Computerworld’s 3G network smackdown: AT&T beats Verizon, Sprint

“Third-generation wireless technology, or 3G, started being rolled out in about 2001, and offers a wide range of services (from music downloads to mobile GPS) and greater efficiency than previous standards. In the U.S., three providers offer 3G service: AT&T, Verizon Wireless and Sprint. The fourth nationwide carrier, T-Mobile, says it will roll out its 3G network starting later this year,” Brian Nadel reports for Computerworld.

“Although 3G can be accessed using many cell phones, road warriors who need to get serious work done will likely want to use their laptops,” Nadel reports. “To gauge the speed and reliability of these three wireless data networks, I used my ThinkPad X300 to collect nearly 500 data points at eight different places in New York, New Jersey and Connecticut, within a 50-mile radius of midtown Manhattan’s urban canyons,.”

“I timed how long it took to establish a connection with each network, followed by speed tests. Using Alken’s bandwidth meter, I was able to gauge download and upload speeds as well as how long it took to load that vendor’s home page,” Nadel reports. “All speed readings — connection time, the Alken speed tests and page loading times — were repeated five times and averaged.”

“AT&T’s network left its competitors in the digital dust, with average download speeds of 755Kbit/sec. and average upload speeds of 484Kbit/sec. The peak download speed was 1.6Mbit/sec. It connected in just 3.0 seconds and loaded the test Web page in 0.228 seconds,” Nadel reports.

“The slowest of the three cellular data networks, the Sprint network delivered on average 494Kbit/sec. download speeds to my notebook. That’s roughly a third slower than AT&T’s network. Its average upload speed was 294Kbit/sec. and the network’s top speed was 1.2Mbit/sec. It connected in 3.7 seconds,” Nadel reports. “Verizon’s BroadBand Connect network averaged 592Kbit/sec. for downloads and upload 232Kbit/sec. for uploads. Peak speed was 1.3Mbit/sec. On the downside, it took a relatively pokey 5.6 seconds to connect.”

Much, much more in the full article here.


  1. 500 data points, 8 different spots around NYC, 5 tests per spot, per device. I would say, this is a decent test. One could argue, 50 different spots would better neutralise any advantages one network might have over another in a particular random spot, but if trends were consisted across these spots, it would be difficult to see how the results could be any different with a larger sample base.

    Your BlackBerry has a really, really crappy browser, which in my experience was always the biggest bottleneck on the device (and not the network speed). Palm is a real computer, with real software, and a Treo probably has more efficient browser (not that any of them are able to hold a candle to the Safari on the iPhone).

  2. Isn’t ATT’s 3G network rather new compared to the others? If so, might it be that they have less utilization since the existing customer base hasn’t saturated it yet and newer 3G subscribers haven’t migrated to it yet?

  3. I doubt they’d say ‘impeccable service’, not even in their marketing. These guys are always careful to say ‘Best customer service according to…’ then some survey, or some market research company, or Consumer Report, JDPower, or whoever picked them in a specific category over their competitor (for example, best customer service for data plans $40 or over). When you are the best, that only means you have edged out your competition; it still doesn’t say anything about the absolute quality of your service.

    There are so many market research companies, surveys and studies, almost any company can claim to be the best and back it up with scores from some of these surveys. Car makers are best example – with so many car magazines, they strategically divide their ‘Best of…’ awards among the car makers, so that every car maker can claim the best in class for every one of their models.

    We have no way of finding who is really the best other than trying for ourselves, as much as it can be done…

  4. Spent the weekend on an island in N.H. – the iPhones had web access while our iBook did not. 🙁
    Played Bridge with Real People instead. ” width=”19″ height=”19″ alt=”smile” style=”border:0;” />

  5. @Predrag

    I tethered the blackberry to a laptop to get internet and you could tell a huge difference so… it’s not just from having a shitty browser. I must say atleast my blackberry doesn’t freeze and crash like the palm treo’s I’ve had. from the 600 to the 755 my blackberry, I cant stand so I’ll be getting the new iphone and putting it on the company plan and using the blackberry as a backup for someone else. reception kinda sucks on it compared to my friends iphone. Cant wait to finally have a damn cool phone that works.

  6. @ Predrag

    Mention a review in Consumer Reports in your advertising, and Consumer Reports will sue you.

    That’s why you more often see something like “according to a major consumer magazine….”

  7. Mention a review in Consumer Reports in your advertising, and Consumer Reports will sue you.

    How can that be? If they reviewed your product in their magazine, how in the world can they sue you for summarizing their results? That’s ridiculous!

Reader Feedback

This site uses Akismet to reduce spam. Learn how your comment data is processed.