OnePlus reportedly cheated on AnTuTu and Geekbench benchmark tests

OnePlus reportedly cheated on AnTuTu and Geekbench benchmark tests
Both Geekbench and XDA say that OnePlus and Meizu are manipulating the results of benchmark tests being run on some of their handsets. Two handsets said to be involved are the OnePlus 3T and the Meizu Pro 6. Both models can reportedly detect when they are being subjected to a benchmark test, and will run differently in order to produce a better benchmark score. OnePlus defends this action by saying that to improve the user experience on some graphically intensive games, it "implemented certain mechanisms in the community and Nougat builds" allowing the CPU to run faster.

As a result, when these apps, which include AnTuTu and Geekbench, are opened, the processor runs faster and benchmark scores are made artificially high. This isn't the first time that the results of benchmark tests have been questioned. In 2014, Samsung accused NVIDIA of patent infringement and also declared that NVIDIA pumped up benchmark scores of the Tegra K1 chipset. This was supposedly being done in order to make the K1 look more powerful than the Exynos 5433 SoC; the latter was being used at the time on some versions of the Samsung Galaxy Note 4. HTC was also rumored to pump up the CPU performance on certain models when a benchmark test was detected.

Considering that results of AnTuTu and Geekbench tests run on the OnePlus 3 and OnePlus 3T could be tainted, we must question whether the OnePlus 3T really deserved to be ranked as the highest scoring Android phone last year on AnTuTu with an average score above 163,000. Also highly ranked was the OnePlus 3 with an average score above 147,000.

For now, we should place a Roger Maris-like asterisk on these scores until OnePlus can prove that they are genuine. To its credit, OnePlus says that it will remove the trigger that causes the CPU to run faster when certain apps are open, with future builds of OxygenOS for the OnePlus 3 and the OnePlus 3T.


source: Phandroid

FEATURED VIDEO

44 Comments

1. NoToFanboys

Posts: 3231; Member since: Oct 03, 2015

What? This is still a thing? smh

2. kiko007

Posts: 7493; Member since: Feb 17, 2016

Literally couldn't have said it better myself.

10. vincelongman

Posts: 5692; Member since: Feb 10, 2013

All that effort cheating, only to get a 2% better benchmark score lol

40. Finalflash

Posts: 4063; Member since: Jul 23, 2013

On benchmarks like Antutu for that matter....who still cares about Antutu. Geekbench is going the way of the dodo as well ever since they got Apple's seal of approval after they weighted their final scores more heavily for single core performance. PCMark will probably still be more useful than both, but they haven't gotten Apple's approval for probably obvious reasons and likely can't be used as a comparison reference.

9. Arch_Fiend

Posts: 3951; Member since: Oct 03, 2015

Who cares, if oneplus says the software used is made for apps that need more out of the soc like games then I believe them, it make sense that said software would see a benchmark as an app that needs the extra power. Boosting scores doesn't really matter when everyone is using the same chipset anyway, I don't think someone would chose one SD821 device over another due to benchmarks of one being a little higher lol.

12. Nathan_ingx

Posts: 4769; Member since: Mar 07, 2012

I agree. If it doesn't affect the overall usage of the phone...there's no harm. And according to them, it's meant for boosting performance, and not benchmarks. Will removing the software being the score down to significant numbers? I don't think so and I don't care. All I know that it's one of the best phones around. So much for benchmark score mongers.

17. NoToFanboys

Posts: 3231; Member since: Oct 03, 2015

Your point is valid, but my problem here is the fact that they are cheating in something so trivial.

44. Arch_Fiend

Posts: 3951; Member since: Oct 03, 2015

My point was is that they aren't really cheating if the software that boost performance works in other apps besides benchmarks. Now if it was like samsung and htc's old cheating where the boosted performance only happened when a benchmark was running then yes it would be stupid an trivial but in this case Oneplus isn't really cheating because the software they use that boost permanence activates when an app such as a graphically intense game is running or any other app that requires more power to run better. You can fault Oneplus for not insuring the performance boost didn't activate when a benchmark was running but still we don't know if that is possible.

38. michaelny2001

Posts: 328; Member since: Aug 01, 2012

well said. I just hope OP doesn't just do this on benchmarks only and not on games. Ihave the OP3 and i can't say it runs hot during gaming. It gets slightly warm but that's it. Regardless if they do this, i am happy with my OP3 and i am sure all users are too, otherwise they would have switched already.

31. ebilcake

Posts: 1229; Member since: Jul 16, 2016

Sadly yes, FYI last time I checked LG still have an option enabled that degrades image quality games and benchmarks for increased performance. But you can turn it off so that makes everything OK apparently.

3. w1000i

Posts: 247; Member since: Jul 22, 2015

When Any phone allow the CPU to run faster at certain load like games and benchmarks > This is a good thing not bad

14. Nathan_ingx

Posts: 4769; Member since: Mar 07, 2012

...if it does not affect the device in any way, then yes. And I bet this also happens in the PC world to boost performance.

27. ebilcake

Posts: 1229; Member since: Jul 16, 2016

Not exactly, this has been done in the past but using software optimizations which degrade image quality. Not something they do anymore to my knowledge, firstly they know they would get found out, and secondly there isn't enough competition to justify it anymore. A decent PC with sufficient cooling never needs to throttle under load so this type of optimization would be pointless for PC hardware.

26. asdadasfff

Posts: 1; Member since: May 09, 2016

Yes it is. But targeting the package name to change the cpu governor or profile isn't.

4. Podrick

Posts: 1285; Member since: Aug 19, 2015

Lmao!

5. Dr.Phil

Posts: 2360; Member since: Feb 14, 2011

This is why benchmarks should not be taken into consideration when determining whether a device is good or not. It's really, in the end, about the end-user experience and real world performance that matters. And it comes down to software optimization for daily tasks that you will use the most on said device.

6. SYSTEM_LORD

Posts: 1168; Member since: Oct 05, 2015

Cheat or not, this phone is blazing fast. Benchmarks are pointless!

32. ebilcake

Posts: 1229; Member since: Jul 16, 2016

Benchmarks as a whole are not the problem, mobile benchmarks need to measure sustainable performance, as well as peek performance. Do you take a device that benchmarks higher but throttles 40%+ after 20 minutes of use or take the device that scores a little less but barely throttles at all even after an hour. Answer should be easy, sadly people don't have enough information to make that call based on most mobile benchmarks.

39. SYSTEM_LORD

Posts: 1168; Member since: Oct 05, 2015

I agree, that's why I don't worry about benchmark scores. When I see that something doesn't throttle after repeated runs like the Kirin and Snapdragon 801, then that bodes well for the chip and battery life usually

42. ebilcake

Posts: 1229; Member since: Jul 16, 2016

Snapdragon 820/821 don't throttle that much, OnePlus 3/3T barely loses any performance at all over time, the Pixel phones are pretty good in that regard also. A few devices sadly let the side some, well mostly Samsung actually. Samsung love throttling.

7. monoke

Posts: 1161; Member since: Mar 14, 2015

What's the point OnePlus? U already have the highest end soc currently available and one of the more real world snappy device. No need to falsify information. Gaining higher points to look good is really meaningless in the grand scheme of things. People already know the hardware is capable.

8. Derekjeter

Posts: 1488; Member since: Oct 27, 2011

Seriously why are these benchmark test still relevant? If you believe, and use them to buy your next phone, your and idiot and Shouldn't own a smartphone. PA and every other site that still report articles about these test should consider shutting down. How can you be taken seriously as a site? Every company has cheated on these and I still don't know why. The highest scoring phone on these test still lags and every phone will lag no matter how much crap they shove in there. Every iPhone to the 7 Plus that I have now lags. Every Note from the first to the fireball (7) lagged. Bottom line these benchmarks should never be brought up EVER!!!

11. fyah_king unregistered

Dude, the iPhone owners are saying their phones are lag free.:)

15. Derekjeter

Posts: 1488; Member since: Oct 27, 2011

Dude, I've own every phone from the OG Moto Droid, iPhone 4, to the Note 7, Pixel Xl and the iPhone 7 Plus I have now. I also have the Nexus 6p as my daily driver and that being my favorite phone of all time I can say it lags less than others but it still lags. Every phone I have ever used lagged. iFans and Shamesung fans swear their phones never lag but those are the ones that lag the most....lol. I still love my iPhone 7 Plus but my Nexus 6p is just the best Android ever made and it didn't score high on these bull spit test.

21. dimas

Posts: 3363; Member since: Jul 22, 2014

Anyone who say that is a liar, especially when you use them with 20+ apps running on background. And not just iphones, every phone lags. Real multitasking tests occur when you use them after a month or so in real life. Even Xperia's and Sense UI, the smoothest UI for me is not 100% lag free.

18. NoToFanboys

Posts: 3231; Member since: Oct 03, 2015

Before iPhone was winning benchmarks: Fandroids care and brag about benchmark scores. After iPhone was winning bechmarks: iFans care and brag about benchmark scores. Simple as that.

22. dimas

Posts: 3363; Member since: Jul 22, 2014

I was never a fan of benchmark tests, phones are not perfect so there will be few quirks and stutter in the long run use. Except s6 edge plus, that was a really s**tty phone for a flagship price.

13. blingblingthing

Posts: 962; Member since: Oct 23, 2012

Wait a minute. Android manufacturers are cheating on this benchmark yet Apple is so far legitimately producing it's scores? Something sounds highly suspect here. Lower clock speed and less cores. How does Apple do that?

19. NoToFanboys

Posts: 3231; Member since: Oct 03, 2015

"Lower clock speed and less cores. How does Apple do that?" Judging a chip's performance based on clock speed/core count is like judging a phone camera's image output based on megapixels. Look up "Instructions Per Cycle"

20. Zylam

Posts: 1816; Member since: Oct 20, 2010

Of course, Death to Apple. All is fine in the world as long as Death to Apple. Yes.

Latest Stories

This copy is for your personal, non-commercial use only. You can order presentation-ready copies for distribution to your colleagues, clients or customers at https://www.parsintl.com/phonearena or use the Reprints & Permissions tool that appears at the bottom of each web page. Visit https://www.parsintl.com/ for samples and additional information.