iPad 3G speed shootout: Telstra vs Optus vs Vodafone vs 3

David Flynn
28 May 2010, 9:24 AM


We put Telstra, Optus, Vodafone and 3 to the test in Australia’s first comparison of iPad 3G providers – and if you expect Next G to be fastest, the results will surprise you!


Apple’s iPad is finally here, and lined up behind it are Australia’s four mobile carriers – each spruiking their own networks with temptingly affordable prepaid 3G plans. But how do they stack up?

With the assistance of each carrier in supplying their respective microSIM starter kits hot off the press, we grabbed an iPad 3G and hit the streets of Sydney to put the telcos toe to toe in Australia’s first iPad 3G mobile broadband shootout.

To typify real-world conditions we tested the iPad 3G on 3, Optus, Telstra and Vodafone, using several Sydney CBD locations ranging from a sidewalk cafe to a meeting room inside a commercial office block. The tests were conducted mid-week in both the morning and afternoon peak usage hours.

The nature of radio networks means that your mileage will vary –  we all have different usage patterns, and tend to frequent different areas with equally varying signal strengths. Our tests aren’t intended to be the definitive guide to network coverage or reliability – they’re more of an indication of general speed and performance.

Across all of our tests, Telstra shone as the most consistent performer – even though Next G wasn’t always the fastest.

One set of cafe tests was Next G deliver an average speed of 2.1Mbps, with the  lowest and highest rates barely 1Mbps either side, while Optus and Vodafone kicked up as high as 3.2Mbps and 3.4Mbps respectively.



Several tests saw Vodafone best Next G for top speed, although Telstra was the most reliable

But Optus, Vodafone and 3 (which peaked at 2.3Mbps) all showed vast variances throughout those same tests.

Optus ranged from a lowest average of 1.3Mbps to a median of 2.2Mbps and up to 3.2Mps; 3 spanned from barely 1Mbps to 2.3Mbps; and our Vodafone connection was a roller-coaster ride which could soar to record highs of 3.4Mbps but then plummet to 250-500Kbps.

(Upload speeds were unimpressive across the board, but unsurprisingly fell into the same pattern as downlinks – Telstra pulled a slow but steady 260Kbps, Optus ranged from 50Kbps to 250Kbps and Vodafone lurched from 45Kbps to 342Kpbs.)



In indoor and outdoor tests around the Sydney CBD, Telstra returned consistently solid results

The indoor tests tend to favour Telstra because Next G uses a lower frequency of 850MHz compared to the 2100MHz of Optus, Vodafone and 3 – and the longer wavelengths of lower frequencies provide better punch and signal soak through walls into buildings and even down through underground carparks.

This played out as expected, with Telstra recording noticeably higher average speeds (and more consistent speeds) of around 2.5Mbps compared to the 2100MHz 3G networks.



3 didn't bend the needle like Optus and Vodafone but it also lacked their oft-insane lows

3 managed a solid 1.3Mbps with little variance during testing, and once again Optus and Vodafone sent the needle bouncing all over the dial. One session saw Vodafone hammering away at 3.8Mbps, even though subsequent tests sat closer to 2Mbps; Optus went from 240Kbps to 2.5Mbps within the space of five minutes.



While Optus hit many high notes during testing, like Vodafone these were offset by sudden drops in throughput


So if you’re looking for speed alone, you might be surprised to see that Optus and Vodafone can deliver some amazing results – but those soaring peaks are usually balanced by deep valleys.

3 turned out to be more like a budget version of Telstra, albeit with a far more modest network map, but was still more predictable and consistent within a given range than Optus and Vodafone.

And while Telstra didn’t exhibit the raw speed with which Next G made its name, it still emerged as the most solid and most reliable choice for your prepaid iPad dollar.


Post your comment



Comments

RSS feed Email alert

Daviddth (New user):

Great stuff David. I hope this helps people make a decision on which carrier to go with. My Telstra Sim arrives today, and the Optus one shold be here in time for next weeks trip away to get some good travelling results on the way to Melbourne

28 May 2010, 7:42 AM (2 years ago)report abuse Send to a friend reply

Gatsby (New user):

A great idea to test this however Speedtest is rubbish. In order to get a decent and accurate test you need to time the speed it takes to download a file from the Carrier's core network to the device itself. Speedtest is hosted by Optus - do you think thats a fair and equitable test?

28 May 2010, 9:08 AM (2 years ago)report abuse Send to a friend reply

Gatsby (New user):

A great idea to test across carriers, however Speedtest is rubbish and not accurate. In order co come up with an accurate result, you need to measure the time it takes for a file to be downloaded from the carriers core network to the device itself. Speedtest is hosted by Optus so do you think that this is a fair a equitable test?

28 May 2010, 9:08 AM (2 years ago)report abuse Send to a friend reply

Pauly (New user):

Several locations in the CBD? Where bugger all people actually live and will use wifi instead.
Seems like a bit of a waste of an article, why not do it in several locations through the Metro Area, close to CDB, edge of metro area? would seem like a far more useful test for your readers

28 May 2010, 11:53 AM (2 years ago)report abuse Send to a friend reply

petert (Advanced Forumologist):

Quoting Pauly:
Seems like a bit of a waste of an article, why not do it in several locations through the Metro Area, close to CDB, edge of metro area?

I believe that the article is a good start to understanding how the iPad will work with telecomm's various groups in major population centres. Nonetheless, if you really want to test how it works geographically, then go bush. See how well Vodafone works, if at all, outside a major metro area :-)




28 May 2010, 12:22 PM (2 years ago)report abuse Send to a friend reply

djsflynn (APC staff):

Pauly, we had VERY limited time - not even a full day, more like five hours - and the sites I chose were ones I've used in previous mobile broadband tests so they're familiar turf that's worked well enough in the past to provide a yardstick of comparison.

I'd love to spend days hitting strategically-selected cafes around all points, I really would - but in this instance time was a luxury.

28 May 2010, 1:04 PM (2 years ago)report abuse Send to a friend reply

anonymous user Anonymous user