Article Info

AT&T Lightspeed Gets More Expensive

All is not well at the Death Star today. AT&T (T ) announced that capex for the U-Verse IPTV & Fiber to the Node initiative (known as Project Lightspeed) would increase from $4.6B to $6.5B. They also announced the scope of the project was being reduced from 19M to 18M homes.

This is a sizable increase (41%) in capex for a project that was designed to minimize cost. It is indicative that the decision AT&T made to substitute advanced technology to deliver an incremental solution in favor of laying fiber isn’t going as planned. The price of mediocrity just went up.

Project Lightspeed involved placing smaller DSLAMs closer to the customer in order to reduce the length of the copper and achieve bit rates of approximately 35Mb/s. This eliminated delayed the need for fiber to the home,  the approach taken by Verizon’s (VZ) FiOS initiative. Instead, AT&T pushed the limits of DSL technology and introduced IPTV as a solution to deliver video.

From the WSJ article:

AT&T blamed the price increase in the cost of adding more servers, which are needed as the company significantly increases the amount of high-definition channels it plans to offer. AT&T is also paying a premium to its equipment vendors to ensure gear is on hand when needed, the company said.

I believe this explanation is nonsense.

It is relatively straightforward to plan for the number of required video servers, they do not scale as a function of channel count – they scale as a function of subscribers.. And I can’t imagine that a project running behind schedule has trouble getting equipment delivered on time. I’m sure vendors would love to ship AT&T whatever they can take.

I suspect AT&T has expended significantly more money on reconditioning existing wires in order to meet the 34Mb/s speed target needed to deliver video and data to a residence. AT&T cannot mention this cost as it would cast serious doubt on their decision to leverage the existing copper infrastructure instead of biting the bullet and putting in fiber.

Regardless of its true cause, this budget slip significantly changes the economics supporting AT&T’s decision to avoid laying fiber. Previously, AT&T could deliver video services at 1/3 the cost of Verizon. With the slip AT&T’s cost per home went up nearly 50% and is almost 1/2 the cost Verizon FiOS FTTH solution.

The truth of the matter is AT&T traded cost for a large amount of operational risk. When they shunned fiber AT&T assumed the risks of inconsistent DSL performance, IPTV software trailblazing, electronics scattered throughout the field, and a sub par experience based on the 35Mb/s limitation.

AT&T is deploying a very complex architecture with major limitations in the interest of saving money. It offers nothing better than what the incumbent Cablecos can provide. And now, this mediocrity comes at a higher price.


Comments are disallowed for this post.

  1. Andrew, It is a disaster waiting to happen.

    Good job. Cooters

    Posted by Cooters | May 7, 2007, 9:12 PM
  2. (Ed: Below comment was interpreted by the software filter as spam, I pulled it out of the spam queue.)

    Who the hell is Andrew Schmitt??? He is brilliant…in his own mind! How do you get to write an article completely based on speculation? Seriously, look at the following quotes from his article…pure speculation. Why believe Andrew over AT&T? So they say it’s a higher cost to do business and he says bull! Boy am I glad Andrew didn’t lace his insider knowledge with facts! (Sounds like a disgruntled employee or former employee) I guess I have to assume that Andrew lives in a part of the country where U-Verse is available, right? And he of course is basing this off his personal U-Verse experience, right? I mean to say that the U-Verse end product is already what cable delivers has to be based on fact right? I mean, my cable has instant channel change and can record 4 channels at the same time…doesn’t yours? How bout that new cable tv feature where you can view your guide and other possible channels in picture-in-picture? Or better yet, that new feature cable offers of being able to program your DVR from your computer (away from your home) or via your cellular phone? That’s right, this product IS NOT mediocre but SUPERIOR! And worth the cost invested…whatever it may be. And the end result, regardless of Andrews opinions, will be a great entertainment product for every home that is PACKED with SUPERIOR features than cable or satellite…and costs less to the consumer.

    “It is indicative that the decision AT&T made to substitute advanced technology to deliver an incremental solution in favor of laying fiber isn’t going as planned. -speculation

    “I believe this explanation is nonsense.” -opinion

    “they do not scale as a function of channel count – they scale as a function of subscribers.” -hmmmm, might more bandwidth actually take more servers…YES!

    “I suspect AT&T has expended significantly more money…” -supposition

    “Regardless of its true cause…” -opinion

    “The truth of the matter is…” -opinion (because Andrew is the voice of the telecom industry and clearly knows all)

    Good day all. Happy TV viewing!

    Posted by AT&T Fan | May 8, 2007, 9:10 AM
  3. Is it really this bad? Yes for greenfield, fiber makes sense. But what about for all the 100’s of millions of homes now with copper “lit up” and running. Your chart shows $4.6-6.5B vs $18B for fiber. Can’t AT&T just wait for 3 years and then go retrofit with the latest technology? Maybe it’ll still be GPON and by then it’s 1/10 the price of today to install?

    Then AT&T get best of both worlds with less risk and more of a linear change vs. paradigm shift.

    Just a thought…

    Posted by Ed Draker | May 8, 2007, 10:14 AM
  4. I think your criticism of the server issue is offbase. The number of server required is dependent on both the amount of content (number of channels and the size of the files, higher for HD) and the number of subscribers (how many streams can a single server support).

    The shift to a stronger focus on HD will create additional server demand due to the increases in total content that needs to be stored/ streamed. However, the increase could (shock!) also be caused by the Microsoft software, which never seems to be as efficient as they claim it will be.

    Your comparison of cost/home is fascinating and makes clear that the cost advantage of FTTN over FTTH may not be what was expected… particularly, if you consider that these totals are based on penetration assumptions. The operational risk, as you describe it, is that penetration of households passed (currently, under 1% according to the article) doesn’t meet expectations, the high fixed costs will be spread across fewer customers, and therefore will further shrink the gap between U-Verse and FiOS.

    Posted by Bruce | May 8, 2007, 10:33 AM
  5. The MSTV configuration requires video servers to achieve their “fast channel change” architecture, and therefore scales proportionally to he number of subscribers AND the number of channels being offered.

    Posted by Francis | May 8, 2007, 10:49 AM
  6. Isn’t it a little odd that the increase in the amount of servers accounts for such a large proportion of the project cost?

    AT&T is obfuscating something here.

    Posted by Andrew Schmitt | May 8, 2007, 1:01 PM
  7. German Deutsche Telecom was rolling out the same FFTN package and VDSL@ package as SBC/ATT. Six months into it they said the products didn’t work, wouldn’t transmit more that 1 HDTV channel; and they reduced their deployment of IPTV from 50 cities down to 10.

    Australia CANCELLED a 4.3 Billion purchase of FTTN. Fiber to the NOde.

    Is anybody in the USA at SBC/ATT listening or are you operating in a VOID?

    Posted by Tony | May 11, 2007, 8:47 AM
  8. The Microsoft IPTV architecture has two types of servers: A-server and D-server. A-servers scale with channels and D-servers scale with subscribers. There are many many more D-servers than A-servers. Server cost increase due to adding HD channels will be much closer to $1.8M than to $1.8B (even on a logarithmic scale). I also don’t buy this timely delivery premium nonsense. Telcos deal with late deliveries by contractual monetary penalties on late equipment shipments. The project cost is most likely higher because, as always, projections were too optimistic (which is also the case regarding project schedule).

    I completely agree with your note of trading cost with risk. Verizon chose to combine a proven CATV video technology with a proven, though costly, standardized FTTH technology. AT&T chose to rely on a proprietary Microsoft technology which was still in development. The IPTV technology is definitely advanced and promising, but when was the last time that Microsoft launched something on schedule?

    Another point I would like to note is that the AT&T data bitrate will not be forever limited to 6Mbps. With the right implementation, it can grow up to the full DSL capability of (best case) 35Mbps when nobody is watching video. Even when full video capacity of 2xHD 2xSD is streamed through the DSL, technologies exist already today to allow average data bitrate of ~15Mbps without any noticeable video quality degradation. Also, since most homes are connected with multiple twisted pairs (multiple phone lines, anyone?), AT&T can use its own version of channel-bonding to go to 70Mbps total bitrate and more.

    Posted by Adi Bonen | May 11, 2007, 2:12 PM
  9. I think people would be willing to embrace the idea of variable DSL bit rate. I think you are on to something.

    Other than Live TV, virtually all content could be downloaded and stored for later viewing during the day while folks are at the office or at night while people are sleeping. Hell, even P2P is an option for distributing the big titles.

    THis means the broadband connection is significantly faster unless someone is watching live TV.

    If AT&T could roll out such a service, and price it aggressively, I think they could take share from cable with a unique service offering.

    AT&T should find ways to leverage their architectural strengths into their business model (low cost high bit rate downloads) rather than spending billions to compensate for it’s weaknesses.

    Posted by Andrew Schmitt | May 11, 2007, 3:06 PM
  10. AT&T Fan, your comment is pulled from the spam queue.

    Posted by Andrew Schmitt | May 18, 2007, 7:15 AM
  11. You make a often missed point about the way DVRs and Tivos have trained the consumer to ‘select record watch’ vs. ‘broadcast/stream and watch’ behavior. CDNs are as much a part of this new equation as carriers are…

    Would love to discuss your take on IPTV vs. cached/managed-p2p/asynchronous video (transfer, store, and view) in depth. Its been a while, it would be good to sync up.

    Posted by Rohit Sharma | May 30, 2007, 12:20 PM
  12. Andrew is more spot on than anyone even knows. Project Lightspeed is internally tearing up the company.

    Posted by John Harrington | June 8, 2007, 12:03 PM
  13. Trackbacks / Pingbacks

  14. FTTHblog » AT&T Lightspeed Costs Rising | May 8, 2007, 11:00 AM