AT&T rolls out 7.2 everywhere

[jav6454]

Yes, for $1 a day, basically unlimited broadband mobile at even 1Mbps is an incredible deal.

We used to pay by connection minute for only ~14Kbps a decade ago. (You sure didn't want to take your time on the web back then!)

Don't forget, you paid the phone company by the minute of the call used to connect to the server, plus the cost of Internet service by the dial-up ISP, also by the minute.

On average I recall it was on average $3/minute to use a theoritical max of 14Kb/s and later 28Kb/s (hell, I still remember using baud modems).

Even though it might seem I didn't use these things because of my relative young age vs age of modems, let me clarify I come from a 3rd world country. Albeit Internet was slowly and painfully adopted. 2003 was first 128Kb/s cable lines at $60/month. Just in 2008 we hit 1Mb/s at same $60/month, that's painful.

 

[thetexan]

Using the speedtest.net application on my Verizon Droid, I Just got 2129kbps down and 732kbps up with a ping of 139ms. That beats anything I ever got on AT&T when I had the 3GS.

 

[SpaceKitty]

The day that AT&T made this announcement, I got my best ever speed test using the speedtest.net app:

2670 Kbps down
238 Kbps up

Since then it's always well above 2200 Kbps every test. Before, I was lucky to get above 1800Kbps. This is on my 3GS and 3G iPhones. I'm loving AT&T lately!

 

[dissdnt]

Everywhere.. *groans* I still don't have 3G, 60 miles from nyc, so boo!

 

[Mad Mac]

I just tested and got 300kbps with 4 bars of 3G here in Silicon Valley. AT&T sucks.

Wanna read something funny? Here's an article from 2008 where AT&T claims that they will have 20 megabits in 2009.
http://www.electronista.com/articles/08/05/14/att.20mbps.3g.in.2009/

We're in 2010 and now they claim 7.2? Fool me once...

 

[Mischka07]

I just tested and got 300kbps with 4 bars of 3G here in Silicon Valley. AT&T sucks.

Wanna read something funny? Here's an article from 2008 where AT&T claims that they will have 20 megabits in 2009.
http://www.electronista.com/articles/08/05/14/att.20mbps.3g.in.2009/

We're in 2010 and now they claim 7.2? Fool me once...

Here's something funny from across the Bay in SF:

• 1/11/10 @ 5:50 AM on Montgomery St = 1.9 mb/s down and 230 kb/s up
• 1/11/10 @ 11:30 AM on Montgomery St = 356 kb/s down and 40 kb/s up

I don't know what's more disappointing, the fact that at 5:50 AM you still can't get the same speed as my Tour on VZW at midday, or that the speed drops 500% by lunch time.

 

[thelatinist]

I don't know what's more disappointing, the fact that at 5:50 AM you still can't get the same speed as my Tour on VZW at midday, or that the speed drops 500% by lunch time.

That's not 500%, that's 81%.

 

[alent1234]

Using the speedtest.net application on my Verizon Droid, I Just got 2129kbps down and 732kbps up with a ping of 139ms. That beats anything I ever got on AT&T when I had the 3GS.

these are my normal speedtest results in NYC at 11:43pm or other off times. i use another test app and the results are always 20% - 30% off each other

 

[Mischka07]

That's not 500%, that's 81%.

That's the "glass-is-half-full" math. I'm using the "glass-is-half-empty" method, i.e. 1.9 mb/s is ~5x (or ~500%) faster than 356 kb/s

 

[Vandam500]

That's not 500%, that's 81%.

ROFL!!!!

 

[thelatinist]

That's the "glass-is-half-full" math. I'm using the "glass-is-half-empty" method, i.e. 1.9 mb/s is ~5x (or ~500%) faster than 356 kb/s

Funny, I would have called my method the "right" math and yours "wrong."
Incidentally, 5x as fast is not 500% faster, it's 400% faster!

 

[Mischka07]

Funny, I would have called my method the "right" math and yours "wrong."
Incidentally, 5x as fast is not 500% faster, it's 400% faster!

Wait, 400 X 5 /= 2000? 2000 kb /= 2 mb?

 

[ToroidalZeus]

Incidentally, 5x as fast is not 500% faster, it's 400% faster!

OWNED

btw speed tests are done in kbps(kilo bits), Mbps (mega bits) or kB/s (kilo bytes).

 

[Mischka07]

OWNED

btw speed tests are done in kbps(kilo bits), Mbps (mega bits) or kB/s (kilo bytes).

Learn something new everyday. This is what I get for sleeping in 3rd grade math I guess.

 

[jav6454]

Learn something new everyday. This is what I get for sleeping in 3rd grade math I guess.

Don't worry, everyone gets bandwidth SI wrong at one point. People think Mb/s means Megabyte per second, when in reality bandwidth has traditionally been measured in bits (ie Megabits per second), any type of bandwidth.

Just remember, x.xMb/s != x.xMB/s.

 

[Mischka07]

Don't worry, everyone gets bandwidth SI wrong at one point. People think Mb/s means Megabyte per second, when in reality bandwidth has traditionally been measured in bits (ie Megabits per second), any type of bandwidth.

Just remember, x.xMb/s != x.xMB/s.

Indeed, thanks for the clarification. Is it still safe to say that running at 356 Kbits/s is a "tad" slow for a 3G network?

 

[ToroidalZeus]

Indeed, thanks for the clarification. Is it still safe to say that running at 356 Kbits/s is a "tad" slow for a 3G network?

I just ran a test at 12:45 PST here in Orange County and i got 1401 kilobits per second and just now at 1:15 PST i got 1703 kilobits per second.

 

[thelatinist]

Wait, 400 X 5 /= 2000? 2000 kb /= 2 mb?

When calculating speed increases of decreases, one must calculate the amount of the increase or decrease always as a percentage of the original number. If speed increases from 400 kbps to 2000 kbps, that is an increase of 1600 kbps. 1600 (the increase) is 400% of 400 (the original speed). Therefore the speed has increased 400%.

By the same token, if speed decreases from 2000 kbps to 400 kbps, the decrease is, again, 1600. 1600 is 80% of 2000, so the speed decrease is 80%.

 

[jav6454]

I just ran a test at 12:45 PST here in Orange County and i got 1401 kilobits per second and just now at 1:15 PST i got 1703 kilobits per second.

Remember that in many places bandwidth speed at the Cell Tower depends on network traffic. It might have been that at 12:45 everyone was in Lunch hours and at 1:15 everyone went back to work.

 

[astarling06]

i just used Speedtest and measured 1800kbps in midtown manhattan at 3pm today. AT&T's routing is still screwed up in NYC and once they finish the upgrade it should be a lot faster. my external IP says it's in Kansas

I wonder if somehow it thinks you're in Manhattan, Kansas where I live. Wait, there's no 3G service here... Nevermind.

 

[ToroidalZeus]

Remember that in many places bandwidth speed at the Cell Tower depends on network traffic. It might have been that at 12:45 everyone was in Lunch hours and at 1:15 everyone went back to work.

ya thats what i was implying

I wonder if somehow it thinks you're in Manhattan, Kansas where I live. Wait, there's no 3G service here... Nevermind.

My ip says i am in the bay area (berkeley) which i guess isn't too bad.

 

[jav6454]

My ip says i am in the bay area (berkeley) which i guess isn't too bad.

I honestly believe some of these bad results are responsibility of the SpeedTest app. More than 10 times have I've been placed (after I allowed it to use my location) in Wichita, KA when I am in New Orleans. Obviously, that gives me the slowest possible speed and I get usual 1.4Mb/s there, but if I manually choose New Orleans or Baton Rouge test servers I get average 2.1Mb/s.

Hell, when I was at Miami, it tested me against a server in Japan, JAPAN! Test resulted in 458kb/s down. Obviously a bad test, and once I manually selected a nearby server (didn't show Miami for some odd reason), I got a decent 1.8Mb/s.

SpeedTest, must test there app's location recognition code a bit more.

 

[ToroidalZeus]

The speed test app unfortunately or maybe fortunately does test you again the server where it believes you are from.

I used a internet web site to check my isp location and on wifi it was accurate but with 3g it gave me berkley.

 

[Black94TSi]

I'm having HORRIBLE speeds in my house.

202 kbps Download
79 kbps Upload

This is on 5 bars and 3g

 

[jav6454]

I'm having HORRIBLE speeds in my house.

202 kbps Download
79 kbps Upload

This is on 5 bars and 3g

Test again using your location server or the nearest one, Speedtest app can test you with a faraway server and it might show slow results, which are untrue.