Galaxy S4 takes pole position in our benchmark comparison

by: Robert TriggsMarch 18, 2013

samsung galaxy s4 vs iPhone 5 6 aa 600

We’ve done a collection of handset comparisons over the past few days, testing out the new Galaxy S4 against a range of top of the line smartphones. But now it’s time to stick all these handsets head to head.

In these benchmarks we used the U.S. version of the Galaxy S4 which was on display at the launch event, so we’ll just be comparing the Snapdragon 600 chip for the time being. Even so, the Exynos 5 Octa model should perform equally as well as the Snapdragon 600, if not perhaps even a little better.

Galaxy S4 vs Android

Firstly let’s look at how the Galaxy S4 compares against a selection of top of the line Android handsets, including the newly released Xperia Z. We ran the same Antutu Android benchmark across all the smartphone to ensure a fair comparison.

Galaxy S4 Antutu benchmark

Very impressive, the Galaxy S4’s 1.9Ghz Snapdragon 600 is clearly a substantial improvement over the previous generation’s S4 Pro chips used in the Xperia Z, Droid DNA, and Nexus 4. There’s an even more obvious gap between the older generation Galaxy handsets which use Samsung’s own Exynox 4 quad-core processors. The performance jump from the Galaxy S3 to the S4 is perhaps the most staggering, so those wondering whether or not to make the switch to the newer handset can be reassured that they’ll notice a substantial performance improvement.

No one can call the Galaxy S4 a slouch when it comes to processing power and it should stack up very well against the new HTC One and LG Optimus G Pro, both of which are also using the same Snapdragon 600 chip.

Galaxy S4 vs Apple & Blackberry

Samsung’s Galaxy S4 tops the table when it comes to Android handsets, but how does it stack up against Apple and Blackberry’s best offerings. To test this we used the cross platform Geekbench benchmarking tool.

Galaxy S4 Geekbench benchmark

Again the Galaxy S4 comes out miles ahead of the competition, making light work of the Blackberry Z10’s S4 Plus dual-core processor, and proving that the Snapdragon 600 has Apple’s dual-core A6 chipset beaten as well. It’s a good performance from the new Android handset, but not an unexpected one considering that Apple and Blackberry are still stuck on older and slower dual-core technologies.

Personally I’m very impressed with these results. The Galaxy S4 has some really decent hardware under the hood, as well as an assortment of software upgrades. I can’t wait to see how the international Exynos 5 Octa model performs against its U.S. counterpart.

  • I’m sold for an S4 even if it scored the lowest scores…

    • joseph pastor


      • Because benchmarks mean nothing…

        • kascollet

          Yes, they do. They just don’t mean everything.
          If you were given the choice, why the hell no choose the fastest architecture ?

          • MasterMuffin

            Because benchmarks don’t show what phone is the fastest in real life

          • Jaime Larios

            And also they don’t count the extra features like Multiwindow and all the S features the Galaxy S brand has come to have. Those are what make the difference in real life.

  • S4disappoints

    I thought Z10 had snapdragon S4 duel. But u say here that it is OMAP.. Which one is real case?

    • Rob Triggs

      Just double checked, we did in fact use the S4 Plus model. The OMAP version is the one I’m used to seeing in Europe. My bad.

    • Hung Nguyen

      Sadly only the Z10 LTE version has S4 Plus.
      Z10 HSPA+ in Asia only has TI OMAP 4470

  • MasterMuffin

    The Snapdragon 600 in sgs4 is going to be faster than HTC One’s and Optimus G Pro’s because it’s clocked higher and Exynos is just going to blow away everything, 27000 is the best score in AnTuTu made by GT-I9500!

    • amine ELouakil

      Sigh…. it’s not the CPU Clock alone, there are other things that contribute to the score, such as RAM, Nand used and so on, and btw the performance increase is not linear with the clock speed the higher the clock is the lower the increase in performace due to heat desipation, electrical resistance increase and so on. not to mention software optimisations

      That’s why you see the One outperforming the SGS4.

      • MasterMuffin

        Everything you said in the first section is correct, but this is for the second section: WTF? I haven’t seen One outperforming sgs4 in any test. Check AnTuTu fir example, all top 10 phones are sgs4.

        • amine ELouakil

          Just check the benchmarks videos available on the internet, you can even see videos where both phones are benchmarked side by side, and not only AnTuTu scores but other benchmarks at the same time.

          • MasterMuffin

            Finished product against unfinished product? Seems fair and sgs4 still wins in for example the AnTuTu which is considered one of the best benchmarking tools for Android. Or the problem may also be in 4.2.2 (benchmark give bad scores to 4.2.2 devices)

          • amine ELouakil

            Not really, You can check the benchmarks of the HTC One from the presentation event they didn’t change at all, not to mention I’ve already spoke about the eventualitty that the SGS4 might score better with some more software optimisation. and no the SGS4 doesn’t score better than the One on AnTuTu, unless I missed something, feel free to show me one -“video”- of the SGS4 where it scores over 26k on AnTuTu (a score that the One achieved)

          • MasterMuffin

            You can also feel free to download AnTuTu and check the scores for yourself, over 27000 :) well I’m not in the mood for searching random videos from web because you’re lazy (:P), so I’ll just link AA’s own video where sgs4 scores 25900 (≈26K)

          • amine ELouakil

            25900 . not 27000 on your video

          • MasterMuffin

            Y U NO read my previous commet

          • amine ELouakil

            commet no read coz no prove yo, the One isn’t even referenced on AnTuTu yet it was benchmarked several times and videos review what ever you want proves it

      • sluflyer06

        Wrong. I couldn’t count the number of times I”ve seen devices and PC’s performance scale linearly with clock speed. You are making a very general assumption that any given device cannot dissipate the heat its generating. Let’s also throw out the nonexistent reasoning that software code is somehow optimized for a specific clock speed.

        I’m not arguing for or against the S4 here, but You’re clearly pulling smoke out of the box and making baseless statements. Now if we look at your statements as a theory and not real world you are correct. Yes, at some point if you kept increasing clock speed through overclocking beyond design specs you will see large problems with heat dissipation and diminishing returns as you start to hit bottlenecks in the memory bus or other traffic areas outside the cpu but these are not issues when your running at stocks speeds, it wouldn’t matter if it was running at 1.2ghz or 2.5ghz if that is the stock speed then the supporting hardware would be designed to support it.

        • amine ELouakil

          It’s only about heat dessipation but also electrical resistance that increase, Also the Mobile SoC don’t cope with overcloacking as much as their PC counterpart as the ladder has deddicated colling system where as mobile SoC does not, Not to mention that you are welcome to prove the fact that CPU performance scale linearly when it comes to overcloacking, espcially when you reach high clock speeds compared to stock.

      • Hung Nguyen

        You’re totally wrong.
        The fact is, HTC One uses LPDDR2 RAM instead of LPDDR3 of Samsung Galaxy S4 (source:

        • amine ELouakil

          Wrong in what aspect please feel free to point it out, Again I never said the HTC One has a faster (or slower) ram, I said there many factors that makes the performance of a phone inclucding the RAM (type or clock speed configuration) and not the clock of the SoC

          And for instance if HTC is using the dual channel compatibility of the One whereas the Samsung is not using it, that would change everything.

          • Hung Nguyen

            So what are “there are other things that contribute to the score” that make “One outperforming the SGS4” if One’s RAM is worse by a huge margin, and about NAND Samsung is well-known as the best player in the current market?
            Do you know that even with the same Adreno 320 GPU, but in GS4 it is notably faster than in One?

          • amine ELouakil

            What are you smooking? “if One’s RAM is worse by a huge margin” if you don’t know what you are talking about you should not do so, Yes top range LPDDR3 is better than top range LPDDR2 as the latter frenquencies range from 200 to 533Mhz where as for the LPDDR3 is from 400 to 800Mhz and has a larger brandwidth, but a 533Mhz LPDDR2 on dualchannel would easly best an LPDDR3 @ 400, unless you know something we don’t know and you know as a fact that Samsung is going with a dual channel configuration ddr3 then just rest your point, I’m sure I’ve seen somewhere 400Mhz I’m trying to find the artical, and I’m sure anandtech will give maximum details about that when they get their hands on the phone.

            “NAND Samsung is well-known as the best player in the current market?” Wrong, they are the biggest producers, and that doesn’t mean anything, to keep cost low and maximise their marging they can choose a lower quality Nand compared to HTC, they did with their cheap plastic construction what would stop them from doing the same with Nand or other component? are you forgeting that Samsung is there to maximise their margings? and unlike HTC, SGS4 will sell no matter how cheaply it is made.

            “Do you know that even with the same Adreno 320 GPU, but in GS4 it is notably faster than in One?”
            And that does make sense to you? not to mention the Vellamo benchmark proving otherwise, but the scores are weird, I know that GSM arena is a serious source, but I’ve seen much higher scores from the One (and I’d bet you did to) while on the other hand I’ve seen lower scores from the Samsung, if you want I can post several videos available proving so.

            Not to mention software optimisations.

            But anyways I’m getting the One, I’ve seen everything that the SGS4 had to offer, and it’s not the phone for me, for me, it’s just stupid to pay premium for a product that is a cheaply built as the SGS4, some people don’t mind that and want sd slot and removable battery I can understand, but for me it’s not the case, and I think alot share my opinion.

      • cycad007

        amine…thanks for the info. I wasn’t aware of all the factors you listed…fairly informative.

        • amine ELouakil

          You are wleocme

    • tegra 4 got 36000 though

      • MasterMuffin

        Benchmark score by nvidia with unknown device, I’d trust that…

  • Balraj

    nice benchmark but all other devices run on low power processor compared to s4 snapdragon 600
    so the comparison is not fair
    results are obvious :-P

  • guest

    wow this must be embarrassing for apple.. just wow

    • kascollet

      It would be if iPhones had some lag, which is still absolutely not the case. Imagine the UI and browsing speed with iOS running on this kind of super powerful platform :-D

      • IncCo

        believe me, there is lag on iphones too, ive seen it myself.

      • Actually, iphones do lag. Iphones not lagging is only a myth.

      • Ciro Morales

        A week ago I saw an iphone 4S lagging with instagram…

      • I’ve briefly used my friends Iphone5’s when not having my own phone with me and I noticed lag in like 2-3 minutes. Yes they run smooth and the “lag” is so marginal it’s really not worth mentioning but it’s the same case for the most new high end android phones. I think we’re pretty much past the part of lagging phones, bad screens and shitty cameras(when comparing the most high end phones from respective manufacturer) it’s mostly about personal preference nowadays.

        • kascollet

          That’s right. The main difference is Apple achieves the global fluidity with less powerfull parts, which allows them to sell a small phone (small battery being possible).
          I have a Nexus 4 and an iPhone 5 in my hands at this very moment. Very interesting !

  • kascollet

    More than 3000 points in GeekBench is a highly impressive number for a phone.
    Think about it : a 2009 Core2Duo @ 2.0Ghz is roughly equally powerful !
    What a beast :-D

  • Henry Léon

    S4 (Samsung S4tisfying Super Smartphone)

  • Henry Leon

    S4 (Samsung S4tisfying Super Smartphone)

  • a22matic

    Robert – can you compare (benchmarks) the U.S. S4 to the HTC One?

    • Yes please :D

    • Rob Triggs

      We covered some preliminary benchmarks a few days ago, ( but the S4 benchmarks there are lower than our own.

      Don’t worry, as soon as we get our hands on the HTC One we’ll do a comparison.

  • George

    Nice results!
    Any vellamo or quadrant runs please? :)

  • Come on guys, the comparison needs to be fair, you can’t compare the new generation chip phones to the old one and claim a pole position!

    Compare it with the HTC One, that would be head to head and lets see who wins !

  • Makes my Galaxy Note 2 look like a sloth….

  • SamerMouj

    Benchmarks don’t mean shit. The blackberry has a dual core processor and it BARELY has any lag. While as the Samsung Galaxy s3 with its quad core processor showed lags all over. Some operating systems don’t need that much processing power and that’s where the innovation is.
    How about you try to run an html5 test on both the galaxy s4 and the blackberry z10 and see which is better?

  • brady

    the iPhone looks stupid

  • It’s obvious that newer tech, faster processors etc will score well in benchmark tests, even dummies would know that. It would be more useful and practical to compare performance in real user situations and see how well the software is optimized to perform the tasks, such as web browsing, unzipping a file, running office documents etc. Android is a great mobile platform but once in the hands of manufacturers it’s performance can vary and so it’s not all about hardware but also how well the software has been written to allow maximization. Apple had been good at that despite its ageing ios.