Share this Story

Even Though Benchmarks Are Silly, the LG G2’s Still Crush

LG G2 benchmarks

While we don’t tend to care about benchmarks as much as we used to, we figured we would at least do a couple to give you a rough idea of what to expect performance wise from the LG G2. As you can see from the two I took above, the results are pretty much off the charts, given that the device is running the new Snapdragon 800 processor clocked at 2.3GHz, has 2GB of RAM and also features that GRAM technology that LG bragged about. 

The benchmarks I ran were AnTuTu and Quadrant, two of the most commonly used tests to pit phones against each other. With a five star rating from AnTuTu and the largest Quadrant score I have probably ever seen, it’s a safe bet that the G2 can perform under pressure.

Any other tests you’d like us to perform before we head out?

  • Kevin

    is there a reason why it’s going up against the HTC One X and Motorola Atrix 4G and not something that was released in at least the past year let alone current flagships?

  • srmalin

    I’m making over $7k
    a month working part time. I kept hearing other people tell me how much money
    they can make online so I decided to look into it. Well, it was all true and
    has totally changed my life. This is what I do, http://www.bar28.ℂom

  • Monk

    Yay back to spec wars…… SIGH

  • Brandon Golway

    [Tyrone Biggums] God Damn! [/Tyrone Biggums]

  • Zachary Manville

    Perhaps show it against some current phones… Galaxy Nexus, HTC One X?

  • Detonation

    The amount of stuff in that status bar makes me cringe

  • Torrey Shelton

    My Galaxy S4 benchmarks a 29000 running CM 10.2 at 2.0GHz. And a 26500 at 2.0Ghz 4.2.2

    • deadpenguins

      Did that comment make you feel like you have a bigger pen15?

      • Torrey Shelton

        No im just dont know why he said the highest scores hes ever seen in the story

        • Fred

          Maybe because it is the highest score he has seen…..unless you think he has seen your phone.

    • rahul

      G2 runs 32K without overclocking . If you play a game then run benchmark tool , certainly values gets lower ( in any phone )

      If you over clock G2 , you will certainly get dream figures ….

  • RoadsterHD1

    Question is what’s the limit to be able to handle the pressure of the memory and processor intensive apps. Is 2 cores enough or do we need 4? Is 1 GB of RAM enough or do we need 2GBs. Is 1.2GHz enough or do we need 2.26GHz. Lets look into that…. Anybody have any test results? Opinions?

  • geedee82

    Why is the navigation bar gray?

  • David Du Buron

    Will it blend?

    • Butters619

      But more importantly, can it run Crysis?

  • Droid-life, I LOVE YOU. You are my favorite android blog. PLEASE TELL ME that you aren’t taking payola to review this phone like techcrunch is reporting is happening. PLEASE. PLEASE.

    • RoadsterHD1

      where did you see that? Link?

        • Fred

          Don’t see the problem….cash for doing a review? Didn’t ask them to lie in the review….in the end, paying for advertisement.

        • rahul

          it seems LG offered less to techcrunch .. Apple or samsung may be owning it . Good to see LG put some money in marketing . Optimus G , G Pro , L9 didnt got the reviewers attention what it deserves.

          S3 vs optimus G
          Grand vs L9
          S4 vs GPro

          now atleast they should market G2 to compete with note3

  • RoadsterHD1

    Tim benchmarks are not that silly. It gives us an idea on how the phone will perform when in heavy use and in apps like Google Earth really weigh heavily on your device’s processor and RAM. For instance let’s take the OG Droid and compare it to the LG G2, the one that handles Google earth better is the one with higher AnTuTu and quadrant scores. Right?

    • Jason B

      Google Earth is hardware accelerated via OpenGL ES, so it’s using the GPU more than the CPU.

      The OG Droid would be limited by its GPU and even memory bandwidth. However, if you put the G2’s Adreno 330 in the Droid paired to the old TI OMAP, it’d be CPU limited (FPS will cap out as the CPU can’t feed the GPU fast enough). That’s not to say you wouldn’t see a substantial increase in performance. The peak performance would just be limited by the slow CPU and RAM.

      So, the CPU needs to be strong enough, but the GPU definitely needs the horsepower more. And since the RAM is shared between the two, the bandwidth needs to be there for both.

      • RoadsterHD1

        Well put and you’re right, I was just trying to make a point and used extremes as a example. Evenso, i have been taught about benchmarks and what they really mean by another member of this wonderful community and as a result am feeling better in my choice on buying the new Droid MAXX.


      • Brandon Golway

        IMO the thing that killed the OG Droid was it’s extreme lack of RAM.

  • RoadsterHD1

    Sunspider please

  • iH8sns

    Drop Test please

    • RoadsterHD1


  • DainLaguna

    You want to know the biggest issues these results are going to cause?

    My threshold for lag. Seriously, if this thing lags in any way, shape or form, it will be 100% unacceptable and further prove why optimization and not specs are more important.

    • starnovsky

      Lag is not necessary caused by CPU underperformance. In many cases it could be a result of UI process waiting on something. So lag is a very complex issue and it’s usually takes optimizing all components and full software stack together to fix it, not just slapping on a super fast CPU.

      • trwb

        things running in virtual memory non-natively can cause lag too, which is an issue the Android OS has had since day one.

        • Brandon Golway

          Yep, I feel that Dalvik is the one thing truly holding back Android since all the apps are run in a Java VM.

          • starnovsky

            I’m pretty sure that all Java byte-code is compiled into native on the fly since 2.2.

          • Brandon Golway

            Ah, but still it’s an extra step that it has to go through.

      • DainLaguna

        So…basically you agree with me then? Optimization is where its at and specs aren’t the end all be all to performance?

        • starnovsky

          100% agree.

  • ConCal

    AnTuTu Should not be used as a benchmark. Its by far the worse one.

  • schlanz

    Benchmarks are hardly “silly”. It’s “silly” to assume today’s processing power is all we’ll ever need.

    Even my GN2 gets sluggish when downloading files, playing music, and web browsing simultaneously. I’d love to have a powerful enough device to push through all that multi-tasking with ease.

    • Ben

      Since benchmarks often have no bearing on a real-world scenario, and also remembering what Samsung was caught with just last week, benchmarks are way less valuable than “how does the device feel?”

      • schlanz

        I see your point, and agree that real world use and experience is more valuable than benchmarks but that doesn’t mean they are silly. It shows potential more than anything else, and that still means something especially when there isn’t enough first hand experience out there.

        • Ben

          When companies do things for specific benchmarks, whether it’s kernel optimizations, or the nonsense Samsung pulled with higher CPU and GPU clocks, the benchmarks become useless at best, and outright misleading at worst.

          • Brandon Golway

            They’re not really misleading/useless because it IS possible to achieve those scores, it’s not like they photoshopped the results. All the did is OC the cores.

          • Jason B

            They OC’d the GPU and locked the CPU cores at the max frequency and activated all cores (regardless of CPU usage).

            So, they disabled all of the power saving features that are in all mobile smartphone SoCs and OC’d the GPU. That is misleading because the SoC doesn’t operate like that normally. It power saves often and throttles up/down (maximum clocks on all cores would only be seen if you were doing something VERY intensive).

            If you ran your phone the way Samsung benchmarks, you’d have a dead battery in 4-6 hours. Maybe less. That’s not relative to the real world. It really serves no purpose other than to show the absolute potential of the phone without regard to battery life.

          • Brandon Golway

            Once again I don’t see the problem with that, they’re not really lying about anything because with a little tweaking the phone CAN do that. Most people don’t care about the these things like we do, they want great battery life. Also, I DO run my phone like that when I’m at home (OC’d to 2 GHz max/min, all cores activated when I do anything, GPU at 545 MHz, etc..). Also do you honestly believe that every advertiser is telling you the absolute truth about their products? If you do, you’re extremely naive because they lie to us all the time.

      • Brandon Golway

        Benchmarks are just a metric for comparing one device to another, how else would you do it, if not for benchmarks? Anyone can say “My Galaxy Nexus with XYZ ROM overclocked to 1.5 GHz FEELS faster than my friends Galaxy S4!” but they have no way to prove this, benchmarks at least give someone without physical access to the phones and idea of how fast it is compared to another device.

        • Ben

          When phone manufacturers spend time optimizing for specific benchmarks, those benchmarks become totally worthless. Antutu, Quadrant, 3DMark, and so on. Your example of the Galaxy S4 is perfect, because Samsung just got nailed for locking the CPU at the highest speed while benchmarks run, and it’s specifically looking for those benchmarks.

          • Brandon Golway

            Of course the results are going to be slightly skewed when they come from the manufacturer because they want you to buy their device over the competition’s. End-users have no such motive.

  • chaoslimits

    Benchmarks aren’t “silly”. They should be taken with a (large) grain of salt and used as background info. Leave it to the users to make those decisions.

  • Nicholas DiChiara

    Crushes……1-2yr old devices? lol

  • Daeshaun Griffiths

    i’d love to see moto get their x8 to work on this. MmmmMmm i smell cookies.

  • JDub

    It’s over 9000!!!

  • calum wilper

    any idea on bootloader? dumb question but i’m keeping ym fingers crossed anyway

    • geedee82

      I’m with ya!

  • Swoops212
    • JoshGroff

      I’m not a part of your system.

      • Weber

        This ain’t my dad! This is a cell phone! What, you think I’m STOOPID?

        • Zachary Manville


    • Steve B

      Love the Lonely Island. Freaking hilarious. Happy birthday to the ground!

    • michael arazan

      The one Lonely island video I somehow missed, thx for posting it

  • Endless

    Its over 9000!

    • JDub

      Damn the, your post didn’t show when I was typing mine. Figured someone beat me to it.

  • Swoops212

    Hulk Smash Test Lol, Happy Birthday to the ground!

    • JoshGroff

      I threw the rest of the cake too.

  • Swoops212

    Just yell out “Price Check on the Verizon Model” Do it!

  • ToddAwesome

    Just another reason for me to give the stink eye to my GNex!

  • mustbepbs

    Of course it’s gonna crush phones/tablets from over a year ago. Did you even look at this before you posted it?

    • JoshGroff

      He was talking about the result as compared to other results he’s seen. (which would, unlike the results screen, include this year’s phones.)

      • Exactly. I am not even referring to whatever they put it up against in the app.

        • mustbepbs


  • SmokeNMirrors

    Can you see if it has MHL / display mirroring? Also heard it had USB host mode, be interested to have that confirmed too. TIA

  • Justtyn Hutcheson

    GFx Bench? Sunspider?

    Edit: And that’s a ridiculous number of icons on the notification bar by the way. Wow.

  • aznmode1

    Can you test GFX benchmark Trex screen on

  • Tyler Durden


    • ßen Murphy

      Not quite.

  • Weber

    Do a Hulk Smash test.

    • Inez_carroll325

      My mothers
      neighbour is working part time and averaging $9000 a month. I’m a single mum
      and just got my first paycheck for $6546! I still can’t believe it. I tried
      it out cause I got really desperate and now I couldn’t be happier. Heres what
      I do, http://www.works77.cℴℳ

      • Weber

        Your mother needs to get Hulk-smashed.

        • Franklin Ramsey

          His mother did get Hulk-smashed. That is where he came from.

          • Weber


    • sirosteen

      I make $82h while
      I’m traveling the world. Last week I worked by my laptop in Rome, Monti Carlo
      and finally Paris…This week I’m back in the USA. All I do are easy tasks from
      this one cool site. check it out, http://www.bar28.ℂom

      • Weber

        Like flies at a mf BBQ. Git!

  • yahyoh

    only 27000 bwhahah

    my GS4 I9500 score 29000 @1.6 not @2.2 LOL

  • Eww. Phone already seems outdated to me.

    • Gnex

      Just die already

    • tyguy829

      I really want to believe you are being sarcastic

  • Droid Ronin

    The Quadrant app needs to be updated. It’s ridiculous to be benchmarking against 2011 phones.

    • joejoe5709

      This was my thought. I had to do a double take on the Samsung Nexus S. My brain thought maybe it said Nexus5! Ha!

      • geedee82

        At first I thought it said Samsung Nexus 5 like wtf!

        • michael arazan

          Surprised the original Nintendo,. Atari, and Macontash 2 wasn’t included in the bench marks, meh

    • srmalin

      Jocelyn. if you,
      thought Denise`s article is inconceivable, on monday I got a top of the range
      Toyota sincee geting a check for $7485 recently and also ten k this
      past-munth. no-doubt about it, this really is my favourite job I’ve ever
      done. I began this eight months/ago and pretty much straight away started to
      earn minimum $72.. per/hr. I work through this website, w­w­w.b­a­r­2­8.c­o­m

  • Kevin Gil

    3D Mark! That’s for performance

    • Kevin Gil

      *graphics performance

  • Kyle Cordiano

    3dmark please.