Samsung Galaxy S III 4G.
Android's latest speed demon pushes Telstra's Next G to the limits.
Plenty has already been written about the Samsung Galaxy S III so I don't want to cover too much old ground here. The key differences with the new 4G model are that it supports LTE networks and has a few changes under the bonnet, so they're the areas I'll focus on. Telstra offers the Samsung Galaxy S III 4G from $66 per month and Optus from $51 (both including handset fees). Otherwise you can buy it outright for $899.
The Telstra versus Optus LTE war is in full swing, with Optus bragging about its wider metro LTE footprint. That's clearly the case in Melbourne and Sydney based on various LTE testing, but it's worth mentioning that Telstra users beyond the reach of LTE do have the advantage of falling back onto Dual-Cell HSDPA for a not too shabby 20 Mbps -- although DC-HSDPA doesn't reach everywhere.
Optus is scoring some impressive LTE numbers across Sydney, but my LTE tests in Melbourne with the iPhone 5 have been varied. Optus took honours in the CBD but Telstra put the competition to shame at the LTE patch surrounding Tullamarine Airport.
Dueling it out on Telstra's Next G network, I've found the Samsung Galaxy S III 4G falls just short of the iPhone 5 on Dual-Cell HSDPA -- the Samsung pulling down 17.6 Mbps, according to the speedtest.net app, compared to the iPhone 5's 18 Mbps. Of course results fluctuated considerably, as you'd expect from any wireless technology.
Head down the freeway to the airport's LTE outcrop and the tables are turned. The Galaxy S III 4G clocked a scorching 57.75 Mbps -- the fastest I'd seen yet on any Australian LTE device or network. Meanwhile the iPhone 5 could only muster 48.98 Mbps, but I've seen it hit 56.65 Mbps in the same spot at the same time of evening. Like I said, results are hit and miss.
These are real-world results, but keep in mind that they're taken outside Tullamarine Airport long term parking at 9 o'clock in the evening on a weeknight -- far from average conditions. It's possible they were the only LTE devices on that cell. I wouldn't expect to pull down those blisteing speeds during the day in a busy location where you're sharing the airwaves with other LTE users. But I was wrong.
The next day, standing in Ascot Vale at 3 o'clock in the afternoon, I finally smashed the 60 Mbps barrier -- clocking 61.36 Mbps on the Galaxy S III 4G while the iPhone 5 couldn't push past 57.18 Mbps. Judging by the coverage maps I was standing around 200 metres away from the nearest tower. After a few speed tests, moving away from buildings and other mobile phones helped me finally break the 60 Mbps barrier.
The Galaxy S III 4G consistently outperformed the iPhone 5 and wins bragging rights but to be honest the small download speed difference is meaningless, especially when mobile data speeds can be so fickle. What's perhaps more interesting are the upload speeds. While the iPhone 5 sat at around 15 Mbps in Ascot Vale, the Galaxy S III 4G uploads were averaging 26 Mbps. Looking back over the test results from various locations, the Samsung's upload speeds over LTE were consistently at least 50 per cent faster than the iPhone 5.
Rather than play a "mine's bigger than yours" numbers game with peak speeds, I'd place more importance on reliability. It will be interesting to see how Telstra and Optus compare in a few months time when they're catering to more LTE users.
As for what's under the bonnet, initially it seemed that opting for the Samsung Galaxy S III 4G meant sacrificing grunt, settling for a 1.5 GHz dual-core processor rather than a 1.4 GHz quad-core powerplant. But the Telstra-issued Samsung Galaxy S III 4G sports the same quad-core processor except it's accompanied by 2GB of RAM rather than 1GB. The extra RAM is a nice touch although it didn't shine through in the benchmark tests. Even with as much bloatware and services disabled as possible the 4G model only came out slightly ahead of the original model under the Quadrant Standard tests -- 5330 v 5252. Meanwhile the older model managed to sneak in ahead on the Rightware Browsermark, scoring 161,223 against the new model's 157,343.
It's worth mentioning that the 4G model comes running Android 4.1.1 Jelly Bean while the original model is still running 4.0.4 Ice Cream Sandwich (and it scheduled to get the Jelly Bean update later this month). As such it's not a like-for-like comparison, but it certainly indicates that very little separates them in terms of performance.
So that just leaves us with battery life, one of the more awkward things to test on a phone. What's really at question is the impact of LTE, so I charged up the Galaxy S III 4G along with a HSDPA-only model. Both feature a 2100 mAh battery. Next I disabled all the background apps on both, turned down the screen brightness and turned off everything but mobile broadband. The idea was to minimise the power load to help distinguish the impact of LTE.
With the phones ready to go, I sat across the road from that same Telstra cell and hammered the internet connection on both Samsung phones for more than an hour. I streamed music from MOG, browsed the web, ran speed tests, downloaded several podcasts and downloaded a 90MB file from Dropbox. Within a stone's throw of the cell, the 4G handset clocked an incredible 67.29 Mbps (while the iPhone 5 in my pocket still couldn't break 58 Mbps).
Both Samsung phones took quite a lot of punishment yet, after an hour, both handsets had chewed through 15 per cent of the battery. Surprisingly LTE had made no discernable difference. After that kind of result I think it's fair to say that LTE isn't too much of an imposition on this handset.
UPDATE: Panatrope makes a going point in the comments, sitting so close to the cell wasn't forcing the phones to work hard. So I went out to the airport, drove down Melrose Drive until I found the edge of LTE coverage and ran the same tests again (the Samsung 4G and iPhone 5 lost the LTE signal in the same spot, to address the reception question). Same result -- the battery on the 4G model didn't drain any faster than the 3G model. The 4G model did get noticeably warmer though. LTE might be more of a drain if your phone is constantly jumping between 3G and 4G cells during the day, but in terms of purely streaming or downloading files it doesn't seem to put extra drain on the battery. I know it's not what you'd expect, but I can only go on my results. As always, your mileage may vary.
There's obviously more to this Samsung Galaxy S III 4G than speed but like I said I just wanted to focus on the key differences to the older model. If you're after raw grunt and blistering speeds, this is quite possibly the droid you're looking for.