25
Apr
htc-one-x-press

With the upcoming release of AT&T's HTC One X, many people are wondering how it compares to the international version, which packs NVIDIA's quad-core Tegra 3 instead of Qualcomm's dual-core Snapdragon S4. One user out in YouTubeLand happened to get their hands on both variants and decided to boot them up simultaneously, as well as run AnTuTu Benchmark to see how they compare head-to-head.

While I realize that benchmarks are definitely not a definitive answer to how well the device performs, they do give a rough idea of what the device is capable of. Plus, they're just fun.

Left: Tegra 3 One X; Right: Snapdragon S4 One X

Right out of the gate, it's pretty clear that the Tegra 3 phone is quicker on bootup. That could be because of the extra AT&T animation in the boot sequence of the S4 phone, or perhaps it has something to do with the "Android is Upgrading" dialog that pops up (which suggests that some sort of update has just been performed). Either way, the Tegra 3 phone is ready to go several seconds before the S4 phone. Then, during the benchmark, the T3 X takes an early lead, and the S4 X just never catches up. At the end of it all, the T3 X's score nearly doubles that of the S4.

Is this a definitive answer to which device is better? Definitely not. Both phones are absolutely killer and offer outstanding performance for top-of-the-line pieces of kit. This is just a look at how the two compare to each other in terms of sheer numbers - remember, the user experience is what really matters. With that in mind, I think that both devices probably offer an exceptional UX, and are both excellent devices.

Cameron Summerson
Cameron is a self-made geek, Android enthusiast, horror movie fanatic, musician, and cyclist. When he's not pounding keys here at AP, you can find him spending time with his wife and kids, plucking away on the 6-string, spinning on the streets, or watching The Texas Chainsaw Massacre on repeat.

  • Shift

    Better on mute.

    • http://twitter.com/SParKlngCyaNide SparklingCyanide

       no way, that music is boss.

      • http://meatcastle.com/ Youre My Boy Bloo

        Mortal Kombat!!!!!!!!!!!!!!

    • http://www.gregpak.com/entries/002275.shtml D.Smithee

       Agreed.

  • TOMMMMMM

    That menu key...
    ಠ_ಠ

    • http://twitter.com/vmlinuz Richard

      Yeah, what's going on with the AT&T model???

      • http://twitter.com/DesignGears DesignGears

        That is a very early AT&T test device.

  • http://twitter.com/ANDROIDlemmon roger l

    you have the phones marked backwards as the tegra 3 one x is on the left and the snapdragon one x is on the right in the video

    • http://www.androidpolice.com/ Cameron Summerson

      Indeed I did. Thanks for pointing that out!

      • http://twitter.com/DesignGears DesignGears

         That is a very early AT&T test device.

  • RyoHazuki

    Cameron, you messed up in your article. It's the reverse. The S4 One X boots up faster.

    • http://www.anivision.org/ Christopher Bailey (Xcom923)

      No, he got it right. if you look the one that boots up slower has the ATT logo on boot that means it's the LTE version IE the s4

      • RyoHazuki

        That's cause he fixed it, genius.

        • http://www.anivision.org/ Christopher Bailey (Xcom923)

          huh? how does that make any sense? Are you saying that cameron "fixed" the video? 'cause each time I see it the S4 boots slower. you exact words were "The S4 One X boots up faster" which just isn't true

          • RyoHazuki

            Nah, it had to upgrade andorid and what not. If they did a fresh boot again it would boot faster.

          • http://www.anivision.org/ Christopher Bailey (Xcom923)

            probably but that doesn't mean it would beat the tegra

          • RyoHazuki

            S4 does beat Tegra 3.

  • Joe

    The one on the right failed to show FPS during the graphic tests too

    • jonathan3579

      It showed it however it was hidden by a faint grey layer that would bring up the menu button.

      • Joe

        Ah yeah I see it now, thanks

    • http://twitter.com/DesignGears DesignGears

       That is a very early AT&T test device on the right. Running this test on a retail unit works properly, lighting and all

      • Yourmom

        Does anyone know if this is a very early AT&T test device on the right?  I wish someone would say.

  • baggy247

    Why is the menu button implementation different on these? The AT&T model is actually overlaying part of the screen and looks like it can be hidden :S

    • Dan

       The menu button implementation is different because they're running different OS versions.

      • baggy247

        Yeah... but why would they implement it differently?

        • Dan

          Because somebody said, "Hey, the menu button looks freakin' STUPID on build 1.26.  Fix it!"  Sometime before build 1.44 it got fixed.

        • Dan

          Because somebody said, "Hey, the menu button looks freakin' STUPID on build 1.26.  Fix it!"  Sometime before build 1.44 it got fixed.

        • Dan

          Because somebody said, "Hey, the menu button looks freakin' STUPID on build 1.26.  Fix it!"  Sometime before build 1.44 it got fixed.

          • baggy247

            To be honest I don't know which is worse, I suppose at least it can be hidden but then its more taps to access the menu

          • baggy247

            To be honest I don't know which is worse, I suppose at least it can be hidden but then its more taps to access the menu

          • baggy247

            To be honest I don't know which is worse, I suppose at least it can be hidden but then its more taps to access the menu

    • http://twitter.com/DesignGears DesignGears

       That is a very early AT&T test device. That is why.

  • Adam Skinner

    The one on the right totally got smoked.  Taking the bootup with a grain of salt (due to the upgrading dialog), the benchmarks don't lie.

    Having said that, we're comparing Tegra3 vs Adreno 225, and not Adreno 320.  Adreno 320 is the real competitor here.

  • fixxmyhead

    as always snapdragon loses again

  • http://www.facebook.com/people/David-Sinor/100002087388095 David Sinor

    I wish this didn't start off with that indicator of some kind of upgrade. I never trust performance of hardware immediately after any upgrading. Really too bad, and it's not like anyone could anticipate that would happen as you setup the shoot. I'd like to think I would restart, but if I didn't notice until reviewing the video just to lazy to go through the whole thing again.   Thanks Cameron for the video.

  • NoUsernamesFree

    The 'Android is upgrading' message happens on every boot on my GNex if it's stock. Even without an update. Just throwing that out there...could be an ICS thing.

    • http://www.androidpolice.com/ Cameron Summerson

      That's definitely weird. I have three ICS tablets and A GNex, and only see that dialog after an OTA.

  • What?

    Not really good comparision to do CPU since ATT has put their own apps and log on phone so that makes phone boots up slower. They should be doing bench test.  Also, if they really want to compare boot times on a fair level. both phones should be rooted.

    • What?

      Not really good comparison to do CPU since ATT has put their own apps and logo on phone so that makes phone boots up slower compared to world phone. Tester should be doing bench tests.  Also, if they really want to compare boot times fairly, both phones should be rooted (clean install), compare boot time then.

  • Zonk7ate9

    Why doesn't the phone have AT&T branding above the screen? I don't think that is a production unit. I also don't trust the test given the OS updating. Most likely the phone had to rebuild caches after an update, it's not going to perform as well, needs a reboot.

    • http://twitter.com/DesignGears DesignGears

      That is a very early AT&T test device. At 1080p you can see the etchings on the top left of the phone on the right.

  • Robert Oliveira

    The AT&T variant of HTC One X is 27 seconds behind the Tegra 3 version. I'd say the UX would have to make up for that in a big way. Interesting also is that the International version had its data connection on all the while whereas the AT&T version was in Airplane mode. Shouldn't there have been fewer "distractions" for the OS on the AT&T phone?

  • wolfkabal

    Anyone else also notice that the ATT version was the most up-to-date with kernel 3.0 even? Not sure how this would effect performance, I thought it was supposed to help. But it still loses. *shrug*

  • Marc

    There is a guy over at XDA that has both the International and Rogers versions (Rogers is the same at the AT&T version).  On the first set of benchmarking, the International version wins (http://forum.xda-developers.com/showpost.php?p=25104485&postcount=17).  But then he re-runs the tests and they are much closer. (http://forum.xda-developers.com/showpost.php?p=25105156&postcount=22)

  • Michael Goff

    I'll stick with my SGS2 Skyrocket with ICS...It scores 5998 on AnTuTu

  • Guest

    Results are definitely not trustworthy -

    1) Addition of AT&T boot animation increases boot up time for sure!
    2) Looks like AT&T just had an OS re-flash and/or Dalvik cache was cleared  (Android optimizing screen) - definitely will slow down things compared to T3 version which wasn't
    3) S4 version was on Airplane mode?
    4) Look at storage numbers, S4 version has horrible SSD speeds - definitely will slow things down. Maybe AT&T chose cheap NAND

    Not an apples to apples comparison. Not that I'm defending S4, but this is a total unfair comparison. Also, we all know that AnTuTu scales well with the number of cores (not an indicator of day to day Android apps), so there was no doubt that T3 version would've won (look at any AnTuTu benchmark on the web, T3 wins when compared to S4).

  • Guest

    Results are definitely not trustworthy -

    1) Addition of AT&T boot animation increases boot up time for sure!
    2) Looks like AT&T just had an OS re-flash and/or Dalvik cache was cleared  (Android optimizing screen) - definitely will slow down things compared to T3 version which wasn't
    3) S4 version was on Airplane mode?
    4) Look at storage numbers, S4 version has horrible SSD speeds - definitely will slow things down. Maybe AT&T chose cheap NAND

    Not an apples to apples comparison. Not that I'm defending S4, but this is a total unfair comparison. Also, we all know that AnTuTu scales well with the number of cores (not an indicator of day to day Android apps), so there was no doubt that T3 version would've won (look at any AnTuTu benchmark on the web, T3 wins when compared to S4).

  • Guest

    Results are definitely not trustworthy -

    1) Addition of AT&T boot animation increases boot up time for sure!
    2) Looks like AT&T just had an OS re-flash and/or Dalvik cache was cleared  (Android optimizing screen) - definitely will slow down things compared to T3 version which wasn't
    3) S4 version was on Airplane mode?
    4) Look at storage numbers, S4 version has horrible SSD speeds - definitely will slow things down. Maybe AT&T chose cheap NAND

    Not an apples to apples comparison. Not that I'm defending S4, but this is a total unfair comparison. Also, we all know that AnTuTu scales well with the number of cores (not an indicator of day to day Android apps), so there was no doubt that T3 version would've won (look at any AnTuTu benchmark on the web, T3 wins when compared to S4).

    • http://twitter.com/DesignGears DesignGears

       To top it off... That is a very early AT&T test device.

    • http://twitter.com/benbaranovsky Ben Baranovsky

      The only thing to do is to go into an AT&T store and test it for yourself.  I think I will do that.  

    • darkxspm

      Also, at 1:25 a device connects to wifi tethering on the phone on the left...

  • Vladimir

    Constanta? Only Constanta I found with Google is here in Romania. (offtopic, yes, I know, but looked too familiar)

  • http://twitter.com/DesignGears DesignGears

    That is a really old AT&T test device, look in 1080p at the top left, and look at the software version at the end.

    IGNORE

    • Killa504killa

       Ok we get it by now "That is a really old ATT test device." Got it. I will ignore all further comments with regard to the ATT devices age.

  • http://twitter.com/rdlf2048 Rodolfo Ferreira

    To put it simple, it's unfair. Why not a Tegra II-powered device against the S4?
    It's obvious the Tegra III processor will outperform anything dual core, name it Snapdragon 4.

    • darkxspm

      Yeah, your statement totally invalidates this test, except when it's not true, which is all the time.

  • http://twitter.com/DesignGears DesignGears

    Cameron Summerson,

    That is a very early AT&T test device. Very poor decision to throw this up as a "definitive answer".

    • http://www.androidpolice.com/ Cameron Summerson

      Did you even read the article? I clearly said twice that this is *not* a definitive answer. 

      "While I realize that benchmarks are definitely not a definitive answer..."
      "Is this a definitive answer to which device is better? Definitely not."

      We all get that you're upset about the S4 losing, but at least make sure you have your facts straight before trying to call someone out for a "very poor decision."

  • Dmassecar

    I wish he had hit the battery test button...

  • Aaron Gascoigne

    very suspect as my evo 3d sores 6000 at 1.5 ghz with a slower process and no dalvik cahch on the xl in the video so that slowed it down also anyone else notice the 2.6 kernel on th tegra 3 ics is supposed to be on 3.0 meaning nvida hacked it together 

  • ga_st

    there must be something strange going on the S4 as it usually beats or it's really close to tegra 3 performances. i mean,S4 usually performs better than that.

    • http://www.facebook.com/SeverAMV Wilson Ngai

      antutu evaluates the performance of the device as a whole. 

      the benchmarks that show the s4 winning are normally benches that only use one or two cores or only evaluate the performance per core, where the a15 cpu cores of the s4 have the advantage.

  • faceless128

    Why is everyone up in arms? AnTuTu is like the only benchmark the International HTC One X beats the International HTC One S in... it's not surprising that the International HTC One X wins this benchmark.

  • Rsoo42

    this test is rigged, the at&t one x has a 1.5ghz krait not 1296mhz. also the sd card is twice as slow and why would the cpu affect the ram which is under half the speed of the tegra version. i recon the krait one x has been underclocked

    • http://www.facebook.com/SeverAMV Wilson Ngai

      you cant replace the SD cards on these devices, which means that those are the memory cards that htc chose to set them up with.

  • Rsoo42

    this test is rigged, the at&t one x has a 1.5ghz krait not 1296mhz. also the sd card is twice as slow and why would the cpu affect the ram which is under half the speed of the tegra version. i recon the krait one x has been underclocked