Samsung’s Galaxy S4 trumps the iPhone 5 to become the UK’s fastest phone

From the Samsung Galaxy S4 to the HTC One, more and more phones are packing in super-powered processors in the hope of becoming the UK’s fastest phone. Of course faster processors, don’t always mean better speeds so we put all the main contenders through our lab testing to see which handset would come out on top.

Despite being crowned the world’s best-selling smartphone in February, Apple’s iPhone 5 has aged badly since its launch last year. We measured it against the latest handsets on the market and it finished in last place, with the Samsung Galaxy S4 proving almost twice as fast. For the full results, cast your eyes below.

Phone reviews – read our verdict on all the latest handsets

Which phone is the fastest?

Fastest phone infographic
The Samsung Galaxy S4 is the fastest phone we have ever tested – and it’s no surprise.The phone’s UK incarnation features a 1.9GHz quad-core processor. In comparison, the iPhone 5 only has a 1.2GHz dual-core chip.

What do the scores mean?

Of course, more processing power doesn’t always translate to a faster mobile; that’s why we put all the handsets through the same industry-recognised Geekbench 2 test. This measures processor and memory performance across smartphone platforms to provide a standard speed rating. If a phone has a high Geekbench 2 score it will perform well when photo-editing, running graphically intensive games and multi-tasking across numerous apps in general.

Each individual test is scored, the scores are then combined together and weighted which eventually adds up to a final numeric score. With Geekbench scores, the higher the number the better. 1,000 is taken to be the base – so all these phones were expected to achieve a score higher than 1,000.

See our league table below for the raw Geekbench scores:

  1. Samsung Galaxy S4 – 3188
  2. HTC One – 2798
  3. Sony Xperia Z – 2173
  4. Google Nexus 4 – 2134
  5. Samsung Galaxy Note 2 – 1950
  6. BlackBerry Z10 – 1698
  7. Apple iPhone 5 16GB – 1664

Fast phones mean high prices

Grabbing second place in our testing was the HTC One. It’s combination of quad-core 1.7GHz processor and 2GB Ram, will provide plenty of punch for phone owners who use Gmail, Facebook, Instagram and Angry Birds all at the same time.

Sony’s Xperia Z also did well to sneak into third place, finishing fractionally ahead of Google’s Nexus 4 mobile. Despite this the Nexus 4 is over £200 cheaper to buy SIM-free than the top three phones in our testing. If you want to buy one the of UK’s fastest phones, you’ll have to pay upwards of £450 for the privilege.

Undoubtedly, Apple will upgrade its next iPhone with an improved processor when it launches later this Autumn. For the moment, Samsung’s Galaxy S4 is the phone to beat when it comes to speed.

Would you buy the Samsung Galaxy S4 because of its speed? Were you expecting Apple’s iPhone 5 to perform better? Let us know in the comments section below.

More like this…

54 replies

  1. The problem is that Geekbench isn’t a real world test, doesn’t take account of differences between mobile operating systems, and doesn’t test graphical performance at all. As such, using it like this mostly just serves to give you a headline rather than meaningfully useful information for consumers about which devices will actually perform better during actual use. I kinda expected better from Which?.

    1. Julian, regardless of how you like/dislike how the iPhone 5 was rated, it doesn’t have a cat in hell’s chance of keeping up with the S4 – which is hardly surprising considering how lowly it’s hardware spec is in comparison, half the amount of CPU cores, 2/3rds the clock speed, half the amount of RAM.

      The only thing that is higher on the iPhone is the price – £529 for 16GB iPhone 5, £455 for an S4 16GB – why pay more for less?

    2. As I said, the results of this benchmark aren’t a good way of determining actual real-world performance, especially when comparing different smartphone platforms. This is partly because platforms vary in terms of how efficient they are, with some delivering far better performance even with less impressive hardware. For example, the iPhone 5 may have a less powerful processor, but that doesn’t stop it from beating the S4 at running SunSpider (a JavaScript benchmark).

      If you only look at smartphones based on a handful of hardware specs, ignoring how well designed they are, battery life, how good the platform/ecosystem is around them, how well you’re supported if you have problems etc., you’re just not looking at the whole picture.

    3. I do agree with Julian that benchmarks do not necessarily represent the user experience.

      HOWEVER, I believe that he is wrong on a point of fact; IT Pro tested the S4 at 855.2ms on SunSpider, whereas the iPhone 5 seems to be widely reported as running the test in 914.7ms – i.e. the iPhone 5 is slower. [This is just from a quick Google, but of course I am happy to be corrected if I haven’t found the correct information! ;-) ]

      This is not as much of a difference as GeekBench would indicate, but like Julian says – what do benchmarks mean anyway?

      And for the record I have a Galaxy S3.

    4. Julian is right. OS must be part of the equation. The fact that Windows runs best on a MAC is proof. I am not an Apple fan or user. As an IT person, I can fully agree. The OS must be part of the equation.


    5. The OS is actually part of the equation. Benchmarks apps runs on the mobile operative system (iOS or Android) and this are the result.
      I don’t think “Which?” guys are running the app on a naked phone, am I wrong?

    6. Thats because its running it virtually, not really on the hardware. I run MAC OSX on my VMware with no issues either… and UBUNTU….. and SuSi…. and even a Android Emulator.

  2. Its a delicate balance with mobile devices between faster performance and maintaining an acceptable battery life.

    How do the devices compare on usability, performance, battery life etc when all taken into account?

    1. Absolutely right Mike. I realise that this report makes a bit of a splash in what is perhaps a slack week for news but speed of the phone is only a very small proportion of the overall judgement of ‘Best buy’.
      On its own it simply misleads.

  3. I spent some time choosing between the HTC One and the S4. In the end; I went into several Vodafone stores, in different towns (I travel round as part of my job) and tested the phones side-by-side. The HTC one consistently opened things faster and smoother every time. Using maps the HTC was on to finding directions for me, whilst the S4 was still trying to open the app.
    The S4 may be a faster phone in some ways but the enormous amount of largely unecessary ‘stuff’ on it and the fact that the HTC seemed faster in real world (or as close as i could get) application, sealed the deal for me.

    Both great phones though and ultimately…..unless you’re a proper tech nerd or live vicariously through your tech devices – it’s all much of a muchness really. They’re all amazing bits of kit that do the job fairly well and all have glitches that leave them less than “perfect”.

  4. “the HTC One. It’s combination of quad-core 1.7-inch processor and 2GB Ram” – presumably a 1.7-GHz rather than 1.7-inch processor?

  5. Where are the handsets running windows phone 8??? Come on Which? It is pretty poor form not to include the flagship Lumia device, which is miles faster than any handset listed in this article (all of the mid to high range Lumias are quicker) The spec sheets for these handsets looks less impressive solely for the reason that wp8 as a operating system doesn’t require 100 core processors to run effectively or efficiently. The lumia’s are the smoothest, quickest ‘real world’ handsets by far.

    1. Just replaced my lumia 900 for a galaxy S3.
      All i can say is it’s like changing from a toy car to a real one. Microsoft’s dead and it’s dragging nokia too.

  6. Hell….some people are soooo infuriating…just cant please anybody these days.
    Your speed tests are a guide and of course individuals will make their own choices .
    Most of us appreciate your reports and information.

  7. Galaxy S4 may be good, but the HTC one has to be one of the worst phones around. It certainly has not been designed to be used. Keeps switching off, reboots itself, turns off Bluetooth, dials numbers from your contact list at random, drops the signal. This phone is a users nightmare. Nobody should buy this phone, trouble is you don’t find out how bad this is until it is too late. So be warned.

    1. Well Chris can only say you are the lucky one. We have over 200 of these phones in my company and they are soon to be replaced as everyone has had major issues with these phones. Perhaps you are only a domestic user and not a business user?

    2. Lol? I thought Samsung stopped paying posters to bash HTC? I and many others have had the HTC One for quite a while now and never had any of the problems you mentioned. Maybe you should ask Samsung to send you some phones instead of just paying you guys so you can stop crying.

  8. What a load of tosh Which. Seems you’re more interested in making a grabbing headline than any real world test. So it’s got the fastest processor but how’s that going to help when I use the phone for everyday tasks. Like rendering photos or playing games isn’t exactly power computing and anyway its only less than half the story in the real world of using a smartphone. How easy is it to use from day one to an inexperienced user. How many quality and tested Apps are available etc. etc. so a lot more parameters to consider before the speed of the processor.
    Have you actually researched what the Geekbench 2 test does and that it isn’t biased to certain processor architectures? or who says it’s an Industry Standard Test ?

  9. Why there isn’t any Windows Phone?

    That’s an easy way to win, by excluding the best competitor…

    A Nokia Lumia 925 would have owned the S4

    1. Yes a Windows phone should have been tested with the others.

      The lumia 925 would have come in at number 5 with Geekbench of 2009. It has a SunSpider score of 922ms. So it is faster than iPhone 5, Z10 and the note 2 on Geekbench and is slower than the iphone 5 with SunSpider.

  10. What this tells me is which phone is the fastest. This isn’t the whole story, for me battery life is important, as well. Still it’s good to see some organisations out there that aren’t just saying buy Apple, buy Apple, buy Apple. It’s not the best at everything. Thanks

  11. Only a hard-hearted cynic would ever say this was a cheap attempt by Which? to get some quick headlines from some lazy tabloids (hello Daily Mail). As someone who works in the industry and gets to play with these phones on a regular basis, running a speed test is about of much-real world use as drag-racing to destruction two family cars at Santa Pod. And I doubt that anyone who drives a Ferrari quickly would have their experience ruined by the fact that a McLaren P1 is half-a second faster. As any fule kno, it all depends on the implementation of the software. And in this Apple has everyone licked. Apple apps, in most cases, will crash less often than anything on an Android device. All the high-end phones tested will react instantly to keyboard touches. Of course the iPhone 5 came out last September, around six months before the other devices tested. It should have been more accurately compared with the Xperia T, S3, and Nexus One. Mind you, if we all followed Which? recommendations we’d be riding around in Toyota Yaris’s.

  12. The graphic at the top of this article is deceptive. It gives the impression that the fastest phone covered fifteen or so times the distance of the slowest in the same time, and hence that the fastest phone is fifteen times faster than the slowest, which is incorrect.

    I suppose the real question is, how fast is fast enough?

    1. “How fast is fast enough?” is the nub of the problem. I sympathise with the Which? testers when they are trying to decide which phone is the best. Each user has a different profile.
      For instance I assume that my Nokia 620 is ‘slower’ that the 925 – but I wonder what that means in the real world. I suspect that anything called a ‘Geektest’ would not match my usage in any case!!
      My two suggestions are that they should concentrate more on categorising phones – perhaps by price? – and, secondly, they should take steps to ensure that their tests are much more up-to-date than they have achieved these last few months.
      I hope that is helpful

  13. NO test of the Lumia 920, wich is the fastest after using it 6 monnth.
    Besides Android is like Windows XP a selfpolluingsystem, wich slows down I, functioning after some 6 months.
    Why WP 8 is not mentioned.

  14. This whole article is complete rubbish, GeekBench doesn’t test the GPU, which is the most important component for games and graphical applications.

  15. This reminds me of the Windows Phone 8 wedding commercial where android and iphone users battle it out. Meanwhile the Wp8 users sit on the sideline wondering if they know about WP8!!!

    Everyone who uses WP8 knows it runs smooth as hell. Clearly WP8 is the winner here.

  16. Is anyone surprised by these results? The iPhone5 geekbench is lower than the quad core Galaxy S3 which was released 3 months before the iPhone5. Of course iPhone is the slowest now too.

    Snapdragon 800 and Tegra 4 devices are due in another month or two and those are clocking in at ~4000 on geekbench. Apple will need to triple the speed of their iPhone just to be competitive. And of course, since Apple only releases one phone a year, that phone will be the slowest thing you can buy in under 6 months, again.

    Here’s what you get with iPhone5:

    -Slow dual core
    -Low res screen
    -Inadequate RAM
    -Scratch prone painted aluminum casing
    -Metal casing to interfere with signal reception
    -Purple lens flare in photos
    -Touchscreen lag because of thinner redesign
    -Dumb antenna that can’t do LTE data and voice at the same time
    -Oddball apps that can’t be migrated to other manufacturers’ phones
    -Apple approved apps only
    -Ridiculously bad Apple Maps
    -Sealed battery (Always on NSA tracking)
    -Highest priced phones in the industry

    Is it any wonder Apple is rapidly losing marketshare and stock value every month?

  17. Useless test.

    Car A has 300HP and Car B has 500HP. Which is faster? Unless you know the weight of the vehicles the HP rating is useless. Just as Geekbench is useless. If you use real-world application tests you find the S4 is only slightly faster than the iPhone 5. Which makes sense – iOS is highly optimized to the hardware while Android is a bloated in-efficient mess.

    iPhone 5 is the 300HP car that weighs 2,500lbs. GS4 is the 500HP car that weighs 4,000 lbs. The GS4 has the power-to-weight advantage, but they are both very close.

  18. These twits have obviously drunk the Android cool aid that more powerful processor = better performance, despite years of first hand, obvious evidence from Android phones that more powerful processors = marketing nonsense (remember all those Android “Quad-core” phones that could only use 2 of their 4 cores!?) and more powerful processors + inefficient OS = mediocre performance.

    In the real world I strongly suspect the assessments would have been very different.

    Oh, and as usual, its not too hard to gin up a pseudo-valid test to make Android shine – as long as you don’t invite a Windows Phone to the test to spoil your pre-determined conclusions. The Lumia 920 leaves the S4 in the dust most of the time, with ‘only’ a dual core processor. Of course, it has a well designed OS, adn in the end that rules the roost.

    1. Which? – Is this correct? If so what was the point of publishing the article.
      We spend a lot of money on the magazine each month and I feel that this ‘speed test’ falls well below your usual standards.
      Am I correct?

  19. iPhone 5 is a last year’s device all these devices were launched this year.. Compare iPhone 5 with S3 that’s a fair fight…

    1. Sure the GS3 wins in floating point scores. Barely. Using 4 cores clocked higher than the iPhones. Yet it loses every real life comparison. Has half the GPU performance (or less) according to benchmarks and has half the speed rendering web pages.

      In short using Geekbench is an incomplete and useless comparison.

  20. This article is flawed. Geekbench scores scale almost linearly with the number of CPU cores and ignore the GPU completely. Real world mobile apps hardly ever use that many cores and the GPU is crucial to perceived performance due to how fluidly the UI gets rendered.

    On top of that; Android apps run in a garbage collected environment where occasional micro-freezes are not uncommon. iOS apps rely on the programmer for memory management which means they’re leaner and more efficient. Find me a user who thinks the iPhone 5 feels slow in normal use and I’ll eat my PC.

  21. When you use a faster phone like an S4 of an HTC, as a result of the higher speed does you voice sound squeaky like when you have inhaled helium?

    It must be me. As I ran my iPhone 4 alongside my 13 year olds £9.99 phone and it did not effect my voice at all

  22. The Galaxy S4 certainly is a very innovative phone with excellent features and a competitive price. As a result it is hardly any surprise that it is favored by most peoople,

Create account

You can leave a reply without having a WordPress account, but if you do register you can upload an avatar. A WordPress account is not connected to your Which? login and cannot be used to login to or any other Which? services.

Sign up