AT&T Wins Gizmodo 3G Bandwidth Test

ink writes "Gizmodo has completed a 12-city test of 3G cellular bandwidth speed. Verizon won four of the twelve, however AT&T scored higher with six: 'Let's get this straight right away: We didn't test dropped voice calls, we didn't test customer service, and we didn't test map coverage by wandering around in the boonies. We tested the ability of the networks to deliver 3G data in and around cities, including both concrete canyons and picket-fenced 'burbs. And while every 3G network gave us troubles on occasion, AT&T's wasn't measurably more or less reliable than Verizon's.'"

Comments

Look at the latency

[mdm-adph]

From what I can see, the latency on the Verizon lines is much better. That's more important to me, at least, considering the amount of VoiP I do on my cell phone.

I mean, uh, browsing I do on mobile networks.

Re:Look at the latency

[Otterley]

I agree, and this is why I have nothing but contempt for typical "best provider performance" conclusions that are driven solely by single-connection TCP transfer tests (e.g. speedtest.net).

In most cases, latency matters more than bandwidth (where bandwidth is roughly the same within an order of magnitude or so). This is why there's a very strong correlation between the provider that had the lowest measured latency and the provider that had the lowest page retrieval time. In the end, real-world page loading is precisely what we use smartphones for, and so we need to know how that application performs, instead of what raw transfer rates are.

I still think the Gizmodo tests are deficient, though, as they are unclear as to whether they repeated the tests at regular intervals over a 24-hour period. Network congestion varies throughout the day, and at any given moment one path may be more congested than another. A valid test, IMO, would take the average (or median) of each metric over a 24-hour period (or even longer, covering both a weekday and a weekend, since usage varies among them).

Re:Look at the latency

[sonnejw0]

Engadget did a similar test a year-ish ago. AT&T was leaps and bounds faster than Verizon's 3G, in fact AT&T's 2G tested as fast as Verizon's 3G, and the latency was measurably lower on AT&T as well, at least where they tested the four providers: [engadget]

Re:Look at the latency

[ircmaxell]

I'd be curious to see the results of a test conducted as follows:

Setup a test rig that automatically tests download speed and latency from a few different servers (globally dispersed) every 30 seconds. Have it geotag each result.

Drive around each of the cities (and metro area surrounding them), and gather the data over a few days for each city.

Then, compute a few different metrics. First, for each carrier, generate a heat map of download speeds (It would wind up looking something like Verizon's 3g Map, but with different colors denoting speed ranges). If there are multiple readings in a given area (Say 100m^2), average them.

Second, figure out the peak speed, minimum speed, average speed (Both mean and RMS) as well as the standard deviation for each carrier.

There's so much focus on peak speed these days (well, at least that's what gets the press), that I think people forget that what matters is the average experience. It's like with stereo amplifiers: Peak wattage tells you nothing about overall power output, RMS wattage (Root Mean Square) tells you about power output...

Just my idea. Feel free to steal or rip it to shreds...

Re:Look at the latency

[Arthur Grumbine]

Second, figure out the peak speed, minimum speed, average speed (Both mean and RMS)...

I always thought RMS and mean were synonymous.

Better Sample Size

[rliden]

While I care about speed, I care more about coverage and reliability. Did the testers cherry pick AT&T friendly locations? I know when I went to Washington D.C. earlier this year that not only was my 3G coverage under AT&T spotty, my ability to simply connect to AT&T and make calls was also poor. It would be a much more interesting test if this was performed in every city over a certain population size. In my opinion 12 cities does not make a good test.

I'll admit I'm biased though. I've been an AT&T mobile customer for a bit over 2 years now and I don't like their service or support.

Re:Better Sample Size

[Darkness404]

Exactly, and also they completely left out the midwest. Well, thanks Gizmodo, if I move to the coasts I'll know what service to get, but since there weren't any tests done even close to where I live the data is totally useless.

They also failed to give a price to service ratio, that would have made things a lot more fair. Yeah, you might get a faster network but if the price difference is $15 a month, many people might reconsider.

This test the way it is, is akin to someone comparing an Intel Atom to a Pentium 4 to a Core i7 based on pure speed and saying that the Core i7 is the better bet, all the while ignoring the fact that a Pentium 4 box is much cheaper and an Intel Atom CPU is going to give you better battery life.

They all suck.

[aussersterne]

Having been through Verizon, T-Mobile, and AT&T over the years (never tried Sprint), my conclusion is they're all way oversold with shitty reliability and doubly shitty and uneven customer service. Typical megacorporations to whom any individual customer matters NOT AT ALL.

Piss off, 3G

[girlintraining]

3G is a joke, but it's not a funny one. The FCC promised that we'd start to see high speed wireless internet now that the spectrum's been auctioned off. But like everything else, they seem to have lied -- shoving costs down the consumer's throat in the middle of a recession, raking in the money with a smile from the auctions... Everything about the so-called digital transition was a scam. Price fixing of LCD TV prices, running out of converter boxes -- and charging twice as much as they were worth in the store to soak up the free money those vouchers gave them... hmph.

Where's the alternatives here? They all have bandwidth caps. None of them are investing in the backhaul infrastructure. The network coverage is a joke, the handsets have disabled tethering, locked in the search engines... I mean, hell -- a pringles can and a wifi card does better than every other solution we have here in the United States for mobile internet. What the hell happened?

Re:It's not the 12 that counts, it's the rest...

[aristotle-dude]

I travel all over (in the US). I usually consult in city that are not the major metropolitan centers (in the US). If you are NOT in the major metros (in the US), Verizon wins 9 times out of 10. Once I got back on Verizon, covered (in the US).

If you don't travel, get the best signal provider in your area. If you travel (only in the US), Verizon is best.

There, I fixed that for you. A CDMA phone is going to be useless outside of the US unless you have a "world" phone which also includes a SIM and GSM radio for the rest of the world outside of the US. If you travel internationally, you are better off with a GSM/HSPA phone.

Wha?

[Itninja]

AT&T's wasn't measurably more or less reliable than Verizon's

So how is this a 'win' exactly? Sounds more like a tie to me.

Re:Wha?

[Bill, Shooter of Bul]

Title of article "AT&T Wins Gizmodo 3G Bandwidth Test".
Not Article Title "AT&T Wins Gizmodo 3G Reliability Test".

Notice any difference in the two?

That isn't the problem with AT&T

[dynamo]

Someone should go do a test of the dropped call quantity and voice quality when in these same areas. That is where AT&T is so difficult to have to use as a primary phone line. The data service is actually much more reliable, and ironically makes Skype average much higher in quality / reliability from the same phone in the same place.

At least, in my experience.

Honest question

[dnaumov]

Why do you, Americans, put up with your mobile operators specifically disabling features (like tethering or bluetooth) on phones being sold via contracts? Here where I live (Finland), such action would be considered blatant fraud, because the operator would be advertising a specific phone model, while in reality, the phone model being advertised in reality has more or better features than the one sold to you under it's name by the operator.

Re:Honest question

[Attila Dimedici]

The thing that Europeans always forget is that the U.S. is more like the E.U. than it is like Finland. Actually, many Americans make the same mistake. When Europe becomes a single cellular market the way that the U.S. is, we will be able to compare the business practices of the providers.

Re:What about sustained transfers?

[AliasMarlowe]

I find I can get a nice burst for the first couple of megabytes then Im throttled pretty badly. Id like to know which carrier doesnt do this. It doesnt look related to reception.

Short answer: it seems that all US carriers do this, either because (i) they underestimated demand and under-invested in infrastructure, or (ii) because they can maximize their revenues while minimizing their costs, and the customers are trapped into long term contracts.
This sort of throttling by carriers is unheard-of in more advanced countries, such as Finland or Sweden. None of the carriers do that here; if they tried it, they'd have no customers left within a month or two (terminating a contract is trivial, and does not entail penalties). There are no usage caps on 3G either - unlimited actually does mean unlimited.

Re:What about sustained transfers?

[yoyhed]

Also, try going to Gizmodo without Adblock on. Note ALL the AT&T "Better 3G Experience" ads - especially the one with fucking Luke Wilson that is masquerading as an ACTUAL FUCKING STORY. Yeah, I'm sure this is a completely unbiased test. And 3G is not slow for me, on Verizon I get sustained 150 kiloBYTE/s downloads. That was my cable internet speed a few years ago, but now I have it on my phone!

Because any state of affairs

[aussersterne]

in which we do not agree to be raped every hour on the hour by corporations is in fact socialism, and socialism we leave for the "evildoers."

Same reason we prefer the poor to starve and the sick to have no medical care.

Re:What about sustained transfers?

[babblefrog]

Actually, it's surprising that nobody else has noticed that this discussion is really about aviation, not telecommunications. Slashdotters are really slipping.

AT&T == Cherry Picker

[erroneus]

Here's the thing -- everyone knows Verizon has a much larger square-mile coverage than AT&T. Wide coverage is important to many people for many reasons. But to accomplish this, you must do less cherry picking. AT&T on the other hand, does not concern itself with wide coverage, but instead focuses on the investments that yield the highest return.

You can decide for yourself which strategy is better. As a customer, I prefer a company that favors customer satisfaction. That said, I am neither an AT&T customer nor a Verizon customer. I am a T-Mobile customer largely because it was convenient at the time and presently have no incentive to change... Sprint burned their bridge with me... I hated that I had to become angry and threaten to leave in order to get anything done with my account. While I am sure there is no shortage of horror stories about T-Mobile, I haven't had any myself.

Re:I use AT&T,

[mjwx]

Yes, I like the iPhone. It's the first smartphone (having used Palm for a long time, then Blackberry for a while) that I really use to its potential

That's easy to do when the potential of a device is so limited.

I on the other hand will say that I will never use the full potential of my Android phone and that is not through lack of trying. I have a phone that sync's my Gmail, ISP mail and work mail into separate programs that can be open at the same time each with different notification settings, can be turned on and off independent of each other and programmed to operate on different schedules (I.E. I want my personal mail on Saturday, not my work mail).

Saying that you use the full potential of anything is not good, this means that you will reach the limitations of the device leaving no room for growth. The Iphone is a closed ecosystem, a completely controlled environment where there is little variation, thus little change (mutation, the iphone is akin to monoculture farming, which tends to degrades the species over time) where as Android is an entirely open ecosystem where new variables can enter easily leading to new functions or enhancements of existing ones.