DSL info for consumers & the industry.
 |
 |
 |
 |
 |
 |
 |
 |
|
DSL Prime is always looking for news. Email Dave Burstein DSL Prime - the trade paper of an Internet community
June
26, 2003
Mike Powell plans Monday to release the Triennial; watch for grave pronouncements and boring pontificating. The bells will claim the part they prefer will incent them to invest, and the areas they dislike prevent it. Reality is cable modems are here and cable telephony coming - the bells will have to fight that, and the D.C. propaganda is a very secondary cause. Telefonica bringing DSL TV to Barcelona, Madrid Alicante trial extending to major markets 20 TV channels over DSL is the kind of new service that Telefonica needs, as firing 5,000 people makes clear. So they are jumping ahead with multiple video channels, video on demand, and other services, aiming first at the over one million DSL users in Spain. Every European telco has trials, but Telefonica may be the first to expand to wide deployment. They just offered 1.7B euro for the 62% of Terra Lycos they didn't own, and presumably will use those alliances for programming. Telefonica has long been involved in video over cable, both at home and in Latin America, and doesn't face the SBC/Verizon problem of acquiring programming. Lars Bengtsson of Sweden's Kreatel, who also supplies Fastweb in Milan, is providing the set top boxes. They will be MPEG-2 initially, but they are working with Sigma Designs on an MPEG-4 system for the future. Telefonica's Spanish division has flat revenues and is planning a 5,000 employee layoff, so the growth is crucial. Is 802.11g robust enough for TV in the home? Brad Kayton of Prismiq is sending me an set top to try, with performance rated to 20 mbps, more than enough to move 4 mbps video streams around a small apartment. 802.11B, with speeds up to about 6 mbps "just doesn't cut it for video" Kayton tells me, but he's found the new .11g units "work great" where there's no interference. Another set top manufacturer, one of the world's largest, has had more mixed results in testing .11a and .11g, and is finding too many dropouts in their testing so far. Bengtsson is supplying wired Ethernet gear to Spain, knowing reliability is ensured. Prismiq is a company to watch, with ex-Polycom CTO Ken Goldsholl as CEO and many of his former colleagues joining him. NTT, Softbank offering 24 mbps ADSL2+ So competitive they can't wait for all the bug fixes ADSL2+ chips, designed to go to 24 mbps, will cost nearly the same as 8 mbps ADSL, ($1-3 difference, dropping rapidly), so it's tempting to start specifying it today. But DSL Forum interoperability testing still has a way to go, and most system vendors aren't yet certain they've solved all the problems. Despite that, the rush in Japan, where most of Yahoo BB's customers are also buying the phone service, means the carriers on jumping on as fast as possible. DSL Prime strongly recommends all carriers switch to ADSL2+ as soon as they believe the gear is reliable, even if you don't anticipate offering that speed any time soon. There should be no downside, with full backwards compatibility and (modestly) improved performance as all speeds and distances. Power down mode should be a saving for everyone, crucial as opex becomes more important. The highest speeds may not be on your roadmap today, but the added cost is repaid without them. Tom Willie of Efficient is encouraged by the improved reach of the new chips, a crucial factor today. Alcatel plans general availability of ADSL2+ early next year, the kind of date that often slips, which is consistent with most other plans. Japan has become the country to watch for the future of competitive telecom, not just DSL. Phone rates have gone down dramatically, and now NTT itself is offering amazingly cheap VOIP calls (about one cent more than Yahoo BB's six cents for three minutes.) With eAccess/Vodaphone a third player, cable a fourth, and strong ISPs like Fujitsu/Nifty also gambling to survive, the entrie structure of the PSTN is under attack. Over 2M have switched to VOIP, and that's accelerating. NTT knows they are cannibalizing $billions in revenue earlier than any other carrier except KT, a risky decision years away in the west. They may be bowing to the inevitable, but a more dramatic scenario is that they are trying to bankrupt Masa Son and Yahoo BB. Son's true financial position is impossible to determine, opaque through a series of related companies; one well informed source is very confident, and says Yahoo BB was headed quickly to a profit despite some of the lowest customer prices in the world. BellSouth Georgia to 100% of COs Why are EU, Australia and SBC still asking for subsidies? France Telecom just announced they'll offer DSL in "every town over 2,000" because new small DSLAMs have proven the service is profitable wherever the telco has fiber. BellSouth has long been the U.S. leader, reaching every CO in North Carolina a while back and now to Georgia. Germany, Korea, Belgium and others reached almost all exchanges last year or before, and Verizon's 80% in 2003 and SBC's 80% in 2004 imply similar. That's little consolation for the many in the U.S. behind unserved remote terminals or more than 15-20K feet from the CO, of course. Many of the remaining 20% can be served profitably, but the bells are still holding out for 30-50% returns on investment and going slowly. Fortunately, the economics of baby DSLAMs and repeaters keep getting better: Verizon's Telus is buying in large volumes, Catena has some sales for upgrading SLICs, and the smaller telcos are far ahead proving repeaters work well. In 2000, I wrote Universal broadband service is practical and affordable The story above, and especially the German and French practice, led me to look back and reprint some of the article. Every telco, not just SBC, should offer it The United States is committed to bringing broadband to all its citizens. SBC, one-third of the US, has made that pledge (2003 - remind Ed Whitacre), so it's clearly possible. FCC commissioner Gloria Tristani explained "We cannot afford to become a society of information haves and have-nots in a world in which the ability to access and manipulate information is the currency of the day." Compared to the $300 or so in marketing it costs to acquire a customer, these techniques are inexpensive, with likely added costs per subscriber of $50-$200 They can extend your market 20-25%, so we think they are good business as well as the right thing to do to serve your communities. Problems remain, but your engineers, if mandated, are good enough to solve them. Here's how to do it: DSL can deliver megabits 25-30,000 feet with repeaters $20 of components, or less, is all it costs to build a repeater that can double the reach of ADSL. Because the primary market for this product is telcos, it won't be produced without your support. Given support, it can be produced and installed inexpensively - and bought in telco quantities the cost will be low. But unless you prove this technology is impractical, you're deceitful telling regulators or customers DSL is limited to 12-18,000 feet. Millions more can be served if you test their lines We were skeptical when Teradyne reported that millions more could be served if the telcos changed procedures and tested lines more thoroughly, but have confirmed the story with one of the largest telcos. Currently, an enormous number of potential customers are being told they can't get DSL although the lines are perfectly able to support the service. Our guess is 5-10% of customers are mistakenly told they can't get service - more than the actual number with a true distance limit. Cleaning up lines will help millions more, as unneeded load coils and unnecessary bridge taps can often be removed and problems solved. Every CO has the right DSLAM; DLC Remote terminals are cost-effective An RBOC vp told us it was "simply too expensive to put a DSLAM in smaller COs" (2003 - "Too expensive" was unlikely then and is certainly untrue today. Small DSLAMs are down to a few thousand dollars, and telcos around the world are profitably wiring their smaller offices. Assuming there's fiber capacity in place, true almost everywhere, there's a right sized unit, typically $2-4,000. For remotes, most telcos are following BellSouth's proven model of upgrading with an inexpensive baby DSLAM, or buying inexpensive upgrade boards from Alcatel, AFC, and Catena. France's 2,000 homes or hundred customer threshold is about right; if your calculations are different, put a better cost accountant to work or ask me to check the assumptions.) Yes, 5% will be tough Some, perhaps 5%, are extraordinary distances away or have lines impractical to clear. But satellite is cheap enough to fill in. (2003 In 2000, I looked at the consumer prices of a satellite connection to justify that claim. Since then, that price has gone up and satellite, while effective, remains disappointing. New satellites with more capacity will help.) Announcements around the world encourage me to hope I won't have to repeat this article in another three years. 6M modems from Ambit/Acer Deutsche Bank predicts 7M in 2004 The label in the west on 2M modems says "Thomson", and in China might say Huawei or ZTE. Yahoo BB is buying almost 2M as well, especially the Trio with VOIP and 802.11, per William Bao Bean. Only a handful of vendors still manufacture their own modems in the west, including Westell and Sagem. Getting the modem price down to $20 Rumors of $20 bids for 800,000 modems for Chunghwa lead me to wonder if that could be true. The lowest price I've previously confirmed was a large European bid at 27 euros, although I haven't been able to get the price on several large deals. Chunghwa prides itself on the most price-effective bargaining in the world, and they face low priced competitors. They created the $70/port DSLAM market a while back with an intense international auction won by Samsung/ADI over Nokia and Alcatel. Comtrend won the modem deal, at a price so low it was for the prestige of a large order, not any prospect of profit. Case, power, discretes, and assembly are now larger parts of the modem cost, but the DSL chips remain a key cost and the chip vendors presumably are alongside the bidding modem maker, considering whether to lower their price for this "special bid". Vendors look to savings generally too small to matter, offering USB instead of combos, and dropping the intelligence now included in most modems. TI has been especially aggressive when their fabs looked empty, Globespan when inventory, ADI supporting a crucial customer like Samsung etc. Several designs are underway in China, possibly in association with the (several) new fabs being built. If they make it to market, they will have to price to win early customers. Most production has shifted to TSMC, UMC, and Chartered, who probably are also playing a direct role in a contract this size, looking to fill a light production schedule. With enough zeros in the contract, price can drop amazingly when the industry has spare capacity. So please don't write me asking "Where can I buy a $20 DSL modem" - unless you are in the market for half a million, and willing to take a bare bones model. Wall street analysts who treat virtually any contract win as good news encourage irrational bidding, although not as much as the "growth at any cost" days. 6 meg for 99 pounds in London Bulldog bringing speed The price isn't right, but the service is on track. Backbone transit and routing has become so cheap that the Japanese offer similar for less than $30, but that requires economies of scale in the millions. Bulldog is virtually the only remaining UK DSL CLEC, and generally is focusing on business customers and voice over DSL. SBC is actively selling 6 mbps in the U.S. for $99; I hope one day London, Houston and my own city, New York, get internet service that matches Tokyo. EFM goes for DMT "VDSL Olympics" test results crucial DMT will be the primary VDSL line code for the U.S., as the Ethernet in the First Mile committee (previously partial to QAM) joined T1E1.4 in supporting the standard. T1E1.4 in parallel will propose a Technical Requirements (TRQ) for QAM, and maintain both going forward. One comment that supported the T1E1.4 resolution that created two documents suggested "One choice got the standards track gold medal. The other got the TRQ, which was described as roughly equivalent to a DSL Forum specification." To me, DSL Forum specifications are pretty significant, as the Forum leads the industry internationally. The dual recommendations (and the careful language) might be read as a compromise supporting both in different ways, but the T1 leadership and the subsequent vote in EFM made the sentiment clear. Even more crucial is the support of the bells, who have not been buying VDSL but may offer it as part of the fiber build. All the big players (Lucent, Nortel, Siemens, Cisco, Motorola) are likely to bid for the Bell fiber deal, so are likely to develop or bundle a DMT board to include in their bid, even if it doesn't go beyond a lab demo unit. Did the U.S. listen to China, Japan, Germany, Korea, Britain and France? A remarkable request came from British Telecom, Korea Telecom, NTT (Japan), China Telecom, China Netcom, Deutsche Telekom, France Telecom and others to allow a free choice between QAM and DMT. My initial opinion was that "T1E1.4 intends that both documents would be maintained going forward" was a respectful acknowledgement of that position, but key members of the committee tell me international opinion had little impact. That amazes me, because international co-operation is more crucial than ever. Until recently nearly everyone looked to the U.S. for telecom innovation, including DSL, but those days are past. Lucent has decimated Bell Labs, killed the Lambda router, and drastically cut R & D. Nortel is a shadow of what it was. Telcordia is scrambling to survive. The Bells have cut research much more than reported, rarely investing in work that only pays off over time. The President of SBC is personally embarrassed by how they've lost the world leadership to Korea Telecom; it was superfluous for me to point out that France Telecom, Telecom Italia, Yahoo BB and NTT had all done better in DSL than his company. Kevin Meyer and others in Electronics News have a strong series on how the world's best engineers are staying in their home countries, no longer moving to the U.S.. TI's most advanced DSL chips are designed in India; China's developing their own video compression and CDMA standards to avoid ridiculous royalties. In this VDSL battle, two key designs came from Israel, one from Sweden and France, and the fourth from Indian born engineers in California. The ITU, perhaps in two weeks, will work towards an international standard. If the major telcos above don't change their opinions, the logical ITU conclusion is to allow both line codes. The politics are very complicated. Now, choosing the best for your deployment No one wants to be dependent on a single source, Ikanos. Their engineering is obviously exceptional (they beat Alcatel/ST to market), but it's a small company whose board members have a history of selling the firms they invest in. The price they are looking for after this decision is presumably very rich, and it's not obvious who will pay that much for a company without earnings. But 2.5M users have proven QAM works well and inexpensively, and many deployments are from remotes and basements where QAM is the obvious current choice for price and field-proven reliability. What the Olympics prove Two DMT chips did better than two QAM chips in the T1E1.4 sponsored "VDSL Olympics", lab testing at BT and Telcordia released June 17 and available to all on the T1 web site. None of the four chips, from Infineon, ST, Ikanos, and Metalink, delivered as much as hoped. DMT chips proved beyond a doubt that they can deliver comparable performance. It's puffery, however, to say this very limited test proved DMT is better. No engineer would accept claims based on only four datapoints loaded with confounding variables. Behrooz Rezvani's team at Ikanos proved their enormous skill by delivering DMT chips far before mighty Alcatel/ST; it's certainty plausible the results came from a better design team, not a better line code. The DMT chips are newer designs, and Francois Crepin of Metalink told EE Times their nextgen chips (with Reed Solomon coding) will more than catch up later this year. Some very smart people tuned the systems for the lab test; a suggestion that C.'s optimizing for the test made a major difference should be taken as a compliment to his ability, but not implausible. It's also always dangerous to extrapolate from the lab to the field, even careful work from the best labs in the world. The dramatic example in our field is that early ADSL chips had the range for far fewer actual customers than the lab claims. I often wish I had independent results to check manufacturer's claims, and hope to see them more often. To encourage that, the next three chips that provide me tested results will get a feature article, even if they are otherwise uninteresting. I hope other reporters - and purchasers - also ask for results. Otherwise, I'm in a bind when I get a press release with extravagant claims. One current release claims improbably more reach. I've asked the company for test results that back up the claim, and will report it when confirmed. Pricing of VDSL $100 for DSLAM + modem is a common Asian price for QAM, as Infineon is selling 10 meg symmetric VDSL chips for $10. The 50 mbps chips are somewhat more expensive, but everything is falling rapidly. Metalink and Infineon know their customers need high volume pricing, while Ikanos initially asked more for the chip than some charge for an ADSL modem. There is no "market price" for DMT gear, because so little has shipped. I believe Ikanos has significantly dropped their price in some bids, and I'd be interested in hearing whether their current effective monopoly is letting them charge more. ST's chips aren't yet deployed in systems visible in the field. Broadcom's likely to make an announcement soon but hasn't given me a date for shipping chips. TI two years ago was promising imminent delivery of current spec VDSL chips, but they pulled the engineering team off VDSL. They cannot give me a date for chip availability to justify their rhetoric. Currently, there's effectively one source for DMT chips if you are deploying in 2003, and no one except ST firm for 2004. So if you want DMT, you'll need to bargain hard. Neither the U.S. nor Europe has seen significant volume, so prices have been much higher. That international discrepancy breaks down rapidly these days, however, and at least two companies have suggested to me they'll offer Asian prices in other markets. Key Bell supplier Alcatel doesn't even have a VDSL card for their U.S. DSLAMs yet, much less price details. Neither fear nor favor "I hope never to hear the words 'line code' after this week" J. told me, and it is boring so many of you I almost killed the piece last issue. Some companies involved currently hate me for not seeing things completely their way, but that usually passes. But I couldn't get an answer on royalties, a key question that exploded with the Globespan TI court case. I couldn't kill the story after that, and got a dramatic letter telling me it was the right move "Private - please do not quote me (you could get me fired). I and many others are anxious to see fair play and I am very reassured that you are covering the proceedings." One email began was "I know your 'opinion' on VDSL (DMT vs QAM)." I was surprised, because I have actually not been sure of my own opinion. Anyone who thinks I didn't report the positive aspects of DMT needs a course in remedial reading. email:
June 17, 2003
"What we really would love to see next is a video iPod. Help us out, Mr. Jobs!" Janice Chen of CNET DSLAM prices close to $50/user (real bid, quantity millions) make "DSL Everywhere" more than a slogan. With costs down, France Telecom promised 90% coverage, something I predict will soon be in the business plan of British Telecom, Bell Canada, and most of the developed world. News just came that the biggest holdout, SBC, is resuming their build and going over 80% next year. Ross Ireland, Mark Wegleitner, and their telco peers have a remarkable opportunity to save millions and keep customer prices down. They can refuse to support DMT as the VDSL line code until they get real assurance DMT royalties will be "reasonable" and not just an empty slogan. One key question about line codes has been answered by the 500 plus pages of lab data released Monday. Far too much to read and analyze to reach a conclusion yet, but DMT results look good. DMT certainly delivered reach as strong or stronger as QAM. The four vendors (ST, Infineon, TI, Metalink) did the entire industry a service by hiring BT and Telcordia to do the analysis. Globespan sues TI, Stanford Royalty issues hit the fan Dropping DSL chip prices have driven the market expansion, but unresolved royalty disputes can change that. Everyone's assumption was that the standards requirement of "reasonable and non-discriminatory" would lead to solutions, but my research indicates the standard is little enforced. So lead DSL chipmaker Globespan took the extremely risky step of going to court. That's extremely costly and even more risky, because it could hang over the company for years. It doesn't make sense unless the demands were very high. TI bought from Amati some basic DMT patents, on which royalties also accrue to Stanford University. Globespan's 10K reports "We have received correspondence stating that Texas Instruments believes that it owns or has exclusive licensing rights to a number of patents that it believes are required for some products to be compliant with certain ADSL industry standards, including the ANSI standard specification T1.413. ... Texas Instruments has offered us licenses to these patents on terms that were not deemed acceptable. ... We cannot assure you that we would prevail in litigation given the complex technical issues and inherent uncertainties in intellectual property litigation ... We continue to dispute the applicability of most, if not all, of these patents to our products and have asserted that Texas Instruments' patents are not infringed, are invalid and/or are unenforceable. Such litigation would be costly ... we believe that we have strong arguments in favor of our position". TI's response Marisa Speziale of TI worked with me to get their response quickly so I can include it in this issue. TI does not comment on pending litigation. TI has a long-standing tradition of protecting TI intellectual property and will pursue those who infringe our patents or encourage such infringement. We feel that some of our ADSL patents are essential to standards-based DSL operation, and we offer non-exclusive licenses to these patents at fair, reasonable, and non-discriminatory royalty rates. This particular licensing program has been in existence for more than 5 years, during which time 10 technology providers have licensed these ADSL essential patents. Freeing France: Speed and Video Free's video combo box headed to 50% availability Michael Bouhobza has a simple strategy for beating France Telecom on DSL; give the customers much more speed at the same price. Currently, that means Free's 30 euro offer is 512K/128K throughout the country, and 1M/128K in Paris where they have their DSLAMs already up and running. FT at that price is only delivering 128K/128K. They intend to add Nantes next month, and Marselles, Nice, and enough cites to over half the population by the end of the year. Bouhobza was very careful not to describe what the video service will be, but the routers they deploy all have video out. LDCOM also racing LDCOM is backed by major investor Louis Dreyfus, and has a 11,000-km fiber-optic network. They've installed DSLAMs in Paris, Lyon, Nice, Marseilles, Bordeaux, Nantes and Bayonne, Strasbourg and Toulouse, with reseller Club-Internet planning 100,000 lines by the end of the year and 9 TELECOM already actively selling 512K/128K for 30 euro. LDCOM has acquired French telephony customers from Belgacom and Telecom Italia in return for stock, two strong allies. France Telecom: Every village over 2,000 90% coverage, 400K quarters planned "We need to accelerate." Thierry Breton proclaims. "Broadband figures at the center of France Telecom's future." He's buying 5,000 DSLAMs to raise coverage from 74% to over 90% in the next 18 months. Any town promising 100 connections will be served as rapidly as possible. SBC to 80% in 2004 Inexpensive gear, not Pronto new builds John Hodulik and the UBS team were briefed by SBC
June 19 or not June 19? Some of the best tech people in our industry are in Anaheim, pouring over hundreds of pages of test results and trying to maintain personal civility and respect as the corporations they represent fight a big money decision. The DMT camp has many members with votes, but not enough to carry the day unless some of the telcos go along with cutting back their own options. That's a hope that less competition will encourage companies to invest more, given opportunities for a larger market. (Natural monopoly, anyone?) BT's surprise comment "We recommend that T1E1 standardize both SCM [QAM] and MCM [DMT] technologies ... However, if T1E1 decides to standardize one technology for VDSL, we strongly request that T1 does not seek exclusive standardization of the selected line-code in the ITU-T." Don Clarke, Kevin Foster. Given that many Asia Pacific operators, the key market, also don't want a decision limited to one line code, the impact of the ANSI T1E1.4 group is much less than I reported previously. Editorial: Facts, please, before a rush to decision Ross Ireland, Dong Wei, Mark Wegleitner, Trone Bishop and the other telco representatives involved in standards [probably] have an opportunity to save their companies and customers millions by insisting on meaningful enforcement of the ANSI rule that royalties be reasonable. I say "probably," because the crucial players refuse to provide answers about what DMT royalties are (ADSL & VDSL). Not answering these questions give credence to rumors that TI's requested royalties are high enough to destabilize the market; it also raised for me a red flag that something was wrong. When things got so difficult that the DMT royalties went to court, the entire standards process is at test. The usual presumption, that companies will honor their promise of "reasonable" royalties breaks down; the usual way of business is no longer satisfactory. This is no any longer a simple technical issue with the business and legal issues presumably satisfactorily handled. Telcos standing firm can drive the process. First step is to get the actual information; the committee per antitrust counsel doesn't want to ask, but individual companies like SBC or Verizon can. If the royalty requests are reasonable, the issue is over; if they are high enough to drive up costs, the telcos can use their votes to temporarily block the standard and do some horsetrading. Alternately, the Chair of ATIS, who is also the CTO of SBC, can find a way to press for "reasonableness." I urge everyone involved to put the facts on the table. The last thing T1E1.4 and the industry needs is more lawsuits, and the first step to avoid them is to end the secrecy. Megan Campbell, ATIS General Counsel, believes there are "many downsides if standards developing organizations like hers get involved in licensing term discussions. We leave it to the companies to bilaterally discuss separate from the committee process and the courts are an available option for disputes over what is 'reasonable' in the rare case where disputes occurs." But ATIS has no consumer representatives on most committees, and an obligation to the public not to create a controlled market with expensive patents that drive up everyone's cost. Everyone in technology knows these are real issues. Qualcomm is facing international opposition to the CDMA royalties they seek. A SIP phone designer was just informed royalties on the standards would add 25% to his bill of materials. The "Television Tax" - $5-15 per year on every future TV to the MPEG4 royalty gang is probably the most outrageous. email:
Stories not written. (suggestions and comments welcome, although I never seem to catch up.)
June 7, 2003
Yet another set of VDSL stories (YAVS) June 18 or not June 18? Some people hope to hold a vote in Anaheim that will establish DMT as the official and only line code. Others hope the committee will decide the carrier can choose either QAM or DMT. Many expect a delay for at least a month, especially because some of the most important questions of price and performance haven't been answered. "We will not be finalizing our position until we have had time to review the test results." I hear from a key player, who hasn't had the results long enough to do that. Another, who usually knows just what is going on, says "We have the votes and it will get decided in two weeks." In a third company, the ultimate decisionmaker didn't even know about the issues, much less have the data to decide. So I hope my gathering the information will help. I heard a suggestion I stay by the pool in Anaheim during the meeting to pick up news, but unless the chair tells me otherwise, I don't think a standards meeting wants a reporter standing by. (Sidenote - I jumped into this story hard when I received improbable claims on the issues, setting off an alarm about a topic I thought had been effectively settled. About 50 hours of research later, I keep finding new aspects, and a wall in front of key answers.) What don't we know
Baby VDSL from Lucent 16 port QAM pizzabox aimed at Asia Quoc Nguyenngoc didn't flinch when I asked him about VDSL port prices dropping below $100, not giving me a precise price but telling me "Lucent will be very competitive." The current model isn't outdoor hardened, but does include splitters. Backhaul includes GigE, with speeds of up to 70 mbps downstream and 40 mbps upstream. Lucent only offers QAM, although says they can make a DMT board for next year if that's what customers want. UTStarcom: ready to go both ways David Howard offers QAM VDSL, but most of his orders remain ADSL, including a "pre-standard" 16 mbps ADSL2+ unit selling briskly in Japan. "We've got the engineering ready if a customer wants DMT." UTStarcom continues to thrive with Yahoo BB, with whom they won an award at SUPERCOMM for video service. (Yes, it's ready.) Sales are also strong to both China Telecom and China Netcom, emerging in India and Latin America, and ready to surprise in the rest of the world. Two prestigious North American customers have UTStarcom on their shortlist. Next Level: We could go either way, but QAM is working great now Huppe of Bell Canada told Vince Vittore they will add 40 more building in Toronto of Next Level gear, Qwest confirmed they are growing the Phoenix launch, and a dozen smaller telcos are in steady production. "If our customers want DMT, we can deliver that as well," they tell me. They've just made a deal with Optical Solutions for fiber, and Motorola has add Quantum Bridge for a complete portfolio (including set tops) ready to bid on the Bell contracts. Corecess: 2,500,000 lines shipped Corecess is barely known in the west, but has been a major supplier in Korea and Japan (they got the early Yahoo BB contracts). Steve Klein, once of Next Level, has been presenting them to North American telcos with good effect, and they are on the short list of one of the most important bids open. They've a full video solution based on QAM, but like others are ready if a customer prefers DMT. Do we need a single line code? "a single line code has the advantages of ensuring that all VDSL DSLAM providers will have the same basic underlying technology, maximizing our flexibility to select suppliers." says one group. The other says "competition is more likely to bring out the best, and the last two years proved it." Relevant news
"Internet time" has passed, but telcos can't afford to wait for ever. They may be making major decisions in the fall, and this decision shouldn't be delayed much longer, if at all. When you have the facts you need I've discussed VDSL lately, by phone or email, with literally dozens of involved experts. The key principals involved have been exceedingly generous with time, and their pr people knocked themselves out to work with tight deadlines. Despite that, I still don't know key information: either the price or the performance of each choice. More to the point, the people at several telcos share that ignorance. Their inclination is to look at the data in light of the prices it inspires in the fiber bid, but several vendors are pressing for an earlier vote, and an exclusive franchise. DSL Prime say to those wanting a vote in June or July, put your chips down not in press releases but in facts. Quote actual prices, royalty rates, realistic delivery time frames. Make a public commitment to full interoperability testing open to all. Release the test results to everyone. You've implicitly promised all of this, to customers, standards bodies, and the press. Deliver all the info, fast, so that fairminded carriers make a decision, fast. Time to get past vagueness and spin, and work from the facts. You'll get more support in the long run, and have happier customers. email:
June 4, 2003
Reply "subscribe" for a free subscription to the trade paper of an Internet community. Unsubscribe if bored. **** ADTRAN's Integrated High-Density, OC-12/OC-3 Optical Multiplexer. In just 2U of rack space, the Total Access OPTI-MX delivers 10/100 Ethernet over SONET, DS3s, DS1s (VT1.5), OC-3s and STS-1s. This flexible and scalable OC-12/OC-3 multiplexer simplifies deployment, improves reliability and sets a new standard for price, performance and size. http://www.adtran.com/common/_adtranpx/Flash/OPTI-MX/OPTI-MX.html (ad) "The networks RBOCs have today are not going to cut it versus cable" Danny Briere, Telechoice At least 35 vendors are involved in showcases of Video over DSL at SUPERCOMM, joining DSL Prime in the belief video is ready. The Bell fiber build has the same goal, because even conservative CFOs like SBC's Stephenson see video speeds as part of the future. VDSL or ADSL2+ will be a crucial component, both in selected installations and while waiting for fiber. Even Seidenberg's million lines a year would take decades to bring fiber across America, while Kirk Ladendorf hears in Austin's SBC land a goal of 20-30 years. DMT and QAM folks are going hard in Atlanta, although the duels are fought with press conferences, and the fights with technical papers rather than fists. The public side features strong statements of support and reasoned manifestos; the private side is loaded with claims and counterclaims, best kept private. T1E1.4 on June 18-19 may resolve the issue, which is why tensions are so high, although they may still delay or reach a compromise accepting both line codes. The telcos, if united, can ask the chipmakers for fundamental changes that will lower everyone's cost. Royalties can be truly "reasonable", just as China has negotiated CDMA savings. All vendors supporting the standard should agree to regular interoperability testing with open results, improving the practical utility of standards agreed. Necessary information must be shared. Wasteful patent lawsuits can be prevented. We cannot recreate the days of a benevolent Bell Labs/AT&T monopoly, but we can achieve some of the benefits. Patent holders deserve a fair return, but like other monopolists incline to extreme greed and market control. Here, users have a ideal lever. Whether telcos choose ADSL2+ or VDSL depends on the services you want to deliver. VDSL is the choice if you want HD TV and serve 3 TVs. ADSL is cheaper if you limit to traditional video quality and generally are satisfied with only a stream or two. One highlight of the last six months is that VDSL gear costs little more than ADSL, with select large customers getting modem and DSLAM for under $100. ADSL is slightly cheaper for the equipment, but the big saving comes from fewer expensive remotes. Another issue coming before I go to VON London this weekend, featuring a deeper look at whether the Bells will really go to fiber. Comments welcome. Not too late to sign up for the conference incidentally - code "DSLP" gets you a nice discount. ** Boost DSL performance with the new, single-chip AR7 solution from TI http://ad.doubleclick.net/clk;5584022;7846380;n?http://www.ti.com/corp/docs/landing/ar7/AR7.html (ad) MCI DSL from the Rhythms COs 1.5/768K as basic service MCI has been quietly selling their own DSL on a wholesale basis, using about 700 COs that presumably cover 20-30% of the U.S. I'm one of the few who can testify from experience to the high quality of the MCI DSL service, which has been supplied to me by Mike Apgar's Speakeasy. Runs well, reliably delivering the promised 1.5 meg download and 700K uploads. But MCI itself has generally only been offering Covad resale. Robert Luke of the Atlanta Journal reports MCI began offering DSL service in metro Atlanta on May 15. What caught my eye was the higher speed of the consumer service, a nominal 768Kbps up, much faster than Covad or the bells offer at a similar price. If the FCC has killed line-sharing, that effectively blocks most DSL services from competing. However, a phone company who is already buying a UNE-L (not UNE-P) from the telco, can offer DSL as well. That means MCI (which has the colos) and AT&T (which wrote off the NorthPoint COs) are ready to go. They can only economically sell combined with voice. MCI is currently selling DSL as $100 bundle with the local/LD Neighborhood. Charlie Hoffman of Covad writes "We believe that over half of DSL will be sold as part of a bundle, which of course would contain voice, by 2000" AT&T just announced three million customers for their bundle, and MCI/ZTel are not far behind. 4% of those customers buying DSL from Covad would double their consumer sales. ***** Our technology will throw you for a loop. Tired of rolling costly trucks to do routine DSL troubleshooting and maintenance? You're ready for the mPhase iPOTS system that lets you temporarily bypass the splitter and perform comprehensive real-time testing, loop monitoring, true bandwidth characterization and troubleshooting of your DSL network from your NOC. Not only do you save money by avoiding those pesky truck rolls, but you serve your customers' needs faster and more reliably. And there's nothing loopy about great customer service The mPhase iPOTS gets past the splitter impasse. For more information, visit our Web site, http://www.mphasetech.com or contact Peter Gasparini at sales@mphasetech.com. (ad) From SUPERCOMM: which VDSL Hoping to win a vote at T1E1 in two weeks, seven chip vendors restated their alliance, presenting DMT with a persuasive announcement, summarized below by John Cioffi. #1 DSLAM vendor Alcatel joined in, with Thomson, Nokia, and Ericsson. Wednesday, Lucent announced their QAM VDSL upgrades, Infineon announced Huawei and new chips, while Samsung's big party, Corecess, and #2 DSLAM vendor UTStarcom all featured QAM as well. The DMT folks still hope to win a knockout in June or July, while the QAM folks will be happy if the standards groups compromise and allow providers to choose either. Ironically, the strength of the support from each camp may make the decision less urgent: both are clearly ready to move ahead, whether they win a quick standard or everything is delayed. Companies like TI, Intel, and ADI that held off development are already rushing to catch the early leaders. 500,000 Ikanos DMT chips Big orders, possibly slow Korean market Hanaro's advertising "the Crazy Speed, maximum 50Mbps VDSL" on their website, which Ikanos suggests will be with their chips. Hanaro has a powerful incentive to upgrade, reporting "Due to KT's aggressive VDSL marketing from 3Q02, there was an increase in our ADSL churn rate, which resulted in a net decrease of 15,776 subscribers." Hanaro hit a devastating 2.9% month churn rate, which the company attributed to KT's VDSL. Competition like that has led Korea and Japan to the world forefront, but KT has fought so effectively Hanaro is struggling and apparently slowing down and their CEO recently resigned under pressure. I hope this merely slows down the "Korean miracle," but an equipment vendor tells me one result is that KT and Hanaro have deferred many upgrade orders. Electriphy jumps in for QAM Connecting with DSM New Santa Clara chip vendor Electriphy broke out of stealth mode, joining the EFM SUPERCOMM demo. Jim Apfel, once at E-14, is CEO and Zhi-Yuan Guan is CTO. They define their products as "high-performance, flexible dual-reach PHY devices. The EFM PHY platforms is highly programmable and optimized for low-power and small die size. The family of devices includes intelligent processors which are capable to perform the extensive algorithms for the upcoming EFM Copper standard." Apfel tells me they are working on both VDSL and SHDSL. They believe in vector processing for multi-line performance, and have given me permission to reproduce their paper on QAM in a DSM environment at http://www.dslprime.com/a/qam_dsm.htm Editorial: T1E1.4, Caesar's wife, and a chance telcos may never have again Must be beyond reproach Massimo Sorbara, Chair of T1E1.4, is respected by everyone, but is on the spot because his own employer, Globespan, has been a supporter of one camp, DMT, for a year. If he doesn't take extraordinary measures to demonstrate fairness, ANSI will lose more than face. This issue should be decided by informed expert consensus, not accidents of organizational rules. A vote dominated by interested parties will undoubtedly be questioned and devalued. Perhaps voters with the strong financial conflicts of interest, like the chip vendors, should be asked to abstain. The practical effect would be to shift the balance of power from chip vendors to providers, probably sensible since they will be make the ultimate buying decision anyway. Sorbara's first step should be complete information. Does DMT actually have superior performance, or did QAM hold its own in the labs at Telcordia and BT? I don't know, unless I ask someone to violate a non-disclosure agreement, which I do not want to do. Everyone should be asked to consent to making the test results available in full. There's no confidentiality being breached, since the principals are all part of the group with the information. No one has threatened lawsuits yet, much less a big bombshell case. Diplomacy may head off the problem. A remarkable opportunity to ensure the future Telcos at T1E1 are briefly in a unexpected position, able to trade their votes for chipmakers' agreements to keep costs down in the future. Strategists at the highest levels should jump in now, and decide what will build the industry for the future. The Bells are working together to lower their costs on fiber, and can do the same on the complementary VDSL package. The telcos or the committee chair should request full disclosure on patents and royalties payable. Rambus and other memory chip makers have spent tens of millions fighting each other, and the DSL industry needs to avoid expensive patent suits. Broadcom and Intel are united this time, but have been twice in high stakes court fights. Experience proves the "reasonable and customary" licenses may be anything but, and chipmakers like Globespan and Centillium have to tell their shareholders they face unknowable patent risk. Infineon, TI, ST, and others should be asked now about on current or possible stealth patents, and what royalties they will charge. All are entitled to a fair return on their research, but patent costs too high block innovation. I don't know what kind of patent deal ADI, Intel and others got from the key patent holders,, but they were in a position to ask something sweet. Must one standard rule any more? A heretical question, because until very recently the consensus was "the winner would take all" and the other line code fade away. VDSL sales were small, and many thought chipmakers would not invest until they had guaranteed control of the market. That is no longer true, as the strong commitments by chipmakers make it clear they will press hard to get to market. John Egan of Infineon argues that competition between those using the same line code and then between the two line code types will more effectively incent the chipmakers to do the necessary R & D and keep costs low. Individual operators can then make the line code decision based on business factors. The endorsement can still be very important, and play a factor in the decisions of major telcos, especially if perceived as well informed and fair. From the QAM side Stakes and emotions are so high, I invited the principals to describe the issues in their own words, promising to print verbatim in a DSL Prime. John Egan of Infineon writes " In all the discussion about line code selection, all involved ought to remember that delivering effective systems technology to Service Providers should be the real focus of standardization efforts. More than 2 million VDSL lines are now deployed and generating broadband service revenues for PTTs and Operators. These lines are practically all QAM VDSL. A dual line code would provide the flexibility to select a mature, cost-effective technology to operators, who need to introduce VDSL services today like Korea. Operators who expect a smoother transition from DMT line code, can decide to wait for cost-effective DMT solutions. The competition between two technology camps will speed up the learning curve." He's promised fuller comments in our next issue. From the DMT side John Cioffi, the Stanford Professor who invented DMT, compiled this on behalf of the VDSL Alliance "Why DMT for VDSL? A large group of major telecommunications vendors, chip suppliers, and service providers have committed support to the world-wide standardized ADSL transmission technology, Discrete MultiTone (DMT), also for Very-high-bit-rate Digital Subscriber Lines (VDSL) for the following reasons: Flexibility – DMT’s flexible adaptive spectrum allows the highest performance in any DSL situation, thereby reducing operational and capital costs for DSL deployments. VDSL DMT solutions offer asymmetric and symmetric digital services. VDSL DMT solutions also allow complete flexibility in contouring spectrum, by design or adaptively, to different DSL data rate requirements, specific deployment situations, national or regional spectrum regulations, and service-provider situations. VDSL DMT’s flexibility allows the lowest possible deployment and continued operation costs of any DSL method. Interoperability - Nearly 50 million ADSL subscribers exist worldwide with many desiring speed increase. VDSL’s DMT enables a consistent incremental backward-interoperable improvement of ADSL’s DMT technology to higher asymmetric and symmetric speeds. Thus VDSL modems, whether at a service provider DSLAM or remote-terminal location or customer-premises location, can readily communicate with existing ADSL modems. The interoperability leads to deployment cost reductions both in reduction of service visits and elimination of the cost of modem replacements where not necessary nor demanded by the customer. Availability – DMT VDSL has large support with many existing products recently entering the market place. Existing DSL modem supplier’s 10-year-plus experience with ADSL enabled their ready migration to ADSL2+ and VDSL, allowing extremely efficient designs of the flexible DMT systems implementing several DSL types from a plethora of suppliers. Chip suppliers committed to the support of only DMT for VDSL already include Intel, Texas Instruments, STMicro, Ikanos, NEC, Broadcom, Marvell, LSI Logic, Analog Devices, and AWARE. Equipment vendors committed to support and/or supply DMT VDSL are Alcatel, Ericsson, Nokia, Nortel, Sumitomo, NEC, Samsung, and Hyundai and a large number of others. Virtually every DSL service provider in the world either supports, tests, or deploys ADSL DMT transmission systems. The VDSL Alliance believes it is in the best interests of the DSL and telecommunications industry to decide now that DMT is uniquely the best transmission technology to be standardized for DSL. " Aidan O'Rourke of Broadcom writes "After supporting Single Carrier Modulation VDSL for the past 6 years, Broadcom has recently decided, subsequent to detailed technical and market analysis, to support Multi-Carrier Modulation for VDSL line-code standardization. This contribution sets forth arguments for the selection of Discrete Multi-Tone (DMT) Multi-carrier Modulation (MCM) as the line code for VDSL" http://www.dslprime.com/a/DMTBroadcom.pdf . The Alliance also prepared an in depth paper. One interesting suggestion is that ADSL/VDSL chipsets will be so close in price to ADSL chips that they can be included in all modems, allowing providers to upgrade customers. http://www.dslprime.com/a/vdsl_dmt_pitch.pdf Paul Spruyt writes "Alcatel has supported DMT from the very beginning because of its many compelling advantages such as optimum performance, support of very high data rates, exceptional flexibility in terms of spectrum use, and excellent resistance to interference and interoperability with ADSL." Alcatel's opinion in depth is http://www.dslprime.com/a/alcateldmt.pdf Questions I'm asking It's easy for everyone to agree moving quickly to a standards decision is a good idea. I believe reliable and full information will make that a better decision, so I'm going to work hard to deliver that in depth. Apologies to readers not directly involved if this gets boring. I'll ask chipmakers what chips they ready to deliver, when and at what price. I'll ask telcos, the key neutrals, how they are reaching a decision. I'll request of all the companies in the process that they make available the results of testing at BT and Telcordia, which currently are under tight non-disclosure. **** Wanforce Qwest Promotion: Order a T-1 through Wanforce, receive a $1,000 equipment credit. Order a Qwest T-3, receive a $2,500 credit. Gear from Cisco, Lucent, Adtran, Juniper, Foundry, Sun and HP, all with 90 day warranties. For quotes, email pete@wanforce.com or call Peter M. Roberts, Wanforce Technologies 636-449-7333 www.wanforce.com (ad) On DSL Prime's bias With so many conflicts on the table, it's only fair to report my own biases and opinions. There is no such thing as truly "objective" journalism, no matter how much one seeks that goal. As editor of DSL Prime, I've tried hard to present both sides on this one, especially when I discovered that smart, neutral parties were undecided and divided. Readers can probably infer my expectations from my invitation to Cioffi to keynote the Fast Net Futures conference, and the extensive coverage I've given DSM and DMT. Since early April, I've believed DMT was the likely but not certain standards winner, and while I've carefully not printed that, I've said so when asked. Texas Instruments is my largest advertiser, and the editor and publisher of DSL Prime are the same person. Infineon has also been an advertiser. Readers will have to judge whether that's affected my reporting. My unqualified, uncertain opinion I'm not an engineer, and not qualified to judge issues like line codes. So I look for experts without a financial interest in the decision. I've previously reported that one of the world's largest telcos was so convinced DSM DMT would work they requested a bid from their primary supplier. One key independent engineer told me he had looked hard at the same, and it should work. Until recently, I believed it better to settle on a single standard, and thought QAM's proven performance a lesser factor than DSM/DMT potential. Now, I'm not sure whether a single choice or a carrier option will best promote the industry. I think the senior engineers at KT, NTT, SBC, Bell Canada, DT, and others like that should decide, not a reporter. If they can generally agree on one line code, that's the right choice for the industry. If they're not ready to agree, forcing a decision merely moves the battle to the sales process in each telco. Parts of the story to come I hope to get out another issue before flying to VON London, with information VDSL from equipment makers including Samsung, UTStarcom, Huawei, Alcatel, Lucent and Corecess. Cisco can play a crucial role, because the Bells will want them to bid on the fiber contract and they may want to offer their QAM based Long Range Ethernet line as part. BellSouth and Cisco just signed another big deal, and they are involved heavily at Bell Canada and SBC as well. Lucent is using QAM in Stinger - will they step in to make sure that can be part of their bid on the contract of the decade? *** Michael Boukobza of Free.fr, Gavin Young of Bulldog, Tim Johnson of Point-topic, Dan Gillmor of the Mercury News, and Ian Keene of Gartner will be on my panels at Jeff Pulver's Voice on the Net London June 9-12. Join Andy Green, CEO BT Ignite, Alan Nunn, Chief Voice Architect, BT, Frederic Potter, CTO, CIRPACK, Bridget Cosgrave, President of the Carrier & Wholesale Business Unit, Belgacom, Gadi Tamari, CEO, RADVISION , Elon Ganor, CEO, VocalTec and many others. Use code DSLP to get a favourable rate. www.von.com (ad) Editorial: Europe is still failing The Net can be much more Six months of growth are encouraging, but not enough. Ewan Sutherland of Intug puts it bluntly: "At the end of 2002 there were more broadband access lines in South Korea than in the European Union. To make matters worse, those lines were up to fifty times faster than European speeds. Since then there has been the roll-out of tens of thousands of VDSL lines, at speeds from 13 to 50Mbit/s; a service unavailable here. In the first quarter of 2003, Japan added 1.4 million ADSL lines at 8 or 12Mbit/s, more than the equivalent growth in Europe and again at very much higher speeds. It is not only that Europe is not in the lead in the adoption of broadband, but that it seems unlikely to catch up for years to come." Four leaders of the Euro telcos were on a panel in Berlin last fall. They spoke with pride about how much they were achieving. I applauded their efforts and reported they were ahead of the U.S. in many ways. But I also asked the question "What will it take to raise Europe to Asian speed levels? Do the French not want as robust an internet as the Japanese?" Answers aren't easy. Intense competition, with significant encouragement by government, worked in Korea, Japan, Hong Kong and Taiwan. With a post-bust shortage of capital, it may need even more vigorous government action. But no country can afford to be left behind on the net, and complacency is unacceptable. *** TI's Broadband Focus is a great way to keep pace with broadband. To read the latest issue, or subscribe for free, http://ad.doubleclick.net/clk;5666873;8086766;h?http://focus.ti.com/docs/apps/catalog/general/applications.jhtml?templateId=955&path=templatedata/cm/general/data/bband_enewsletter (ad) Correction:
|