Tuesday, February 01, 2005
SBC buys AT&T for wha?
The pundits are saying that SBC is buying AT&T for its three million business customers. Ha!
Any business that can afford to hire a telecom manager should weigh building its own net -- and will if the telecom manager is doing his or her job. Says here that Ford Motor Company, Bank of America, Bausch & Lomb, Gannett Co. and Google, among others, are building their own fiber networks.
The article continues
Thanks to Steve Guich for pointing to this!
Any business that can afford to hire a telecom manager should weigh building its own net -- and will if the telecom manager is doing his or her job. Says here that Ford Motor Company, Bank of America, Bausch & Lomb, Gannett Co. and Google, among others, are building their own fiber networks.
The article continues
"Analysts have long said it's too expensive and requires specially trained optics engineers to build and run these networks," said Gary Gunnerson, IT architect at Gannett, which has already built metro fiber networks in three cities where it has newspapers. "I've found just the opposite to be true. The fiber and the equipment are so cheap now, and anyone who is familiar with IP networking gear can handle a short-distance optical network. I could teach them how in half a day." . . . Gunnerson said companies that spend between $7,000 and $10,000 per month on telecommunication services should consider building their own fiber networks.and
"We could have gone to a local service provider to lease an OC48 (2.5 Gbps) ATM or a Gigabit Ethernet service," said Larry Schaeffer, senior vice president of network services at Bank of America. "But we would have gotten less bandwidth for more money. And the infrastructure would have been shared with the provider's other customers. It made more sense to build it ourselves."Compared to fiber, building wireless networks is even cheaper and easier. Now that the fences are down, how long is SBC expecting AT&T's three million cash cows to stay in the pasture?
Thanks to Steve Guich for pointing to this!
Comments:
There have always been companies that build their own networks. I would submit that there are few companies that can affort to dig trenches and lay fiber to every location they have. OK, so they lease dark fiber - but is there dark fiber laying around to the location they want to go to? (The deep dark secret of the "fiber glut" is that while 80% of the fiber in the ground may be unlit, that aggregate percentage is because there are some routes with 95% excess capacity and other routes with none -- and chances are, you need capacity on the route with none...)
On a broader level, a communications network is like any other service a business needs -- it can choose to build and run it internally, or it can pay someone else to. There are positives and negatives to either choice; there will be companies that choose to do both; and the popularity pendulum will swing back and forth. ("Focus on core competencies!" "Control your resources and your destiny!")
Additionally, while the equipment may be cheap, it's not clear to me that a lot of companies want to make the investment to run it at the level needed for "mission-critical applications". Sure, you can afford to hire a telecom manager to build a network -- but you also have to hire the second guy to do the software upgrade out of business hours, and the third guy to cover when the first is on vacation and the second is home sick, and who do you call when the box goes down at 6 AM Monday? And is it efficient for three million business customers to all build these network operations departments? (Though a lot of out-of-work telecom network operations folks would be thrilled...)
On a broader level, a communications network is like any other service a business needs -- it can choose to build and run it internally, or it can pay someone else to. There are positives and negatives to either choice; there will be companies that choose to do both; and the popularity pendulum will swing back and forth. ("Focus on core competencies!" "Control your resources and your destiny!")
Additionally, while the equipment may be cheap, it's not clear to me that a lot of companies want to make the investment to run it at the level needed for "mission-critical applications". Sure, you can afford to hire a telecom manager to build a network -- but you also have to hire the second guy to do the software upgrade out of business hours, and the third guy to cover when the first is on vacation and the second is home sick, and who do you call when the box goes down at 6 AM Monday? And is it efficient for three million business customers to all build these network operations departments? (Though a lot of out-of-work telecom network operations folks would be thrilled...)
Well, as DG Lewis already commented on the "fiber glut" myths... Allow me to deconstruct the "wireless myth" a bit.
Running point to point, or even point to multi-point wirless links between locations in a metropolitan environment can be a serious pain in the backside if you don't own the building. You're assuming companies would want to pay for rights to either run their own conduit (or a shared conduit if the building has a decent wire/phone closet management plan) up to the roof. Pull power and fiber. Get approval to mount and aim equipment after determining they do, indeed, have line of sight. (Ok, you'd likely do this step first.) Then service said equipment.
I did this kind of thing for the financers who purchased E-xpedient.com out of bankruptcy. There was a lot of red-tape and paperwork nonsense to deal with regarding roof rights and rights-of-way inside the buildings. Additionally, I spent far too many cold winter days 20-60 stories above the street replacing equipment that was mounted in difficult to reach spots even if it hadn't been windy/snow covered/ice covered.
Last I heard, they've moved away from the wireless model and instead found less labor intensive ways of bringing connectivity into their lit buidlings... like traditional TDM circuits or longhaul gig-e.
I invite anyone who believes wireless is the answer to all our WAN needs to take a drive down just about any US Interstate and admire all the pretty, vacant AT&T Microwave towers. (In fact, if you really want one, they auction them from time to time...)
Or were you just assuming to use 802.11g across town? :-) -DougDever
Running point to point, or even point to multi-point wirless links between locations in a metropolitan environment can be a serious pain in the backside if you don't own the building. You're assuming companies would want to pay for rights to either run their own conduit (or a shared conduit if the building has a decent wire/phone closet management plan) up to the roof. Pull power and fiber. Get approval to mount and aim equipment after determining they do, indeed, have line of sight. (Ok, you'd likely do this step first.) Then service said equipment.
I did this kind of thing for the financers who purchased E-xpedient.com out of bankruptcy. There was a lot of red-tape and paperwork nonsense to deal with regarding roof rights and rights-of-way inside the buildings. Additionally, I spent far too many cold winter days 20-60 stories above the street replacing equipment that was mounted in difficult to reach spots even if it hadn't been windy/snow covered/ice covered.
Last I heard, they've moved away from the wireless model and instead found less labor intensive ways of bringing connectivity into their lit buidlings... like traditional TDM circuits or longhaul gig-e.
I invite anyone who believes wireless is the answer to all our WAN needs to take a drive down just about any US Interstate and admire all the pretty, vacant AT&T Microwave towers. (In fact, if you really want one, they auction them from time to time...)
Or were you just assuming to use 802.11g across town? :-) -DougDever
# posted by 
: 2/04/2005 5:54 AM
: 2/04/2005 5:54 AM
The deep dark secret of the "fiber glut" is that while 80% of the fiber in the ground may be unlit, that aggregate percentage is because there are some routes with 95% excess capacity and other routes with none.
Capacity issues were relieved when Wave Division Multiplexing hit the scene several years ago. If you have a single pair of fibers, you can now light it with up to 64 OC-48 wavelengths. That's like taking your existing fiber in the ground and suddenly having 64 more times the capacity.
Why wouldn't the cost come down and why would any fiber owner not want to go from 100% capacity to 16% capacity by using this technology if there was a buying public for those services.
GigE services cost practically nothing to light up a segment so why are they still priced so high?
I agree 100% that it is *very* costly to put new fiber in the ground. However, if fiber owners were actually "selling" 100% of their xWDM capacity, they'd have good business models for adding more fiber.
I'll also conjecture that if local governments run the numbers, they'll see that they can create a brand new revenue stream using xWDM for a small town or city that easily supports the cost of installation and becomes the one-stop-shop for all communications services (triple play, video, voice, data) in the region while bringing better Internet connectivity and lower costs for all communication services to the taxpayers.
-- ggunners
Post a Comment
Capacity issues were relieved when Wave Division Multiplexing hit the scene several years ago. If you have a single pair of fibers, you can now light it with up to 64 OC-48 wavelengths. That's like taking your existing fiber in the ground and suddenly having 64 more times the capacity.
Why wouldn't the cost come down and why would any fiber owner not want to go from 100% capacity to 16% capacity by using this technology if there was a buying public for those services.
GigE services cost practically nothing to light up a segment so why are they still priced so high?
I agree 100% that it is *very* costly to put new fiber in the ground. However, if fiber owners were actually "selling" 100% of their xWDM capacity, they'd have good business models for adding more fiber.
I'll also conjecture that if local governments run the numbers, they'll see that they can create a brand new revenue stream using xWDM for a small town or city that easily supports the cost of installation and becomes the one-stop-shop for all communications services (triple play, video, voice, data) in the region while bringing better Internet connectivity and lower costs for all communication services to the taxpayers.
-- ggunners
