As someone whose first home broadband experience was a 256 kbps broadband connection from Verizon’s (s vz) grammy Bell Atlantic, I have always retained a soft spot for DSL technology. Sure, I was jealous of my friends who got @Home cable-based broadband and its 1 Mbps service, but in Manhattan of the nineties, DSL was the only game in town. If you saw the cables in my East Village apartment block, you too would feel incredulous – how do these creaking, aging old copper wires bring fast broadband. As time went by, the speeds increased.
Cable broadband suffered from too much popularity — too many people shared an infrastructure and as a result the speeds delivered to the home were actually a fraction of what was advertised. And when I moved to San Francisco, I decided to stick with DSL and used Pacific Bell’s (now AT&T) (s t) connections. However, somewhere in the mid-2000s, things start to change.
DSL speeds, though nearly 15 times faster than my first connection, started to fall behind the cable broadband speeds. DSL performance became spotty. And I switched to Comcast (s cmsca). Today, I live in the future — I have a 200 Mbps fiber connection, thanks to my local independent ISP, WebPass. It costs a lot less money than what the cable company wants from me. And it is a heck of a lot faster than what AT&T has to offer.
Like me, a whole bunch of people have switched from the creaking DSL offerings to faster connections. I have been writing about the slow migration away from the classic DSL offering for a long time. People have switched in big numbers to cable companies, particularly those who offer better quality, higher speeds such as Comcast and Cablevision (s cvc).
DSL owners have switched to faster offerings from their own phone companies — Verizon’s FiOS for example — as the demand for consumer bandwidth has exploded thanks to growing popularity of web services such as Netflix, (s nflx) social networks like Facebook(s fb). The growing number of in-home devices has started to increase our need for bandwidth.
[dataset id=”582557″]
AT&T and Verizon, two of the largest DSL providers in the world didn’t really keep up with the times, and the speeds like their European peers did. The reasons were complex — our geography was a disadvantage compared to very compact cities in Europe, for starters. But most importantly, the Baby Bells wanted sops from the elected officials.
Whatever the reasons, we didn’t really see speed bumps on DSL like we saw from the likes of Free in France. AT&T built U-Verse, a hybrid fiber-copper network and Verizon built FiOS, but mostly for their richer constituents — the people who could afford to pay couple of hundred a month for a triple-play service. That focus on higher-end customers meant that the classic DSL was left to die on the vine.
[dataset id=”582578″]
The market too was speaking loudly — the people were switching away from AT&T and that did indeed threaten AT&T’s whole existence. As DSL sales swooned, AT&T customers went to Comcast and Cox and Time Warner. AT&T couldn’t sell switchers a phone service, a declining business to begin with. It couldn’t sell them a television connection. The lure of a wireless connection packaged neatly with everything wasn’t a reality anymore.
Today, AT&T essentially put the nail in the coffin for DSL technology when it announced that it was going all-in on IP-based networks and IP-technologies. As Stacey Higginbotham reported earlier this morning, Dallas-based AT&T is spending nearly $14 billion to completely switch from last century’s technologies and put old copper-based network out for pasture. Here is what the company said in a press release:
- U-verse. AT&T plans to expand U-verse (TV, Internet, Voice over IP) by more than one-third or about 8.5 million additional customer locations, for a total potential U-verse market of 33 million customer locations¹. The expansion is expected to be essentially complete by year-end 2015.
- U-verse IPDSLAM: The company plans to offer U-verse IP-DSLAM service (high-speed IP Internet access and VoIP) to 24 million customer locations in its wireline service area by year-end 2013.
- Speed Upgrades: The Project VIP plan includes an upgrade for U-verse to speeds of up to 75Mbps and for U-verse IP-DSLAM to speeds of up to 45Mbps, with a path to deliver even higher speeds in the future.
These investments are a realization of a harsh reality AT&T and to some extent Verizon is living in — everything is going IP. Voice is an app. Video is an app. And even the thermostat is an app. The puny Internet speeds they continued to offer via the old DSL has no part of this bandwidth-hungry future. And even with these upgrades, AT&T is still lagging behind its fiber-based competitors. The need for bandwidth isn’t going away – and for Ma Bell, that is the reality. It needs to figure out how to live with it.
For me, it is a bittersweet moment — for I can only remember being blown away by the 256 Kbps speeds and dreaming of a future when I could have 100 times the speed.
Related Stories
AT&T Goes All IP: We are parsing the news in a series of posts, for we believe this is an end of an era. Here are our two stories on the topic so far:
As someone who can only get 1mbps DSL from ATT and no other broadband alternative (out of range of everyone, including ATT…), I really hope this means faster connectivity for me in the near future.
How is it that you define U-verse as not DSL? It’s a VDSL technology delivered over a copper pair to the home. When did VDSL stop being a form of DSL?
Bernie
As I pointed out it is classic DSL that we should be saying good bye to. U-verse is fiber+VDSL.
Om, could I impose upon you for a more persuasive explanation of this? For me, uverse just means I have to spend $100 for a new DSL modem, which will be connected to the same copper wire pair, leading back to the central office. Uverse places mini COs in cabinets throughout the neighborhood, but there’s no appreciable difference in the customer’s connection other than that ATT grants permission to receive faster speeds. The protocol on that wire pair is the same DSL as this supposedly obsolete “classic DSL”. Let me repeat: I cannot order “classic DSL” faster than 3Mbit just because ATT won’t let me, not because the circuit won’t support it. It’s fine for ATT to use uverse to extend the range of faster service, but they’d force me to switch to uverse just to make adoption stats look better.
Om,
I guess I’m probably splitting hairs here, but most DSL in the field today is fiber + DSL. That is, a fiber fed DSLAM, with some flavor of DSL (ADSL2+, VDSL, VDSL2, etc) in the ‘last mile’ to the home. The only difference with U-verse is they use shorter loops to get better DSL performance. But at the end of the day, it’s still a DSL based technology.
And a little Googling will show that this magical “IP-DSLAM” which makes better use of copper wire is *NOT* the name of an advanced transmission protocol. It’s the name of those boxes ATT places around the neighborhood where it connects customers to the fiber cable. From the IP-DSLAM to the customer, the link is xDSL. It’s faster because ATT is willing to unthrottle the modem more, in light of the probable increase in reliability.
We’re only 6000′ from our central office. Traditional DSL should be reliable in the double digits over that distance. Furthermore, the last time we had a service tech over here, he said that our lines indeed ran a “straight shoot” to the CO, and that it went through the curb side box 2500′ from us. He had to run out there to check it. It’s an IP-DSLAM. So what we have is a perfectly good 24Mbit DSL modem, which is only granted 3Mbit (2.57Mbit, actually) because we won’t buy a new DSL modem for $100, that uses the same protocol, wires, and joins the ATT network at the same DEMARC. Oh, and to let them call it uverse.
Uverse is just old technology in a topological arrangement that is more expensive, but much more reliable. Om, I’m a bit surprised that a journalist of your prominence is still caught up in the hype.
ATT is going to take 3 years to expand their U-Verse reach by about 33%. Wow! Those guys are amazing (heavy sarcasm). Uh, but that only brings them to 57M customers. What about the rest?
$14B is not a lot of money for a company the size of ATT, and also for the size of their customer base. It’s a seemingly large number designed to impress people, but is woefully short of what should be invested. And they’re not putting last century’s copper networks out to pasture; they’re building out a hybrid fiber/copper network, instead of going all out and building a fiber network. Because fiber would cost more now than hybrid, they will leave the job of upgrading to all fiber to their successors in 15 years or so, long after the current management has retired with nice bonuses for conserving capital during a time when interest rates are near zero.
ATT does not care about making customers happy, they only focus on doing the minimum possible. Which is what they are doing now. Which is why when there are other alternatives (either a greenfield fiber network or cable 3.0), people will leave them in droves.
$14B absolutely IS a “lot” of money when there’s no guarantee you’ll get any customers.
Most high-speed Internet is provided by cable companies, meaning AT&T gets whatever and whoever is left over. Would YOU spend millions to lay fiber in a neighborhood knowing 75% of the homes there already subscribe to cable modem service?
I’m constantly amazed at how people talk about how “cheap” broadband costs “truly” are, when they are talking about someone else spending the money . . .
DSL is like adding special chemicals to a horse’s feed to make the horse run 65 MPH. It only works for awhile.
I remember demoing Covad’s ADSL for a reporter from BYTE in 1997. He downloaded a Star Wars poster. It took about 3 minutes. Then he asked, “Who needs this much bandwidth?”
Ehh you DO realize that, save for wireless LTE, the other future plans of action laid out by AT&T both use some form of DSL technology, delivered over the SAME twisted pair copper infrastructure that’s been serving your home for a good part of the past century? Also, a major reason our DSL tech falls behind the Europeans is the fact that we don’t even use the technology to its fullest potential. In major cities for example, AT&T could uncap its ADSL2+ network to offer the full 24 Mbps, and throw in Annex M to allow for 3 Mbps uploads, instead of the pokey 6/768 connections they sell to people without U-verse. Also, uncap U-verse lines with no video subscriptions… a lot of people routinely sync at between 35 and 52+ Mbps down, making it a fairly decent competition with cable (for most mid-range users anyways.) As revenues flow in by better leveraging the EXISTING infrastructure, AT&T can start extending U-verse fiber to the home over the next decade (paving the way for micro-cell backhaul for the wireless division), to allow for long-term growth. Meanwhile, use the newly-acquired WCS spectrum for either WiMax or fixed LTE (maybe some kind of line-of-sight system?) service to more rural areas, using a dynamic microwave backhaul network to reduce recurring costs of running cell sites.
I am a fellow victim of ATT, with a “Pro” DSL plan which stays rock-solid and unmoving at only 82% of our promised “up to” speed, no matter how bad the weather, despite being only 6000 feet from the central office. Distance vs speed charts suggest that they could provide speeds an order of magnitude higher, or this speed an order of magnitude farther. They’ve configured our modem to throttle too low. And once a year they “accidentally” re-throttle us below half speed, despite the modem showing the correct profile. They’re throttling us transparently someplace upstream! Don’t let ATT tell you they don’t do that! One call to L2 by the on-site tech, reboot the modem, and full speed is returned!
Our modem is capable of 24Mbs, but the only way we can get speeds faster than 3Mbs from ATT is to order uverse, and buy a brand new modem for $100. This supposed “fiber+VDSL” technology, Om, will consist of a new copper-line DSL modem connected to fiber in the same CO 6k’ away, and will be no more “fiberous” than our current connection.
I’d like to see some journalistic uncovering of the reasons ATT preferred to provide a poor customer experience for everyone, even before uverse rolled out and needed justification, rather than to unthrottle all those DSL modems and reap increased revenues from higher speed plans. I can’t believe that it’s to protect them from shame when it’s revealed that 3Mbs was never “high speed”. More likely it’s to prevent a suspicious wave of customers from testing their speeds and pressuring ATT to provide what they promised.
My point is, equipment is already deployed which provides decent speeds, but ATT has always kept them turned down to “low”. Those horses were bred to run 65MPH, but hobbled down to 3 their whole lives. Great swaths of territory, certainly not everywhere, have always been capable of “uverse-like” speeds, but ATT let us think they were crap instead of responding to customer demand. Those slow, expensive speed tiers have irreversibly given them the reputation of being unable to deliver! Like Anonymous above points out, the technology was never used to it’s full potential. Why? A looooooong term conspiracy to drive up prices through a deliberately planned shortage?
Your understanding of DSL isn’t apparent until the end of the article. I’m scratching my head, wondering if you really understand DSL.
Corporate greed and non-regulation have resulted in the US falling woefully behind in Internet connectivity.
Unfortunately, U-Verse customers are handed a singular Public IP address that NEVER changes.
I’ve had the same IP for the past 18 months, even after calling useless tech support people at att TechConnect – nobody seems to be able to change a customer’s public IP. This is one of the largest security holes I’ve ever heard of… ever.
You have a static address. Most people would like that. I have to pay extra to get static addresses from Verizon. It’s not a security hole, it just means it’s not as easy for you to hide from the police.
I understand that some users might find this appealing, especially if hosting a website… a dynamic IP would be more difficult to contend with in that situation.
However, we have to consider the security of a ‘static’ IP solution; my firewall is being inundated with requests, many of which could be potentially malicious. The fact that Uverse is “unable” to make such a change seems bizarre… and the fact that the traffic is nearly constant means that I’m unable to take advantage of the bandwidth that I pay a premium for.
I maintain that a customer’s public IP address should be changeable, even if it’s not 100% dynamic. I should be able to protect my family from identity and information theft, and nobody should be at the mercy of corporate bureaucracy.
Nikoli976, I don’t think having a dynamic address woud make a difference. Those addresses get poked all the time. As for making it easy to change, it’s either static or dynamic, and if it’s the former, the protocol doesn’t allow for occasional re-assignments. It would have to be done manually, and that’s an unfair burden to place on ATT (although they deserve it).
You can set your firewall to ignore all requests from the internet, that should stop most, if not, all of your problems.
I mean no disrespect, but your IP address will likely change if you turn off your modem for a few minutes or hours. Have you tried turning it off and back on again? 😀
Erm FTC still uses DSL to get from the Cab and this road map was described (using SDSL or VDSL) decades ago when I recall reading the BT telcoms journal special issue they did around the time ADSL was launched.
I rember our in house Mag in the article abotu the first 50 prototype modems thay had made and tested in ipswitch its tone ws “F^&k me it actulay works”
When I see these speeds, they typically only indicate what best possible rate will be in the download direction (Internet to home). What is expected in the upload direction (home to Internet)? While most people surf or stream videos and need the download speed, times are changing with larger megapixel photos, amateur digital video, email attachments, etc. that the upload speed needs to keep pace as well. Even on the television side as more and more content is being offered as 1080p, Blueray, even the Ultra HD 4K is not that far around the corner, there have been numerous reports on tests done with different service providers (cable and phone companies) on picture quality. You can only compress the video so much to fit in your bandwidth without seeing a degradation on the TV screen. TV screens just keep getting bigger and that extra detail lost in the compression becomes very noticable. As a ATT stockholder, I certainly want to see company do well, but can’t come to the realization on why they just don’t proceed to begin fiber to the premise like Verizon did. The technology that ATT is deploying will never be able to keep up. Verizon deployed fiber when the technology was more expensive and alot of lessons were learned along the way. The electronics and associated costs have come down significantly – I just don’t get it.
In a nutshell, ATT is maintaining a false shortage of bandwidth to keep the price up. It’s an embargo, and they’re part of a cartel. They’re acting like it’s an expensive, precious resource that is abused by some of it’s customers. This keeps the pressure from regulators and the public down. They bill by the byte, as though bandwidth were kilowatts of power, when this model bears no relation to the real cost of providing Internet connectivity. ATT’s service isn’t actually worth anything near what they charge for it. It’s critically overvalued. Your stocks will plummet the moment ATT has real competition and must charge a fair price. Until then, they exist only to place barriers between the customer and the Internet, which the customer must pay dearly to evade.
The more common really fast connections become, the less valuable they become in the eyes of the customer, and thus represent less leverage to the provider. Those $50/mo 3Mbit tiers will have to be discounted to $5/mo. ATT and other ISPs prefer to pretend that they’re working hard, pushing the technology to it’s limit, to bring you your cutting edge service. It would be a bit inconvenient for the public to learn that everyone’s throttled way down. A small configuration change, and everyone has 10x more speed, at very little cost to the providers.
Residential Internet service is asymmetrical, so that the ISP can squeeze more customers onto the frequency range they use (cable), or maximize the download speed (DSL). Originally it was due to various technical considerations, but it will likely remain the rule as it also limits customers hosting server applications or file sharing. An asymmetrical connection is a second rate connection.
elfonblog, as speeds increase, interference between pairs becomes a bigger problem. And of course as you add even more customers in a binder group, the interference becomes greater still. But I’m sure all those near and far-end crosstalk issues are just a result of corporate greed — physics has nothing to do with it . . .
“AT&T essentially put the nail in the coffin for DSL technology”… Uh, no it didn’t. It may have put the nail in the coffin os ADSL2+, their old DSL offering. But they replaced it with VDSL2, the latest deployed version of DSL technology. It’s hilarious to read that AT&T is killing DSL technology by deploying IP-DSLAM’s. DSL – Digital Subscriber Line. DSLAM – Digital Subscriber Line Access Multiplexer .
U-verse, like Bell Fibe, is still DSL (VDSL2, specifically). It’s not a “hybrid fibre” solution just because they put the DSLAM closer to you to reduce loop lengths. That does definitely enable much faster service (I’ve got a bonded 50 megabit down, 20 megabit up VDSL2 connection), but changing the relative lengths of the fiber and copper segment doesn’t magically make it any more hybrid than it was before.
If anyone wants to understand why DSL lagged cable modem in customer popularity, all you have to do is consider the FCC’s record of regulating (or not regulating) the two services.
From its beginnings, the FCC has taken a “hands-off” approach to cable modem service, to the point of classifying it as a Title I information service, a decision upheld by SCOTUS in the “Brand-X” decision.
OTOH, the FCC regulated telco DSL to the hilt; consider the roadblocks put up by the FCC to SBC’s “Project Pronto” as far back as 2000:
http://www.channelpartnersonline.com/articles/2000/12/regulatory-news-fcc-reviews-sbc-s-project-pronto.aspx
With regulatory interference, is it any wonder that telco DSL lags cable modem service? Thanx to the heavy hand of the FCC, the telcos will likely NEVER catch up to the cable companies.