Benchmarking the iPhone 5

Benchmarking the iPhone 5

Summary: Apple's latest iPhone outperforms its 4S predecessor by a significant margin, but is matched in most areas by Samsung's Android flagship, the Galaxy S III.


ZDNet (in Germany) has examined the iPhone 5's performance using multiple benchmark tests. By way of comparison, we've included the leading Android-based competitor, Samsung's quad-core Galaxy S III (which has just received an update to Android 4.1 Jelly Bean), along with the previous-generation iPhone 4S and the iPad 3 tablet.

On test (from left): iPhone 5, iPhone 4S, iPad 3, Samsung Galaxy S III.

When you buy a new smartphone, the first hint as to its performance will be the device's start time, although it's barely relevant in general use. Still, comparing a device's start time to its predecessors can indicate how a manufacturer's product line is improving. The iPhone 5 starts in 20 seconds — noticeably faster than the 4S, which takes 35 seconds. The iPad 3 takes 44 seconds to start up, while Samsung's Galaxy S III, at 23 seconds, is comparable to the iPhone 5:


Geekbench bases its overall performance score on four tests: two determine the speed of the processor when handling integers and floating-point numbers (Integer, FPU) and two assess memory speed and bandwidth (Memory, Stream).

This cross-platform benchmark is well suited to analysing the performance of iOS devices, but caution is required when comparing the results on other platforms. For example, in the past, the Windows version differed from the Mac version by not using all CPU extensions, with the result that Mac OS systems delivered significantly higher scores than comparable Windows systems.

The overall Geekbench results show that the iPhone 5 is more than twice as fast as its 4S predecessor:


PassMark Software's PerformanceTest Mobile also uses four tests, but unlike Geekbench includes 2D and 3D graphics benchmarks. It also examines the I/O performance of the disk subsystem. The iPhone 5 delivers very good results in Passmark's suite, with nearly twice the overall performance of the iPhone 4S:


At first glance, the iPad 3's results appear particularly poor. However, this is due to the fact that the 3D graphics tests are performed at the device's native resolution, and the iPad 3 has to move over four times as many pixels (2,048 by 1,536) than the iPhone 5 (1,136 by 640). Since both devices' GPUs have similar capabilities, the iPad delivers a worse result thanks to its higher resolution — in the individual 3D Graphics test (below), the iPad 3 scored 587 points to the iPhone 5 's 1,752. On the other hand, this does reflect the performance that's actually experienced.


A different picture emerges if 3D performance is assessed at a predefined resolution (see GLBenchmark result, below). In this case, the iPad 3 delivers comparable performance to the iPhone 5. However, this finding is of little relevance in practice, as most 3D applications run with the device's native resolution. In this respect, the iPad 3 is 'suffering' for its high-resolution Retina display.


Modern websites regularly use JavaScript to present content, so this needs testing in any balanced performance analysis. It should be noted that the results obtained can be influenced not only by the device hardware, but also depending on the operating system. OS developers usually work hard to improve their built-in JavaScript engines. The iPhone 4S is a case in point: under iOS 5 the iPhone 4S delivered a BrowserMark of just under 90,000, a score that rose to over 100,000 under iOS 6:


SunSpider JavaScript Benchmark and Futuremark Peacekeeper, another JavaScript test, show the iPhone 5 delivering approximately twice the performance of the iPhone 4S. Note that the Android flagship, Samsung's Galaxy S III, cannot quite match the performance of the iPhone 5 in these tests:


Apple's iPhone 5 offers excellent performance across the board, and is significantly faster than its iPhone 4S predecessor. The Android-based Samsung Galaxy S III delivers comparable performance in all areas except 3D games. Having said that, it beats the iPhone 5 in PassMark's PerformanceTest Mobile 2D Graphics test by some distance (see graph on page 2).

Version 4.1 of Google's operating system is largely responsible for the excellent performance of the Android smartphones: Jelly Bean has put an end to the screen freezes of the past — even on far less powerful devices than the Samsung Galaxy S III used for these tests. Unfortunately for Android users, few manufacturers have so far released Jelly Bean for their devices. The update process for the Android platform is too long and device support is restricted. By contrast, the latest iOS version is available for the three-year-old iPhone 3GS. Technically savvy Android fans can, of course, use Jelly Bean images from the developer community. If you're wary of an Android version that isn't from your phone's manufacturer, you should purchase a Google phone, for which the latest OS updates are quickly available.

This article originally appeared on ZDNet Germany as iPhone 5 im Benchmarktest.

Topics: iPhone, Apple, Reviews, Smartphones

Kick off your day with ZDNet's daily email newsletter. It's the freshest tech news and opinion, served hot. Get it.


Log in or register to join the discussion
  • Only a couple of weeks ago ...

    there were headlines that Iphone5 BLASTS S3 OUT OF THE WATER!!
    Even then the figures showed comparable performance on both. Good to see that the hyperbole at launch subsided and made way for more sober assessments.
    • 1752 versus 854 is not comparable

      On the 3D test, the iPhone 5 is twice as powerful as the S3, that is NOT "comparable". This means a 3-D game will run at 30 fps on iPhone 5 but 15 fps on the S3. That's the difference between smooth and unplayable.

      Depends if you like 3-D games of course...
      The Star King
      • ... i'm sorry your mom dropped you when you were young...

        just because iphone5 blew the sgs3 out in ONE or TWO scores doesnt mean it blows it out in everything. there are thousands of benchmarks out there. I can go out and test each phone for all of them and im sure there will be a few where the sgs3 blows out the iphone5. they are stupid tests that are only supposed to be used for guidelines.

        another note: just because a score is doubled does not mean it has double the frame rate prcessing speed. that is not how these scores work. so take your ignorant assuming a$$ back to school and do your homework.
        • Typical


          When you have to resort to insults, it's clear that you've already lost the argument.

          It would seem that only in very rare situations that make heavy use of parallel processing can any advantage be seen on the Samsung device. Otherwise, across the board and especially anything GPU related, the iPhone 5 blows the Android flagship device away. Also, this is the very best case scenario (SG3 running Jellybean... really, how many of them out there?).
          • typical of what?

            Technically, i wasn't name calling. i asked @The Star King a legitimate question. He ranted off a paragraph's worth of garbage thinking he knew what he was talking about and i questioned whether he may have fell and hit his head in his infancy.

            You on the other hand bring up good points about the iPhone GPU. It is better than the GS3, but it doesn't blow it away by any measure. And as i said before, benchmarks are only guidelines. Not a single one of them measure what a phone can do in the real world. Android has proven time and again it is better at multi-tasking than any apple device ever, and that is the true test of a phone. How much can it do at one time before it is bogged down and you can actually see diminished quality?
          • Name calling...

            "Technically, i wasn't name calling. "

            Right, and was the subject title for your post? Allow me to remind you. "... i'm sorry your mom dropped you when you were young...". Honestly, this is simply insulting and childish. I really takes away from any potential point you have to make.

            "It is better than the GS3, but it doesn't blow it away by any measure. "

            Well, I suppose it's hard to find a technical definition for "blows it away". To me, being better than another product by 10%, 20% maybe even 30% is just beating it. Beyond that though, you pretty much fall into the "blows away" category. The differential here, as evidenced by these benchmarks is clearly well beyond that measurement.

            "Android has proven time and again it is better at multi-tasking than any apple device ever, and that is the true test of a phone."

            For starters, it would seem that the true test of a phone would be to make phone calls, but I suppose that's a different argument. Second, when you make blind statements like "better at multi-tasking than any apple device ever", then that would also include Mac computers running OS X. Do you still contend this is the case? If so, please illustrate how this is true.

            As for mobile devices, good designs are about trade-offs. Nothing comes for free. Having multiple applications comes at the price of battery performance, the need for more memory and overall sluggishness as resources become low. iOS devices don't suffer from any of this due to design decisions which limit user land multitasking. To that end, you speak of practical benefits, so knowing that iOS does allow services to multitask, I would ask you to provide a few practical use cases whereby Android provides a clear advantage over iOS in this regard. Having used both devices myself, I'm not clear on what those advantages would be. Please enlighten me.
          • this is getting old...

            ... so i just wrote a very lengthy response and then my computer crashed. stupid work computer with IE7

            so here's the basic recap of what i had said...

            instults --> something about cry me a river. Star King said something silly, i called him out on it.

            "To me, being better than another product by 10%, 20% maybe even 30%" --> Where in the world did you find anything besides that one benchmark test (whichagain, means next to nothing) that says iOS6 is >30% better than Android at anything?

            "phone" and "apple device" --> I am sooo sorry. i thought that since this article was covering the iphone 5 and the samsung galaxy gs3 that we intelligent folk would understand i was talking about SMARTphones and iPhones in my original arguement. Alas i must be more thoughtful to being specific so that you can't demerit my arguments with semantics.

            as far as multitasking. I am sorry i did not cite sources in my previous homework assignment. I thought it was a pretty well known fact that Android whoops the iPhone in multitasking. But since you seem unconvinced. Here some others that agree with me:

            "Android has always been better at multitasking then iOS"

            "In fact, almost nothing in the multitasking bar is truly “multitasking.” "

            "Android tends to do more true multitasking than iOS"

            This one isn't even a Jelly Bean Comparison
            "Ice Cream Sandwich offers a much better and more advanced multitasking option for users than iOS"

            Also an ICS comparison
            "Android is much better at managing multiple apps than iOS"

            This one is comparing IOS5 to Gingerbread... that's G...H...I...Jelly Bean... that's 3 versions ago and this article says multitasking is about even (which is a stretch of the truth in favor of apple, in any case... wow)
            "Multitasking Capabilities – Although iOS 5 and Gingerbread 2.3.3 fair almost equally in the parameter"

            "Apple made a big deal about [multitasking], it’s still not up to par with Android’s"

            "IOS has fake multitasking where you can't download while doing anything else and apps are put into suspension when you leave them and stop running."

            This guy doesn't necessarily say Android is better, but he says apple needs help with multitasking and includes android as a viable alternative
            "Is iOS 6 the time for Apple to revamp multitasking fast app switching for the iPhone? For the iPad? For mobile?"

            "Sure iOS has multitasking now, but Android has real multitasking the same type of multitasking which is present in the PC"

            "While iOS never claimed to have true multitasking, Apple's avoided it primarily due to battery life and performance concerns"


            "Android utilizes the entire screen with screenshots of the running applicaitons compared to iOS, where multitasking is only done using the dock at the bottom. iOS wastes the entire screen space when it comes to multitasking."

            Shall i continue?
          • Get a life!!!

            You can take you tin hats off people, it's going to be alright...
          • nope!

            Benchmark scores are silly. Check actual head to head 'real-world' comparisons of these two phones on YouTube and you'll see what I mean.
      • It doesn't mean anything of the sort

        The GPU that apple uses is a Raw Horse Power Beast but doesn't feature as many optimization as some competitors...

        This is why the Tegra 3 has actually produced better looking effects in most games and still manages to run them at 60 FPS.

        Bottom line, once you cross 30 FPS the game is going to be pretty smooth and Android caps games at 60 FPS via a v sync mechanism. And the superior FPU of the S3 means that you can offset the difference by using the CPU to draw polygons and the GPU to shade them.

        The truth is, Benchmarks are good for bragging rights but that's about it.
        • that's what ive been trying to say

          but no one listens to me
        • How you know?

          This is only a theory.

          "Android must be this, Android must be that"

          In fact, you have absolutely no clue whether iOS does not provide APIs to use both the CPU and the GPU for certain tasks. Judging by what is available in OS X (and the underlying OS is the same), that might well be the case.
          After all, Apple themselves designed both the CPU/GPU and the software that operates it.

          Besides, it's funny that nobody recognises the importance of memory bandwidth. This makes for *huge* difference. It doesn't help if your CPU is plenty fast if it only can achieve that speed running (small) loops.
          Remember that Cray computers in the past? Their CPUs weren't much that fast anyway, but compared to other computers at the time they had *huge* bandwidth bus between the CPU and the rest of the system, resulting in unbeatable performance.

          Anyway, performance and benchmarks are two different things.
          What would you prefer? To know your device benchmarks better, or know your device runs better? I would care less about benchmarks.

          By the way, what about battery life?
          • What?

            It isn't the OS man... These things have to be built in to the hardware! It's like when Sony would brag about having RAW Horsepower numbers from the PS2 and neglected to mention the Chip didn't support FSAA... The result was with software based FSAA enabled the PS2 was only marginally stronger than the Sega Dreamcast.

            These RAW Numbers are why Imagination Technology has really only had limited appeal with anyone outside of the Mobile and Tablet sectors...

            As for Battery Life, both the Note and the SGS 3 are said to be better with Battery Life.
  • Quadcore?

    The US version of SGIII is only dual core right?
    Will there be a difference?
    • Yes.

      Yes from what I've seen the dual core S3 is not as fast as the iPhone 5. I assume they tested the international version.
    • Yes and YES!!!!

      Yes, the U.S. model of the SG3 is only dual core, but supports LTE unlike the European model. And YES, the U.S. dual-core SG3 is absolutely trounced by the iPhone 5. The dual core iPhone 5 handily beats even the quad-core SG3 in almost every benchmark shown here.

      My guess is this article was sponsored by Samsung. If it was truly independent, then I'd expect to see cellular data benchmarks, too, but the article doesn't even mention data services, which is where the European SG3 would fail to compare at all to the iPhone 5. Heck, the article doesn't even mention that the iPhone 5 has only dual cores or that the model of SG3 isn't available in the U.S. The fact that the SG3 is described as having quad cores would seem to be the author's lame attempt to appear impartial.
      • No fair comparable tests done yet with US version of SG3

        Because the US SG3 hasn't been updated to Jellybean yet. Once it gets updated to Jellybean, then we'll see how they truly compare at least in terms of benchmarks which are not reflective of real world use anyways, but it's interesting to see how they perform.

        Though there is some evidence that the dual core snapdragon does perform better than the quadcore exynos in certain benchmarks. Doesn't the US version have more ram and isn't the dual core chip run at a higher rate? We'll see soon. Maybe by next month.
        • Well?

          What's holding up the Android update????????
      • real world tests

        I conducted some real world tests:
        bluetooth file transfer speed:
        Iphone5: 0
        SGS3: much more than 0

        time to assign song as ringtone:
        iphone5: about 2 minutes with itunes, about 5 mins if I have to boot PC and start itunes. much more than that if I haven't downloaded itunes.
        SGS3: under 1 second.

        attaching a file to email
        iphone5: wtf is a file?
        SGS3: around 10sec

        downloading some pr0n
        iphone5: there's no pr0n on the net
        SGS3: download in progress

        battery swaps
        iphone5: about 10 mins if you are good with tools
        SGS3: about 30 seconds with blindfold on

        Swapping SD cards:
        iphone5: what sd card
        SGS3: fiddly but possible with just one hand and a chin

        sharing a URL via barcode
        iphone5: highlight and copy URL from browser, open barcode app, paste URL in barcode app. about 20seconds or so.
        SGS3: menu--->share--->barcode. about 3 seconds if you pause 1 second between clicks.

        I could write a book on this!
        • yep!

          Your examples are exactly the sort of things that really shows which smartphone is superior.