Very slow

Hello, autismdefensedotcom!

Thanks for posting. I'm sorry to hear about your internet speed issues. We would be happy to look into this for you, so click here to send our U-verse experts a private message.

In your message, please include your name, phone number, email address, and the best time to reach you. Keep an eye on the little blue envelope icon in the top right corner of your screen for a response.

In the meantime, please feel free to message me with any other questions or concerns.

-Mariana

---

@julywashere wrote:
Your maximum upload speed might be limited by the test server's upload speed. If you live in a densely populated area, that server can also be overwhelmed with test requests and be limited to a certain upload speed. Here in Dallas, I've noticed that in most cases the preferred server for mobile phones is typically the same, and it doesn't always give me an accurate reading, while on PC the preferred server is selected based on ping (in my case, also the same server.) which doesn't always give me an accurate reading. Did you try various servers? How about more distant servers?
Xfinity's Tests are powered by Ookla, but the servers designated are meant to handle faster speeds. (Which is why they're ideal for fibre speed tests.)

---

The server that SpeedTest.Net chooses can definately make a difference.  Last night I did a speedtest.net and it chose a server in a small mountain community 60 miles to the west and it told me my link was 4 Mbps/1.6 Mbps.  I asked for a different server (this time it picked one with a larger company 50 miles to the north) and got 12.7 Mbps/1.65 Mbps.

FTR, this is my speed tonight (on a 12/1.5 HSIA account):

I've tried serveral times, but I can't get it to take me back to the site in the Georgia mountains.

I'm on 18Mbps with 2 recording in progress. I thought the first test was wrong, restarted  the computer and hardwired, only got about .21 difference. 

Here's a speedtest of the server determined by ping.

Here's a speedtest of my preferred server

You are getting the 18Mbps you're paying for as you stated yourself, you saw speeds up to 22Mbps. You are testing the connection from a server on the public internet to your local machine, not local network. I'll explain how speedtests work, and perhaps you'll get a better understanding of what they demonstrate. 

(This is specifically how Speedtest.net handles their Speedtests, but many different server handle them similarly.) 

1. When you request a speed test, your local device sends a request to the speed test server.
2. The speedtest server responds by sending very small binary files to your device. (This process also determines the ping.)
3. The server measures the transfer speeds of these small binary files to gain an idea of your connection speed. Based of this result, it determine how much data to send to your device. (You won't send hundreds of MB for a speedtest run on dial-up or 2nd generation wireless networks, it'll take too long. You don't send hundreds of KB to a Gigabit Fibre connection, it'll go to fast for the server to accurately determine how fast your speed is.) The idea is for your connection to download the data in about 10 seconds, give or take a few for fluctuations.
4. The download is stored in the temporary files of your computer (cache). As such, to prevent the cache from recognizing that the file has been downloaded, a random string of data is attached to each download to differentiate it. 
5. Two to four HTTP threads are used to saturate the connection to get a truly accurate measurement. (This is to ensure that the data-pipe to your device is full.) 
6. There are about 30 samples sent per second. All the samples collected are aggregated and divided into 20 sections (each containing a small percentage -- 5%). From these 20 sections the fastest and slowest sections are discarded to avoid the results from protocol overhead, cpu bursts, buffering between layers, etc. (They take off the slowest and fastest due to certain limitations that may occur to give you a more accurate result.)
7. They grab the remaining sections, average the numbers and display a final result. 

All in all, this tells you that the speed that is being tested, isn't the speed to your local network, it's the speed to your device. Speedtest.net on desktop tends to use the server who responds the fastest (lower ping.) While mobile Speedtest.net uses the closest server. (Which sometimes may be the same server, especially if you don't have many servers in your area.) The reason that different areas offer different speedtests has to do with complex factors. One such factor is the maximum upload connection speed of the server. 

For example, hypothetically if you're capable of downloading at 20Mbits per second, the server has the ability to upload up to 100Mbits per second, there are only two people currently downloading, you can get the max read of 20Mbits per second, because that's all AT&T will let through.  If 20 people are downloading from that server, however this leaves on average about 5Mbits per second, per user. AT&T can allow up to 15Mbits more but the server is only capable of delivering the 5Mbits per second. Now this isn't exactly an accurate representation, there are plenty of factors that can have an effect, but this is to give you an idea of why different servers may offer different speeds. 

It also has to do with local availability. Dallas' network infrastructure may allow for multi-Gigabit connections, while middle of nowhere in the mountains may be limited to only a few hundred Mbits. (Cabling limitations, and distance still apply.) 

---edit---

To reiterate, the maximum speed is determined by the lowest speed. If you can only download 20Mbps and the web-server can only provide 5Mbps, you're limited to 5Mbps. Think of a small tube draining water into a bigger tube, there's no way the bigger tube will push it's maximum capacity of water because the small tube can't provide that much water. 

In contrast (which is the case with more servers, especially those transfering large files) if the server can push 1Gbps, but your connection can only handle 20Mbps, you'll be limited to 20. In this case, think of a large tube dumping into a funnel. 

We just got Uverse a couple of weeks ago, and it is MUCH slower and less reliable than our old DSL connection with a Netgear wifi router.

We have internet and phone, not tv. I have chatted with three techs and this is probably not going to improve. The speed test, at 3AM, showed 7.4Mbps download and 1.5 Mbps upload. That stinks. I would like to get out of my contract. I have a suspicion that AT&T wants our TV, and if we gave up and let them be our monopoly of choice, we would get the better modem. I hope I am wrong about that.

On my Mac I get decent coverage, but my other devices (iPad, 2 iPhones, Kindle Paperwhite, Kindle, and iPod), wifi is intermittent, not showing up, or asking repeatedly for the password. I PROFOUNDLY regret changing to U-verse.

Even when my devices are in the same room with the modem, they have intermittent wifi. Even in a hotel I get better wifi!

@ElleCoyote , if you're getting 7 Mbps/1.5 Mbps, I have trouble believing that is slower than your DSL service was before.

You didn't mention what service level you purchased.

You didn't mention if you got poor speeds both wired and wireless, or only wireless.

One option you could give consideration is to purchase your own wireless router or wireless access point and use it to either replace or supliment the wireless coverage from your AT&T Residential Gateway.  Many do this and have great result.  It allows you to select the wireless features that are important to you (e.g. dual-band, MIMO, etc.)

If you could tell us what brand/model Residential Gateway you have, and even capture the Broadband statistics from http://192.168.1.254/ and paste them here would be helpful to us advising you.

---

@ElleCoyote wrote:

We just got Uverse a couple of weeks ago, and it is MUCH slower and less reliable than our old DSL connection with a Netgear wifi router.

 

We have internet and phone, not tv. I have chatted with three techs and this is probably not going to improve. The speed test, at 3AM, showed 7.4Mbps download and 1.5 Mbps upload. That stinks. I would like to get out of my contract. I have a suspicion that AT&T wants our TV, and if we gave up and let them be our monopoly of choice, we would get the better modem. I hope I am wrong about that.

 

On my Mac I get decent coverage, but my other devices (iPad, 2 iPhones, Kindle Paperwhite, Kindle, and iPod), wifi is intermittent, not showing up, or asking repeatedly for the password. I PROFOUNDLY regret changing to U-verse.

 

Even when my devices are in the same room with the modem, they have intermittent wifi. Even in a hotel I get better wifi!

---

See SomeJoe's great wireless interference thread:

http://forums.att.com/t5/Features-and-How-To/How-to-Minimize-Wireless-Interference/m-p/2303003#M1138

Also see the SJ post:

Please Read for a Potential Solution

For many people, the DHCP problems with the 2Wire routers can be solved by configuring the RG as follows:

1. Go to http://192.168.1.254/xslt?PAGE=C_2_1 which is the RG's wireless configuration page.

2. Change the setting "Wi-Fi Protected Setup" to Disabled.

3. Change the setting "Authentication Type" to WPA2-PSK (AES).

4. Click the "Save" button at the bottom.

5. This procedure may ask for your RG's password. The default password for your RG is printed on a white sticker on the side of the RG (3600HGV, 3800HGV-B, and 3801HGV) or on a white sticker on the backup battery (i3812V).

Disabling WPS should help with the DHCP timeout issue and general wireless connection problems.

Setting authentication specifically to WPA2-PSK/AES improves connection reliability on Macs and other Apple products. 😉

Chris
__________________________________________________________

Please NO SD stretch-o-vision or 480 SD HD Channels
Need Help? PM ATT Uverse Care (all service problems)
ATT Customer Care(all other problems)
Your Results May Vary, In My Humble Opinion
I Call It Like I See It, Simply a U-verse user, nothing more