Sidebar Sidebar
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.

Samsung Again Caught Inflating Benchmarking Scores, Phil Schiller Calls 'Shenanigans'

bbeagle said:
Did Apple get off scott-free with that? No. They took a lot of hurt in the press, Consumer Reports said they could not recommend the iPhone because of this (which no doubt caused a lot of lost sales) and Apple gave away free bumpers to everyone who owned an iPhone 4.

Samsung still needs it's punishment.
Click to expand...

Punished for what - people's phones still work the way they should

Apple would have gotten away scott-free if they had their way. And they still tried to spin the public saying they did nothing "wrong." That's why they took a hit. That's why they gave bumpers. And it was their choice to "cover up" the issue. They weren't mandatated.

Samsung NEEDS a punishment? Oh please. For what? And wouldn't that be for Samsung's customers to determine if they felt they have suffered. Let me ask you - do YOU own a Samsung phone that was tested higher than it should have? :rolleyes:
 
iAlphard said:
Wow, it's not cheating people. It runs within specification, not above. And it's not modifying any geekbench 3 data file. It is called power management so the processor can run at maximum performance within specification to get maximum score. What is so wrong about that?
Click to expand...

I don't think you understand what's going on here exactly. They programmed their device to only boost the processor speed when running specific benchmarks. So any comparisons to real world performance would be distorted. It's literally looking to see if a binary called "Geekbench" (or others) is running, and then, and only then, does it increase speed higher than an enduser would ever see when not running a benchmark. You don't see any issue with that?
 
I heard that Charter was doing this when people tried to benchmark their internet connections...

If you can't beat your opponent, cheat. Lessons taught in only the best business schools in America...:eek:
 
gpat said:
Does this seem more of a feature only to me?
When a particularly demanding task, like a benchmark, is executed, the CPU releases more power. I call this "efficiency".
And these numbers came out from the CPU's calculating power, not from nowhere.. how is this fake?
Click to expand...

It's fake because this performance only becomes available when running these Benchmarks. They don't become available when running anything else. The benchmarks are meant to help to display the phones performance to another phone, when this phone lies it makes the other phones it's being compared to look worse when they may not be.

For example in the article this phone is compared to the LG phone, they share an identical CPU and without this meddling by Samsung score the same in the benchmark. But with the meddling the Samsung gains 20% that you'll never ever get in actual applications making the LG look worse when in actual fact it is identical.

This is not a new concept. NVIDIA and ATI used to do this on PC Benchmarks years ago and probably still do today.
 
tanyet said:
They didn't seem to have a problem with it and I don't either. Seems to me they are just making sure it's running at full speed at certain times. What's the big deal. Jeez
Click to expand...

lolwut

You don't run your phone at full tilt all day long, your battery would last 30 minutes and be hot. They did this to spike CPU benchmarks taken by tech reviewers and not impact battery or heat for all other uses, which DO NOT get this boost. The review on Ars is littered with performance complaints and janky UI, ridiculous features, and a gallery app so bad the recommendation is to wait for an update.

Perhaps samsung should have worked on quality rather than trying to spike specsheet hunters.
 
While majority of ppl don't buy a device by comparing benchmark, some does and Samsung just don't want to miss any customers

I don't know why but some people do love to show off using benchmark - this is why some people assemble super over clocked desktop with liquid cooling system plus over powered graphic card just to achieve a very high score - I am not sure whether those people are simulating nuclear weapon test lol

Samsung would probably say it is not their fault as it is the job of benchmarking app designer to detect anyone who cheat. This is Samsung, what can you expect anyway,,,,
 
blackcrayon said:
I don't think you understand what's going on here exactly. They programmed their device to only boost the processor speed when running specific benchmarks. So any comparisons to real world performance would be distorted. It's literally looking to see if a binary called "Geekbench" (or others) is running, and then, and only then, does it increase speed higher than an enduser would ever see when not running a benchmark. You don't see any issue with that?
Click to expand...

Umm why is it wrong to boost the processor speed (withing processor specification and not overclocked) when running a specific app?
It is the same like when you run a game. Any device will go hot, because the processor usage is higher than normal to ensure no lag present.
The way i see it, here benchmark is like a sport for phone. Some people like it, some not. But when you do sport, you like to compete and giving your best. And is it wrong to do that? Unless if that person takes a drug or modified any component in that sport so he can get any advantage.
 
Rafterman said:
I would say a phone that has poor signal quality is a lot worse than shenanigans over benchmark tests.
Click to expand...

I would disagree completely. Apple was never at fault, only assumed to be. Samsung here is clearly guilty.

Somehow the loyalty and obedience of Samsung fans to Samsung products is far more than any Apple fans love for Apple ever was. Samsung is the 2nd coming according to all it's followers and can do no wrong. I don't get this ludicrous frenzy mentality.
 
ActionableMango said:
I am astonished that there are people who would pick a phone based on synthetic benchmark results. That is in no way excusing Samsung's deception, I never said any such thing, and I don't appreciate the implication.
Click to expand...
Like I said, it's doubtful that the benchmark is the primary criteria, but it is a criteria in the decision making process.

Sorry if I mistook your displeasure with people seeking 3rd party comparisons on a phone's performance in their decision making process, with an excuse of Samsung. But you have to see how easily that mistake could have been made.
 
dugbug said:
lolwut

You don't run your phone at full tilt all day long, your battery would last 30 minutes and be hot. They did this to spike CPU benchmarks taken by tech reviewers and not impact battery or heat for all other uses, which DO NOT get this boost. The review on Ars is littered with performance complaints and janky UI, ridiculous features, and a gallery app so bad the recommendation is to wait for an update.

Perhaps samsung should have worked on quality rather than trying to spike specsheet hunters.
Click to expand...


I've had a Note 3 for almost a week now and I've noticed none of this. I've been using an iPhone for 5 years now and loved it but I wanted a bigger screen. I was really worried about lag and micro stutters because I was used to IOS but I've noticed none. It actually seems smoother than my iPhone 5 running IOS 7.
I've also seen no other reviews that were so critical. I'm sure they are out there but I haven't seen them.
 
gpat said:
Does this seem more of a feature only to me?
When a particularly demanding task, like a benchmark, is executed, the CPU releases more power. I call this "efficiency".
And these numbers came out from the CPU's calculating power, not from nowhere.. how is this fake?
Click to expand...


Because.... Anybody could do what Samsung did, Apple included.

Real question is, why don't they turn this special turbo mode on all the time. For a couple of reasons, heat and battery life.

A real benchmark is one where you're measuring during normal use.
 
Piggie said:
I'm pretty sure either AMD or Nvidea have been caught....

Few years ago, but one of them (I think AMD) was caught sneaking graphics effects down during certain benchmarks.

Perhaps someone can clarify this?

found some:

AMD Cheating?

http://www.pcauthority.com.au/Featu...s-and-degrading-game-quality-says-nvidia.aspx

INTEL Cheating ?

http://techreport.com/review/17732/...loy-questionable-3dmark-vantage-optimizations

NVIDEA Cheating?

http://www.geek.com/games/futuremark-confirms-nvidia-is-cheating-in-benchmark-553361/

APPLE Cheating ?
http://www.theregister.co.uk/2003/06/24/apple_accused_of_cheating_over/


Perhaps Samsung is not so different after all huh ?
Click to expand...

Hey! this is an anti Samsung thread! No facts allowed here!
 
iAlphard said:
Wow, it's not cheating people. It runs within specification, not above. And it's not modifying any geekbench 3 data file. It is called power management so the processor can run at maximum performance within specification to get maximum score. What is so wrong about that?
Click to expand...

Read every post response to me when I made a similar point.

It's impossible to have an unbiased opinion when it comes to Samsung/Apple. You're one side or the other on this forum.
 
gpat said:
Does this seem more of a feature only to me?
When a particularly demanding task, like a benchmark, is executed, the CPU releases more power. I call this "efficiency".
And these numbers came out from the CPU's calculating power, not from nowhere.. how is this fake?
Click to expand...

Uh, you're missing the point. The CPU does NOT release more power at this level when the device needs it. It ONLY does it when a particular benchmark software is running. The CPU looks for these benchmark softwares and jack up performance ONLY then. And under those circumstances battery life is absolutely atrocious and the device gets so hot that it gets in damage territory and you can't even hold it in your hand. Playing HD videos, powerful games, or literally no other CPU intensive task will put the CPU into that mode and for obvious reasons. ONLY running "x" benchmark will.

So they're trying to get battery life and everything else tested under one limit while CPU tests get an entirely different limit that the battery/heat/life-threating device tests can't see.
 
First they make fake iPhones and iPads and now they make fake benchmarks.

I told you Samsung was fakers.
 
stynkfysh said:
Anyone else notice that the Note 3's 'normal' mode has a higher overall score than the iPhone 5S? Per the results on MacWorld.com. I didn't buy Geekbench app and test myself. Only like 2 points though.
Click to expand...

Geekbench hasn't been updated to for ARMv8 so benchmark is running in 32bit mode.

Should be noted, and retested when Geekbench is updated to reflect ARMv8.
 
A few months ago both Kia and Huandai had to admit for inflating gas mileage ratings in their cars. I wonder if this is a nation wide practice?

----------
 
You must log in or register to reply here.
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.
Top Bottom
Back