• Samsung Caught Inflating Benchmarking Scores Once Again, This Time for the Note 3

    by
    Akshay Masand
    Published on 2013-10-02 06:52 AM
    19 Comments Comments


    Similar to what it did with its Galaxy S4 smartphone, Samsung has been caught lying about benchmarks once again. The South Korean electronics manufacturer has been caught artificially increasing CPU speeds on its Galaxy Note 3 phone when benchmark apps are running according to a report from the folks over at Ars Technica.

    Samsung has once again been found using special code inside its operating system to identify benchmarking apps by name to boost CPU clock speeds and prevent CPu cores from entering low-power modes. This process results in the Galaxy Note 3 benchmarks report CPu performance roughly 20% faster than most apps will actually run on the device. The following was mentioned regarding the performance difference:

    The difference is remarkable. In Geekbench's multicore test, the Note 3's benchmark mode gives the device a 20 percent boost over its "natural" score. With the benchmark boosting logic stripped away, the Note 3 drops down to LG G2 levels, which is where we initially expected the score to be given the identical SoCs. This big of a boost means that the Note 3 is not just messing with the CPU idle levels; significantly more oomph is unlocked when the device runs a benchmark.
    One of Apple’s executives, Phil Schiller, the senior vice president of Worldwide Marketing and the most prolific tweeter among Apple’s senior staff linked to the original Ars Technica report on the matter and said “shenanigans.”

    What do you think of the whole ordeal?

    Source: Ars Technica, Phil Schiller (Twitter)
    This article was originally published in forum thread: Samsung Caught Inflating Benchmarking Scores Once Again, This Time for the Note 3 started by Akshay Masand View original post
    1. Categories:
    2. Apple
    Comments 19 Comments
    1. jose060789's Avatar
      jose060789 - 2013-10-02, 07:24 AM
      So Samsung cheated because they knew they couldn't keep up with the iPhone when it came to the benchmark test... Tusk, tusk, tusk...
      Keep talking more s**t android users, it's hilarious when you guys claim that android is better than iPhone. This is proof android sucks and iPhone is better. SMH
    1. Kaisei's Avatar
      Kaisei - 2013-10-02, 08:34 AM
      Anyone have a screenshot of the Geekbench scores?
    1. Akshay Masand's Avatar
      Akshay Masand - 2013-10-02, 08:45 AM
      Quote Originally Posted by Kaisei View Post
      Anyone have a screenshot of the Geekbench scores?
    1. Jato_BZ's Avatar
      Jato_BZ - 2013-10-02, 09:00 AM
      iPad mini retina will probably blow it to pieces! #LetsGo
    1. lmk's Avatar
      lmk - 2013-10-02, 09:40 AM
      Wonder the Samsung chip in my iPhone 5S will do the same stunts while being tested?
    1. reddemonic's Avatar
      reddemonic - 2013-10-02, 11:53 AM
      does samsung make their own chip? if not how come they don't make for their devices?
    1. bisayakid07's Avatar
      bisayakid07 - 2013-10-02, 12:00 PM
      Quote Originally Posted by reddemonic View Post
      does samsung make their own chip? if not how come they don't make for their devices?
      Yes they manufacture chips but they don't design it not like Apple.
    1. tridley68's Avatar
      tridley68 - 2013-10-02, 12:33 PM
      Quote Originally Posted by jose060789 View Post
      So Samsung cheated because they knew they couldn't keep up with the iPhone when it came to the benchmark test... Tusk, tusk, tusk...
      Keep talking more s**t android users, it's hilarious when you guys claim that android is better than iPhone. This is proof android sucks and iPhone is better. SMH
      Not surprised by this one bit that they would do this
    1. hogcia's Avatar
      hogcia - 2013-10-02, 01:56 PM
      Plastic junk.
    1. Faresismail's Avatar
      Faresismail - 2013-10-02, 02:10 PM
      Benchmark tests are like horse races, the result only differs by 1 to 2 secs sometimes even less which doesn't usually affect any of the overall user experience. Now im not an android fan boy and i think that samsung getting caught is a huge scandal, but i just felt like giving my honest opinion on this matter 😊
    1. Zokunei's Avatar
      Zokunei - 2013-10-02, 02:23 PM
      Quote Originally Posted by Faresismail View Post
      Benchmark tests are like horse races, the result only differs by 1 to 2 secs sometimes even less which doesn't usually affect any of the overall user experience. Now im not an android fan boy and i think that samsung getting caught is a huge scandal, but i just felt like giving my honest opinion on this matter 😊
      I don't know, I saw a video of the 5s and S4 running benchmarks and the iPhone finished quite a few seconds sooner. I'm not sure if the S4 had the benchmark cheating code running though.

      4:11 in this video:
    1. bigbaba's Avatar
      bigbaba - 2013-10-02, 04:14 PM
      Quote Originally Posted by bisayakid07 View Post
      Yes they manufacture chips but they don't design it not like Apple.
      Actually they do design the exynos chip and it is a powerhouse. It manages to compete with the Atom series of processors you find in your netbooks. Unfortunately they can't seem to get it to work with LTE.
    1. Jahooba's Avatar
      Jahooba - 2013-10-02, 05:35 PM
      Quote Originally Posted by hogcia View Post
      Plastic junk.
      For the sake of accuracy I would amend this to say "Malware-filled plastic junk"

      It will be nice once the iPhone gets a bigger screen, then Android users will have nowhere to hide. At this point the tiny screen on the iPhone is definitely a hard criticism to defend.
    1. 8lias's Avatar
      8lias - 2013-10-02, 06:23 PM
      Shamesung shenanigans.
    1. lmk's Avatar
      lmk - 2013-10-03, 02:45 AM
      Quote Originally Posted by bigbaba View Post
      Actually they do design the exynos chip and it is a powerhouse. It manages to compete with the Atom series of processors you find in your netbooks. Unfortunately they can't seem to get it to work with LTE.
      The Exynos Octa core (8 core) works with LTE, and even the faster and newer LTE-A but only available in Korea. From reviews, they run it slower at 1.8 Ghz in order to save battery life.
      What i am wondering is why do they need to "cheat" to run faster on benchmark tests? Rooting the phone and overclock it can boost the speed tremendously, yet the battery can last for more than a day usage. LIke the S4 and New Note3, impressive is the battery life. My iPhone 2, 3, 4 and now 5S cannot last for a day although it is nippy and quick on apps.

      It is just optimization of applications and boosting up processor speed. Turn off power saving mode to run faster.

      It is time that Apple should run away from Samsung made chips.
      So that Samsung cannot copy, steal and worse, saying... "Apple, i made chips for you!"
    1. Moving_Target32's Avatar
      Moving_Target32 - 2013-10-03, 03:36 AM
      You guys need to stop hating on android in comparison to iphones for the iphone 5s is newer lol
    1. magoo657's Avatar
      magoo657 - 2013-10-03, 03:20 PM
      Quote Originally Posted by lmk View Post
      Wonder the Samsung chip in my iPhone 5S will do the same stunts while being tested?
      No, but because Samsung are a bunch of dicks, they might have mad it lower. When it comes to benchmarks.
    1. lmk's Avatar
      lmk - 2013-10-04, 04:34 PM
      Quote Originally Posted by magoo657 View Post
      No, but because Samsung are a bunch of dicks, they might have mad it lower. When it comes to benchmarks.
      LOL...SS made slower chips for Apple so that iPhone runs slower!
      I am so guilty!
    1. alanjf's Avatar
      alanjf - 2013-10-04, 05:27 PM
      Quote Originally Posted by Faresismail View Post
      Benchmark tests are like horse races, the result only differs by 1 to 2 secs sometimes even less which doesn't usually affect any of the overall user experience. Now im not an android fan boy and i think that samsung getting caught is a huge scandal, but i just felt like giving my honest opinion on this matter 😊
      I really don't understand why everyone treats this like a scandal, when it's the point of most CPU benchmarks to test a CPU at highest rated capability, which usually means throttling up to 100% of what the CPU is rated for, otherwise how can you reliably test a CPU's full rated capability?

      So what I'm saying is, it is usually the app itself doing this, whether from an API call or other means. I don't know the specifics of Samsung's hardware, but keep in mind that they don't make the operating system, Google does. I'd like to see what happens on a stock Android firmware swap. It wouldn't surprise me if that doesn't change anything, if it's the benchmark app and not the OS that's initiating the throttle up.

      CPUs running on mobile devices, be they laptops, phones, or tablets, usually run below the rated speeds as to converse battery power and also partly for heat management. Apple's mobile CPUs are no different in this regard. Many benchmarks that are either designed for mobile devices or are at least mobile device aware (at least the ones I've used in the past) will usually test the CPU in both it's normal usage mode and at the full (rated) throttle for that CPU, or at least have this ability. That said, I haven't actually run very many phone or tablet based benchmarks recently myself.

  • Connect With Us

  • Discord

    Join our Discord

  • Twitter Box

  • Facebook