousands nearly two decades ago. He works hard and intelligently for his country's development. This tech reporter, 7,000 miles away in New York, is not the one to judge.

Policy U.S.
I don't believe the $1,759,530.79 per unserved home figure that jumps out from the spreadsheet NTIA sent on my FOIA request. But even after adjusting to reality, the results of spending $1.2B are scandalously poor. Barack promised broadband available for all. The stimulus money, if not wasted, could get us to 98-99%. It's hard to understand why some really great appointees are failing so miserably, but the facts suggest that the remaining $3B for NTIA should be mostly frozen. The results aren't coming, either in jobs or broadband deployment. 
    Meanwhile, D.C. is planning a massive and unneeded broadband tax with most of the money going to the telcos instead of helping poor and rural consumers. Well-informed lobbyists are sure the $150B Woman will recommend putting a $5-10B tax on broadband and hand 9/10ths of the money over to the carriers. Verizon is now demanding a huge increase despite WP reporting V & T are already collecting $billions from USF. Mostly behind closed doors and really, really ugly.
    On the other hand, I'm reporting what looked like another scandal - the Windstream proposal - is not that bad. The public information was incomplete and the company insisted on holding back the facts. Both Qwest and Windstream are asking substantially more than they need in subsidy, but if Adelstein negotiates well he can prevent anything inappropriate. I wish RUS would go out and solicit cost-effective projects rather than funding expensive ones, but the system doesn't know how to work that way.
    In D.C. this week slews of folks will embarrass themselves saying things like NN will produce significant investment reductions, that we need $100B or $300B for subsidies for broadband after the planners carefully calculated $15B will get us past 99%; that networks are getting slower and congested when actually the opposite is true; some of the most profitable companies in the world need massive subsidies or they'll have no choice but to shut down the phone networks; and that Julius is delivering affordable broadband.
    I'll try not to laugh out loud.

Wall Street Confirms: Modest if Any Investment Effect of Net Neutrality http://bit.ly/aCfAPn
Net neutrality/reclassification opponent Thomas Seitz (Height Analytics and previously Barclay) today joined the parade of top analysts doubting the claims that Net Neutrality rules would produce a serious cutback in broadband investment. Washington is inundated with claims NN will clobber investment, but the carrier CFOs are telling Wall Street NN won't be a determining factor. Seitz joins John Hodulik of UBS (voted #1 telco analyst), Craig Moffett (voted #1 cable analyst) and Michael Rollins of Citigroup as well as several others who haven't gone on the record.
    There are plenty of sensible arguments against Net Neutrality, voiced by Dave Farber and others. Governments frequently muck things up when they get involved. But the "investment will be clobbered" doesn't stand up to the facts. A panelist added "competition is a far more important factor in driving investment. ... the market usually adjusts to things like this successfully." Columbia's Bob Atkinson, who led a factfinding report on future broadband investment for the plan, at the NYLS seminar said he didn't think NN - or most other policy measures - would have a major impact. Atkinson and Schultz's paper is used by all sides in D.C. as a primary reference.
    Seitz thought "Title II reclassification" to enable NN would be bad for investors, but probably not NN itself. "It's the nose of the camel in the tent. The real investor fear is that the laws' other provisions - including 'reasonable and fair prices' - might be enforced. It's those other issues that make people fear reclassification." Moffett is even blunter. "Importantly, the regulatory issue here has nothing to do with 'net neutrality.' The section of the Communications Act mandating 'just and reasonable rates'  opens the door to broader price regulation." I refrained from asking Seitz and Moffett whether they believe in "excessive and unfair rates." Their job is to find companies with the market power to do just that, not decide what's good public policy.
   Hodulik a while back researched the technical side of the issue and discovered that at modest cost most large networks had avoided congestion. Glen Campbell of Merrill Lynch looked at wireless costs, and even on wireless the cost of additional bandwidth was significant but not enough to change the economics of the service.

No Scandal in Windstream RUS Grant After All http://bit.ly/baxii2
Windstream wants $238M in 75% grants from the RUS program for broadband the unserved, but their application was challenged by 39 companies serving the areas they want funding. All publicly available information suggested Windstream would reach so few unserved it would be a scandal to give them the money. When I checked with them and pressed, they gave me no reason to believe otherwise. Only after I spent the July 4th holiday researching and writing the article would Windstream make public even the most basic facts of their application.
       Fortunately, there is no looming scandal other than the company's and RUS refusing information about how public money is being spent. Windstream now tells me they will reach 160K-190K unserved, which means it's a reasonable project. That's a total cost of perhaps $2,000 per unserved home reached, very expensive but not out of line with what many extreme rural builds costs. They plan 7,000 miles of fiber, which at $20K/mile will cost $140M.
      I urge RUS and Windstream to go over the maps and find a way to stretch the $238M to reach 60-80% of the unserved Windstream homes, not 30-40%. That could be done by re-allocating part of the money from currently served territories to the unserved. In going over the details with me, Schueneman made clear he understands the ways DSL can be extended further, including mini-DSLAMs, small pedestals, pole-mounting, and bonding. With guidance from RUS, I think he can reach more homes.
      Alternately, I hope RUS goes closely over the proposal with an eye to saving public money. The rules specify that RUS should give priority to projects asking less than 75% grants, a powerful lever to negotiate the grant percentage down. (This is even more important in the larger Qwest application. Both are reasonable projects, but according to the publicly available data many of the subsidies could be reduced 30-60% and the companies would still have a good deal.)
   The article I killed was "Windstream: Say No! to our RUS Grant." It turned out Windstream VP Frank Schueneman had spoken out when NTIA funded a competitor to them, the notorious North Georgia project. My research did find that without a subsidy, Windstream could and should do more to serve their customers. They are currently running with capex 30-40% lower than depreciation, close to the bare minimum to keep the system running. The capex gap is about half a billion, more than enough to cover costs of broadband if regulators had courage and competence. They are using the cash temporarily saved to maintain one of the highest dividends on the stock exchange, 9.4%. In 2009, their dividend was $100M higher than their earnings.      More, including the original article at http://bit.ly/baxii2

New Leaders in DC http://bit.ly/baxii2
Doug Sicker of the University of Colorado is taking over as FCC Chief Technologist.  I wrote Doug Sicker: Ethical Researcher  http://bit.ly/Qq1CQ because he clearly revealed on his web site his funding sources. Doug's first official act should be to bring back the Technical Advisory Committee, because the lawyers at the top of the FCC desperately need to hear from some alpha geeks.  Editorial: End FCC Technical Impoverishment http://bit.ly/cVPocU. Bruce Gottlieb is leaving the FCC to be general counsel for Atlantic Monthly and related publications, a major change from all the FCC staffers who've left to become high-paid lobbyists the last few years. Rick Kaplan is taking over. As Chief of Staff for Mignon Clyburn, Rick has done solid work on issues like how the FCC actually can support minority media. Mignon, who came to the FCC as a great unknown, is proving to be forceful and thoughtful. Jim Newby is taking over from Lisa Zaina as Chief of Staff at RUS. He's a veteran of the electricity side of RUS.

Unbelievable NTIA Figure: $1,759,530.79 per Unserved Home Passed http://bit.ly/cEYrnx
39 of the first 40 projects Larry Strickling approved in the first stimulus round at NTIA do not aim to bring broadband to people who can't get it. NTIA spent $1,759,530.79 of public funds per unserved home according to the spreadsheet I just obtained using the Freedom of Information Act. Yes, that's millions, although I sure the figure would be reduced to under $200,000/home and probably under $100,000/home if the agency carefully analyzed the projects funded for other reasons.
      I'll have a more later in the week when I return from Washington and I will post the data. Since NTIA is currently considering spending another $2B or so of public money -- much I believe wasteful -- I wanted to get the basic information out before I left.
      The one out of forty projects actually designed for people without service is 682 homes at Williamstown, Kentucky.  Some of the other projects cover areas where the broadband take rate is low, while many are for backhaul overbuilds or school connections where service is already widely available.
     The Obama platform promises "We will ensure every American has access to highspeed broadband," which should have come first. A half dozen Congressmen of both parties at an oversight hearing castigated NTIA and RUS for not reaching the unserved which they affirmed was the primary goal of the $7B they allocated.
    In May I wrote "Strickling amazed me by giving the House Committee a big 'eff you, saying he would use the money not for people without service but instead for faster connections to community colleges, etc. ... . Boucher and Markey have backed off criticizing NTIA because of the election."
     I have three times asked Strickling "How many unserved have you reached and how many jobs directly created?" He always refused to answer, and I now have only a very partial answer a year later in response to my FOIA request. In addition, NTIA does not even have an estimate for "jobs directly created," instead using some figures that were seriously discredited by Raul Katz of Columbia and other scholars as well as the work of the broadband plan.
   The $7.2B could easily have brought megabits to 98-99% of the U.S. if spent well, but will only reach about 20% of what was practical. With hindsight it becomes apparent the program was based on a totally false premise: Julius Genachowski's belief that 20% or so of the U.S. could not get broadband. The real number is about 5%, most of them in such small groups they can't be reached economically by anyone without local facilities. In most places, only the local telco and cableco have facilities, but the program was built on an assumption many others would bid.
        When the applications came in, nearly all the insiders knew the program was likely to fail. One top FCC official in August told me "I've given up on the broadband stimulus having any effect at all." Most of the money should have been frozen and everything rethought. Adelstein at RUS had the courage to say "I was wrong" and change the rules for round 2 but NTIA didn't.

Say Hello at MMTC in D.C.
Feed a story to me and meet the irrepressible Jennie Bourne and her camera. A slew of D.C. officials will  attend. Five FCC Commissioners, Jonathan Adelstein of RUS and Tom Powers of NTIA. Blair Levin of the broadband plan will also be on hand, and is always worth listening to. FCC people tend to be informal and open to introductions and questions at events like this while rarely holding press conferences, so this is a good opportunity for reporters to pin people down. Among the other speakers are three who had hopes of being FCC Chair (Leo Hindery, Julia Johnson, and the vivacious Debbie Tate,) as well as the usual predictable lobbyists who will repeat sometimes true things they've said far too often before. The Minority Media and Telecommunications Council has long played a role connecting people to financial opportunities. MMTC has cultivated the closest possible ties with the leading companies and provides strong support of many corporate positions I disagree with. One advantage of those ties is they have connections that back up their sessions on how to access procurement opportunities and raise capital. The fee is $375 for two days and it's a good way to connect with official D.C.

Held for next issue
Free.fr $20 Dedicated Web Host Server http://bit.ly/9atRgk
U.S. Congress: 4 Meg Sucks http://bit.ly/aHfwld
Goldman Sachs on Wireless Carveouts http://bit.ly/cbiBz7
Killer Apps: Le Monde, Al Jazeera, ?Jerusalem Post TV  http://bit.ly/94RhMf
Julius & Frontier West Virginia: The Very Ordinary Facts http://bit.ly/a2FJB0

popular articles
$13/100 Meg Profitably Sweeping Hong Kong http://bit.ly/9oMIso
Switches Getting Cheaper; "Don't Bet Against Moore's Law" http://bit.ly/b8464u
John Stankey Is Not a Big Fat Liar  http://bit.ly/blEGpP
Cisco VNI: Internet Traffic Growth Rate Falling by Half in U.S. http://bit.ly/cqZkaX

Reply "subscribe" to be added, "un" to be dropped
Volume 11, #11 July 17 2010

Jacob Vigdor and Helen Ladd "examined computer use among a half-million 5th through 8th graders in North Carolina. They found that the spread of home computers and high-speed Internet access was associated with significant declines in math and reading scores. This study, following up on others, finds that broadband access is not necessarily good for kids and may be harmful to their academic performance." (David Brooks NYT)

     Broadband is a good thing, but the promotion is over the top. Multiple honest academics have found the effect on jobs and the economy is so modest it can barely be measured.  Telehealth expert Richard McManus, whose TASMINH2 hypertension trial was just reported in The Lancet, emails me "We used a very simple dial up internet access which did not need broad band." Broadband's exploded to 60-70% penetration but telehealth is growing slowly, suggesting factors other than broadband are the key drivers.

    Don't believe the hype.
---------------
Saul Hansell's NY Times article last year was critical in killing a $1B pure giveaway to Verizon that was added to the stimulus by the Senate Finance Committee.  Saul's moved on, and Cecilia Kang at the Washington Post can become a leader in U.S. telecom reporting if she follows up her story on universal service.
    David Cameron has killed off the despised broadband tax that was really a giveaway to British Telecom, but the FCC is considering a multibillion dollar levy on broadband that will almost all go to the companies in return for very little investment. The big telco lobbyists expect a windfall of at least $1B/year and more likely $3-4B from ICC & USF changes.  (policy at end)

*** ASSIA Dynamic Spectrum Management. The crucial first step to better DSL performance. (ad)

DSLAM 15 Miles from Power: Adtran in Alaska http://bit.ly/ctcKcD
Mitch Vieu, plant manager, Copper Valley Telecom likes the Adtran 1124 because it works "70 miles from the nearest office and 85 kilofeet from the nearest commercial power source,” The local mountains get 700 inches of snow and there's not a town on the map within 50 miles of Valdez.  CVT is a co-op and wants to provide service everywhere, a tougher problem around Valdez than in most territories.
     Adtran traditionally was a Bell supplier but is now moving aggressively after independent carriers. They have a reputation as reliable but high-priced, but they've been cutting prices drastically to win new customers. Their big win was in the ex-Verizon territory of Frontier, where the better part of 1M lines are being deployed.
      While there are a few places that are brutally expensive to serve, the broadband plan found truly high costs at fewer than 1/2 of 1% in the U.S. The Adtran is a line-powered 24 port DSLAM that goes almost anywhere. Ericsson has units the size of a book for 12 ports.  There's nearly always a cost effective way to get to customers, despite what you hear from the beggars in business suits demanding subsidies.  The pr http://bit.ly/ctcKcD

*** Kevin Werbach's Supernova Forum 2010: Perestroika is a superior gathering of articulate Internet folk. July 29-30, 2010 at the Wharton School in Philadelphia, PA. An ideal chance to share ideas with some of the best, including key startup entrepreneur John Borthwick (Twitter, among others), Jeff Jarvis (most interesting thinker in new journalism), danah boyd (Microsoft), Beth Noveck (White House), and several dozen others. (psa) Code sn75 to attend for just $75

2 Terabytes to West Africa http://bit.ly/aUtEOt
"You go girl!" headlines the web page for Engineer Funke Opeke at the Queen's School Old Girls Association. Ms. Opeke went on to Columbia University and Verizon before returning hometo Nigeria. She's now brought massive connectivity to Lagos and Accra (where one's heart belongs) via the 7,000 kilometer Main One cable. Working with Tyco and Alcatel she's already planning the extension to meet SEACOM in South Africa. The cable is operational today, delivered on time and on budget.
 A terabyte is enough for 10M broadband connections at today's rates for DSL, and probably for 50M if much of the information is sourced or cached locally. They've only lit 30 gigabits per section but are able to expand that rapidly if there's demand.
 “Today is a historic day for West Africa. The arrival of the Main One cable proves that much good can be done by Africans for Africans. We are pleased to realise the fruit of our dedication and commitment in the past 30 months. More importantly, we are happy to be a channel for driving growth in Africa and changing the status quo for the average African as reliable internet connectivity becomes easily accessible and affordable for all” Fola Adeola

*** ASSIA -- the world leader in dynamic spectrum management.  Our performance and reliability are unmatched. http://www.assia-inc.com/ (ad)

Components Currently Are Short http://bit.ly/d1lNud
Except possibly in consumer wireless, end-user demand in the continuing recession is not enough to logically create shortages. Computers and mobile phone sales are up 10-20% because of demand from Asia, but other sectors are not growing that fast in our economically troubled time. There are clearly shortages of many components along the supply chain, but it's hard to know how much of that is over-ordering for protection. Production levels are strong and most delivery commitments are being met. Those who want additional quantities often must wait two to six additional weeks, however.

 This story was inspired by Rick Pierson's report "Severe Shortages Impact Key Commodity Components." Rick is with iSuppli, a reliable analyst firm, and his findings correspond to what I'm hearing informally. "Across the board, lead times are longer than forecasts indicated a month ago. The lead time in June was 20 weeks for power MOSFETs and small signal transistors, and 18 weeks for bipolar power devices and rectifiers. Availability remains extremely tight for widely used analog Integrated Circuits (IC) and memory ICs. The supply situation is even more critical for standard logic ICs and power management discretes such as low-voltage MOSFETs and tantalum capacitors, which now are experiencing shortages and effectively are on allocation status.”

   The Semiconductor Industry Association (SIA) reports worldwide sales in May were $24.7 billion, a sequential increase of 4.5 percent from April when sales were $23.6 billion. That's a year-on-year increase of 47.6 percent from May 2009, when orders were exceptionally low because of fear of financial collapse. For the whole year George Scalise predicts 28.4 percent growth. Part of that 28% growth is price increases, so the effective unit growth is significantly lower. A 10-20% underlying growth rate should not be causing shortages this severe so much of this is probably companies protecting themselves by over ordering. That is unproven and is little consolation for companies that can't ship orders because they can't get parts.

*** ASSIA DSL Expresse:The choice of the most demanding carriers in the world http://www.assia-inc.com/ (ad)

New Service: Common Sense Reading for Errors http://bit.ly/a4ky8f
I've decided to experiment offering a "common sense read" of presentations, filings, white papers, briefs or even analyst reports. (One of the most interesting presented a bad calculation to an FCC workshop.) I am neither an engineer nor a lawyer, but in 12 years on this beat I've learned a fair amount. I've earned almost nothing from my policy reporting because I don't go soliciting the lobbyists for ads. That gives me credibility, but until the MacArthur Foundation anoints me I have to earn a living. I give absolutely no guarantee, but I'll probably find some things you prefer to clarify.  Prices are modest and I'm willing to do pro bono. Details http://bit.ly/a4ky8f or just respond to this email (daveb  dslprime.com)


Email
From a distinguished Professor after I wrote about government regulators setting advertising rules for unicorn meat: “You may know that the dragon is the symbolic animal of Wales. This fact inspired a Welsh meat processor to market a product called Dragon sausage. He was told by the UK regulator that this was misleading and unacceptable. When the regulator was asked if he actually thought the consumer might expect to be getting dragon meat, the answer was of course not. His view was that the consumer might expect it to be vegetarian, while the product actually had pork in it.” Unlikely as it may sound, I confirmed the story.
Briefs
Yiannis Argyropoulos of AT&T Wireless is skeptical about femtocells for massive offloading of data traffic.  Two years ago AT&T had decided they needed to massively deploy femtos to reduce traffic on their mobile network, but I haven't heard anything lately and wonder if they've reduced their plans.
Softbank Japan, on the other hand, is now giving a femto and a DSL line for free (really) to their mobile customers. The femtos are not locked to the customer's phone and can service several outside callers. The result: instant reduction of traffic to the cell towers.
Occam's User Group meeting in D.C. had a generous spirit of support and co-operation. It was a pleasure to join.

Press
TiWho? Earl Zmijewski of Renesys finds the ex-Tiscali International division among the top 10 international providers of transit and not that far behind AT&T. “Tinet has been on an absolute tear in 2010 with wide ranging customer wins including India's Bharti (AS 9498), Europe's LambaNet (AS 13237), Venezuela's CANTV (AS 8048), Viet Nam's Saigon Postel Corporation (AS 7602) and many others.” Bought last year by Italy's BS Private Equity for 47M, they are growing in the highly competitive transit market while others, especially Sprint, are retreating.
Roy Rubenstein followed up my article on 100 gig connections by pointing to his well-done technical article on the topic bit.ly/bOS7Y6. It's far beyond what I could have written on optical connections, but I'm studying. I'm working my way through Rajiv Ramaswami Optical Networking, A Practical Perspective to better understand. The book is clear, well-written and a pleasure to read. It's been updated for 2010 so is current. Next optical story I write I'll understand the difference between DPSK and DP-QPSK. Ramaswami has left Cisco for a senior job at Broadcom, yet one more brilliant engineer they have lured.
Wall Street
Mauricio Fernandes of Merrill thinks the pro-competition attitudes of new Mexican regulator Mony de Swaan “result in further noise against AMX and TMX.”  Fernandes goes on to say “We wouldn't expect to see major regulatory changes in the foreseeable future.” Wall Street analysts' job is to find the reality beneath the hype and this is likely a good call.
Craig Moffett, the most visible bear on the bells, wonders if maybe the price has gone down so much they are currently a smart investment. Moffett, like John Hodulik, believes AT&T's wireless cap/price increase will dramatically raise the earnings from wireless data.
People
Mark Rumer, co-founder of Occam Networks, also is a founding member of rock band Anti-M. He plays keyboards on their first album, No Waves in Hell http://www.amazon.com/exec/obidos/ASIN/B000002NPH/antimfeaturinron  in which a surfer discovers a profound lack when he arrives below.
Larry Summers spoke for Obama's wireless expansion and referred to the incumbents as “sclerotic entrenched interests.”

Policy

Can Kang Conquer Capitol Confusion? http://bit.ly/9oNlAt
Saul Hansell's NY Times article was critical in killing a $1B pure giveaway to Verizon that was added to the stimulus by the Senate Finance Committee.  Saul's moved on, and Cecilia Kang at the Washington Post can become a leader in U.S. telecom reporting if she follows up her story on universal service. David Cameron has killed off the despised broadband tax that was really a giveaway to British Telecom, but the FCC is considering a multibillion dollar levy on broadband that will almost all go to the companies in return for very little investment. The big telco lobbyists expect a windfall of at least $1B/year and more likely $3-4B from ICC & USF changes. 
    “Over the past three years, AT&T received $1.3 billion in funds to deploy phone lines to rural areas [and] Verizon got $1.27 billion,” Kang reports. V&T have invested virtually nothing in expanding broadband. They are treating rural customers the way the Romans treated the Sabine women while they try to sell off the lines. Bell broadband coverage in rural areas is among the worst in the developed world. 
        There's plenty of muck to rake if Kang's editors give her enough time to do some investigation. There's the $91K/line giveaway to Sandwich Islands, part-owned by Democratic politicians. There's $B or so in savings possible in the schools and libraries program if they efficiently and publicly sought the lowest prices. There's the “for-profit” subsidiary of NECA, a huge conflict for a non-profit that distributes what's ultimately public money. It's rumored to be making the officers rich, but refuses to give me any information.  An FCC Chief of Staff told me the ICC calculations are obviously illegal because they are not based on the costs of terminating the calls as the law requires.  No one reporter can cover all of this, and I'll happily share files and sources with anyone interested.
      If any of my readers know Marcus Brauchli at the Post, can you forward this to him and suggest it's time to make stars on the technology desk? D.C. needs that reporting. As far as I know, no D.C. publication has looked closely at where the new subsidies will go. The $150B Woman, Carol Mattey, is doing her best to rationalize the system, but the political pressures are brutal. It's literally impossible to satisfy the $20B+ of demands with the $7B available.  Everyone in the system knows these issues. Without press coverage, lobbyists overwhelm the public interest.
     Truth to power, anyone?

Canada: Smart Regulation, Not De-regulation http://bit.ly/a2G7wH
Canada's CRTC http://bit.ly/ap7a3c isn't as dumb as U.S. regulators who are considering ruling that the law doesn't apply where the telcos oppose it. (Title II deregulation) Canada just decided wireless needs to follow the rules. In turn, the CRTC intends to make sure the rules are reasonable.  Rather than saying “never any rules,” they instead try to write sensible ones.
    They've required the carriers to clearly inform customers about throttling. That's under consideration in Britain and the U.S. as well, but probably in a meaningless fashion. The minimum disclosure should inform users about how many hours/month they are affected and how much users are slowed down, but the FCC is allowing telcos to say “sometimes we throttle” without the details needed to determine if it's negligible or abusive.
    Comcast currently throttles far fewer than 1% of users and then only to a speed of 7 megabits  faster than most DSL connections. They rarely do that for more than 15 minutes. If that were clearly disclosed, most of the criticism would disappear. But there is a problem if a carrier slows down, for example, Netflix streaming video. The current “complete information” doesn't let me distinguish. How often are users throttled? How much are they slowed down? Typical disclosure hides what we really need to know. Dale Hatfield, are you listening?
  The minimal Comcast throttling is possible because they have essentially solved the p2p upstream congestion issue. Comcast top technical people have described how they've made inexpensive upgrades to 10-20 megabits from the 2 megabits of older systems. Explaining why they are moving slowly on DOCSIS 3.0 bonding, 4 CTO-level execs told Cable Show audiences they have essentially no upstream congestion these days. The speakers were in charge of several of the largest networks in the world. It's amazing how many in D.C., including just about all the cable lobbyists, don't seem to know the progress they've made.

War Over U.S. Public Safety http://bit.ly/a5sUWh
Takeaway: D.C. insiders tell me Verizon is willing to sabotage the public safety network to prevent D block spectrum being auctioned. Julius fears V+T will become even more dominant so might put Harbinger-like "no V+T" on the D Block. Verizon's opposition is an ugly move, and disastrous for the broadband plan, which hopes competition enabled by more spectrum will bring down prices.
People died on 9/11 because they couldn't get information over the radio and the cops and firemen couldn't talk to each other. It's a scandal the problem still hasn't been solved. Carlos Kirchner and Stagg Newman focused the Broadband Plan on this and the plan proposed a single national network.
     To bring down the cost, public safety would piggyback on one of the LTE networks being built by Verizon or AT&T and use public safety radios that are compatible with the LTE networks. This would reduce costs by 50-80% by most accounts and would be particularly important in the gravest of emergencies.
   The radios could work over the regular LTE networks when needed, and the plan included a provision that public safety could use as much bandwidth as they needed on request. That probably never will be needed. The existing public safety spectrum can carry all the voice likely needed and plenty of video as well.
     The current "public safety" operators, led by the cops in New York, hit the roof. They insisted on maintaining control of the network and spectrum. It felt like a turf war, especially when the "public safety" folks started circulating claims about what their plan could do. Two engineers I respect have looked closely at the claims and don't believe them. One said the throughput claimed would require building twice as many towers as they propose at a correspondingly higher cost.
      Cecilia Kang (WP, http://bit.ly/aMbokw ) broke the story that Verizon, AT&T, and Motorola are funding the campaign to block or delay the network. She noted V&T expect local deals with public safety that will get them the spectrum, often with lucrative turnkey contracts to run the network. She reports Motorola has 80% of the market, often charging $5,000 for each radio. My research suggests the U.S. radios are very similar to the "tetra" radios sold to public safety around the world for $500-800, one of the reasons the costs in the U.S. are so high.
    One constructive step the FCC can and should take to insist the standard for LTE chips include the ability to cover the public safety frequencies. That will add little or nothing to the chip costs, but might not happen otherwise because the "public safety" market is small and often ignored.
     I'm not sure it should be funded by a “broadband tax.” Bringing down the price of broadband is the only proven method to increase adoption. Julius often cites "affordable broadband" as the key goal. Something is profoundly wrong if the net effect of the broadband plan is prices going up.
    It's also a mistake to automatically extend the contract for this or any other any spectrum auctioned beyond ten years. The renewal process is generally a rubber stamp and the law doesn't require it. Adelstein's “Use it or lose it” renewal policy would solve nearly all the broadband coverage problems in a few years without government money, and should have been central to the plan.

Failure of the Broadband Plan?   bit.ly/baU3V9
Craig Moffett sees this as I do: “If LTE networks are going to be usage-capped, then the last pretense that LTE can be positioned as a substitute for terrestrial broadband would seem to be gone.” The heart of the U.S. broadband plan is to release more spectrum. 300 megahertz is  enough for 10-20 networks like Verizon's LTE now building. Julius is praying that will be enough competition in five to seven years to check price increases.
   In Indonesia, India, Pakistan and most of Africa, the scarcity of landlines mean "wireless broadband" will be dominant. "Wireless broadband" - especially LTE - could become a player in developed countries if priced right. Rob Pegoraro (Washington Post) finds that Clearwire WiMAX could be serious competition to broadband sold by incumbent phone and cable companies. He's getting a consistent connection of about 5 meg down, 500K up, on the current, lightly loaded network. He sometimes has to look to see which wireless network his computer is connected to, Clearwire or his 15/5 FiOS. Large uploads are painful because of Clearwire's slow upstream, but 4G will do better in time. However, wireless speeds are likely to fall if many people watched quality video over the net.
    How much wireless could compete with landlines, especially as all cable connections are moving to 50 meg, was a crucial question for the broadband plan. The consensus of several good engineers is that 4G competes fine with DSL iff not many people expect video or other high-bandwidth apps. Wireless certainly can't keep up if many people want to watch their TV over the net, so it's only a partial substitute.
   Making wireless an important substitute for DSL requires raising bandwidth caps from today's typical 5-10 gigabytes to several times as high as LTE makes the cost reasonable. If Verizon follows AT&T with an abusively low cap of 2-5 gigabytes and Sprint etc. don't clobber them, the whole broadband plan falls apart because that's not enough for competition in the future.
   I doubt Julius understands this, because he would be doing everything in his power to avoid low caps. It's just one more strike against “affordable” broadband, like the recent Comcast and Verizon price increases. People need to laugh out loud when Genachowski says “affordable” while tolerating continuous price increases.

Notes from the U.S. Stimulus http://bit.ly/baxii2
Marquette-Adams Telephone Cooperative will get $20 million in grants and loans to run fiber to 4,600 unserved homes in Wisconsin, http://bit.ly/9SUvP4 New fiber will cover 498 route miles, most of which will be labor costs. The fiber will cost $10M at the typical $20K/mile, so the price is not out of line. Jobs and newly served homes are precisely what the stimulus is about, so this is on target. But is it right to spend $4K of public money for each home, or the $10K that RUS rules allow?
A reader of Rick's article pointed out how broadband destroys jobs as well as creates them, as every print reporter or record executive knows. “The $20 million in grants & loans may have a negative affect on the Adult Video Stores. Will we now have to bail out the owners of Adult Stores?”
GTA TeleGuam has begun installing VDSL2 from Occam to speed up their network. When I interviewed Dan Moffat for the story about stimulus funded overbuilds, I urged him to fight back by delivering the best possible network. Glad to see some people on Guam will now have the choice of 50 megabits and higher.
For the record: Vermont Tel received a grant from NTIA. I didn't have anything to do with writing the proposal, but I previously worked on a statewide wireless project for them and they remain friends. I spent a day or so going over what I thought would help with D.C., including following the rules and giving the Feds a proposal that fits their program rather than trying to persuade the Feds to stretch to fit what VTEL thinks is better. In total, my fee was four figures. Years back, I did a little work for VTel as they extended their DSL network to 98.5% of a rural territory. They and others  including British Telecom  proved 98-99% coverage was practical in rural areas. So I've long been skeptical of those who say they can't go beyond 88-93%. That field experience, confirmed by dozens of other U.S. rurals, informs my comments about Windstream and others.  More http://bit.ly/baxii2

*** ASSIA -- competitive advantages:  DSLAM vendor independence, field experience, high-speed automatic line re-profiling, and scalability are unmatched. http://www.assia-inc.com/ (ad)

Held for next issue
Broadband's Great Middle: France, Germany, UK and (yes) U.S. http://bit.ly/aeO8ep
Pakistan Taking Off http://bit.ly/9S4J6h
Free.fr $20 Dedicated Web Host Server http://bit.ly/9atRgk
No Scandal in Windstream RUS Grant After All http://bit.ly/baxii2
U.S. Congress: 4 Meg Sucks http://bit.ly/aHfwld
Goldman Sachs on Wireless Carveouts http://bit.ly/cbiBz7
Killer Apps: Le Monde, Al Jazeera, ?Jerusalem Post TV  http://bit.ly/94RhMf
Julius & Frontier West Virginia: The Very Ordinary Facts http://bit.ly/a2FJB0

popular articles
$13/100 Meg Profitably Sweeping Hong Kong http://bit.ly/9oMIso
Switches Getting Cheaper; "Don't Bet Against Moore's Law" http://bit.ly/b8464u
John Stankey Is Not a Big Fat Liar  http://bit.ly/blEGpP
Cisco VNI: Internet Traffic Growth Rate Falling by Half in U.S. http://bit.ly/cqZkaX

Reply "subscribe" to be added, "un" to be dropped
Volume 11, #10 July 10 2010
June 24
Q1 2010 DSL: China Remarkable, India, Russia, Philippines Surging http://bit.ly/8YivBc
$13/100 Meg Profitably Sweeping Hong Kong  http://bit.ly/9oMIso
Alcatel, Infinera 100 Gig Closer Than You Think http://bit.ly/c5o6Tj
Cisco ASR9000 Winning Raves, Sales http://bit.ly/c5o6Tj
"The War is Over" 90% of Australia Getting Fiber Home After $10B Telstra Deal http://bit.ly/cwyVU8
For the Record: Bandwidth Costs http://bit.ly/dchYmP
Briefs: Vint Cerf on growth slowing, cheap high speed broadband is largely TV release 2.0. rather than innovation, component shortages, Bill Lehr, Steve Bauer, Stephen Fry calls iPhone 4 “an object of rare beauty that leapfrogs the competition,”Infineon selling wireless, Jia Lynn Yang, California's Utilities Commission, Kevin Fitchard at Connected Planet/Telephony, Susan Crawford, Pete Chow
Reply "subscribe" to be added, "un" to be dropped

“Cease and desist from calling unicorn 'the other white meat,' our client's trademark.” Lawyers for the National Pork Board, who didn't check the April 1 date on the Thinkgeek story. Unicorn meat? 

Carlos Slim of Telmex was in New York proposing to make broadband a universal service. “High penetration, good quality, high speed, low cost.” Sounds great, but someone should have asked him why Telmex and American Movil prices are so high they are holding things back.

     In 2008, I wrote Carlos Slim's Patriotic Choice http://bit.ly/bWs3DR urging him to drop prices in Mexico from then current $25-60 to $10-35. Worldwide experience has proven two profitable models: high prices, low take rate or low prices, high take rate. Both can work for the business, but the low price/high take rate is better for the country. Some DSL Prime readers I believe are close to Slim. I urge them to convince him that lower broadband prices are good for the company, long term. France and now Hong Kong are proving low priced carriers can make money. Marketing costs plummet, voice business improves, and competitors are not able to win customers with lower prices.

--------------

Fiona Vanier calculates broadband will reach 500M homes sometime between now and August.  Our lives are different. We're always connected with what we care about, many of us so closely we live with constant messages and interruptions. Moore's Law has produced 10G connections to the home and 100G connections everywhere else. There's plenty of capacity for the next 500M connections as well.

     The Point-Topic figures Fiona compiles show 300M+ DSL lines. Chronicling that rise from almost nothing in 1999 has been an honor. 

-----------

Say hello to the round fellow with a beard in D.C. Wednesday or Thursday at Occam Networks' User Group meeting. Lots of practical information to be shared.

*** ASSIA -- the world leader in dynamic spectrum management, is now supporting over 35 million lines of DSL.  Our performance and reliability are unmatched. http://www.assia-inc.com/ (ad)

Q1 2010 DSL: China Remarkable, India, Russia, Philippines Surging http://bit.ly/8YivBc
Q1 results from the ever invaluable Point-Topic show that countries 60% or more served with broadband are growing slowly. Several with low penetration are catching up. China continues to dominate growth, likely to continue as the cablecos jump into the market with DOCSIS and EPON. Fiona Vanier sums up "China contributed 40% of the total broadband lines added in the first 3 months of 2010.... Relatively poor performance in the American markets and flat numbers in Europe."

    Some observations:

    * Most of the growth is from less wealthy companies, with China, India, Russia and Mexico among the leaders. Turkey, Indonesia, Vietnam, and Egypt also grew.
    * In IPTV, France currently leads as triple play over DSL is the primary product. The U.S. is catching up, led by AT&T U-Verse. China is growing rapidly and with government support for "convergence" is likely to quickly pull ahead.
    * China has 25M more broadband lines than the U.S. and the gap is widening at about 10 million lines per year.
    * Where fiber is available, it draws subscribers from DSL. Korea and especially Japan are seeing absolute drops in DSL with Japan losing 1/7th of the DSL lines in the last year.

From Point-Topic, The top 7 countries in DSL and those with the best DSL growth in Q1 DSL.
 
Country     2010 Q1 Total DSL  Growth From Q1 2009
   
China     89312920     17.96%
USA              32570284     1.89%
Germ     23272600     6.02%
France     17964500     5.90%
UK         14767100     5.87%
Italy         13119000     6.90%
Japan     9861000     -14.30%
Brazil     7288400     6.12%

But much of the growth is elsewhere.
Country     DSL Total     Q1 DSL Net Additions

China     89312920     5416920
Germ     23272600     635500
India     6875318     405328
Russia     7997449     299449
UK         14767100     281300
Mexico     6897087     261622
USA         32570284     245307

More, including Broadband Forum tables of overall leaders and IPTV http://bit.ly/8YivBc

*** Kevin Werbach's Supernova Forum 2010: Perestroika is a superior gathering of articulate Internet folk. July 29-30, 2010 at the Wharton School in Philadelphia, PA. An ideal chance to share ideas with some of the best, including key startup entrepreneur John Borthwick (Twitter, among others), Jeff Jarvis (most interesting thinker in new journalism), danah boyd (Microsoft), Beth Noveck (White House), and several dozen others. It's priced like an academic event so it's an incredible bargain. Early registration is $100 at bit.ly/aroZOa (psa)

$13/100 Meg Profitably Sweeping Hong Kong  http://bit.ly/9oMIso   
City Telecom/HKBN doubled their dividend as profits rose 54%. They added 73K broadband subscribers over six months while everyone else was flat to down. Chairman Wong Wai Kay has set a "10-year Big, Hairy, Audacious Goal (BHAG) of becoming the largest IP provider in Hong Kong by 2016. We are not only aiming at providing better service than the incumbent, but also services that they cannot match with the legacy network." They've now doubled up, offering a gigabit (really) for $27 (U.S.), an incredible story.  Xavier Niel in France proved a low price strategy (30 euro triple play) can make money and HKBN is actually profitable enough to pay corporate income taxes.

    ARPU for broadband fell from $23 to $17 (U.S.) but that was more than compensated by customer growth and sale of phone and other services. City is doing well enough to apply for an over the air free TV license and continue expanding the network. 100 meg is available to1.68M apartments and 1,300 corporate buildings, with 1 gig only in those with fiber all the way to the apartment. By the end of 2011, they expect to reach 2.1M homes, about 90% of Hong Kong. Their cost per apartment passed is about $100 (U.S), crucial to the business model.
The prices for voice and video are competitive, but there are a few limits to the service. In particular, the speed out of country is capped at "only" 20 megabits. That's less of an issue than it would seem, because most services are available from in-country Google, Akamai, and other servers. Most p2p is fulfilled locally.

   CFO NiQ Lai explains the low-cost model only works after the network is in place. "HKBN started the investment ten years ago, and we did not get a glorious result until two years ago. Most CEOs or management team in other companies would probably have been sacked in the middle of such a project before reaching the harvest stage. If you had HK$3 billion, instead of building our Fibre Network, you could have turned around a number of mid-sized profitable real estate projects in the past ten years. Why invested in the broadband business when there were so many other attractive businesses to make money in Hong Kong? In our case, our decade of focus and persistence is only just being rewarded today. However, now that we have an established “edge”, we believe will be here for the next generation,”
 
  There's no reason that another carrier in a dense urban area (Chicago, London) couldn't do similar. (Special thanks to Glen Campbell of Merrill Lynch who first introduced me to the company.)

*** ASSIA -- competitive advantages:  DSLAM vendor independence, field experience, high-speed automatic line re-profiling, and unmatched scalability http://www.assia-inc.com/ (ad)

Cisco ASR9000 Winning Raves, Sales http://bit.ly/c5o6Tj   
Mike Wheeler of NTT America is delighted with his new Cisco ASR9000s. Within 12 months, he expects to upgrade to 100 gigabits connections. Brad Hokamp of Telx has just announced a nationwide "Ethernet Exchange" along with Neutral Tandem, all running the Cisco ASR9000. Brad echoed the praise. The "9K" was working well.
  Wheeler expects the 9K to enable his large customers to jump from 10GigE to 100GigE within twelve months, a market he expects will boom. Per bit, 100GigE ports will be much cheaper than 10 or 40GigE. Very large customers are paying as little as $1.5/megabit for transit, according to actual bids. Smaller customers, anywhere near a major city, can usually find prices well under $10/megabit.  Cisco's "edge router" market share is 5% higher this quarter, at 47% according to Dell'Oro.

   Wireless backhaul has pushed Tellabs up to #4. Tellabs sunk today as Barclay/Lehman analyst Jeff Kvaal drop his price target because he thinks Cisco and Alcatel will take all the business in AT&T's LTE generation. (Barron's)

    Virtually every carrier worldwide is upgrading from T-1's for mobile voice (a few megabits) to either fiber or microwave, usually at 40-125 megabits. AT&T alone is upgrading thousands of sites. China has already done cell site backhaul upgrades last year when building 3G networks, so Huawei has lost sales and market position.

Alcatel, Infinera 100 Gig Closer Than You Think http://bit.ly/c5o6Tj
Basil Alwan of Alcatel "will be production shipping 100G in the next few months," he emails. They've delivered 100G that Softbank is demonstrating in Japan and Verizon in the U.S.  Alwan claims "full service routing including all edge services (VPNs, QoS etc)."and proving that "already in several customer labs. ...  We build our own silicon and have been shipping 100G packet processing / TM chips for awhile.  In fact we first ramped this silicon around two years ago (introduced as FP2) and we used it for our 50G full duplex cards.  Now that 100G optics are becoming available we simply re-configured the same deep touch chipset in simplex mode."
 
   Rick Dodd of Infinera explained to me why 40 gig was late and disappointing while 100 gig is drawing enthusiasm. "The step from 10 gig to 40 gig was particularly tough because the modulation format changed to phase shift keying and that wasn't easy. 100 gig is also PSK so there are fewer obstacles. We expect 100G will be very big in 2012."

    10G required only three optical components: a laser, a modulator, and a photodetector. 40G's change to PSK and coherent detection literally quadrupled the number of components, from 3 to 12. The same components are used for 100G, running faster as Moore's Law improves performance. Most components, such as line system amplifiers, are little changed from 40G to 100G.

    Alwan is proud that Alcatel is ready to deliver both the 100G router and 100G long-haul optics.” Both sides of the fiber need to support 100G for it to be useful  the device that is the on-ramp to 100G (the IP/MPLS router) and the long haul optical 100G system.  ALU has both. A little more at http://bit.ly/c5o6Tj

*** ASSIA Talk to the DSL experts.  Critical enabling technology and DSL network best practices http://www.assia-inc.com/ (ad)

Other recent stories:
Moffett: Caps "very likely spell the end of mobile video" http://bit.ly/cLHNqP
$100 for GPON or EPON http://bit.ly/d4RF7S
Three month archive http://bit.ly/as2gre

Corrections: In the initial web posting of the Alcatel story above, I confused Alcatel's 100G long range optical unit with their 100G edge router. Alwan pointed out to me this may be the first time one company had the first units in both categories for a new generation. The biggest favor you can do any reporter is point out an error.

Email
Vint Cerf made a comment about "growth slowing. Not as dramatic as a few years ago, but still growing substantially and will continue to do so." I asked him for a comment about the Cisco projection of the U.S. dropping to an 18% growth rate in 2013 (which I think is a little low.) Vint replied "The growth rate appears to be about 40% over the last 15 years - with the addition of smart grid and a lot of video, I am not sure the 18% figure is right. CISCO may be right about US traffic growth since we are now about 75% penetrated. Worldwide, the number of users continues to grow as we are only about 25% penetrated, world wide, as of now. Mobile phones are providing access to hundreds of millions who did not have access before and there is continuing increase in the speeds for these devices as 3G and 4G propagate."
A wall street guy asks "Can you forward to Julius at the FCC my notion that cheap high speed broadband is largely TV release 2.0 and that it will further comatose society. Very very few people are using it to help the 'economic innovation'" I'm not as sure about that, but he's certainly right that video is the primary use of high speeds at home. Not very exciting if you don't watch much TV. (I don't know if Julius reads DSL Prime these days. I send it to most officials.)
A senior exec writes "I read the note in your latest DSL Prime about component shortages.  It is a real issue in telecom and IT - every company I speak to is seeing shortages of components.  Many times these are the popcorn parts like resistors and caps.  One of Broadcom fabs is being pressed so hard by Apple for components that they are slipping BCOM orders. In late 2008 and early 2009 many factories shut down expecting a long deep recession.  Now they are not totally convinced this economic period is not a double dip, so are slow to invest in restarting facilities." It's really hard to tell whether this is temporary over-ordering because of the fear of a shortage or a problem that will take longer to resolve. DSL chips themselves seem to be in good supply. 
Bill Lehr writes about a paper he wrote with Steve Bauer. They looked at some of the popular speedtests and discovered many sources of error. "Significant bottlenecks arise in home networks, end users' computers, and server side systems and networks. ... Many testing methodologies are inappropriate for the purposes of assessing the quality of a broadband network." He's more respectful of the hardware testing systems used by SamKnows for OFCOM in the UK and now coming to the U.S. In particular, delays typically assumed to be the carrier network often are the PC or server, leading to an overestimate of "congestion." http://bit.ly/d1yVcU
Briefs
Stephen Fry (yes, the actor) calls iPhone 4 “an object of rare beauty that leapfrogs the competition.” It really is that good, unless you want to use it as a phone. Hint: don't wrap your hand around the phone but hold it gently by the edges. Leapfrogs in software and some features, including the gyro, but it has the same processor as the soon to release Samsung Galaxy, an ARM Cortex 8 running at 1 gigahertz. That doesn't mean the Galaxy, an Android phone with a wonderful AMOLED screen, will be as good as the iPhone, but when T-Mobile releases it soon it will surprise people.  http://bit.ly/ceJxeN TechInsights has a full teardown of the iPhone4 worth a look at http://bit.ly/cVldKF In particular, click on the circuit boards to the right of the screen to see how much they've included.
Chips
Infineon hired JP Morgan to sell off their wireless chip division, where the $billion or so of sales included the iPhone. (FT) They apparently don't have the volume to compete in this space, with massive R & D required. It's painful to see what's happened to the former Siemens chip division. They split off their huge memory chip division as Qimoda and it soon went bust. Fortunately, the DSL and related chip division spun off as Lantiq is profitable.
Press
 Jia Lynn Yang filled in for Cecilia Kang at the Washington Post briefly and showed an open spirit more typical of the web than print. She picked up with credit Ed Wyatt's NY Times story that some local public safety guys want control of the safety network rather than a national network and Joe Flint's LA Times http://bit.ly/9cyDzl piece about how the FCC is allowing Comcast and Hulu to blackout the news. She included Niraj Sheth's quote from the usually invisible Randall Stephenson that he might cut U-Verse even more if he doesn't get his way at the FCC, as well as Matt Lazar's Ars Technica item making clear Randall is deliberately misleading. The web has a spirit like that.  I'd rather point you to any of those writers rather than retell the story unless I have important additional information. (That's what this press section is for.) Most telecom reporters are friends and freely share information, and I freely share files and sources with other reporters. Print is often different. The Wall Street Journal and Telephony both told me the editors decided they “couldn't quote me because I'm competition.” Both Times and the Chicago Tribune don't care and have been generous with credit.    
A valuable summary of the proceeding of California's Utilities Commission is offered as a free newsletter by the law firm of Cooper, White and Cooper http://bit.ly/8Xj3YM
Kevin Fitchard at Connected Planet/Telephony has a thoughtful article “T-Mobile expanding '4G' to cover 75M pops” http://bit.ly/8Xxkm4 in which he notes that T-Mobile's HSPA+ “3G” is often getting speeds of 5-8 megabits/second downloading. That's more than the 3-6 megabits of Clearwire's WiMAX, called a “4G” technology. That's not so far behind the 5-12 megabits Verizon's Dick Lynch expects from Verizon's LTE network in a few months. Kevin Tofel at GigaOm reports he got speeds of 4 megabits. T-Mobile's network is very lightly loaded and there are other ways the comparison is not exact, but the HSPA+ “21 megabits shared” is doing remarkably well.  I would guess that LTE achieves about 40-50% more total throughput in equivalent spectrum, but I haven't seen any careful testing to compare. LTE is not magic or even breakthrough technology. The “E” in LTE stands for “Evolution”; it's essentially a combination of the most advanced techniques wireless engineers expected to soon develop about 5 years ago. Many of those same techniques can be adopted for HSPA and dramatically increase speeds. AT&T has upgraded the “3G” to 7.2 meg down and will soon go to 14 and 21 as well. That's part of Stankey's plan to solve the congestion problem without a ridiculously low cap, now irrelevant.  
People
Susan Crawford, one of the very best, is moving back to New York from the University of Michigan and a year in the White House. She'll be teaching at Cardozo Law School. She's working on a book about the Comcast-NBCU merger, but it will be hard for her to publish it in time to kill the merger. She's also taking a role with Ed Felton at Princeton. Susan has only good things to say about Michigan but wants to be on the East Coast. Felten is also extraordinary. He was the one who taught me that video over the Internet doesn't need quality of service unless the network is marginal. 
Pete Chow, one of the original Amati engineers and an industry leader at TI, has rejoined John Cioffi at ASSIA. Good guy.


policy

"The War is Over" 90% of Australia Getting Fiber Home After $10B Telstra Deal http://bit.ly/cwyVU8
The first stage of Australia's NBN is about to go live in Tasmania. A price war has brought the price of 25 megabits (low cap) down to $26 (U.S.). Meanwhile, Telstra and Mike Quigley's NBN have cut a deal. For about $10B, NBN gets ducts, other facilities, and an agreement Telstra will move customers to NBN and ultimately shut down the copper network instead of competing. The price is a little high, but the government is paying to silence opposition.

     NBN started because Telstra wanted so much government money for DSL/FTTN the government asked "Why not run fiber all the way home?" The logical next question was "If public money is paying for the network, why should we give it away to a private company?"

   The NBN is a campaign issue across the front pages of the newspapers. The politicians will make some more noise, but without Telstra pulling the strings they are unlikely to stop the build. The last 10% will probably be wireless, a sensible compromise. Grahame Lynch at Commsday asked for some comments and I wrote back about some of the things we had learned from large projects like this elsewhere.

NBN deal a good thing but don’t overstate the benefits
Some perspective from the other side of the world in New York City. Ultimately, having a great Internet is a good thing for any country. The cost to Australia of this deal is high enough to create opposition today, but looking back after a decade I'm sure almost everyone will think it was the right move.

    Experience from other countries points to issues ahead. Britain’s separation has worked marvelously on the retail level and brought British prices down 20-30% from where they likely would have been. There are some great deals. If you’re already a SKY customer for TV, they will give you a low end DSL line for free and a pretty good connection for $10-20 - plus the eleven pound line charge to BT.

    BT has a monopoly in wholesale across half the country and a weak cableco in the other half, giving them market power they take advantage of. Implicit in the NBN proposal is that it will have strong, perhaps monopoly control over the wholesale part of the network. NBN costs will have to be watched by an outside body very closely. NBN will be a quasi-governmental monopoly likely to become wasteful or bloated unless done right. Government bodies can be run well, but far too many aren't.

   Both the economic and social benefits of broadband are wildly overstated almost everywhere. There’s a social return to better broadband, but it’s far, far lower then the hype suggests. Most of the numbers thrown around are from shills and zealots. Honest academics looking for the effects find only modest ones.

   Tasmania getting faster net connections will not transform the local economy, although it’s likely to help a little. Kids will not do markedly better in school because their home Internet is ten times faster. Rather few net jobs are created by broadband. Some sectors do better, but othersnewspapers, bookstores, local TV programmingdo worse. If efficient distance learning takes off, it will displace some local teachers. Again, a mix of true believers and shills for the companies wanting government support has wildly exaggerated the likely benefits.
 
    It is absolutely essential not to try to do too much, too fast. Operating procedures and billing turn out to be much harder to get right than anyone expects and take time. Literally thousands of crews will need to be trained, which takes 6-18 months. Untrained crews take twice as long to do an install (literally) and drive up costs. We learned this during the DSL Hell years at the turn of the century. Five years later, AT&T re-learned that lesson. They pushed too hard and U-Verse came in two years late and billions over budget. Verizon FiOS on the other hand went very slowly by plan the first three years. It consistently met the schedule and budget.

    Ultimately, a better Internet is of important value, but there are many lessons learned the heard way of what can go wrong.



For the Record: Bandwidth Costs http://bit.ly/dchYmP
In the spirit of disclosure I like to report important comments I've made to FCC and other officials. There's no news here, but it was important to pass on some facts because a senior lobbyist was presenting what I believe is bad data.

    He filed a claim of "annual traffic growth rates of between 100% and 114% through 2014." In addition, the projected the cost of handling that bandwidth "decreases at 10% per annum over the 2009 to 2014." Cisco and nearly everyone I know predict growth rates of 20-40%, with Cisco's respected VNI predicting U.S. growth will fall to less than 20% in 2014. (Hard to tell.) Senior officials of both Cisco and Alcatel tell me they expect declines in the cost of equipment needed for bandwidth (switches, routers, DWDM, etc.) will "continue on the historical trend or possibly decline slightly faster." (25-45%.) Combine the two errors and the projection is off by a factor of about 10 and distorting any discussion of "managed services."

   I sent a note pointing out that "Bandwidth is not free but it's cheap enough to almost never be an issue at large DSL, Cable or fiber carriers. I think that's mostly true on wireless with careful planning, but I'm not a wireless expert.

Quick numbers:
Current provision per customer, no congestion noticed: 100K (per Alcatel and about on target.)
Current cost per megabit, large carrier, all in. ~ $10.
Net cost of bandwidth per customer, today: $1/per month.
Equipment costs are going down at 25-45% per year, bandwidth demand probably going up more slowly.
Result: Cost per month per customer should be flat to down as far out as we can look."

More details http://bit.ly/dchYmP Because this is going to policy people, I'd appreciate a note from anyone with contrary data.

Related:
Cisco VNI: Internet Traffic Growth Rate Falling by Half in U.S. http://bit.ly/cqZkaX
Switches Getting Cheaper; "Don't Bet Against Moore's Law"  http://bit.ly/b8464u

New policy articles:
War Over U.S. Public Safety http://bit.ly/a5sUWh (I pointed to Verizon's desire to prevent auctioning the "D" bloc as a major issue and I know was passed around D.C.)
U.S. Congress: 4 Meg Sucks http://bit.ly/aHfwld
Goldman Sachs on Wireless Carveouts http://bit.ly/cbiBz7
Killer Apps: Le Monde, Al Jazeera, ?Jerusalem Post TV  http://bit.ly/94RhMf

Reply "subscribe" to be added, "un" to be dropped
Volume 11, #9 June 23 2010

DSL Prime June 14

• AT&T's John Stankey is Not a Big Fat Liar bit.ly/blEGpP
   
• Cisco VNI: Internet Traffic Growth Rate Falling by Half in U.S. http://bit.ly/cqZkaX
   
• Moffett: Caps "very likely spell the end of mobile video" http://bit.ly/cLHNqP
   
• $100 for GPON or EPON http://bit.ly/d4RF7S
   
• Switches Getting Cheaper; "Don't Bet Against Moore's Law" http://bit.ly/b8464u
   
• G.hn Do-Everything Home Networking Approved. Working? Ubiquitous? Cheap? To be Followed  bit.ly/cE8sXd
   
• Briefs:  China Telecom GPON/FTTH figures, Jim Baller, Zhone, Adtran, Graeme Lynch, John Tanner on TD-LTE doing well, Dana Mattioli

Reply "subscribe" to be added, "un" to be dropped

"I was wrong." Barack Obama, a most unusual politician 
Breaking news: FCC Chair Genachowski is likely to take the teeth out of Net Neutrality at  Thursday's meeting. I'm 90% sure the rules have loopholes wide enough to drive through 500 TV channels. The reality will likely be buried in plausible spin about the actual result of definitions of "reasonable," "managed services" and other details. I hope I'm wrong. Ultimately, NN is important but the short run effect is likely minor because the possible political backlash will prevent the carrier's doing anything abusive. Meanwhile, watch closely the details of broadband "USF" and lifeline. The Bells have enormous interest in seeking “limited law change” that allows them about $20B in subsidies, which might have been held back by the Title II court fight.

Bigger news about to break: China is about to allow cable companies to offer broadband to 90M customers, breaking the China Telecom/China Netcom monopoly and driving down prices. “Triple Play” “Convergence” will also provide hundreds of channels across the country over the next few years. It's quite possible China will require more set top boxes than the entire rest of the world and come to dominate the set top industry.

 Arielle Sumits of Cisco predicts Internet traffic growth will drop from today's 42%/year to 18%/year U. S. and 31%/year worldwide in 2014. P2p growth is dropping faster than video is growing, probably because regular downloaders already have more music than they can listen to. With data from 50% of the routers in the world and 20 large carriers directly, her conclusions are definitive for today's growth and very interesting for 2014. below, a story not to skip

AT&T President John Stankey is a big fat liar, nearly all the reporters and talking heads seem to believe. “Everyone” is saying AT&T needed the 2 gig cap because of congestion  except AT&T.Stankey personally promised thousands of investors on the quarterly call AT&T's congestion problems will be solved within months. He's restored the capex cuts and they are upgrading their network. Two of the best engineers I know expect him to succeed.

     I believe Stankey is right and there's absolutely no operational reason for a cap at a low 2 gigabytes  about 90 minutes of quality TV each week. 2 gigabytes would be even more ridiculous for Verizon's soon-to-launch LTE network running at 5-12 megabits. At 5 megabits, you'd run past 2 gigabytes in less than an hour a month. LTE is 2-4 times as efficient as 3G, which is enormously profitable with an implicit 5 gigabyte cap, so the natural cap in the LTE generation is 10-25 gigabytes for basic service.

    Wall Street thinks the low cap is about higher prices and price discrimination. “These new wireless bundles from ATT seem demonically brilliant! See the cap rates? ATT is going to be flowing in extra money and quickly,” one of the best wrote. People inevitably will use more mobile bandwidth and consumers will ultimately pay as much as 50% more because of the change. The low cap also protects the telcos' lousy and over-priced video packages.

     Stankey's cap, if successful, is a dagger in the chest of the National Broadband Plan, especially affordability. The heart of the plan is releasing more spectrum in the hope that will yield increased competition, but a 2 gig capped wireless offering will have minimal effect on DSL and cable prices. (policy, at end or http://bit.ly/blEGpP )

*** ASSIA -- the world leader in dynamic spectrum management, is now supporting over 35 million lines of DSL.  Our performance and reliability are unmatched. http://www.assia-inc.com/ (ad)

Cisco VNI: Internet Traffic Growth Rate Falling by Half in U.S.
In 2014, Cisco estimates Internet traffic growth in the U.S. will be less than 18%, far less than most previous estimates. Worldwide, they measure the current rate at 42% and expect that to fall to 30% in four years. Actual numbers at http://bit.ly/cqZkaX

VNI is the definitive source on Internet traffic today because they have direct relationships with carriers from China Telecom to AT&T. Their future estimates are the most carefully done publicly available.

      Arielle Sumits of Cisco warns to be cautious with future predictions because "we try to be conservative in our estimates." One reason to consider is that China Telecom's estimates for growth the next few years are higher than Cisco's. NTT America's Mike Wheeler is seeing 70%+ growth rates currently. He's seeing exceptional demand from Latin America, where broadband is booming.

      Sumits expects p2p growth to fall to 16% in some markets, corresponding to many reports over the last three years. AT&T actually had a quarter where p2p traffic fell in absolute terms, although that was an exception. People who download typically already have more than they can listen to for the remainder of their life. In the U.S. and Britain, most of the current popular shows are available to watch for free at Hulu and similar. It's easier to listen to a few commercials than go through the hassle of downloading.

     Ironically, this drop in p2p growth is occurring when the actual speed of downloads has been going up dramatically. I did some tests on Bittorrent and was typically able to get 3 or 4 times the speed two years ago. On a ten meg Time Warner Cable connection, popular downloads came in at 2 meg to 8 meg most of the time. The typical hour long TV show, encoded at 700K (not great), took 10-20 minutes. Set it running overnight and 22 episodes of Lost are there in the morning.    

   Even 10-25% growth in bandwidth per user is still a great deal in absolute terms and will require thoughtful network planning. 

More, including a chart with crucial data and highlights from Cisco's report http://bit.ly/cqZkaX

 

*** Kevin Werbach's Supernova Forum 2010: Perestroika is a superior gathering of articulate Internet folk. July 29-30, 2010 at the Wharton School in Philadelphia, PA. An ideal chance to share ideas with some of the best, including key startup entrepreneur John Borthwick (Twitter, among others), Jeff Jarvis (most interesting thinker in new journalism), danah boyd (Microsoft), Beth Noveck (White House), and several dozen others. It's priced like an academic event so it's an incredible bargain. Early registration is $100 at bit.ly/aroZOa

 

(psa)

Switches Getting Cheaper; "Don't Bet Against Moore's Law"
Equipment makers including Alcatel tell me routers and switches are getting 25-40% cheaper per bit for the next five years, almost certainly comparable or higher than the increase in traffic per user. That means that the total cost to deliver bandwidth to the user - other than building fiber in the last mile - will go down at most carriers if Cisco‘s traffic estimates are correct.

    The industry standard figure for bandwidth costs is $1/month per user at a large, efficient carrier today. It will almost certainly be similar in 2015. Great news for consumers anywhere competition is strong.  "I don't see an end to the historical trend of dropping per bit prices. I wouldn't bet against Moore's Law. In fact, there's potential we will do even better," a Cisco Vice-President confirms.

     Currently, very large users are paying $1.50-2/megabit for transit, based on actual bids to one I know. Cogent advertises $4/megabit for slightly smaller users, towards the low end of a market for a Gig-E at $3,000-10,000. 100 GigE prices will probably be high until competition develops, but will drop under $1/megabit reasonably soon.

     50 years later, that little graph by Gordon Moore never seems to stop.

$100 for GPON or EPON
Leiping of China Telecom holds competitive tenders for 10M lines in order to bring down prices. In March, the bids for both GPON and EPON came in at around $100, end to end including a basic network ONU and home ONT. Korea reports similar prices. DSL, cable modems and simple Ethernet (100K or gigabit) remain cheaper, $20 to $50 in million quantities, but the difference over a ten year or even three year service life is small.

     A world-class technologist explained to me last September what it takes for GPON to match EPON in price. "The major component difference is that standard GPON is designed for a 20 kilometer reach and requires a more powerful laser. If the customer doesn't need that reach, the manufacturing cost is similar." Most of China is so densely populated the range isn't necessary and Wei is going for the savings.

    Verizon is pressing to actually extend GPON range to 100 kilometers so it can eliminate many central offices for huge opex savings. Like KPN in Holland, they expect windfalls from real estate sales. Around 2002, Verizon developed a strategy of closing most exchanges. Paul Lacouture told me that one large soft-switch can serve what are currently 10 typical exchanges. A single outside terminal can serve into the thousands of users and is all you need in many locations. Bill Smith, rebuilding BellSouth after Katrina, considered replacing many of the destroyed exchanges with terminals.

   Few other carriers can expect a price as low as China Telecom. Most order in much smaller quantities, are dependent on a single supplier, and require expensive add-ons such as third wavelengths for video and substantial batteries for backup.
 
   China Mobile's FTTH build is forcing China Telecom to go faster. In year 2010, China Telecom will have 0.5-1 million GPON subscribers. http://bit.ly/d4RF7S

 

*** Ikanos, Infineon, Trendchip, Calix, Zhone, Ericsson, Alcatel. It's time for your ad to go here. DSL Prime's 10,000 readers include a remarkable percentage of the technibal leaders who buy your gear and the people who influence them.

G.hn Do-Everything Home Networking Approved. Working? Ubiquitous? Cheap? To be Followed
G.hn chips will carry data at great speed around the home over powerlines, coax, or phone lines. All previous chips will become obsolete and G.hn will take over the world. Ten companies will make the chip and quickly drive down the price of the chip. Over powerline or coax, speeds will be dozens of megabits a second or maybe hundreds, enough for three HD channels.

    Or so the chipmakers claim, several of whom have no doubt they will achieve all these goals. Three have promised to ship samples before the end of the year. Without working chips and carriers who have tested them, I have no way to verify any of the claims. When anyone has test results, let me know how they are doing. 
 
    Friday the ITU approved the standard after secret discussions.  http://bit.ly/cE8sXd

 

*** ASSIA -- competitive advantages:  DSLAM vendor independence, field experience, high-speed automatic line re-profiling, and scalability are unmatched. http://www.assia-inc.com/

 

(ad)

Correction:
China Telecom's Wei Leiping apparently said they will have between 500,000 and 1M GPON lines this year. I mistakenly reported that as the total of all fiber lines, based on a report in the Chinese press. They have already installed millions of lines of GEPON, a competing technology.         

Stories in the works - Info welcome

• Cable upstream bonding working but not deploying
• 100 gigabits from Alcatel, Cisco, and Infinera
• U.S. Public Safety Network tied up by Verizon maneuvering and turf battles
• $13/100 megabits making money in Hong Kong
• The $250M fund Larry Strickling established
• Goldman Sachs on why incumbents should be limited in spectrum auctions
• Broadcom's delayed DSL chips (I need confirmation from anyone with information.)
• Massive Comcast influence-buying
• Will Sandwich Islands' political clout force USF giveaway?

Email

• Jim Baller, a favorite D.C. lawyer, writes "I thought you might find interesting the complaint that the City-Parish of Lafayette, Louisiana, filed yesterday at the FCC against the National Cable Television Cooperative and the cable operators represented on NCTC’s Board of Directors." Cox Cable apparently blocked the city fiber system from joining the

NCTC programming co-op to weaken the competition. Commercial and municipal small carriers outside the co-ops are generally screwed, a problem the FCC never fixes.
Briefs

• Zhone shipped 65,000 “ports” of fiber in Q1, much of it to the Emirates. A U.S. company, their latest customer win are du in Dubai and Bosnia Telecom.
• Adtran won the DSLAM contract for hundreds of thousands of lines Frontier is upgrading after taking over from Verizon. That interrupts an extraordinary string of wins by Calix in the mid-sized U.S. telcos.

Press
Grahame Lynch of COMMSDAY isn't happy with government  national or international  taking control of ICT and wrote a strong article at bit.ly/ag3hOo

 

John Tanner of Telecom Asia has an important article about TD-LTE expanding out of China, including a prediction that TD-LTE will win 40% of the market. It has the crucial advantage of working well in 2 gigahertz spectrum already controlled by many operators.  http://bit.ly/d7pew4

 

Dana Mattioli's WSJ article 'Telecom Supply Strained" may become a self-fulfilling prophecy. There certainly is a minor shortage, with new foundry orders typically delayed 30 days. With shortages possible, sensible buyers build inventories, exacerbating the problem. The reach of a WSJ article presumably caused companies around the world to re-double efforts to protect themselves. Amitabh Passi of UBS is looking closely, trying to determine if the delays are an induced shortage (which passes quickly) or overwhelming true demand and tells me its almost impossible to be sure. 

policy
John Stankey is Not a Big Fat Liar
AT&T President John Stankey is a big fat liar, nearly all the reporters and talking heads seem to believe. “Everyone” is saying AT&T needed the 2 gig cap because of congestion  except AT&T.Stankey personally promised thousands of investors on the quarterly call AT&T's congestion problems will be solved within months. He's restored the capex cuts and they are upgrading their network. Two of the best engineers I know expect him to succeed.

      I believe Stankey is right and there's absolutely no operational reason for a cap at a low 2 gigabytes  about 90 minutes of quality TV each week. 2 gigabytes would be even more ridiculous for Verizon's soon-to-launch LTE network running at 5-12 megabits. Even at the low end 5 megabits, you'd run past 2 gigabytes in less than an hour a month. LTE is 2-4 times as efficient as 3G, which is enormously profitable with an implicit 5 gigabyte cap, so the natural cap in the LTE generation is 10-25 gigabytes for basic service.

     Wall Street thinks the low cap is about higher prices and price discrimination. “These new wireless bundles from ATT seem demonically brilliant! See the cap rates? ATT is going to be flowing in extra money and quickly,” one of the best wrote.

    People inevitably will use more mobile bandwidth and consumers will ultimately pay as much as 50% more because of the change. The low cap also protects the telcos' lousy and over-priced video packages.

     Stankey's cap, if successful, is a dagger in the chest of the National Broadband Plan, especially affordability. The heart of the plan is releasing more spectrum in the hope that will yield increased competition. That was a long shot, the planners knew, but possible. With this low cap almost everyone will need to maintain their landline broadband. There's nothing wrong with a cap at a reasonable level at least as high as today's 5-10 gigabytes. A 2 gigabyte cap would be destroyed by the market if Verizon + AT&T weren't marginalizing wireless competition.

     Julius Genachowski needs to think clearly and act decisively. He's invited me in to talk, and affordability and rural availability will be my focus, along with avoiding waste of public money. Unless I write something he doesn't like and he rescinds the invitation.

  iPhone connections are still struggling in New York and San Francisco, but 90% of AT&T's 3G networks are reliably delivering megabit speeds. AT&T has massive amounts of unused and underused spectrum in most of the U.S. although it takes 6-18 months for some of it to come online.

Moffett: Caps "very likely spell the end of mobile video"
Craig Moffett (a friend) is a great "big picture" guy, a top tier Wall Street analyst always interesting to read. His job is to help investors find companies with improving profits, not the public and consumer interest. Like most of us, he's concluded that if Verizon follows, low caps will substantially increase profits in the long run by making consumers pay more for net connections.

      "[Investments over the next few years,] including the provision of 4G networks, will drive down the cost of wireless data dramatically," Moffett reports, citing estimates of capacity growing from 2.9 to 20 times with a capex level that will be declining after the current catch-up.

     He thinks AT&T raising profits is great, even beyond the 300 basis point margin increase AT&T recorded last year.  I think consumer prices going up is terrible, and Julius Genachowski just testified to Congress making broadband affordable is his top priority. With enough competition, supply and demand usually reach a fair compromise.

      There's no room under such a low cap for video. Netflix "content is currently encoded at a minimum of 500 kilobits per second and a maximum of 3.4 megabits. At 500 kilobits, a 200 MB plan user would burn through their monthly data allotment in just 12 minutes per week." At 1 megabit, still a lousy picture by today's standards, even the top tier is less than an hour a week. People will need to keep their landlines for video, preventing an erosion of the DSL business.

    Verizon has "signaled six ways to Sunday that they like the idea of metering." That's how to achieve monopoly pricing in a way that's hard to prosecute. The companies, Wall Street, and even the Department of Justice know exactly what's going on. V&T have 60% of the U.S. mobile market, and a much higher % of mobile data. They have such a commanding lead the two of them alone may dominate the market; Sprint and T-Mobile have been losing money and falling further behind. 

      Moffat is off-base in thinking that a 2 gigabyte cap is needed to prevent degrading network quality according to slews of technical experts who've analyzed the problem as well as AT&T's President John Stankey.  There is some point at which demand becomes economically impractical to serve, but that's 10 to 25 times the 2 gigabyte cap AT&T is imposing.
 
     Moffett's last thought is also worth re-examining. "[With] Lower prices would come lower investment returns… and therefore lower investment" is almost a tautology. Anything that raises profits will induce some amount of additional investment, but few think we should double corporate profits by taking the money from consumers.

    The question is how much of the profit decrease/increase would go to investment and how much to higher dividends? I believe, based on past experience and the structure of the industry, most of the money would go to the shareholders. That's particularly true because "investment" in networks is rarely from "investors" making theoretical asset allocations.

      Virtually all carrier investment in telecom comes from the profits of the existing companies, not "investors" putting money in. Most carriers devote 60-80% of cash flow to dividends and buybacks, and the rate would probably be higher with incremental income from a price hike. In theory they could decide to stop putting part of the profit back into the business, but with even weak competition there would be massive losses while the companies could go out of business.  At first pass, most of the increased or decreased profit in this business affects the stock price far more than the level of investment.

     "AT&T is running their business with a profit motive," Craig adds. They prefer monopoly pricing.  A little more at http://bit.ly/cLHNqP

 

For the record
 In the spirit of disclosure, I like to tell readers comments I have made to D.C. officials. (Their comments remain confidential unless “on-the-record.”) While factchecking a routine item with the Department of Agriculture, I also wrote “Thanks for getting an accurate number for jobs directly created in round 1. I've looked in depth at the figures for indirect and induced job creation and the studies circulating in D.C. are wildly inflated. Jobs directly created is a much lower figure and less impressive, but it's much more accurate for comparisons.”  In Qwest's Sensible Stimulus Request: 173K Unserved I made a point of concluding “the $466M cost is appropriate” based on all information public.  However, I was very disappointed they would not provide a figure for jobs directly created. Instead, they pointed to some research that wasn't based on builds like Qwest and was seriously inflated.  There's an even bigger issue. Qwest is asking for a $350M subsidy although the build would be profitable if that were cut in half. 14 states' worth of Governors and Senators are presumably demanding Adelstein give Qwest what they ask, making it politically difficult for Jonathan not to waste $100-150M by giving Qwest too much.

--------
A personal note about Gaza & Palestine because I cannot stay silent. My family has been deeply involved.  My father smuggled guns past the British blockade in 1947. His brother Samuel Burstein was one of the first pilots in the Israeli Air Force, a volunteer when pilots in war did not expect to survive.

   6,000 miles away in New York, I have no special wisdom about how to make peace.  I do know about technology and how it constantly gets better. With the technology available in Palestine today, Israel feels threatened. With tomorrow's technology, the damage will be much worse.  Some readers of any technology publication, including DSL Prime, can presumably improve on improvised rockets. I do not want to even think of what can be produced in a college biology lab, either today or in the near future.

        Technology's advance inevitably will make the struggle far more costly for both sides. The deaths inevitably will increase, probably horrendously. A way must be found.  

Reply "subscribe" to be added, "un" to be dropped
Volume 11, #8 June 14, 2010