Overclock.net › Forums › Industry News › Hardware News › [FP] Samsung Caught Cheating On Benchmark Tests
New Posts  All Forums:Forum Nav:

[FP] Samsung Caught Cheating On Benchmark Tests - Page 15  

post #141 of 145
Quote:
Originally Posted by steelbom View Post

Firstly the majority of benchmarks that exist on a smartphone ARE about real world performance.
Based on WHAT?
Quote:
Originally Posted by Dimaggio1103 View Post

Lol oh really?
Sure seems like it. rolleyes.gif
Misinterpreting. Not my problem. lol
Alienware 17 R2
(10 items)
 
  
CPUGraphicsGraphicsGraphics
i7 4710hq NVIDIA GeForce GTX 970M Intel HD 4600 Zotac GeForce GTX 1070 on Alienware Graphics Am... 
RAMHard DriveOSMonitor
16GB DDR3-3200 750GB Crucial MX300 SSD Windows 10 Home 64-bit 27" ASUS  
KeyboardMouse
Logitech G710 Logitech G602 
  hide details  
Alienware 17 R2
(10 items)
 
  
CPUGraphicsGraphicsGraphics
i7 4710hq NVIDIA GeForce GTX 970M Intel HD 4600 Zotac GeForce GTX 1070 on Alienware Graphics Am... 
RAMHard DriveOSMonitor
16GB DDR3-3200 750GB Crucial MX300 SSD Windows 10 Home 64-bit 27" ASUS  
KeyboardMouse
Logitech G710 Logitech G602 
  hide details  
post #142 of 145
Quote:
Originally Posted by Gigatel View Post


It's unfortunate OEMs feel the need to boost results like this. We"d all appreciated honest tests.

 

 

Quote:
Originally Posted by Gigatel View Post


It's not the same thing at all. If it was, there wouldn't be reports on it. When I overclock my PC, I see those results system wide. Android oems are "overclocking" their phones ONLY when benchmarks run to skew results on review sites and they don't tell you about it.

:thumb:

 

Quote:
Originally Posted by Gigatel View Post


The only debate you're having is on my credibility and side-stepping the obvious facts at hand: OEMs are cheating. It's plain as day. I'm not arguing against documented facts like you, nor am I vehemently attacking anyone, nor am I presenting my opinion as fact, or trying to muddy the waters by bringing up points about Intel or other companies that are simply not relevant.

 

 

Quote:
Originally Posted by Gigatel View Post


I've tried making this point, it's just not sinking in for some people.

This is blatant cheating and nothing more.

 

 

Quote:
Originally Posted by Gigatel View Post


Is that "performance mode" limited to just benchmarks being run?

:thinking:

 

Quote:
Originally Posted by Gigatel View Post


It only kicks in when specific benchmarks are run, which is why this is all so silly.

You'd have zero battery life if it were constantly on.

:thumb:

 

Quote:
Originally Posted by quakermaas View Post
 

 

 

You keep implying this to back up your argument, prove it

 

Now, if it was true, then it really would be an  issue.

 

 

 

 

It is nothing new

 

From this post

http://www.overclock.net/t/1430896/anand-state-of-cheating-in-android-benchmarks/90_30

 

 

 

Quote:
Originally Posted by Gigatel View Post


Yep, secretly run at that speed only during benchmarks.

:thumb:

 

Quote:
Originally Posted by Gigatel View Post


Huh? I didnt start this thread. I was hoping for a rational discussion based in logic and facts, instead we got your trolling comments. Good work. Its typical though...bash the user when youve got no talking points left.

True, but how about these

 

 

:doh: 

 

You're now on block for being the back end of a small horse. 

Money pit
(12 items)
 
  
CPUMotherboardGraphicsRAM
SB-E 3930K @ 4.5GHz & 1.36v Rampage IV Formula R9 290 OC CF DDR3 1866 Dominator Platinum 4 x 4GB (9-11-9-27... 
Hard DriveCoolingMonitorKeyboard
2 x 128GB SSD + 2 x 500GB + 1 x 1TB Heatkiller rev. 3, EK motherboard VRM, Aquacomp... Catleap Q270 27" 2560 x 1440 @ 65Hz Ducky Mini (Snake) 
PowerCaseMouseAudio
Corsair HX850W lian li v2000b plus (black) G400 in MX518 case Audioengine A2 + 10" Sub 
  hide details  
Money pit
(12 items)
 
  
CPUMotherboardGraphicsRAM
SB-E 3930K @ 4.5GHz & 1.36v Rampage IV Formula R9 290 OC CF DDR3 1866 Dominator Platinum 4 x 4GB (9-11-9-27... 
Hard DriveCoolingMonitorKeyboard
2 x 128GB SSD + 2 x 500GB + 1 x 1TB Heatkiller rev. 3, EK motherboard VRM, Aquacomp... Catleap Q270 27" 2560 x 1440 @ 65Hz Ducky Mini (Snake) 
PowerCaseMouseAudio
Corsair HX850W lian li v2000b plus (black) G400 in MX518 case Audioengine A2 + 10" Sub 
  hide details  
post #143 of 145
Quote:
Originally Posted by KSIMP88 View Post

Based on WHAT?

The benchmarks? Most of them are Javascript based benchmarks.

Kasuf
(9 items)
 
  
CPUMotherboardGraphicsRAM
Core i5 6600K ASRock Z170 Pro4 ASUS Radeon RX 480 ROG Strix Corsair Vengeance LPX 32GB 3000MHz 
Hard DriveCoolingMonitorPower
Samsung 850 EVO Noctua NH-D15 LG 34" Ultrawide (LG34UC98) Corsair HX750i 
Case
Silverstone FT05B-W 
  hide details  
Kasuf
(9 items)
 
  
CPUMotherboardGraphicsRAM
Core i5 6600K ASRock Z170 Pro4 ASUS Radeon RX 480 ROG Strix Corsair Vengeance LPX 32GB 3000MHz 
Hard DriveCoolingMonitorPower
Samsung 850 EVO Noctua NH-D15 LG 34" Ultrawide (LG34UC98) Corsair HX750i 
Case
Silverstone FT05B-W 
  hide details  
post #144 of 145
Setting the CPU to be 100% of it's capacity for clock speeds during a benchmark isn't cheating.

I cannot state this enough but: this is exactly how 99% of benchmarks submitted to threads on XDA are performed.

Are those benchers cheating? No. They're ensuring that their scores are the best they can get.

Are Samsung intentionally deceiving anyone? Nope! Your phone will run at 2.3GHz. Games don't need 100% clockspeeds because they aren't a performance stress test.

I can say this with all certainty too, if review websites were allowed to root the test devices they're given, they'd also use a similar method to test. Any review site that wouldn't would be unreliable.


There isn't a single person around who's going to bench their phone on Quadrant and then say "hold on a minute... my real world performance is 20% slower?! That window animation just performed 10% slower. what the deuce?!"
Edited by That Mew - 10/15/13 at 1:06am
post #145 of 145
we are done here.
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Hardware News
This thread is locked  
Overclock.net › Forums › Industry News › Hardware News › [FP] Samsung Caught Cheating On Benchmark Tests