Despite its mundane and nondescriptive name, the new iPad (third generation) is actually the first 4G device from Apple.
I do, however. How, you might ask? Well, it's simple: 4G is fast, and putting a Ferrari logo on your Honda Civic won't make it a racing car. No offense to the Civic; it's great on gas, but it's speed we're talking about here. So let me say this once again: the new iPad is the first 4G device from Apple.
A couple of months ago, I had a chance to roam around San Francisco with the new iPhone 4S from all supported carriers and found that none offered cellular Internet speeds even close to those of the slowest 4G device. Now, I just did the same thing with the new iPad, both AT&T and Verizon versions, and it was quite the opposite story.
But first, let's again go over how I did the testing.
How it was tested
It's necessary to say that cellular Internet speeds vary greatly from one location to another. This is because cell towers are scattered, and when many devices are connected to a tower, each will get a smaller portion of the service. This means that even at the same location, the speed might also change at a different time of day.
That's why it's very hard to have a good sense of how fast a cellular connection is. In an effort to get the most representative samples, I picked three well-populated locations around San Francisco: CNET's headquarters near the Financial District, the Pier 39 area, and Union Square.
At each place, I did the testing three times, but unlike with the iPhone 4S, I tested the new iPad each time about 30 minutes apart from another by driving around those places between each round of testing. While this was more time consuming, (yes, I did go up and down the building a couple of times), it helped make the average number similar to what you'll likely experience, since different times of day are factored in.
For the testing, I used the Speedtest.net mobile app, which is not designed for the iPad's screen, but since it's not a game, that won't be a problem. For each test, I picked the same and closest server for each device.
In addition to the Verizon iPad, I also used an iPhone 4 to test the speed of the hot-spot feature, which allows the iPad to share its cellular 4G connection with other Wi-Fi devices. The AT&T version of the new iPad doesn't offer this feature.
Keep in mind that these tests only evaluate data speeds for these devices in San Francisco and are not designed to be representative of data speeds you'll find in your area. However, they at least should show how the data speeds compare between each carrier's version of the iPad, as well as the speed the Verizon iPad's hot-spot feature offers.
Unlike the iPhone 4S, where the connection speeds changed significantly at the same location, the iPad showed much more consistent speeds at one location. They did change a great deal from one location to another, however.
Here are the changes broken down at different places via three rounds of tests:
I tested the devices at the CNET office on the sixth floor where the iPhone's 3G signal has been notoriously bad. This turned out not to be the case with 4G. Both devices had good signal and offered quite consistent download and upload speeds that averaged 13Mbps down and about 5.5Mbps up for AT&T, and 12.4Mbps down and 7Mbps up for Verizon.
When working as a mobile hot spot, the Verizon iPad also offered very good speed, with the connected iPhone 4 averaging 11Mbps down and 12.9Mbps up. Note how the upload speed was actually slightly faster than the download speed and faster than when the Verizon iPad was not working as a hot spot. This happened consistently in my trials and was rather puzzling.
Union Square was by far the prime spot for AT&T; the new iPad showed unreal cellular 4G speed that averaged 30.4Mbps for downloading and around 24.5Mbps for uploading. These are really fast speeds that are normally only seen in the upper tiers of residential cable connections.
The Verizon iPad, on the other hand, still offered similar speeds to those it did at the CNET office, but just slightly faster, at 15Mbps for download and 9.6Mbps for upload. When working as a hot spot, these were lowered to 10.7Mbps and 7Mbps for download and upload, respectively.
Pier 39 was packed with people when I did my tests, and as expected was the place both devices showed the slowest connections. The AT&T iPad, however, was still able to beat its counterpart by a few megabits per second, by averaging 10.3Mbps down and 8.9Mbps up, compared with 8.7Mbps down and 6.3Mbps up for the Verizon iPad.
When working as a mobile hot spot, the Verizon iPad's speeds were slightly lowered to 8.5Mbps down and 4.6Mbps up.
You are the winner
Unlike the fight between 3G devices, both versions of the iPad were really fast in my testing. While the AT&T version was generally faster, both have such a starting point that from a user's point of view, there's likely no difference at all in terms of speed.
In fact, in my trials, regardless of where I was around town, all Internet applications worked instantly for either of the iPads. There was no lag or buffer time for video streaming. It was really a big difference moving from a 3G device. For this reason, as far as the connection speed is concerned, the winner this time is actually you, the user.
Note that since the new iPad is sold at the full price, it's likely that they are not locked. It's been confirmed that the Verizon version can actually use a generic SIM card, including one from AT&T. It's my guess that you might be able to use a local SIM with the AT&T version, too, which is really a bonus for those who travel internationally.