dslreports logo
 
    All Forums Hot Topics Gallery
spc
uniqs
51
BiggA
Premium Member
join:2005-11-23
Central CT
·Frontier FiberOp..
Asus RT-AC68

1 recommendation

BiggA to elray

Premium Member

to elray

Re: I am not the most up-to-date on upgrading UVerse

The problem is that AT&T built their system to barely keep up what cable is today. Add node splits, SDV, 1ghz plants, the next DOCSIS standard, etc, etc, and U-Verse will look just like DSL does today. The only way that AT&T can compete in the long run is FTTH, and that's what they should be looking to do.
tanzam75
join:2012-07-19

tanzam75

Member

said by BiggA:

The problem is that AT&T built their system to barely keep up what cable is today. Add node splits, SDV, 1ghz plants, the next DOCSIS standard, etc, etc, and U-Verse will look just like DSL does today. The only way that AT&T can compete in the long run is FTTH, and that's what they should be looking to do.

The reason that AT&T didn't adopt FTTH is the same reason that Verizon gave up on FiOS expansion. It just costs too much, relative to the take rate. The headroom of coax gives the cable companies a cost advantage over phone companies moving to FTTP. The cable companies can survive any price war that they choose to fight against FiOS.

Verizon has to pay all of the cost upfront. Before the first customer pays a penny, they have to have the entire neighborhood wired up with fiber. Then, each new subscriber requires a truck roll, to install the customer drop. Truck rolls are expensive.

The cable companies, by contrast, can pay for the upgrades incrementally, as they're needed. Move to DOCSIS 3.1? Mail the customer a new modem. Node split? Transparent to the customer.
elray
join:2000-12-16
Santa Monica, CA

elray to BiggA

Member

to BiggA
I agree that CableCo has done a better job at investing in plant, and they usually do a better job at virtually all aspects of the last-mile compared to AT&T.

But speed isn't the consumer's primary or sole concern.
Google should already know this, but then again, they've proven before that the smartest guys in the room can be pretty dumb.

AT&T can compete without delivering cable speeds.
There is no application yet that requires them - until 4K/8K UHD happens, and that's a long time off.
davidhoffman
Premium Member
join:2009-11-19
Warner Robins, GA

davidhoffman to tanzam75

Premium Member

to tanzam75
What if Verizon had used its rights as a utility provider, where they already provided "regulated" POTS, to take the fiber-optic cable to all the premises on a city block during one short period of time? They could have rolled a few trucks and gotten all the premises on several city blocks in one day. They would have only been going to the exterior wall, and probably could have terminated the fiber-optic cable at or very near the existing NID. That would have reduced the final connection and installation costs for those who signed up for actual service.
c4junk
Premium Member
join:2004-05-08
Orlando, FL

1 edit

c4junk to BiggA

Premium Member

to BiggA
Fiber is expensive in large part due to the labor of splicing but I don't see why Telco could not run coax down property line on the pole from the RT (VRAD), it is easy , in my younger lineman days 2-3 guys could/did pull and hang a few blocks in a day of 50-100 pair copper ca. If they ran coax (lighter than copper ca.) you could put in a run and feed just the houses that had or could be sold VOIP, TV etc and leave the copper POTS alone for now.
As homes went VOIP they would not be allowed back on copper (after a period of trial time) like Verizon did with FTTH.
elray
join:2000-12-16
Santa Monica, CA

elray to davidhoffman

Member

to davidhoffman
said by davidhoffman:

What if Verizon had used its rights as a utility provider, where they already provided "regulated" POTS, to take the fiber-optic cable to all the premises on a city block during one short period of time? They could have rolled a few trucks and gotten all the premises on several city blocks in one day. They would have only been going to the exterior wall, and probably could have terminated the fiber-optic cable at or very near the existing NID. That would have reduced the final connection and installation costs for those who signed up for actual service.

It doesn't happen in "one short period of time".
It doesn't involve "a few" trucks.
It is a massive event.

And that's just to wire to the premise.
Wiring the building is another massive undertaking.

Why should they go through all the effort and expense, if in the end, the customers aren't willing to buy the service?
BiggA
Premium Member
join:2005-11-23
Central CT
·Frontier FiberOp..
Asus RT-AC68

1 recommendation

BiggA to tanzam75

Premium Member

to tanzam75
The reason they didn't do it is because they have impatient, technologically stupid investors who don't understand that in the long run, AT&T will go FTTH, or they will die. At least the landline division. Aside from the stock market, the rational business decision would be FTTH coverage nearly everywhere, and strategically building FTTH into markets with other telcos that haven't upgraded to fiber.

Yes, cable companies have a much easier upgrade path, but the telcos, in the long run, can't afford not to upgrade. The problem is, they have management and investors who just don't get it.

The longer AT&T waits, the more costly it will be in the long run to go to FTTH.

Cable could easily compete with FTTH if they decide they want to, but copper telephone lines will never compete.
elray
join:2000-12-16
Santa Monica, CA

elray

Member

That's simply inaccurate.
AT&T does just fine without upgrading to FTTH, and according to the peanut gallery and Dane Jasper and Google, the cost has come down dramatically, so AT&T is proven quite wise to wait.

It may turn out that indeed, they can choose to never upgrade, especially if the customer base continues to refuse to pay the price for higher speeds. Management and investors do "get it". Insulting their intelligence doesn't change the math.

Mind you, I think AT&T is a horrible company that does nothing right by their customers, and I take great pleasure in permanently disconnecting them at every turn. But the business case is pretty clear, and they're not a charity.
BiggA
Premium Member
join:2005-11-23
Central CT
·Frontier FiberOp..
Asus RT-AC68

BiggA

Premium Member

The guys at Verizon who did FIOS get it. AT&T doesn't. They can't stay stagnant forever, and what they did is create a system that will be forever stagnant. Cable will keep upgrading with D3.1, small nodes, 8-channel bonding, etc, and U-Verse will have nowhere to go. The investors are driving AT&T into the ground with short-term thinking. They need leadership who actually has balls to go in there and do what's right for the company (lay as much fiber down as they physically can until they can serve nearly everyone) in the long term, not just for this quarter or next quarter, or the next 10 quarters, but for the long haul. When consumers realize how much U-Verse sucks, they'll switch back to cable or satellite, and AT&T will have to upgrade again, after having wasted millions on the kludge that is U-Verse.

AT&T should just cut their losses now, and wire their territory with GPON FTTH with IPTV, and then right-size the copper plant by reducing it's size and complexity significantly, scrapping a significant amount of the wiring out there, making as much of it as passive as possible, and using ADSL2+ to serve customers still on copper, as well as extremely rural customers with RDSLAMs where FTTH wouldn't pan out. They could differentiate their fiber and generate more profit by hitting the $70/1gbps price point, as opposed to bottom scraping like they sort of are now.

What would be even bolder is if they built an HSPA+/LTE microcell and public WIFI router into every home router, giving them a huge advantage in the wireless space.

Instead, they are throwing money down the drain by upgrading their copper system and making an already kludgy and overcomplicated system even worse.

The business is clear if you look at a couple of years. It's also very clear that FTTH is the right move for the long term business model.

OTOH, the government should allow them to start dismantling their copper plant in exchange for offering gigabit fiber service to each and every customer in each municipality where they want to dismantle copper.
elray
join:2000-12-16
Santa Monica, CA

elray

Member

Consumers aren't going to realize how much AT&T sucks and switch.
They don't care as much as you or I do. AT&T's speed, services and gizmos are more than adequate.

The "leadership" you refer to doesn't assure profitability, only a race to the bottom with cable, and it ain't gonna happen so long as they are publicly traded.

$70 FTTH doesn't sell.
tanzam75
join:2012-07-19

tanzam75 to BiggA

Member

to BiggA
said by BiggA:

... AT&T will go FTTH, or they will die. At least the landline division.

Business is about maximizing your returns. Sometimes this means strategic retreat from an area where you cannot be competitive. In AT&T's case, this means opting for FTTN and choosing to keep only the customers without high bandwidth requirements.

If the landline division has to shrink, then that might be a better outcome for AT&T than pouring money into it and getting only negligible returns.

The problem is that the cable companies have an unassailable cost advantage all the way up to the 10 Gbps capacity of the coax. Until that point comes, the telcos cannot compete on equal terms with the cable companies.

If you do what Verizon did, then you find yourself paying off an expensive new physical plant, while the cablecos only have to maintain their already-paid-for coax. The cablecos can therefore afford to run customer-retention promotions, which limits your take rate, which makes it even harder to pay for the fiber.
said by BiggA:

The longer AT&T waits, the more costly it will be in the long run to go to FTTH.

The longer they wait, the less costly it will be. First, because the equipment will cost less. Second, because a dollar in the future is worth less than a dollar today.

There will be a window of opportunity that opens up in another decade or so, when the cable companies max out the bandwidth of the coax. At that point, the cablecos will also have to deploy FTTP. And it'll be an even playing field.

AT&T landline will be a smaller division then, because they'll have shed customers. But they will still have the right-of-way to build upon. And there's no real advantage to incumbency -- you're really competing for customers all over again when you deploy fiber. Verizon, Cincinnati Bell, etc. are getting roughly the same take rate for fiber as any cable overbuilder would.

tschmidt
MVM
join:2000-11-12
Milford, NH
·Consolidated Com..
·Republic Wireless
·Hollis Hosting

tschmidt

MVM

said by tanzam75:

The problem is that the cable companies have an unassailable cost advantage all the way up to the 10 Gbps capacity of the coax. Until that point comes, the telcos cannot compete on equal terms with the cable companies.

During the early days of residential broadband I though DSL would win since the cost of rolling it out was lower than what the MSOs had to do to modify HFC cable plant.

However DOCSIS advances have far outstripped DSL. Being copper, both have first-mile have distance limits, but it is much cheaper to add another cable node then a VRAD to improve performance. As mentioned DOCSIS channel bonding, while a kludge from an architectural perspective, is a cost effective way to add capacity. Throw in switching rather than broadcast to conserves cable bandwidth and Cable has a massive advantage over DSL today.

I’m a DSL subscriber about 13,000 feet with 6Mbps service. Unless there is some wonderful new physics/DSP magic that is the end of the line for DSL. That speed works fine for my family today but I’m sure in a few years it will be woefully inadequate.
said by tanzam75:


The longer they wait, the less costly it will be. First, because the equipment will cost less. Second, because a dollar in the future is worth less than a dollar today.

Fiber construction cost is decreasing but the bulk of the cost to roll out FTTP is labor. Waiting does not really help. The problem is first-mile CAPEX is high so ROI takes a long time, which is why it has been such a contentious issue all these years. As long as the Cablecos are able to offer ever increasing speed there is very little marketing advantage for a second player, too much risk and too little payback.

I was excited when Verizon aggressively rolled out FIOS. They could offer a compelling advantage and FIOS OPEX is much lower than copper reducing long term cost. But they, like most other publicly traded companies, seem to have gotten caught up in the profitability of the quarter club and walked away from wired networks all together. That may be good business for Verizon but bad for the rest of us. Long term wired fiber networks are a great investment, but one needs a reasonable time horizon.

Personality I'd love to see a wholesale first-mile network, sort of like what has happened in the electric utilities, but that is unlikely to happen.

/tom
davidhoffman
Premium Member
join:2009-11-19
Warner Robins, GA

davidhoffman to elray

Premium Member

to elray
My bad. After your comment I did the math based on a standard Chicago city block from my childhood. There are approximately 8 blocks per mile with 50 houses per block. A 8 block segment would comprise 400 houses. The alley between the houses holds the telephone poles. I only need to take the fiber from the alley to the exterior wall. The fiber between the alley poles is already done. Assuming 6 hours per shift per day of actual wiring time, that leaves one house wired per minute for one wiring crew. Obviously not enough time. Increase that to a more reasonable 30 minutes, and you need about 32 wiring crews to do 400 houses. That would be 4 crews per block. My statement of a few trucks was inaccurate.

I still think it would be worth it to have a capped off fiber optic cable at every premises or house. If a request for service came in, the technicians would not have to do the big labor of running the fiber from the alley. In its most basic form for telephone and internet, the technician could install the Uverse NID and then connect into the existing telephone line infrastructure of the house. VDSL2 technology should then allow for 100Mbps symmetrical service over the typically under 300 meter in house wiring runs.
BiggA
Premium Member
join:2005-11-23
Central CT

BiggA

Premium Member

If they wanted to, they could do what Verizon did and use MoCA. MoCA is fast enough to support IPTV over coax, so they would have a gentle upgrade path. However, to get to gigabit in-house speeds, you need CAT 5.