CraigBee
join:2004-07-23
Overland Park, KS |
Is 24/3 the end of the line for U-Verse Internet?Are there any plans to offer Internet speeds beyond the 24/3 service offered today? |
|
zed2608
Premium Member
join:2007-09-30
Cleveland, TN |
zed2608
Premium Member
2011-May-30 9:00 am
it probably is the end of line for most users of the service what is your lines stats |
|
 Folk
join:2001-05-09
Fletcher, NC |
to CraigBee
If they could fatten the pipe they most surely would, so that they could provide more TV bandwidth. I could see where they could offer faster Internet-only speeds (no TV) but that would get messy from a logistical standpoint, with someone getting > 24 Mbit Internet speeds and then deciding they now want TV too which would then be impossibile to provide. |
|
| |
to CraigBee
Unless some miracle DSL standard comes out then I can't see how AT&T is going to go any further with their antiquated copper wires. |
|
CraigBee
join:2004-07-23
Overland Park, KS |
to zed2608
I had U-Verse Internet for about 2 years. I remember my line stats would support something around 50/5 speeds. The local cable company upgraded to DOCSIS 3.0 in the Kansas City area and offer 30/5 service, which is what I have now. I'm disappointed that U-Verse probably won't offer faster Internet service. |
|
zed2608
Premium Member
join:2007-09-30
Cleveland, TN |
to CraigBee
well you can get faster speeds then if theyed put in profile
even if they offer it though do you want it if they have caps |
|
 ·T-Mobile
|
to CraigBee
I knew at&t was short sighted by doing FTTN instead of FTTP, I didn't think it would be this quick.
They should have done it right from the beginning like Verizon did and laid fiber to the home. Verizon doesn't have problems with bandwidth and aren't capping their users. In fact, FiOS users can get a 150Mb tier, if they want it (it costs $200 a month). They also don't have overly compressed HD feeds. Their lowest internet tier is 15Mb, while at&t's is 3Mb.
at&t went cheap and it's going to cost them more in the long run.
I think at&t is going to have to admit that FTTN was a bad strategic move and start running fiber. That is if they want to stay competitive with cable.
This is what happens when companies look to short term quick fixes instead of long term investments. Now at&t is going to either have cable eat their lunch or have to spend billions to run fiber to people's homes. |
|
| |
WhyMe420
Premium Member
2011-May-30 3:59 pm
It's like AT&T is on heroin (copper) and each new DSL standard is a hit, until that runs out, and AT&T is still stuck with the addiction. While on the other hand Verizon cut the copper and has been clean for a few years now. |
|
45612019 (banned)
join:2004-02-05
New York, NY |
to CraigBee
It really doesn't matter if they roll out FTTH or not if they continue with these bandwidth capping plans. 24 Mbps or 150 Mbps, the speed is utterly useless when you're effectively limited to using it at an average speed of under 1 Mbps for your billing period. |
|
dave006
join:1999-12-26
Boca Raton, FL |
to CraigBee
said by CraigBee:Are there any plans to offer Internet speeds beyond the 24/3 service offered today?
Nothing published yet but a current single pair is limited to 64 Mbps at the VRAD. Before this was capped, there were often reports of customers syncing above 90 Mbps on a single pair. The upstream Max sync is not published by the 2Wire but the current service profile is 5 Mbps. The base VDSL supports 64 / 8 as a valid profile. Pair bonding will raise the Max Sync rates even higher. Here is a recent post from a user that is syncing above 90 Mbps to the VRAD using a pair bonded install. » Re: Question about upstream 1 vs upstream 0As previous posts indicate, not sure that with caps you would want to move to a higher HSI profile if offered, but the technology supports it. Dave |
|
 trparkyCYA! I'm gone!
Premium Member
join:2000-05-24
Cleveland, OH |
trparky
Premium Member
2011-May-30 10:35 pm
This assumes that everyone has a second pair to be used for pair bonding but that's not always the case, especially in the highly neglected areas of AT&T in which there are users who are lucky they have a dial-tone let alone a second pair. |
|
| |
WhyMe420
Premium Member
2011-May-30 10:50 pm
said by trparky:This assumes that everyone has a second pair to be used for pair bonding but that's not always the case, especially in the highly neglected areas of AT&T in which there are users who are lucky they have a dial-tone let alone a second pair.
Also, Max Rate doesn't mean that they can sync at that speed. FEC overhead needs at least 20% of the Max Rate, but think about it, for every 1 person with a good Max Rate, there are hundreds of people that have borderline Max Rates which either cannot even support the 32/5 or barely support the 25/2. |
|
 maartenaElmo
Premium Member
join:2002-05-10
Orange, CA |
to CraigBee
There are some technologies being developed that could increase the speed and bandwidth.... for instance, if you use 1 pair to receive 24/3, there is no real technical issue to use 2 pairs, and combine 2 connections of 24/3 into 48/6, or something like that.
My downstream is about 60 Mbps down, and 4.5 Mbps up over 1 pair, according to U-Verse realtime, so IN THEORY I should be able to get say.... 2 pairs, each with 50 Mbps down and 4 Mbps, totaling into 100/8, provided that I don't use any other services.
The reality is however, that it is probably not worth it to replace the current technology with technology that will milk copper just a little bit longer. Upload speeds are not going to change without an infrastructure change to fiber.
The sooner AT&T starts to invest in this, the better. DOCSIS 3 is very expensive still, often more then a hundred dollars for the fastest connection. It will get very dangerous for AT&T once the cable companies have started to recoup some of their DOCSIS 3 investment costs, and they can bring the price down to an acceptable level.
I see AT&T making some radical changes in about 2 to 5 years from now. But I think that until then..... 24/3 is probably what it will be. MAYBE they will open up a 30/4 internet connection, that will just fit in a 32/5 profile with voice, for non-tv customers.... but that will probably be it until they roll out fiber. |
|
zed2608
Premium Member
join:2007-09-30
Cleveland, TN |
zed2608
Premium Member
2011-May-31 12:07 am
cross talk limits the use of multiple pairs though |
|
trparkyCYA! I'm gone!
Premium Member
join:2000-05-24
Cleveland, OH |
trparky
Premium Member
2011-May-31 12:23 am
said by zed2608:cross talk limits the use of multiple pairs though That's another thing that they have to fight. Multiple VDSL lines in the same cable bundle plays havoc with the frequency usage on the lines. Like other people in this thread have said already, the faster AT&T wakes up and smells the fiber the better off they will be and the more capable of handling the oncoming DOCSIS 3.0 storm. |
|
| |
Having the 24/3 here would be great. They just upgraded our lines here to finally support the 12/1 this month. They aren't even gonna give us the 1.5 people are getting elsewhere with the 12mb downloads. Sometimes the modem resyncs and only syncs at 9.5 or something like that. |
|
 MOWAA
join:2010-03-25
Fort Lauderdale, FL |
to CraigBee
I have no faith that AT&T will ever achive faster speeds than what we currently have. |
|
| |
to WhyMe420
said by WhyMe420:Also, Max Rate doesn't mean that they can sync at that speed. FEC overhead needs at least 20% of the Max Rate, but think about it, for every 1 person with a good Max Rate, there are hundreds of people that have borderline Max Rates which either cannot even support the 32/5 or barely support the 25/2. That's not correct. The UVRT line stats database allows me true visibility of how AT&T's U-Verse deployment functions in the field. Right now, I have 9000 entries in the UVRT database and they break down as follows: Low profile ( 19/2): 8.6 % 19/2 profile: 5.5 % 25/2 profile: 21.9% 32/5 profile: 64.0% First, nearly 2/3's of AT&T's U-Verse customers have the 32/5 profile, and a large number of those have additional room to go higher. Second, most of the low profile people are Internet-only self-installs with 3600HGV modems that are limited by procedure to a 13/2 profile. Most of their lines could actually handle a lot higher. Third, approximately 1/3 of the 25/2 customers lines could handle 32/5 but they're not on it. So this statement you made that for every person with a good max rate, there's "hundreds" who are borderline or can't support high rates is 100% pure FUD. I know you have some sort of personal vendetta against AT&T and the U-Verse product, but unless you have a 9000-entry database that backs up your assertions, your FUD stops here. Furthermore, I recently logged the following in my database: 1. There are approximately 15 installations in my database that are showing a single-line 36/5 profile that's in use. This may be a test of an upcoming profile that may be deployed. 2. There is one installation showing a single-line 55/8 profile. This may be a stress test of U-Verse's VDSL2 implementation's maximum capability. |
|
 NormanSI gave her time to steal my mind away
MVM
join:2001-02-14
San Jose, CA |
to WhyMe420
{Changed my mind.} |
|
|
| |
to SomeJoe7777
Any way you can tell me what the max profile is for my area? And, what is the max the Motorola 2210-02-1ATT can handle? Both my brother and I had uverse installed last week and both our lines shows max of 13xxx download and it seems odd that since we live in different parts of town that it would show almost exactly the same figures. Also we are only getting 12mb download/ 1mb upload. I read that it's supposed to be 1.5.
thanks |
|
| |
My figures above are only for VDSL-based U-Verse using a 2Wire 3600HGV, 3800HGV-B, 3081HGV, or i3812V gateway.
If you have a Motorola 2210 modem, you are using ADSL2-based U-Verse. None of these figures apply to that.
I don't have data on the ADSL2-based profiles, but a 13M down seems approximately correct, and I believe 12/1 is the correct internet speed. I believe 12/1.5 is only available on VDSL-based U-Verse. |
|
| |
Thanks for the quick reply. I wasn't sure what version we was getting. Not too knowledgeable when it comes to to the different ones and how to tell. They said that the 18mb wasn't available yet and it might be later this year. |
|
| |
to SomeJoe7777
said by SomeJoe7777:-blah-
So this statement you made that for every person with a good max rate, there's "hundreds" who are borderline or can't support high rates is 100% pure FUD.
I know you have some sort of personal vendetta against AT&T and the U-Verse product, but unless you have a 9000-entry database that backs up your assertions, your FUD stops here.
-blah-
Haha! You think you're so smart, implying that your measly 9000-entry database is so special, do you honestly think your very small sample size could even possibly represent 16,500,000 U-verse customers? Haha! I laugh! Try again. I know you have some sort of AT&T-wannabee-employee vendetta going on but hey, who's judging here? I love it when people bring in their personal judgements about people as if that's some sort of viable argument against their claims. Besides, regardless of your little study, that still doesn't change the fact that AT&T is at a dead end. You know just as well as I do that even with things as they are, hell the 32/5 profile is made possible with smoke-and-mirrors (Interleaving.) Maybe if they ramp up the interleaving we can all get 42Mbps profiles with even higher latency! Having a system that operates on the same frequency as AM radio stations, on unshielded copper lines (basically giant antennas,) on an old system that was never even designed for high-bandwidth applications, is inherently flawed. Hell all you have to do is listen to an AM radio station during a thunderstorm, or pass under some high-powered power lines while listening to an AM radio, to know that the potential for interference on those frequencies is quite high. Without the smoke-and-mirrors interleaving, hell U-verse couldn't make it through a single thunderstorm. |
|
trparkyCYA! I'm gone!
Premium Member
join:2000-05-24
Cleveland, OH |
to maartena
said by maartena:DOCSIS 3 is very expensive still, often more then a hundred dollars for the fastest connection. Actually, if they could implement DOCSIS 3.0 across their entire footprint it would be cheaper in the long run. The biggest problem that often cable has is congested nodes. If the system was DOCSIS 3.0 all they would have to do to solve a congested node issue is to just turn up 1 or 2 more downstream channels and the congestion is gone. Beats the crap out of having to split the node. Plus with the advent of SDV this makes it so that they have even more bandwidth on the cable to handle more DOCSIS channels for downstream data. Most DOCSIS 3.0 cable modems can handle up to 8 downstream channels and 4 upstream channels. More than enough room for cable to expand cheaply and fast. Which this comes back to cable being able to beat the crap out of AT&T with DOCSIS 3.0. Quite simply, cable is going to kick their asses. |
|
WhatNow
Premium Member
join:2009-05-06
Charlotte, NC |
WhatNow
Premium Member
2011-May-31 4:25 pm
If Uverse and Fios had not come along do you think anybody would be getting DOCIS 3.0. For many of the posters here Uverse is slow but for most of their customers the 6 or 12 Meg is all they want or need. I agreed the speed world moved faster then most people thought and Uverse will be playing catchup. The other side of the coin is your area might not have Uverse if they had gone FTTP. You may be in the same boat a lot of Verizon customers are with no Fios and none planned so their area. Verizon sold off all the states that were thinly populated. We are seeing the same thing for Uverse if you don't get it by the end of next year you may not get it. |
|
| |
to CraigBee
Yes.
The only way at&t can get faster speeds is full FTTP, like FiOS. Twisted pair has reached it's viable limits. |
|
| |
to WhyMe420
said by WhyMe420:Haha! You think you're so smart, implying that your measly 9000-entry database is so special, do you honestly think your very small sample size could even possibly represent 16,500,000 U-verse customers? Haha! I laugh! Try again. As you wish. Basic statistics can specifically calculate the standard error and confidence intervals from any sample that is much smaller than the population. First, the number of IPTV U-Verse users (the population) is about 3,200,000 according to AT&T's last press release at the end of April 2011, not the 16,500,000 figure you cited. While there are Internet-only VDSL users who add on to that figure, UVRT is primarily used by IPTV users, thus we can consider the population from which the sample is taken to be the same as the number of IPTV users. Thus, the population size here is 3,200,000, and the sample size is 9000. I will not bother with the equations, but instead simply go to the calculator at: » www.custominsight.com/ar ··· ator.aspIn the #3 block towards the bottom, put in 3200000 for the population size and 9000 for the sample size, and click Calculate. You will see that any figures you derive from that sample will be accurate to within 1% with a 95% confidence. So yes, my 9000 samples absolutely represent, within 1% accuracy, all U-Verse customers as a whole. said by WhyMe420:I know you have some sort of AT&T-wannabee-employee vendetta going on but hey, who's judging here? I love it when people bring in their personal judgements about people as if that's some sort of viable argument against their claims. That statement had no intention of disputing your claims. I already did that mathematically. That statement's purpose was to illustrate that you're letting your personal feelings get in the way of facts. It renders you unable to construct a well-formed argument that's backed up with evidence. As for the rest of your comments, I make no claim of any form as to the future roadmap of AT&T's or anyone else's technology, and I have no favoritism or cynicism. My post was simply to refute your FUD, which has been done. |
|
| |
WhyMe420
Premium Member
2011-May-31 6:53 pm
said by SomeJoe7777:said by WhyMe420:Haha! You think you're so smart, implying that your measly 9000-entry database is so special, do you honestly think your very small sample size could even possibly represent 16,500,000 U-verse customers? Haha! I laugh! Try again. As you wish. Basic statistics can specifically calculate the standard error and confidence intervals from any sample that is much smaller than the population. First, the number of IPTV U-Verse users (the population) is about 3,200,000 according to AT&T's last press release at the end of April 2011, not the 16,500,000 figure you cited. While there are Internet-only VDSL users who add on to that figure, UVRT is primarily used by IPTV users, thus we can consider the population from which the sample is taken to be the same as the number of IPTV users. Thus, the population size here is 3,200,000, and the sample size is 9000. So yes, my 9000 samples absolutely represent, within 1% accuracy, all U-Verse customers as a whole. You assume that of these 9000 entries, there are 9000 unique users. Did you even bother to take into account that the same person could use your program twice? Also, your entries are not a simple random sample, but rather from a specific demographic i.e. people who have seen your posts and downloaded your program. Admittedly, my figures were off, but gathering from this: » AT&T's Stankey: U-verse Build Virtually Over [111] comments article, U-verse will reach 30-million homes by the end of this year, thus, a sample of 9000 is (and will be) an extremely small sample size, less than 1 percent. Even with your 3 million figure, it is still less than 1 percent. With such a small sample size, you would at least need another group of samples to compare, between the two. |
|
| |
said by WhyMe420:You assume that of these 9000 entries, there are 9000 unique users. Did you even bother to take into account that the same person could use your program twice? Also, your entries are not a simple random sample, but rather from a specific demographic i.e. people who have seen your posts and downloaded your program. Admittedly, my figures were off, but gathering from this: »AT&T's Stankey: U-verse Build Virtually Over [111] comments article, U-verse will reach 30-million homes by the end of this year, thus, a sample of 9000 is (and will be) an extremely small sample size, less than 1 percent. Even with your 3 million figure, it is still less than 1 percent. With such a small sample size, you would at least need another group of samples to compare, between the two. UVRT specifically generates a random but unique ID at the client end, and uses this ID to ensure that everyone's line statistics are only submitted to the database once. If their line statistics change, the record is updated, not resubmitted. Thus, 9000 samples represents 9000 unique users, not repeated submissions. The 30-million figure cited by Stankey is number of homes passed, not number of installations. (In other words, the number of homes that could get U-Verse service, not the number of homes that have it.) The demographic of those who use UVRT is primarily those users who frequent the U-Verse forums, which is where UVRT is discovered by them. Overall, the users who come to the forums to specifically use UVRT represent a segment of U-Verse customers who have more problems than the ordinary user (that's why they're coming to a forum and seeking help). Thus, the argument can be made that my sample is actually skewed with worse service than the general population, therefore my figure of 64% on 32/5 is actually low. I have already shown that 9000 samples, even though it represents a fraction of the installed customers, is more than enough to draw very meaningful and accurate statistics. Another group is not at all required. All that another group would highlight is whatever non-random biases remain in the sample, which as we just discussed, would tend towards showing somewhat better service than what my figures currently indicate. |
|
| |
WhyMe420
Premium Member
2011-May-31 7:21 pm
That's why I said it will reach 30 million users. Not all people that have U-verse even know what forums are. So yes, a lot of people visit forums, because they have problems, but these people are also more likely to get their problems fixed than others, who don't even know what stats are, yet are set up with crummy installs. Even with 9000 unique users, it still (as I keep saying) represents a ridiculously small portion of the total user base. |
|
·
