Asus, HTC and LG also caught boosting Android benchmark scores

The HTC One's processor speed goes up when certain benchmark programs are used.

Tuesday's report on Ars Technica that showed the processor in Samsung's Galaxy Note 3 boosted its speed when it detected a benchmark program was running was apparently just the tip of the iceberg. A follow-up investigation on AnandTech shows that devices from Asus, HTC and LG all have been found to do the same thing.

Specifically, tests on the Asus Padfone Infinity, the HTC One, the HTC One mini, and the LG G2 showed that the processor speed went up when those devices detected the AnTuTu and Vellamo benchmark programs running on their hardware. The HTC One and HTC One mini also showed increases in speed when the 3DM and GFXB 2.7 benchmarks were used.

However, the end effect of all these changes was pretty small in terms of skewing the results. The scores went up about five percent in CPU data and up to 10 percent in GPU tests. The good news? Google's phones such as the Nexus 4 and 7, and its Motorola Moto X and RAZR i showed no such increases in chip speed while running benchmarks. You can bet that reviewers that use these programs to test mobile phone performance won't take the results they generate at face value from now on.

Source: AnandTech

Report a problem with article
Previous Story

Dell announces upcoming Windows 8.1 notebooks and tablets but no new Windows RT products

Next Story

Microsoft is selling out of Surfaces, backorders starting to pile up ahead of retail launch

157 Comments

Commenting is disabled on this article.

I've called so many other people idiots for using their substandard phones, that I can't admit that this matters, I'd look like a fool.

Aren't users supposed to change the governor to "Performance" (which holds the cpu's at max frequency) when running benchmarks?
What's the difference here?

At the end of the day your not going to do 3D modelling on your phone... i would personally be more bothered about how long i can actually use the phones smart features for, not if it scores a few more points in a benchmark.

All you've done is prove that people are indifferent to which companies are cheating the tests. It was frowned upon then, and it's being frowned upon now.

The saying "two wrongs don't make a right" comes to mind.

dead.cell said,
All you've done is prove that people are indifferent to which companies are cheating the tests. It was frowned upon then, and it's being frowned upon now.

The saying "two wrongs don't make a right" comes to mind.


But the funny thing here is that Samsung haven't cheated.

If you want to know why, read my reply where i wrote what another user on the Ars Technica's website about the same case said.

It's pretty obvious on why Samsung and the others do this.

This was posted by a user on the article about this on the Ars Technica's website it self.

BehindTheTruth wrote,

For the first time in my life I see how Ars and its valued commentators are full of sh*t.

CPUs in most modern smartphones under ARM/Linux are managed by frequency governors like ondemand or powersave - those governors have a certain reaction time - i.e. they don't instantaneously raise the CPU speed when there's a 100% cpu load - they do it with a certain delay.

What this whole situation shows us, is that most if not all benchmarks do NOT properly heat up the CPU before performing computations on it, thus the CPU can run at lower speeds.

Samsung does not cheat here - they just fight with CPU governor's latencies and benchmarks' faults.

So, f*ck you Ars for being unprofessional and not properly researching the topic before posting your "results".

// vonikhsaT .S metrA

So yeah, Samsung, HTC, LG and ASUS is not at fault here.

Edited by Exynos, Oct 3 2013, 4:06am :

Now having said all that, this test doesn't affect my buying decision. I have a Note II and I plan to get the Note III because of how I use the phone like most, none of those tests scores mean a cotton-pickin thing. They are just numbers on paper like the binary code.

If you showed those scores to people the vast majority wont even care.
Apple typically have an upperhand on Sunspider and Antutu simply because of how they optimize their software to deal with Java and other software application tweaks.

In those tests, raw CPU/GPU power really has no effect other than it pushed through the code faster so it scored better.

The reality is, on many web pages, java can hog up a CPU. Even on my Note I have loaded pages that pop up a message that the java routine is still loading and is taking a while would I like to stop it. If the CPU was clocked faster, it would simply move through the script faster.

The issue is in Samsung native browser, thy simply haven't found a way to skip code that isn't even needed in Java routines. Apple has which is why they get better performance. if the benchmarking was based on raw speed, I am sure a quadcore phone would dust a dualcore phone, especially a quadcore phone with 3GB of RAM vs a dualcore phone with only 1GB of RAM

@Exynos @Serrasonic - The fact is both of you are right and both of you are wrong.

There are different types of benchmarks. In the area that Exynos is trying to say, he is right for the definition of a - Component Benchmark/ micro-benchmark. Those tests are solely based on the hardware configuration and power.

I believe the benchmark that was performed on these devices was a - Synthetic Benchmark

Since Sunspider and Antutu tests are all browser/software based they would be synthetic. In those tests having faster hardware just means the hardware will push through the code faster.

PCMark is a component benchmark which pushes hardware to its most possible capability to render a score.

In software tests in this case I would say and I am not sure I am right is this...Basically when Samsung java script forced all 4 cores to run at 2.3Ghz, it basically was able to complete the routine faster thus scoring better. Since natively you wouldn't be running those routines at full speed, the results are not fair and thus it is cheating because the result are artificial.

However, if this had been a hardware test and Samsung forced the CPu to run at 2.3Ghz which is full speed and the other phones failed too, then it wouldn't be cheating or faking numbers because the Samsung device would have been forced to fun at full speed under heavy load which is what you yourself described.

The fact is you are right if this had been a PC Mark test where hardware capability was what was being benchmarked. In this case that is not what it was. Sunspider and Antutu are not component based benchmarks.

Sunspider is a javascript benchmark. Antutu I am not sure. I do know, PC Mark is a component benchmarking tool much like those used on PC's. Geekbench os a benchmarking tool that is suppose to perform real world tasks. Which means when it calls a routine it is suppose to let the hardware perform that routine as if you actually loaded the app. Which means for a certain the chances the CPU would run all 4 cores at full speed is not going to happen. However, having said that there are exceptions. Samsung made the GS4 so that any app that takes full screen will run the CPU at its full 533 GPU speed. In the tests done above, those were not GPU tests.

However, this is where Samsung didn't cheat. As a fact, which Samsung also stated about the GS4, the GPU is designed that if any app has to run in full screen the GPU will automatically run at full speed no matter what the app is. This is true of games and something as simple as a browser. So since benchmarking software run in full screen which means the status bar is not visible, that means the phone is going to run at full speed anyways in many tasks.

I do notice on my Galaxy Note II, that if I open a browser in full screen, my phone does get equally as warm as it does when I play games in full screen over time. So I would say Samsung is likely being truthful.

The fact is however, in those tests the phone was design to run a full screen app at full speed. The tester is saying this is cheating. But again if Samsung designed their phone to do this with every app then it is not. The problem in this case, Samsung whitelisted certain apps in the java script that made sure when they were opened the phone ran the CPU at full 2.3Ghz and normally this would not happen in normal use.

Much as I hate to admit it, they are right for this test. That is cheating.
I think what you should so is read the different types of benchmarking tests and you will see why in this case they are right.

For those who I said were wrong in the previous article, my apologies. In this instance you are right as these tests are suppose to be based on real world usage.

If this had been a PC Mark test and Samsung wanted to make sure the hardware was being used to its full potential, then the script would have been acceptable.

HipHopSinceFriday said,
Much as I hate to admit it, they are right for this test. That is cheating.
I think what you should so is read the different types of benchmarking tests and you will see why in this case they are right.

For those who I said were wrong in the previous article, my apologies. In this instance you are right as these tests are suppose to be based on real world usage.

If this had been a PC Mark test and Samsung wanted to make sure the hardware was being used to its full potential, then the script would have been acceptable.

Thank you.

My only surprise is people didn't know this is how the game works. Skip any and all marketing and only pay attention to actual independent testing. Really you can skip benchmarks for phones completely. It's really not going to make any difference in your texting.

I agree. Benchmarks were rather irrelevant anyway, but with all these mixed up standards of how to test, they're completely useless now.

For anyone who has not seen the actual tests, here are screenshot - http://www.slashgear.com/galax...igans-accusations-01299884/

In the tests what it shows is Samsung simply made sure all the cores ran at the native 2.3Ghz speed in each of the tests. Even without this the Galsaxy Note III still beat all the other devices. Insteading of winning by a nose, they won but a full car length. They didn't cheat. Benchmarks are suppose to push the hardware to its full potential. That is actually not happen in the tests as you can see that in some tests 2 or 3 cores were totally idle in the middle of the test which is wrong. All cores should be active. Also you can see in the tests that not all the cores were running at full speed. Again this is misleading because the whole point of a benchmark is to push the hardware to its full potential.

If you only want user results, just place a phone next to each other and load apps and browser web pages. You wouldn't even do a benchmark to begin with.

The fact is the vast majority of people here don't even understand benchmark tests, so they can't possibly be saying Samsung cheated because they didn't. The images speak so and if I was Samsung I would be suing ARC for slander.

It's not the majority of people here that don't understand benchmarking, it's editors on anandtech and other sites.

So clearly a LOT of people see it as samsung and other android manufacturers cheated.

So far the exceptions are Motorola and Apple.

HipHopSinceFriday said,
For anyone who has not seen the actual tests, here are screenshot - http://www.slashgear.com/galax...igans-accusations-01299884/

In the tests what it shows is Samsung simply made sure all the cores ran at the native 2.3Ghz speed in each of the tests. Even without this the Galsaxy Note III still beat all the other devices. Insteading of winning by a nose, they won but a full car length. They didn't cheat. Benchmarks are suppose to push the hardware to its full potential. That is actually not happen in the tests as you can see that in some tests 2 or 3 cores were totally idle in the middle of the test which is wrong. All cores should be active. Also you can see in the tests that not all the cores were running at full speed. Again this is misleading because the whole point of a benchmark is to push the hardware to its full potential.

If you only want user results, just place a phone next to each other and load apps and browser web pages. You wouldn't even do a benchmark to begin with.

The fact is the vast majority of people here don't even understand benchmark tests, so they can't possibly be saying Samsung cheated because they didn't. The images speak so and if I was Samsung I would be suing ARC for slander.

Yes, it is the benchmarks job to push the system to 100%, WITHOUT HAVING THE SYSTEM ALTERED SPECIFICALLY FOR IT!

The problem is that benchmarks do not run consistently on these devices now as some benchmarks will have the full 100% cores, while others will be throttled since they aren't in the list of special apps. Why do you think it's fair for these manufactures to pick and choose how to represent their device performance over how the devices actually work?

stevan said,
I think iOS 7 is better than android.

Why?

And did you actually look at the link i posted?

It's crazy beyond anything that an OS in 2013 have that many limitations.

Exynos said,

Why?

And did you actually look at the link i posted?

It's crazy beyond anything that an OS in 2013 have that many limitations.

I did, there's definitely things on there that could be useful on the ios but nothing that I care too much.

I guess I was just trying to say that we all have our opinions on which is better. And last time we checked, opinions were not facts.

LOL, what can i say?

Welcome to the benchmarking world Ars Technica. It's pretty obvious that Ars Technica doesn't have a single clue about what the main purpose of a benchmarking app is and how it's supposed to run on the hardwares.

Edited by Exynos, Oct 2 2013, 9:07pm :

Exynos said,

Because?
Because benchmarking is to test the devices speed in the state that the all apps run at, if there are special exceptions made for the tools that test the speed of the device, then those apps are running faster than any other app on that device will run.

If benchmarks were to test ONLY the hardware, they wouldn't be run at the OS level, they would work before any software is loaded and run all calculations based on hardware alone.

Since these benchmarks run at the OS level, to be fair they should run at the same performance that all other apps run at. Otherwise they are falsely reporting that apps will run faster than they will.

don't bother, he wont get it or understand it. He owns a g4.

SierraSonic said,
Because benchmarking is to test the devices speed in the state that the all apps run at, if there are special exceptions made for the tools that test the speed of the device, then those apps are running faster than any other app on that device will run.

If benchmarks were to test ONLY the hardware, they wouldn't be run at the OS level, they would work before any software is loaded and run all calculations based on hardware alone.

Since these benchmarks run at the OS level, to be fair they should run at the same performance that all other apps run at. Otherwise they are falsely reporting that apps will run faster than they will.

SierraSonic said,
Because benchmarking is to test the devices speed in the state that the all apps run at, if there are special exceptions made for the tools that test the speed of the device, then those apps are running faster than any other app on that device will run.

If benchmarks were to test ONLY the hardware, they wouldn't be run at the OS level, they would work before any software is loaded and run all calculations based on hardware alone.

Since these benchmarks run at the OS level, to be fair they should run at the same performance that all other apps run at. Otherwise they are falsely reporting that apps will run faster than they will.


No, doing benchmark is not to test the speed of what other apps might run the CPU and GPU at. It's to determine how powerfull your device / computer is independent of what other apps are doing with your device / computer.

If i run the latest PCMark app on my computer, it wont do the benchmarks after how others normal apps use my hardwares, it's there to gain information and test my computer to see how powerfull it is at full speed (100%) on the CPU, GPU and the RAM.

And no, benchmarks are still there to test your hardware while you are in the OS. Ofc if there had existed benchmarking tools that let you run some benchmarks from lets say your BIOS or your UEFI, i wouldn't mind those. That would be absolutely fine by me.

But that doesn't exist, so the benchmarking apps have to be runned from the OS it self to be able to do the different GPU / CPU tests.

So no, no benchmark apps should run at the same power as normal apps. You have clealy misunderstood the whole concept of doing benchmarking if that's the case.

Edited by Exynos, Oct 2 2013, 10:43pm :

rippleman said,
don't bother, he wont get it or understand it. He owns a g4.


Yeah, i own a Galaxy S4, but i own a pretty good computer to that i often do benchmarking on. So just because i own a Galaxy S4, it doesn't mean i don't know how benchmarking works.

Again, benchmarking is there to determine how powerfull your hardware is, not to see how your computer run at moderate usage lol.

Exynos said,

No, doing benchmark is not to test the speed of what other apps might run the CPU and GPU at. It's to determine how powerfull your device / computer is independent of what other apps are doing with your device / computer.

If i run the latest PCMark app on my computer, it wont do the benchmarks after how others normal apps use my hardwares, it's there to gain information and test my computer to see how powerfull it is at full speed (100%) on the CPU, GPU and the RAM.

And no, benchmarks are still there to test your hardware while you are in the OS. Ofc if there had existed benchmarking tools that let you run some benchmarks from lets say your BIOS or your UEFI, i wouldn't mind those. That would be absolutely fine by me.

But that doesn't exist, so the benchmarking apps have to be runned from the OS it self to be able to do the different GPU / CPU tests.

So no, no benchmark apps should run at the same power as normal apps. You have clealy misunderstood the whole concept of doing benchmarking if that's the case.

OK, name one PC Benchmark that changes your SpeedStep, Power Profile, and Raises the thermal Limit of GPU. If you can show me ONE that does everything that the Samsung phone does to make it MORE POWERFUL than YOUR SYSTEM REALLY RUNS AT, then I will believe you.

You won't find one though, because all benchmarking software benchmarks YOUR CURRENT SETUP.

SierraSonic said,
OK, name one PC Benchmark that changes your SpeedStep, Power Profile, and Raises the thermal Limit of GPU. If you can show me ONE that does everything that the Samsung phone does to make it MORE POWERFUL than YOUR SYSTEM REALLY RUNS AT, then I will believe you.

You won't find one though, because all benchmarking software benchmarks YOUR CURRENT SETUP.


And my current setup runs at the maximum speed when i run a power intensive app. It runs the CPU and GPU to the maximum speed (100%) then in the same way as Samsung, HTC and LG's scripts does the same with their smartphones, but with another twist to go around the speed throttling that makes the benchmark results to be false.

So my question is, is there any differences that the benchmarks apps runs my speed at 100% on my CPU and GPU on my computer than it does the same on the smartphones but with a script?

Why should benchmarks be different on smartphones than benchmarks on computers are?

Benchmarks are benchmarks no matter how you want to twist it.

Exynos said,

And my current setup runs at the maximum speed when i run a power intensive app. It runs the CPU and GPU to the maximum speed (100%) then in the same way as Samsung, HTC and LG's scripts does the same with their smartphones, but with another twist to go around the speed throttling that makes the benchmark results to be false.

So my question is, is there any differences that the benchmarks apps runs my speed at 100% on my CPU and GPU on my computer than it does the same on the smartphones but with a script?

Why should benchmarks be different on smartphones than benchmarks on computers are?

Benchmarks are benchmarks no matter how you want to twist it.

The point of the Benchmark is to stress the unit, you are correct, but isn't the built in limit how far you want to stress the device as any program won't be able to surpass that limit anyway? IF you choose to remove the limit yourself by rooting or changing the rom so be it, that will give you the extra performance all around. But since these phones remove the limits ONLY FOR BENCHMARKING this is not the true limit of your device in its current setup as no program will run at it.

Benchmarks test your current system setup, that's all they ever have tested, the second anything else alters that setup, you get false results, as they may not be fairly matched up against another setup, such as phones that don't change their settings just for benchmarks.

Here's the difference, your CPU and GPU and RAM all max out on your PC because your PC is SETUP to do that, your phone ISN'T until these manufactures decided to deceptively alter the way their phones work, but only for these programs. If your PC benchmark software turned all your computers settings to the fastest, but with the lowest quality while also removing the thermal limiters from your PC and gives your equipment extra voltage you would think your computer is faster than it is at it's current state.

If they came forward, and warned the users, the benchmark makers, and everyone pretty much, they would know that the benchmarks are less accurate and they would LOSE FAITH in these results for these devices. I am sure eventually these benchmark apps will just warn users about deceptive tactics and/or not accept benchmarks on these devices to their verified servers.

SierraSonic said,
The point of the Benchmark is to stress the unit, you are correct, but isn't the built in limit how far you want to stress the device as any program won't be able to surpass that limit anyway? IF you choose to remove the limit yourself by rooting or changing the rom so be it, that will give you the extra performance all around. But since these phones remove the limits ONLY FOR BENCHMARKING this is not the true limit of your device in its current setup as no program will run at it.

Benchmarks test your current system setup, that's all they ever have tested, the second anything else alters that setup, you get false results, as they may not be fairly matched up against another setup, such as phones that don't change their settings just for benchmarks.

Here's the difference, your CPU and GPU and RAM all max out on your PC because your PC is SETUP to do that, your phone ISN'T until these manufactures decided to deceptively alter the way their phones work, but only for these programs. If your PC benchmark software turned all your computers settings to the fastest, but with the lowest quality while also removing the thermal limiters from your PC and gives your equipment extra voltage you would think your computer is faster than it is at it's current state.

If they came forward, and warned the users, the benchmark makers, and everyone pretty much, they would know that the benchmarks are less accurate and they would LOSE FAITH in these results for these devices. I am sure eventually these benchmark apps will just warn users about deceptive tactics and/or not accept benchmarks on these devices to their verified servers.


Again, you are misunderstanding what i'm saying. What i'm saying is that Samsung does like everyone else. They just forces the CPU and GPU to run at the full speed while heavy apps are running. What the apps will use out of the CPU or GPU shouldn't have anything to do with how Samsung allows the some apps to go around the speed throttling.

When i sit idle on my computer, then my CPU speed on my Intel Core i5-2500k CPU goes down to like 1.6-1.8 GHz to save power. The second i run any apps, my CPU speed goes up to 3.7 Ghz while the apps loads or when i use the apps actively.

If i'm starting a GPU benchmark application on my computer, then both of my CPU and GPU will speed up to it's maximum default speed, witch are completely normal.

And by this, why is it cheat or wrong that Samsung allows some heavy apps to do the same by speeding up the speed on the CPU and the GPU up to the maximum default speed to prevent the speed throttling to take action on them?

Why is this wrong when every benchmarking apps on computers speeds up the CPU and GPU to the maximum when they are being run?

Is that because someone are extremely butthurt over that Samsung does a better job by optimizing it's TouchWiz ROM's so it runs really good over what other brands does?

Again, every benchmarks apps to computers are designed to use the full potential of the hardwares in our computers, and it's the same for every smartphones to. It's just that Samsung, HTC, ASUS and LG have to use this script to overrun the speed throttling systems that are in the smartphones today.

Yes, the speed throttling on smartphones is a bit different than it is on CPU's on computers, and that's why they have to do this to not let the speed throttling downgrade the benchmarks results because the benchmark apps couldn't run at full speed because the speed throttling system prevented it.

This is something you should understand.

If this is wrong on smartphones, it should also be wrong for every benchmark apps out to computers that forces the CPU's and GPU's to go on full power when they are run.

It's that simple.

SierraSonic said,
Because benchmarking is to test the devices speed in the state that the all apps run at, if there are special exceptions made for the tools that test the speed of the device, then those apps are running faster than any other app on that device will run.

If benchmarks were to test ONLY the hardware, they wouldn't be run at the OS level, they would work before any software is loaded and run all calculations based on hardware alone.

Since these benchmarks run at the OS level, to be fair they should run at the same performance that all other apps run at. Otherwise they are falsely reporting that apps will run faster than they will.

So you would be wrong.
The true definition of a benchmark is - Benchmarks are designed to mimic a particular type of workload on a component or system. Synthetic benchmarks do this by specially created programs that impose the workload on the component.

Exynos said,
What you are confusing here is what Samsung is doing and what PCs do is completely different. It's funny that you call this cheat an optimization, if that was true it would work for all heavy processing apps and not just benchmarks, just like PCs. You said your computer does this for all applications, what these phone manufacturers did is make it strictly for benchmarks.

I'm not arguing what they did, of course they made the device the fastest it could be for the benchmark. The reason this was done was to simply give the device a better score than the competition. Realize only a select few devices change their behavior for benchmarks, and they are all in the wrong.

Benchmarks are designed, and designed in mind for testing your current system setup. If something changes your current setup just for benchmarks, then the benchmark is tainted no matter how you look at it. There is a reason why PCMark started rejecting certain windows 8 setups, and why other benchmarks rejected other results from certain devices and/or setups.

Why don't you get that this was an underhanded way about it?

Edited by SierraSonic, Oct 3 2013, 1:55am :

HipHopSinceFriday said,
So you would be wrong.
The true definition of a benchmark is - Benchmarks are designed to mimic a particular type of workload on a component or system. Synthetic benchmarks do this by specially created programs that impose the workload on the component.
http://dictionary.reference.com/browse/workload
work·load [wurk-lohd]
noun
the amount of work that a machine, employee, or group of employees can be or is expected to perform.
Also, work load.

Yes, now where does it say here that the device the workload is on should be altered specifically to boost the way this workload is done?

SierraSonic said,
What you are confusing here is what Samsung is doing and what PCs do is completely different. It's funny that you call this cheat an optimization, if that was true it would work for all heavy processing apps and not just benchmarks, just like PCs. You said your computer does this for all applications, what these phone manufacturers did is make it strictly for benchmarks.

I'm not arguing what they did, of course they made the device the fastest it could be for the benchmark. The reason this was done was to simply give the device a better score than the competition. Realize only a select few devices change their behavior for benchmarks, and they are all in the wrong.

Benchmarks are designed, and designed in mind for testing your current system setup. If something changes your current setup just for benchmarks, then the benchmark is tainted no matter how you look at it. There is a reason why PCMark started rejecting certain windows 8 setups, and why other benchmarks rejected other results from certain devices and/or setups.

Why don't you get that this was an underhanded way about it?


Running a benchmark on a smartphone shouldn't be any different that it is on a computer. A benchmark still have the same purpose on both things. So i can compare Samsung and computer as much as i want then.

And secondly, just because some others doesn't do it the same way as Samsung, HTC, LG and ASUS, it doesn't mean it's the wrong thing to do. Because it would be more wrong to let the benchmark give you false results on 60% of what the hardware is capable of instead of letting them give you the results from when it ran at full speed (100%).

So yes, any benchmarks apps are made to let the tests runs at full speed.

However, just because Samsung and the others opens up the apps to use 100% of the CPU and GPU speed, it doesn't automaticly means that the benchmarks apps will use 100% of the speeds available then constantly.

Even Ars Technica's article about this shows that the CPU speed changes alot under the benchmarks. But like it have been said many times by now, the script that is used just makes the CPU and GPU to be open to be used at full speed without having the fear of letting the speed throttling system downgrade the available speeds for those apps.

Why you don't know this easy and basic stuffs isn't good to say, but i'll guess someone is a totally noob when it comes to this.

So what are all the whining and crying about when there is only one purpose of a benchmarking tool and when it's quite obvious that any benchmarking tools are supposed to run at the hardwares full speed regardless of the fact that some scripts are used or any other methods that just makes sures the maximum default speed can be used for several apps that just needs that to achieve what they are made for.

Exynos said,
*some insults*.

The benchmark works exactly the same as it does on the computer... The thing that is different is the way the device is setup. The default way of doing it is by using no special tweaks on any platform as it always was frowned upon. Like when ati/nvidia did it, like when other companies did it, they all got shamed and quickly changed their ways.

Apple and others do not have to stoop to using a fake profile just for benchmarks, why do these companies have to use something like this? Is it really worth the bad publicity that usually comes from this? All for a 4.4% gain?

The point is to be more consistent and fair it should fall victim to the throttling, as that is the true active profile of the devices. The scirpt is the problem, and mainly because it unfairly changes the way the device actually runs, and only for these tests, not for other apps.

SierraSonic said,

The benchmark works exactly the same as it does on the computer... The thing that is different is the way the device is setup. The default way of doing it is by using no special tweaks on any platform as it always was frowned upon. Like when ati/nvidia did it, like when other companies did it, they all got shamed and quickly changed their ways.

Apple and others do not have to stoop to using a fake profile just for benchmarks, why do these companies have to use something like this? Is it really worth the bad publicity that usually comes from this? All for a 4.4% gain?

The point is to be more consistent and fair it should fall victim to the throttling, as that is the true active profile of the devices. The scirpt is the problem, and mainly because it unfairly changes the way the device actually runs, and only for these tests, not for other apps.


If you read my post before this post, you will see why Samsung and the other brands are using this type of script. It's a good reasons for it.

It's basicly better that they use that type of script than not using it for benchmarking stuffs.

Exynos said,

If you read my post before this post, you will see why Samsung and the other brands are using this type of script. It's a good reasons for it.

It's basicly better that they use that type of script than not using it for benchmarking stuffs.

We know why, to boost their benchmark score to make it seem like the device will run apps faster than it really will, thus giving it a slight edge in numbers over the competition.

If they really cared about the actual performance of the device, they wouldn't boost ONLY the benchmark scores and give every program that requires all the power the same boost.

You say there is a good reason, the reason is to misrepresent the way apps will run on the device.

SierraSonic said,
We know why, to boost their benchmark score to make it seem like the device will run apps faster than it really will, thus giving it a slight edge in numbers over the competition.

If they really cared about the actual performance of the device, they wouldn't boost ONLY the benchmark scores and give every program that requires all the power the same boost.

You say there is a good reason, the reason is to misrepresent the way apps will run on the device.


How is that to make it seems faster when it's still inside it's default speeds?

How is it cheating by using it's default CPU and GPU speeds?

Exynos said,

How is that to make it seems faster when it's still inside it's default speeds?

How is it cheating by using it's default CPU and GPU speeds?


How is overriding the default thermal limits, over-volting the GPU, and changing the performance characteristics NOT cheating?

Edited by SierraSonic, Oct 3 2013, 8:35pm :

SierraSonic said,

How is overriding the default thermal limits, over-volting the GPU, and changing the performance characteristics NOT cheating?

Running default speeds are within it's thermal limits. You are from another planet if you think that Samsung would give out a phone with so high speeds on it's CPU and GPU that it would melt.

Dream on i can just say.

Edited by Exynos, Oct 3 2013, 10:47pm :

Exynos said,

Running default speeds are within it's thermal limits. Youare from another planet if you think that Samsung would give out a phone with so high speeds on it's CPU and GPU that it would melt.

Dream on i can just say.


1) On the Exynos 5410, Samsung was detecting the presence of certain benchmarks and raising thermal limits (and thus max GPU frequency) in order to gain an edge on those benchmarks, and

2) On both Snapdragon 600 and Exynos 5410 SGS4 platforms, Samsung was detecting the presence of certain benchmarks and automatically driving CPU voltage/frequency to their highest state right away. Also on Snapdragon platforms, all cores are plugged in immediately upon benchmark detect.

I never said they would melt, I said that they are pushing the GPU harder just for these benchmarks than it runs any other program.

Do you need to be spoon fed every bit of information every time you make a counter point? It is a little tiring repeating myself over and over and over.

SierraSonic said,
I never said they would melt, I said that they are pushing the GPU harder just for these benchmarks than it runs any other program.

Do you need to be spoon fed every bit of information every time you make a counter point? It is a little tiring repeating myself over and over and over.


LOL, isn't that the whole point by a benchmark, that it will make things go hotter under heavy load?

Does it looks like i can get any normal apps like Firefox, Google Drive, Dropbox and such apps to run on full power output?

Why do they have to run at full speed just because some benchmark apps does it?

And, have you ever done benchmarking on a normal computer?

Clearly not. Because you are raising the temprature on your CPU and GPU over what's normal under normal usage by running benchmark apps there witch is normal. It would be something wrong if the components didn't had got any hotter under heavy load.

Again, what are the differences?

Edited by Exynos, Oct 3 2013, 11:31pm :

Exynos said,

LOL, isn't that the whole point by a benchmark, that it will make things go hotter under heavy load?

Have you ever done benchmarking on a normal computer?

Clearly not. Because you are raising the temprature on your CPU and GPU over what's normal under normal usage by running benchmark apps there witch is normal. It would be something wrong if the components didn't had got any hotter under heavy load.

Again, what are the differences?

The difference is, the system normally won't get this hot under load, as the system normally for any other app would reach a thermal limit and lower the clock speed/pull back power to preserve the device. At the cost of the devices life, Samsung has forced the device to stay in a hotter state, and overvolted the unit to preform better falsely.

SierraSonic said,
The difference is, the system normally won't get this hot under load, as the system normally for any other app would reach a thermal limit and lower the clock speed/pull back power to preserve the device. At the cost of the devices life, Samsung has forced the device to stay in a hotter state, and overvolted the unit to preform better falsely.

No, Samsung have forced the device to run at it's normal and default speeds with isn't there to kill your phone. And if the heat doesn't kill the phone, then what are the problem that the benchmarks raises the temprature on the device under benchmarks?

Isn't that quite obvious by everybody that any apps that are heavy will use alot of the hardwares potential and make things go hotter?

And where is your sources or evidences that Samsung raises the speeds to the default speed limits on the phones and then at the same time raises the tempratures over what the thermal limits for the CPU and GPU are designed to handle at default speeds?

Yeah OMGWTF, they makes the temprature on the phone to go skyrocketing (but still inside the thermal limits) under heavy load. That's news @ 11 everyone.

Exynos said,

No, Samsung have forced the device to run at it's normal and default speeds with isn't there to kill your phone. And if the heat doesn't kill the phone, then what are the problem that the benchmarks raises the temprature on the device under benchmarks?

Isn't that quite obvious by everybody that any apps that are heavy will use alot of the hardwares potential and make things go hotter?

And where is your sources or evidences that Samsung raises the speeds to the default speed limits on the phones and then at the same time raises the tempratures over what the thermal limits for the CPU and GPU are designed to handle at default speeds?

Okay, read the article this news post is about. You are obviously not well informed in this situation.

SierraSonic said,
Okay, read the article this news post is about. You are obviously not well informed in this situation.

I'm just pointing out as long as the tempratures are all inside it's thermal limits and as long as the CPU and GPU speeds are as they are specified in the spec sheets, then it's not a good damn problem with that.

Exynos said,

I'm just pointing out as long as the tempratures are all inside it's thermal limits and as long as the CPU and GPU speeds are as they are specified in the spec sheets, then it's not a good damn problem with that.
I don't care about the temperatures, I care that they are cheating by forcing benchmarks to work better than the device does by default.

SierraSonic said,
I don't care about the temperatures, I care that they are cheating by forcing benchmarks to work better than the device does by default.

So you say it's a cheat that the benchmark apps on a computer are making so the computer are working better than it does by default over running Firefox, Goodle Drive, Google Earth and so on?

Alirighty then, we know what makes you so sad. Simply because you don't get what CPU's and GPU are made to do and how they can run.

Exynos said,

So you say it's a cheat that the benchmark apps on a computer are making so the computer are working better than it does by default over running Firefox, Goodle Drive, Google Earth and so on?

Alirighty then, we know what makes you so sad. Simply because you don't get what CPU's and GPU are made to do and how they can run.

It is the benchmarks apps job to stress the system, it is not Samsung's job to create a script to rig benchmarks in their favor, which is called cheating.

SierraSonic said,
It is the benchmarks apps job to stress the system, it is not Samsung's job to create a script to rig benchmarks in their favor, which is called cheating.

No, it's Samsung's job to fix the problems no others want to fix.

If the developers doesn't update their apps to go around the restrictions on the governors, or if Google doesn't update their own kernels and governors for Android so it can let the heavy apps run without getting speed throttled, then Samsung and the others have to make a script to help us get the performance the device can deliver under benchmarking.

i must chime in again... quote: "If the developers doesn't update their apps to go around the restrictions on the governors"

LOL... so let me understand what you are saying.... Its the programmers fault that they don't work around limitations put in place by the handset manufacturer for reliability issues?

Really?

Really???????

rippleman said,
i must chime in again... quote: "If the developers doesn't update their apps to go around the restrictions on the governors"

LOL... so let me understand what you are saying.... Its the programmers fault that they don't work around limitations put in place by the handset manufacturer for reliability issues?

Really?

Really???????


Yeah, it's a problem via the governors that the benchmarking apps can't run at full speed because they gets speed throttled. And getting speed throttled under a benchmarking isn't really the way to determine how powerfull or fast the device actually are at full speed.

Exynos said,

Yeah, it's a problem via the governors that the benchmarking apps can't run at full speed because they gets speed throttled. And getting speed throttled under a benchmarking isn't really the way to determine how powerfull or fast the device actually are at full speed.
Full speed is the max the system supports by default.

Let me put it this way, some cars come with governors, they cant go above 110 mph. Is it okay for them to advertise that their cars are gear limited at 145 mph if they are computer limited to 110?

Here is the same thing Samsung is doing. If their devices are limited by throttling, is it okay to advertise how the device works unthrottled? Especially if the common trend and/or regulation is based on not advertising unthrottled benchmarks?

By the way, it is NOT a problem that these devices are throttled, they are built to do that! Samsung isn't fixing a problem, they are fixing their score!

SierraSonic said,
Full speed is the max the system supports by default.

Let me put it this way, some cars come with governors, they cant go above 110 mph. Is it okay for them to advertise that their cars are gear limited at 145 mph if they are computer limited to 110?

Here is the same thing Samsung is doing. If their devices are limited by throttling, is it okay to advertise how the device works unthrottled? Especially if the common trend and/or regulation is based on not advertising unthrottled benchmarks?

By the way, it is NOT a problem that these devices are throttled, they are built to do that! Samsung isn't fixing a problem, they are fixing their score!


Normal apps that doesn't needs the full power of the devices are getting speed throttled to save battery instead of running at full speed with no benefits of doing that.

Benchmarks apps on the other hand are made to run at full speeds with the benefits of the full potential of the hardware without having to be speed throttled.

Would you like to see 3DMark gets throttled down by the GPU it self because it thinks it can't run at full speed?

Exynos said,

Normal apps that doesn't needs the full power of the devices are getting speed throttled to save battery instead of running at full speed with no benefits of doing that.

Benchmarks apps on the other hand are made to run at full speeds with the benefits of the full potential of the hardware without having to be speed throttled.

Would you like to see 3DMark gets throttled down by the GPU it self because it thinks it can't run at full speed?

If my computer was setup to do it, then yes, I wouldn't want something to override the settings of my system to show me my computer is running faster than it is. Especially if it means my computer will run hotter and faster using more voltage all due to one company trying to make their product seem like it is better than it really is.

If I was looking to see my computers/mobile devices true maximum potential, I would personally disable all running tasks/apps/etc, tweak all my profiles and settings for optimum performance, overclock and over volt as far as i can go to barely pass the benchmark without crashing.

But since all that score is show me an inflated perception on how my computer could possibly run as fast it could without care for stability or quality, I wouldn't care for that score as all it does is serve me to make my epeen look bigger.

I care about benchmarks for setups where the device is used in that condition, and not over-rated lies.

What you fail to realize is that there are programs out there that do demand 100% from current devices, you don't see them getting this extra boost, because well it doesn't make Samsung look good so they weren't added to the script.

Edited by SierraSonic, Oct 4 2013, 2:34am :

SierraSonic said,
If my computer was setup to do it, then yes, I wouldn't want something to override the settings of my system to show me my computer is running faster than it is. Especially if it means my computer will run hotter and faster using more voltage all due to one company trying to make their product seem like it is better than it really is.

If I was looking to see my computers/mobile devices true maximum potential, I would personally disable all running tasks/apps/etc, tweak all my profiles and settings for optimum performance, overclock and over volt as far as i can go to barely pass the benchmark without crashing.

But since all that score is show me an inflated perception on how my computer could possibly run as fast it could without care for stability or quality, I wouldn't care for that score as all it does is serve me to make my epeen look bigger.

I care about benchmarks for setups where the device is used in that condition, and not over-rated lies.

What you fail to realize is that there are programs out there that do demand 100% from current devices, you don't see them getting this extra boost, because well it doesn't make Samsung look good so they weren't added to the script.


If an application can use 100% of the potential of the hardwares, it's by no means cheats in any possible ways. It's called utilizing the hardware of what's available.

As long as it's running withing the default speeds and doesn't breaks the thermal limits as written in the specifications to the CPU and the GPU and as long it need to use the full power, there shouldn't be anything that should prevent that app from running at full speed with lots of heat generation. But as it have been said a million times already, the current governors are making the benchmark apps to not run at the full speed like they should. So that's why Samsung, LG, HTC and ASUS are using this script to go around that problem.

Why are you whining and crying about this?

Edited by Exynos, Oct 4 2013, 2:21am :

Exynos said,

If an application can use 100% of the potential of the hardwares, it's by no means cheats in any possible ways. It's called utilizing the hardware of what's available.

As long as it's running withing the default speeds and doesn't breaks the thermal limits as written in the specifications to the CPU and the GPU and as long it need to use the full power, there shouldn't be anything that should prevent that app from running at full speed with lots of heat generation. But as it have been said a million times already, the current governors are making the benchmark apps to not run at the full speed like they should. So that's why Samsung, LG, HTC and ASUS are using this script to go around that problem.

Why are you whining and crying about this?

The didn't get around any limitation, they just aren't representing their devices as they should.

My problem with this is that they are cheating! All of them, on every device that has a special script or setting that changes the way the device behaves.

If I ask you how fast your car is, I'm not asking after you change your cars performance, I'm asking about NOW.

SierraSonic said,
The didn't get around any limitation, they just aren't representing their devices as they should.

My problem with this is that they are cheating! All of them, on every device that has a special script or setting that changes the way the device behaves.

If I ask you how fast your car is, I'm not asking after you change your cars performance, I'm asking about NOW.


Can you please explain WHY it's cheating to use the full potential of a computer or a smartphone?

You sounds very very veeeeeeeery unexperienced with this according to how you talks here. I mean seriously?

I simply don't care what one single guy are saying when every devices are made so apps can use their full potential regardless of what other apps might use in power on the devices. PERIOD.

EDIT: Just to take a very good example on your rabble. Quadrant was some years ago only meant to be runned on a single core CPU on our smartphones. Then suddenly the dual-core CPU's came out and Quadrant was still only using one core under the benchmarks.

A little later Quadrant was updated to take advantage of both cores and the full potential of the smartphone hardwares at that time. Was that cheating by them then?

Edited by Exynos, Oct 4 2013, 3:30am :

Exynos said,

Can you please explain WHY it's cheating to use the full potential of a computer or a smartphone?

You sounds very very veeeeeeeery unexperienced with this according to how you talks here. I mean seriously?

I simply don't care what one single guy are saying when every devices are made so apps can use their full potential regardless of what other apps might use in power on the devices. PERIOD.

EDIT: Just to take a very good example on your rabble. Quadrant was some years ago only meant to be runned on a single core CPU on our smartphones. Then suddenly the dual-core CPU's came out and Quadrant was still only using one core under the benchmarks.

A little later Quadrant was updated to take advantage of both cores and the full potential of the smartphone hardwares at that time. Was that cheating by them then?

No, but it's different when the benchmark supports the max hardware, vs the hardware itself changing just for the benchmark.

For example, when ati and nvidia changed their driver settings specifically for benchmark software, they were called out on it and changed their ways. Samsung and others are just now being called on it, if they wont change, I am sure the benchmarks will.

It really doesn't matter at this point, does it? If writing scripts to turn their phones into bottle rockets or one-pump chumps is what they need to do, then so be it. At least we all know that their benchmarks aren't worth a damn.

SierraSonic said,
No, but it's different when the benchmark supports the max hardware, vs the hardware itself changing just for the benchmark.

For example, when ati and nvidia changed their driver settings specifically for benchmark software, they were called out on it and changed their ways. Samsung and others are just now being called on it, if they wont change, I am sure the benchmarks will.


The hardware didn't in any ways change as you think it does. It was a software script that changed something with governors that made the heavy apps like benchmarks to be able to run at full speed as the phone is made to be used at, witch is by todays defenition completely normal and legal to do.

And also, what ATI and NVIDIA did was not to let the benchmarks run faster, because they couldn't run any faster than they did, but they did trick with the benchmark results in the end witch has nothing to do with how the benchmarks runs.

But in Samsung's case, they are opening up the speed on the CPU to 2.3 GHz that isthe default speed so the governor doesn't force the speeds down under any of the benchmarks that can give you false results because the benchmark app wasn't really allowed to test the full potential of the device. So Samsung just made sure that the benchmark apps actually did run at full speed to give us the more true results rather than just letting the Android governors throttle the speeds down and gives us false results on a half baked test because the CPU / GPU messed with the benchmark apps.

Again, this is FACTS and something you can't deny because you can't counterargument it. So please stop with the rabble before i'm forced to send in a report of a troller.

Edited by Exynos, Oct 4 2013, 4:22am :

Exynos said,

The hardware didn't in any ways change as you think it does. It was a software script that changed something with governors that made the heavy apps like benchmarks to be able to run at full speed as the phone is made to be used at, witch is by todays defenition completely normal and legal to do.

And also, what ATI and NVIDIA did was not to let the benchmarks run faster, because they couldn't run any faster than they did, but they did trick with the benchmark results in the end witch has nothing to do with how the benchmarks runs.

But in Samsung's case, they are opening up the speed on the CPU to 2.3 GHz that isthe default speed so the governor doesn't force the speeds down under any of the benchmarks that can give you false results because the benchmark app wasn't really allowed to test the full potential of the device. So Samsung just made sure that the benchmark apps actually did run at full speed to give us the more true results rather than just letting the Android governors throttle the speeds down and gives us false results on a half baked test because the CPU / GPU messed with the benchmark apps.

Again, this is FACTS and something you can't deny because you can't counterargument it. So please stop with the rabble before i'm forced to send in a report of a troller.


Did you miss the quote I posted that stated that samsungs gpu gets a overclock? Why are you ignoring these facts?

Also by not using default values and not letting the system throttle we are getting fake results.

SierraSonic said,
Why are you ignoring these facts?

typically people always want to believe what they own/have is superior, remember, he has g4. Its a great phone, but he needs to believe its superior, for him to do that, he must justify the tests slanted numbers. Its not his fault, its just part of being human.

SierraSonic said,

Did you miss the quote I posted that stated that samsungs gpu gets a overclock? Why are you ignoring these facts?

Also by not using default values and not letting the system throttle we are getting fake results.


Well, let me see..........

First i will make sure to do this: HAHAHAHAHAHAHAHAHAHA. Secondly, the official specs on the Qualcomm Snapdragon 800 are rated up to 2.3 GHz as default speed from the official Qualcomm webpage: http://www.qualcomm.com/snapdragon/processors/800, so no overclocking here.

Thirdly, the Adreno 330 are rated at 2 different speeds on the Wikipedia infopage about the GPU. The first speed is 450 MHz and the second speed is 550 MHz. So if the Galaxy Note 3 runs the GPU at 533 MHz, it's still lower than the 550 MHz maximum speed that the GPU is made to handle all day long without any problems: http://en.wikipedia.org/wiki/Adreno

So, how about you defend that huh?

EDIT: If the GPU is getting overclocked from 450 MHz to 533 MHz, then it's not just the benchmark apps who gets that benefit. Everything that is heavy on the phone will get the GPU to run at 533 MHz. So even if it would be overclocked, it still doesn't change the fact that the Benchmark apps should use the same speed on the GPU as the other heavy apps and on top of that get the benefits of not getting speed throttled under a benchmark run.

Edited by Exynos, Oct 4 2013, 3:06pm :

Exynos said,

the official specs on the Qualcomm Snapdragon 800 are rated up to 2.3 GHz as default speed from the official Qualcomm webpage:

So what you are saying is samusung is gimping their phones right out the box?

rippleman said,

So what you are saying is samusung is gimping their phones right out the box?


No?

What i'm saying is that some of the Android kernel governors (witch Samsung doesn't have anything to do with) are overtaking some apps and are speed throttling them down in speed where they shouldn't really do that. Samsung have actually prevented that from happening by creating this script.

So, have you seen the light yet on this?

Exynos said,

Well, let me see..........

First i will make sure to do this: HAHAHAHAHAHAHAHAHAHA. Secondly, the official specs on the Qualcomm Snapdragon 800 are rated up to 2.3 GHz as default speed from the official Qualcomm webpage: http://www.qualcomm.com/snapdragon/processors/800, so no overclocking here.

Thirdly, the Adreno 330 are rated at 2 different speeds on the Wikipedia infopage about the GPU. The first speed is 450 MHz and the second speed is 550 MHz. So if the Galaxy Note 3 runs the GPU at 533 MHz, it's still lower than the 550 MHz maximum speed that the GPU is made to handle all day long without any problems: http://en.wikipedia.org/wiki/Adreno

So, how about you defend that huh?

EDIT: If the GPU is getting overclocked from 450 MHz to 533 MHz, then it's not just the benchmark apps who gets that benefit. Everything that is heavy on the phone will get the GPU to run at 533 MHz. So even if it would be overclocked, it still doesn't change the fact that the Benchmark apps should use the same speed on the GPU as the other heavy apps and on top of that get the benefits of not getting speed throttled under a benchmark run.


It is the fact that the over clock only happens for benchmarks that makes it a cheat, also why are you still debating what an benchmark is. A benchmark tests current system setup performance, if the current setup is changed and changed only for benchmark apps, it is wrong no matter how technical you get.

SierraSonic said,

It is the fact that the over clock only happens for benchmarks that makes it a cheat, also why are you still debating what an benchmark is. A benchmark tests current system setup performance, if the current setup is changed and changed only for benchmark apps, it is wrong no matter how technical you get.

Didn't other Samsung apps use the same speed as the benchmark apps to as it was explained in the Galaxy S4 article about the same?

So no, it's not just the benchmark apps who gets that benefit.

Exynos said,

Didn't other Samsung apps use the same speed as the benchmark apps to as it was explained in the Galaxy S4 article about the same?

So no, it's not just the benchmark apps who gets that benefit.


No it stated it that that the script only works for benchmark apps, they said that the inflating the results is almost pointless as its on for a 4.4% gain, still cheating though.

SierraSonic said,

No it stated it that that the script only works for benchmark apps, they said that the inflating the results is almost pointless as its on for a 4.4% gain, still cheating though.

I think you should read the article about the Galaxy S4 about the same case as here one more time. They clearly stated that other Samsung apps was noted down in the script to be allowed to use the maximum speed of the CPU or GPU.

Exynos said,

I think you should read the article about the Galaxy S4 about the same case as here one more time. They clearly stated that other Samsung apps was noted down in the script to be allowed to use the maximum speed of the CPU or GPU.

Look up the difference between some and all, since its not all its cheating.

SierraSonic said,

Look up the difference between some and all, since its not all its cheating.

It's still not cheating as the speed on the CPU and GPU is still being used as it's described on the official specs on the Galaxy S4 / Galaxy Note 3.

Exynos said,

It's still not cheating as the speed on the CPU and GPU is still being used as it's described on the official specs on the Galaxy S4 / Galaxy Note 3.
its cheating because it is not benchmarking default performance.

Exynos said,

They clearly stated that other Samsung apps was noted down in the script to be allowed to use the maximum speed of the CPU or GPU.

nope. only benchmark apps on the note 3

rippleman said,

nope. only benchmark apps on the note 3


ahh, even better. The point is this isn't a max hardware benchmark anyway, read what hiphopsincefriday wrote about 5posts down. And remember current setup, not max.

SierraSonic said,
its cheating because it is not benchmarking default performance.

The default benchmark method is to use the maximum performance out of the CPU and GPU of what's AVAILABLE. And what's available on the hardware out of box is default.

rippleman said,

nope. only benchmark apps on the note 3


And your evidences on that are where?

If the Galaxy S4 use the other Samsung apps on maximum performance, then why shouldn't the Galaxy Note 3 do that?

Exynos said,

And your evidences on that are where?

If the Galaxy S4 use the other Samsung apps on maximum performance, then why shouldn't the Galaxy Note 3 do that?


It doesn't matter anyway.

if all other apps can use it, they why have "code" to boost what's already avialible? lol

Exynos said,

And your evidences on that are where?

If the Galaxy S4 use the other Samsung apps on maximum performance, then why shouldn't the Galaxy Note 3 do that?

SierraSonic said,

It doesn't matter anyway.

Ahh, so it doesn't matter anylonger when i just gave you an argument that you can't counterargument. Nice move.

rippleman said,
if all other apps can use it, they why have "code" to boost what's already avialible? lol

Not all apps can do it. Like Dropbox for Android to take an example, it will never use more than 20% of the performance potential out of the Galaxy Note 3, so it doesn't need the full speed ofthe device to run prefectly. So i'm only talking about those apps who really needs the full potential out of the hardware, but at the same time gets interupted a little by the speed throttling.

so benchmarks (since they are the ONLY things programmed in the code) are the only things that need the full potential of the hardware? Nothing else is coded for it right?

Exynos said,

only talking about those apps who really needs the full potential out of the hardware,

rippleman said,
so benchmarks (since they are the ONLY things programmed in the code) are the only things that need the full potential of the hardware? Nothing else is coded for it right?


Like i have said, other Samsung apps are also allowed to use the full potential of the hardware in the Galaxy S4 and Galaxy Note 3. And more apps will for sure be allowed for that soon when they are needing that much power to run.

Exynos said,

Like i have said, other Samsung apps are also allowed to use the full potential of the hardware in the Galaxy S4 and Galaxy Note 3. And more apps will for sure be allowed for that soon when they are needing that much power to run.
Who cares, what about all the apps that could use all that power and then some but aren't allowed to?

Exynos said,

Ahh, so it doesn't matter anylonger when i just gave you an argument that you can't counterargument. Nice move.
I said it doesn't matter as all apps do not have the option to run full speed, I am not saying they need to be forced at full speed, but even the ones that would be able to use the full power of the device aren't allowed too.

This isn't that hard of a concept, I have literally destroyed all your points one by one.

"Benchmarks are for testing max speed" No they are not, they are for the current setup.

"There's no overclock" Yes there is, in addition to the forced full cores, raised throttle maps, and the overvolting of the GPU. It overclocks from the default values to ones only allowed by the script.

"This is not a cheat" When you alter your device to have certain apps run faster than all other apps, especially when those apps are there to test the devices default fastest speed available to all apps, you are cheating.

"PC benchmarks run at 100%" That is only because your PC is setup to do so, by default or by changing values yourself. I am certain there are devices out there that have different power plans that limit their max potential by default, and I am also certain that those pc makers won't stoop as low as creating a script to alter the performance only for certain apps.

"Other apps run too, not just benchmarking apps" So? It doesn't matter anyway.

"It's really only using 100% of whats available" This is irrelevant, as only certain apps get this special mode. This mode is only to make the device seem faster than it generally runs at.

"Everybody is doing it" No, only certain vendors on certain devices, usually their flagships to make their devices seem better than they are by default.

"It's not their fault, blame Google and their governors" These device makers can make their own profiles, as this is what the script does, but they choose to use it only for certain apps.

"Certain other illogical comment that makes this okay" It isn't.

"Insult" Great point there.

Again, not complaining that all apps aren't forced to run at 100% with all 4 cores, I'm complaining that all apps aren't treated equally to all the hardware if they need to. The problem created here is that even if there are apps not in the script that want and can use that power, they wouldn't be able too.

SierraSonic said,
Who cares, what about all the apps that could use all that power and then some but aren't allowed to?

Then those apps should be allowed. Nothing more to say about that. But just because some apps aren't allowed (witch should change), it doesn't mean Samsung are cheating with the script now.

SierraSonic said,
This isn't that hard of a concept, I have literally destroyed all your points one by one.

No you haven't. All you have been doing is crying over that Samsung allows some apps to use the power the phones are able to deliver.

SierraSonic said,
"Benchmarks are for testing max speed" No they are not, they are for the current setup.

And the current setup is what you get out of the box and what the phone can deliver with the speeds it have on the CPU and GPU.

SierraSonic said,
Yes there is, in addition to the forced full cores, raised throttle maps, and the overvolting of the GPU. It overclocks from the default values to ones only allowed by the script.

The GPU on the Galaxy Note 3 are running the GPU at 533 MHz, witch is the speed the phone comes with out of the box and every apps that require all of that power from both of the CPU and GPU should be able to use that. It doesn't matter if the GPU are overclocked or not by Samsung over what the default GPU speed is. The point is that every apps that needs it should be able to use the power the phone comes out with.

SierraSonic said,
When you alter your device to have certain apps run faster than all other apps, especially when those apps are there to test the devices default fastest speed available to all apps, you are cheating.

When you makes some apps to use the CPU / GPU better in your hardware as the hardware is out of the box, it's not called cheats. It's called optimizing the apps. Thus no cheating.

Is it cheating that Google optimizes Android 4.3 to run better on our devices and use the hardware better over what Android 4.2.x does?

SierraSonic said,
That is only because your PC is setup to do so, by default or by changing values yourself. I am certain there are devices out there that have different power plans that limit their max potential by default, and I am also certain that those pc makers won't stoop as low as creating a script to alter the performance only for certain apps.

And a smartphone is also setup to use 100% of the hardware under any benchmarks. Benchmarks on a computer is exactly the same as running benchmarks on a smartphone.

SierraSonic said,
So? It doesn't matter anyway.

It doesn't matter to you, but for me and ALOT of others it does.

SierraSonic said,
This is irrelevant, as only certain apps get this special mode. This mode is only to make the device seem faster than it generally runs at.

So letting the apps run at the specified speeds as the CPU and GPU are able to use out of the box is cheat to make apps to use that?

How is that cheat?

Is it right to use a wrong type of fuel on my engine on my car so the car only did 220 km/h instead of doing 250 km/h (witch is the speed limit) with the right performance fuel type?

Would it be cheat to allow the car to perform like it should?

SierraSonic said,
No, only certain vendors on certain devices, usually their flagships to make their devices seem better than they are by default.

ASUS, LG, HTC and Samsung is reported to do it. And maybe others are doing it to. And why do you think everyone is doing it?

Are they doing it to let the device perform like it should where the apps needs that performance, or is it to cheat inside theit default specified specifications out of the box?

SierraSonic said,
These device makers can make their own profiles, as this is what the script does, but they choose to use it only for certain apps.

And do you think every apps that don't even needs that performance just will get that fix just because they can?

That's like wasting alot of battery for nothing if that's the case. But sure, more apps will need the same fix as the benchmark apps needs, so it will come soon most likely if some apps needs that.

And because there exist toooooons of different apps out there on the Play Store, it will be a longer process of letting those apps to use the full potential out of the hardware in those devices with the script Samsung and the others are using. I'm pretty sure more and more apps will be updated in the near future to use the current monster hardwares in our smartphones today much better.

SierraSonic said,
Again, not complaining that all apps aren't forced to run at 100% with all 4 cores, I'm complaining that all apps aren't treated equally to all the hardware if they need to. The problem created here is that even if there are apps not in the script that want and can use that power, they wouldn't be able too.

The apps aren't treated equally because it's so many apps out there that needs this fix that it will take alot of time to get that fixed.

It will most likely be that Google fixes or releases a new kernel governor that will allow some apps to use the full potential out of the devices if they reach a certain point in speed on the CPU and GPU. Or maybe Samsung will add more and more apps to their script on each of the new Android releases with fixes and stuffs.

Edited by Exynos, Oct 5 2013, 12:58am :

Exynos said,

Then those apps should be allowed. Nothing more to say about that. But just because some apps aren't allowed (witch should change), it doesn't mean Samsung are cheating with the script now.
No, the script shouldn't exist and all apps should be allowed to use as much power as they need without sacrificing the default power profile.


Exynos said,
No you haven't. All you have been doing is crying over that Samsung allows some apps to use the power the phones are able to deliver.
Which is exactly the problem!


Exynos said,
And the current setup is what you get out of the box and what the phone can deliver with the speeds it have on the CPU and GPU.
Which is why stealth benchmarks report the proper speed of the device. And not the falsely boosted one.


Exynos said,
The GPU on the Galaxy Note 3 are running the GPU at 533 MHz, witch is the speed the phone comes with out of the box and every apps that require all of that power from both of the CPU and GPU should be able to use that. It doesn't matter if the GPU are overclocked or not by Samsung over what the default GPU speed is. The point is that every apps that needs it should be able to use the power the phone comes out with.
Exactly, I'm going to go in more detail about your exact words further below.


Exynos said,
When you makes some apps to use the CPU / GPU better in your hardware as the hardware is out of the box, it's not called cheats. It's called optimizing the apps. Thus no cheating.
Samsung added the script to make these apps perform better than the out of box experience, just because the script is there by default doesn't mean that this is default speed.

Exynos said,
Is it cheating that Google optimizes Android 4.3 to run better on our devices and use the hardware better over what Android 4.2.x does?
No it isnt, but adding a script to change the profile of the device only for certain apps is.


Exynos said,
And a smartphone is also setup to use 100% of the hardware under any benchmarks. Benchmarks on a computer is exactly the same as running benchmarks on a smartphone.
Exactly, then Samsung added a script to make certain benchmarks run faster than stealth benchmarks, the exact same benchmarks with just different names for your information.


Exynos said,
It doesn't matter to you, but for me and ALOT of others it does.
It doesn't matter because it doesn't prove your point.


Exynos said,
So letting the apps run at the specified speeds as the CPU and GPU are able to use out of the box is cheat to make apps to use that?
It's a cheat when it's not the default profile, be more specific, you meant to say CERTAIN apps correct? Again that is the problem.

Exynos said,
How is that cheat?
Because it only allows certain apps to use this speed.

Exynos said,
Is it right to use a wrong type of fuel on my engine on my car so the car only did 220 km/h instead of doing 250 km/h (witch is the speed limit) with the right performance fuel type?
Use whatever fuel you want, the car will compensate the best it can. If you have a premium only car, it is recommended you use that fuel to prevent damage to your car more than it affecting performance.

Exynos said,
Would it be cheat to allow the car to perform like it should?
No, which is exactly MY point. What you are thinking of here is similar to the F1 style overboost button. Everyone in F1 gets a certain amount of time with the use of that setting, if someone uses it longer, or has it give more power during the same period of time, it is cheating.


Exynos said,
ASUS, LG, HTC and Samsung is reported to do it. And maybe others are doing it to. And why do you think everyone is doing it?
I didn't say everyone, I said everyone that IS doing it, but again you misunderstood, you are the one claiming its okay because others are doing it. I am merely stating that if your with a group of people and they all started to rob a store it would still be stealing if you joined them or not.

Exynos said,
Are they doing it to let the device perform like it should where the apps needs that performance, or is it to cheat inside theit default specified specifications out of the box?
To cheat, why else are only recent flagship type devices doing it?


Exynos said,
And do you think every apps that don't even needs that performance just will get that fix just because they can?
Not they will not get access to that cheat. The will just use the common profiles maximum.

Exynos said,
That's like wasting alot of battery for nothing if that's the case. But sure, more apps will need the same fix as the benchmark apps needs, so it will come soon most likely if some apps needs that.
You do realize that apps only use as much speed and power as they need. There are apps that require more power than any phone can output right now, but they will run the best they can using the default settings.

Exynos said,
And because there exist toooooons of different apps out there on the Play Store, it will be a longer process of letting those apps to use the full potential out of the hardware in those devices with the script Samsung and the others are using. I'm pretty sure more and more apps will be updated in the near future to use the current monster hardwares in our smartphones today much better.
The problem with this line of thought is who is testing every app, what are the guidelines to get access to this profile, and will there be an approval process?


Exynos said,
The apps aren't treated equally because it's so many apps out there that needs this fix that it will take alot of time to get that fixed.
LOL? Care to explain this one some more? I assume you mean it will take time to get the apps into the profile, if so I again point to what I said previously. What is the process to get approved, is there a process, are there and what are the requirements? Why has no statement been issued to app developers about this "fix"? Can I submit my apps now?

Exynos said,
It will most likely be that Google fixes or releases a new kernel governor that will allow some apps to use the full potential out of the devices if they reach a certain point in speed on the CPU and GPU. Or maybe Samsung will add more and more apps to their script on each of the new Android releases with fixes and stuffs.
Or you are full of **** and just making things up and trying to argue with what ifs more than facts.

Exynos said,
The point is that every apps that needs it should be able to use the power the phone comes out with.
Exactly!

Anyway it doesn't matter now as most benchmarking apps are going stealth, changing their package name with every update, and most review sites are going to list both boosted and unboosted results.

SierraSonic said,
No, the script shouldn't exist and all apps should be allowed to use as much power as they need without sacrificing the default power profile.

But the problem is that the apps isn't powerfull enough to actually use the same power as the benchmark apps do. 95% of all of the apps on PlayStore are apps that doesn't even needs 50% of the full potential of todays smartphones. So the apps doesn't needs more power than the apps are made to use. Thus there is no need to let them use any more power than they already do.

SierraSonic said,
Which is exactly the problem

Yeah, you are crying that Samsung are letting some apps use the full power of the device when they aren't allowed to do it by Android it self. And it makes you crying because Samsung make the other powerfull smartphones looks cheap and poor because the Galaxy Note 3 just pumps out awesome performance because Samsung have optimized their ROM's to perform extremely well over what the other brands have done.

SierraSonic said,
Which is why stealth benchmarks report the proper speed of the device. And not the falsely boosted one.

So you mean it's a cheat that apps that needs the full power of the device can use the power the phone delivers out of the box?

LOL, you must love to buy a car that are announced with 250 hp and when you comes home with it you realizes that it have only 200 hp because the fuel pump wouldn't get enough fuel to the engine, and you would still love it and absolutely not fix it.

SierraSonic said,
Exactly, I'm going to go in more detail about your exact words further below.

Then what exactly are the problems by letting the apps use the announced speeds on the Galaxy Note 3?

As long as they use the speeds as the phone is announced with, it's by no means overclocking or cheats. Just because a GPU CAN be overclocked by default from some brands, it doesn't mean that it's overclocking when you run an application on it, becaise it's already overclocked by default. So it means it's running with the default speeds the phone comes with.

SierraSonic said,
Samsung added the script to make these apps perform better than the out of box experience, just because the script is there by default doesn't mean that this is default speed.

But when Android speeds down the apps where they shouldn't, wouldn't you say that the apps, like those benchmark apps should be allowed to run at the maximum speed of every devices to see what the performance is on the phone?

Because i don't think there are anyone in the world who does benchmarking that would like to see the final results of how his computer performs at half of the power the computer can deliver. Right?

SierraSonic said,
No it isnt, but adding a script to change the profile of the device only for certain apps is.

I don't care if it's a script or an app or whatever, all the case is about is that Samsung optimized their software to run several apps at the power output the phone can deliver out of the box. Just because Samsung uses a script to let the apps that needs that to run like they should, it doesn't mean they are cheating or running things faster than it normally wouldn't. Because the CPU are rated at 2.3 GHz, and the CPU are still running at 2.3 GHz under any benchmark runs. It's the same with the GPU. The GPU are rated at 533 MHz out of the box as default speed and are still 533 MHz under any benchmark runs.

So again, how is it cheating to let some apps run at the speeds the phones are made to run at as normal?

And how is it cheats to optimize something in the ROM so some apps runs better and to use the hardware as it is better?

Basicly, it's cheats for anyone who releases an app and then later updates it where he have optimized the app to run better on our hardwares and softwares according to you, witch is hilarious and something to LOL hard over.

Your definition on how these things should work is waaaay way to funny and to stupid.

SierraSonic said,
Exactly, then Samsung added a script to make certain benchmarks run faster than stealth benchmarks, the exact same benchmarks with just different names for your information.

Again, read my answer over. You have the facts there. You are way to obsessed over that Android downclocks some apps and games to 480 MHz on the phone over that the apps shouldn't really be downclocked if they shouldn't. You should really care about letting the apps using the full potential of the phones rather than crying over that Samsung optimizes their software instead of realizing that the problems is that Android downclock several apps that shouldn't be downclocked that way.

Samsung is making things to work more like it should.


SierraSonic said,
It doesn't matter because it doesn't prove your point.

It doesn't prove a point to YOU, but for me and many others who are a benchmarker on computers and smartphones, it proves everything that you are on the wild wild west with your empty crying arguments.


SierraSonic said,
It's a cheat when it's not the default profile, be more specific, you meant to say CERTAIN apps correct? Again that is the problem.

Just because some certain apps so far have got opened to use all of the power the Galaxy S and Note phones can deliver, it's still a beginning to let more apps do the same. You know, Samsung have to start somewhere.

So it's still not a cheat to let some apps that needs that much power to use the available speed of the device. The 533 MHz GPU and the 2.3 GHz Quad-Core CPU are there to be used. And as long the apps use that and doesn't makes the GPU or CPU go faster than the original specs out of the box, it's no cheating. It's the same with the RAM to, it's not there to not be used, it's there to use as much RAM as possible.

SierraSonic said,
Because it only allows certain apps to use this speed.

Maybe because most of the other apps doesn't needs that speed to begin with?

So just because it's just those ceratin apps that gets this so far, it doesn't mean it's a cheat for the other apps that doesn't needs that much power available at this moment.

SierraSonic said,
Use whatever fuel you want, the car will compensate the best it can. If you have a premium only car, it is recommended you use that fuel to prevent damage to your car more than it affecting performance.

You can use normal fuels on every of those performance cars. It's just that you will use the race fuel when you are on the speed circut. It's the same with the Koenigsegg Agera R. On normal fuel it have a little over 900hp on 95 octane regular fuel, and more than 1100 hp on E85 bio fuel.

Is that cheats to you to that the engine gives you the maximum power witch is the default hp on that engine with the right fuel (script in Samsung's case)?

It's specs says the engine have 1115 hp as you can see here: http://www.supercars.net/cars/5261.html

Does that mean it's a cheat because someone on the race field use E85 bio fuel (script) to get 1115 hp over those others who are using it on normal streets with 95 octane regular fuel (no script)?

SierraSonic said,
No, which is exactly MY point. What you are thinking of here is similar to the F1 style overboost button. Everyone in F1 gets a certain amount of time with the use of that setting, if someone uses it longer, or has it give more power during the same period of time, it is cheating.

Again, you fail totally to see what i'm talking about. What's happening here is that by using that button as you are talking about, you are raising the default horsepower over what's the default horsepower. There is under no circumstances that the script Samsung uses will raise the clock speeds over what's written in it's original specs sheet. Again, this is optimizing of the current default setup.

SierraSonic said,
I didn't say everyone, I said everyone that IS doing it, but again you misunderstood, you are the one claiming its okay because others are doing it. I am merely stating that if your with a group of people and they all started to rob a store it would still be stealing if you joined them or not.

If you read my comment over about the case that many of the supercars that are out today can get different horsepowers out of their cars by using different fuels then you know that this is all fine by doing, because everyone who have a supercar have an engine where you can use different fuel types on and switch between the power outputs.

This with the phones are no different. Samsung is just using race fuel on it's phones to open up all of the available power the phone have as described in the spec sheet.

SierraSonic said,
To cheat, why else are only recent flagship type devices doing it?

Because the flagships today are way way more powerfull than the earlier phones from them witch isn't so powerfull that it needs somekind of a script to let some of the apps use the available speed the phones have.

SierraSonic said,
Not they will not get access to that cheat. The will just use the common profiles maximum.

Because they don't need something extra access to some more performance profiles to run perfect?

Those apps runs as good as they do now without needing something extra that opens up for more performance.

SierraSonic said,
You do realize that apps only use as much speed and power as they need. There are apps that require more power than any phone can output right now, but they will run the best they can using the default settings.

Yeah and benchmmark apps needs all the power the phones have to deliver, witch is a 533 MHz GPU and a 2.3 GHz CPU on the Galaxy Note 3.

SierraSonic said,
The problem with this line of thought is who is testing every app, what are the guidelines to get access to this profile, and will there be an approval process?

No, if Samsung sees that some apps will struggle with what the apps are made to do because of the speed throttling in Android are messing them up, i'm sure Samsung will add that app to their script. Because you know, Samsung want's us to have as good experience with our phones as possible.


SierraSonic said,
LOL? Care to explain this one some more? I assume you mean it will take time to get the apps into the profile, if so I again point to what I said previously. What is the process to get approved, is there a process, are there and what are the requirements? Why has no statement been issued to app developers about this "fix"? Can I submit my apps now?

Look at my reply over this one, you have the answer there.

SierraSonic said,
Or you are full of **** and just making things up and trying to argue with what ifs more than facts.

Says the guy who doesn't know what benchmarking and optimizing of the software is, lol.

SierraSonic said,
Exactly!

Anyway it doesn't matter now as most benchmarking apps are going stealth, changing their package name with every update, and most review sites are going to list both boosted and unboosted results.


And this might be a shocker for you, but Samsung is in fact allowing this for apps already.

And no, this tactic will be more used because more and more phones will be more optimized to use the hardware better and like it should.

Why do you think certain benchmark app developers are making changes to their apps to avoid being activated by this script?

This may be a shocker to you, but I think the app developers should have more say than a company trying to boost their own score.

I would think that the developers of the app should decide if they want their apps boosted or not, why did Samsung force their apps to be boosted?

We've been working with all of the benchmark vendors to try and stay one step ahead of the optimizations as much as possible. Kishonti is working on some neat stuff internally, and we've always had a great relationship with all of the other vendors - many of whom are up in arms about this whole thing and have been working on ways to defeat it long before now. There's also a tremendous amount of pressure the silicon vendors can put on their partners (although not quite as much as in the PC space, yet), not to mention Google could try to flex its muscle here as well. The best we can do is continue to keep our test suite a moving target, avoid using benchmarks that are very easily gamed and mostly meaningless, continue to work with the OEMs in trying to get them to stop (though tough for the international ones) and work with the benchmark vendors to defeat optimizations as they are discovered. We're presently doing all of these things and we have no plans to stop. Literally all of our benchmarks have either been renamed or are in the process of being renamed to non-public names in order to ensure simple app detects don't do anything going forward.

SierraSonic said,
Why do you think certain benchmark app developers are making changes to their apps to avoid being activated by this script?

This may be a shocker to you, but I think the app developers should have more say than a company trying to boost their own score.

I would think that the developers of the app should decide if they want their apps boosted or not, why did Samsung force their apps to be boosted?


So basicly, those few benchmark app developers doesn't want to optimize their apps so it can run better on the new smartphones?

So they'll rather use a horribly optimized benchmark app for todays smartphones that doesn't work like it should just because they can't be arsed to update their own apps to run at the full speed on the smartphones today like all of us want since we want to see the true power of our device on what it can deliver us and not just some half baked results of a speed throttled benchmark run.

Think if Futuremark had been lazy as that?

That would be LOL. We would be sitting here with 3DMark 2003 on our computers today witch is not optimized to run on todays computers.

And let me ask you this. Why did the Quadrant developer update / optimize his app 2-3 years ago so the benchmark app could run much better on Dual-Core CPU's after smartphones with those CPU's was released then?

Edited by Exynos, Oct 5 2013, 7:38am :

Exynos said,

So basicly, those few benchmark app developers doesn't want to optimize their apps so it can run better on the new smartphones?

So they'll rather use a horribly optimized benchmark app for todays smartphones that doesn't work like it should just because they can't be arsed to update their own apps to run at the full speed on the smartphones today like all of us want since we want to see the true power of our device on what it can deliver us and not just some half baked results of a speed throttled benchmark run.

Think if Futuremark had been lazy as that?

That would be LOL. We would be sitting here with 3DMark 2003 on our computers today witch is not optimized to run on todays computers.

And let me ask you this. Why did the Quadrant developer update / optimize his app 2-3 years ago so the benchmark app could run much better on Dual-Core CPU's after smartphones with those CPU's was released then?


they are not, not updating there apps, in fact they are updating specifically to block this cheat. See what you call optimizing the makers of the apps call cheating. Since they are the ones that make the apps I'll take their word over how it should run over the maker of devices that only want to boost their score with cheap cheating tactics.

Updating to include dual core support is different than forcing the device to run maxed out and then some.

SierraSonic said,

they are not, not updating there apps, in fact they are updating specifically to block this cheat. See what you call optimizing the makers of the apps call cheating. Since they are the ones that make the apps I'll take their word over how it should run over the maker of devices that only want to boost their score with cheap cheating tactics.

Updating to include dual core support is different than forcing the device to run maxed out and then some.


I'm pretty sure you are preety clueless about what you are talking about.

Because adding dual-core support in Quadrant is optimizing the apps to run better on the new hardwares. Or what do you call it?

So if the benchmark app developers are gonna block their apps from getting optimized, then i'm not sure, but they don't really want us to use their apps, because there is not a point to use a benchmark app who only tests 70% of the power the device can deliver.

Exynos said,
because there is not a point to use a benchmark app who only tests 70% of the power the device can deliver.

glad you are starting to see exactly what we are meaning, only took 100 posts lol!

Exynos said,

I'm pretty sure you are preety clueless about what you are talking about.

Because adding dual-core support in Quadrant is optimizing the apps to run better on the new hardwares. Or what do you call it?

So if the benchmark app developers are gonna block their apps from getting optimized, then i'm not sure, but they don't really want us to use their apps, because there is not a point to use a benchmark app who only tests 70% of the power the device can deliver.

Optimize
to make as perfect, effective, or functional as possible

They will support it to run as optimized as possible, but they also have to add support for new features.

Feature
an interesting or important part, quality, ability, etc.

They have to add the ability to use multiple cores otherwise it wouldn't even detect them. The difference is adding support to use cores is different than forcing the system to work harder for certain apps.

cheat
to break a rule or law usually to gain an advantage at something

The majority of app developers are united to not allow the use of these type of tactics to cheat the system.

SierraSonic said,

Optimize
to make as perfect, effective, or functional as possible

They will support it to run as optimized as possible, but they also have to add support for new features.

Feature
an interesting or important part, quality, ability, etc.

They have to add the ability to use multiple cores otherwise it wouldn't even detect them. The difference is adding support to use cores is different than forcing the system to work harder for certain apps.

cheat
to break a rule or law usually to gain an advantage at something

The majority of app developers are united to not allow the use of these type of tactics to cheat the system.


Still doesn't matter, because there doesn't exist any cheats by allowing apps to use the devices true power.

So you can say whatever you like, but that's the fact.

Exynos said,

Still doesn't matter, because there doesn't exist any cheats by allowing apps to use the devices true power.

So you can say whatever you like, but that's the fact.


What do you mean allowing, they are forcing apps to work differently than the devs want them to work. If I was a truck driver with a governed truck, and my boss set it up to drive no faster than 75 due to his insurance policy. Then we took the truck to a mach servicer and they removed the governor while it was in for service, my boss would be ****ed when gets fined for failing to comply to the policy. He expects the truck to go no faster than 75 as that is how he set it up. You are basically saying its ok for this kind of practice to happen. Even if it is not what the owner of the truck wants.

Edited by SierraSonic, Oct 5 2013, 6:42pm :

SierraSonic said,

What do you mean allowing, they are forcing apps to work differently than the devs want them to work. If I was a truck driver with a governed truck, and my boss set it up to drive no faster than 75 due to his insurance policy. Then we took the truck to a mach servicer and they removed the governor while it was in for service, my boss would be ****ed when gets fined for failing to comply to the policy. He expects the truck to go no faster than 75 as that is how he set it up. You are basically saying its ok for this kind of practice to happen. Even if it is not what the owner of the truck wants.

If i buy a car with 500 hp, then i'm allowed to use that. However, if i'm on a normal road, i'm not allowed to use those hp's. I have to take a trip to a race track to be able to use that.

Do i whine that ohhh boohooo, the others who are on the race track (running benchmark) are allowed to use the true power of the car while i'm driving on a normal road where we have speed limits?

Think about it.

Exynos said,

If i buy a car with 500 hp, then i'm allowed to use that. However, if i'm on a normal road, i'm not allowed to use those hp's. I have to take a trip to a race track to be able to use that.

Do i whine that ohhh boohooo, the others who are on the race track (running benchmark) are allowed to use the true power of the car while i'm driving on a normal road where we have speed limits?

Think about it.


The law doesn't make your car slower, you just avoid driving that fast to avoid a ticket. Nissan may be limiting the car from using max power to avoid liability, but you have to accept this as this is the default performance. And no, a benchmark would be that dyno that's not at the track but in a auto shop far away from the track.

Lets go back to the GTR reference. You are basically saying that even though the driver may want to use the comfort settings on the transmission and shocks, it is ok for Nissan to remotely force his car into the racing+ settings as that is the max of the car, even though the guy doesn't want Nissan to control his car at all.

Edited by SierraSonic, Oct 5 2013, 8:56pm :

Exynos said,

If i buy a car with 500 hp, then i'm allowed to use that.

you are getting closer and closer to understanding what we been saying all this time. you are correct in saying that If you buy a car with 500 hp, you SHOULD be able to use it if the situation calls for it!

rippleman said,

you are getting closer and closer to understanding what we been saying all this time. you are correct in saying that If you buy a car with 500 hp, you SHOULD be able to use it if the situation calls for it!


And if i buy a phone with 2.3 GHz CPU and 533 MHz, i'm gonna be allowed to use that to if the situations calls for it. And Samsung allows me to achieve that for those apps who really needs that power.

So i'm not sure, but it seems like you have never understood what i have been saying all the time. But what you are saying now, is exactly what i have been pointing out the whole time.

So, i don't want those heavy apps to use 480 MHz on the GPU and the CPU at 1.8-1.9 GHz on the Galaxy Note 3 if i bought it with a 533 MHz GPU and 2.3 GHz CPU as standard in the phone. Then i want the really heavy apps to take advantage of those standard speeds that comes with the phone that is 533 MHz GPU and 2.3 GHz CPU.

So are we done yet?

Edited by Exynos, Oct 6 2013, 2:09am :

Exynos said,

Samsung allows me to achieve that for those apps who really needs that power.

So, i don't want some apps to use 480 MHz on the GPU and the CPU at 1.8-1.9 GHz on the Galaxy Note 3 if i bought it with a 533 MHz GPU and 2.3 GHz CPU as standard in the phone.

Since you don't get the full power (but yes, benchmark apps do), i am sure you are mad about that. that's what we been saying this whole time from the very beginning.

rippleman said,

Since you don't get the full power (but yes, benchmark apps do), i am sure you are mad about that. that's what we been saying this whole time from the very beginning.


No, i'm still taking about those apps who are so powerfull that it needs the full power of the device to run smoothest as possible and not the half baked speeds while running those powerfull apps.

Those other apps that is nowhere near to even use 50-60% of the power of the device, i wouldn't care about that as they runs as good as they can to begin with.

Edited by Exynos, Oct 6 2013, 4:41pm :

yes, you are talking about the benchmark apps, i know... its just crappy that "only the benchmark apps" get the full power, i can see why you are upset. I would be too.

Exynos said,

And i'm still taking about those apps who are so powerfull that it needs the full power of the device and not the half baked speeds.

rippleman said,
yes, you are talking about the benchmark apps, i know... its just crappy that "only the benchmark apps" get the full power, i can see why you are upset. I would be too.


I'm not upset. I'm just happy that the light apps who doesn't use much power isn't allowed to use the CPU and GPU on full power as they don't need it to be able to run smoothest and fastest possible. Otherwise it would just be waste of battery for those apps. Do you understand that?

And those apps who actually use ALOT of power and such, should be allowed to use the full potential of the hardware if they reach the CPU and GPU usage at a certain point to make sure those apps runs as smooth and fast as possible.

Apps shouldn't be restricted. I wouldn't be worried about "light" apps using too much power unless they were coded poorly.

Putting a restriction in place for "light" apps would just be a bandaid for poor coding.

Exynos said,

I'm not upset. I'm just happy that the light apps who doesn't use much power isn't allowed to use the CPU and GPU on full power as they don't need it to be able to run smoothest and fastest possible. Otherwise it would just be waste of battery for those apps. Do you understand that?

And those apps who actually use ALOT of power and such, should be allowed to use the full potential of the hardware if they reach the CPU and GPU usage at a certain point to make sure those apps runs as smooth and fast as possible.

You are wrong again though, the apps won't use any more power than they need unless an outside force changes the environmental settings just for them.

Do you understand how apps work? They will always use as much power as they need until they hit the limits of the device. It is irrelevant to the apps weather the limit is hardware or software.

It only matters in this case because certain apps were targeted, the developers disapprove because they consider these deceitful. The developers simply weren't even contacted to find out if they want these apps to be altered to run like this. It really matters as these apps were designed to test the general environment, not the maximum, even according to the developers.

The thing is developers assumed all apps will run the same, and since that's is what the apps were developed for it is ruining the reliability and removing the trust from these rankings.

Edited by SierraSonic, Oct 6 2013, 5:29pm :

i would figure you would be upset knowing "only benchmarks" apps got the full power of the device and all other apps only got most of it... odd... I know i would be feeling cheated had i owned a phone that told me something was "this" powerful when really all it was "just this" powerful. I would feel like i was mislead.

Exynos said,

I'm not upset. I'm just happy that the light apps who doesn't use much power isn't allowed to use the CPU and GPU on full power as they don't need it to be able to run smoothest and fastest possible.

And those apps who actually use ALOT of power and such, should be allowed to use the full potential of the hardware if they reach the CPU and GPU usage at a certain point to make sure those apps runs as smooth and fast as possible.

rippleman said,
i would figure you would be upset knowing "only benchmarks" apps got the full power of the device and all other apps only got most of it... odd... I know i would be feeling cheated had i owned a phone that told me something was "this" powerful when really all it was "just this" powerful. I would feel like i was mislead.


Again, the reason most of the other apps isn't running the CPU and GPU on full power isn't because they aren't allowed, it's because they don't need that full power to run optimal, so they don't need any special needs or roles on the CPU and GPU.

Doesn't matter if you mean some games are overheating the device, because it's still a fact that the device is still way inside it's thermal limits even if it's getting overheatet.

Yes, the device can get really hot, but that's just normal considering what kind of hardware that is under the hood on those devices.

Exynos said,

Again, the reason most of the other apps isn't running the CPU and GPU on full power isn't because they aren't allowed, it's because they don't need that full power to run optimal, so they don't need any special needs or roles on the CPU and GPU.

Doesn't matter if you mean some games are overheating the device, because it's still a fact that the device is still way inside it's thermal limits even if it's getting overheatet.

Yes, the device can get really hot, but that's just normal considering what kind of hardware that is under the hood on those devices.


If the apps can't access the power unless they are in the list, then that is something other than the hardware that is restricting it. That is what I like to call limited. Normally apps access as much power as they need.

Exynos said,

Again, the reason most of the other apps isn't running the CPU and GPU on full power

you keep saying "most"... don't you mean "all"? remember, only the benchmark apps have access to the full power. Doesn't that seem odd to you?

Apparently so, but honestly Nokia doesn't have much to brag about when it comes to performance, they concentrate more on features. The rest of the OS runs quite smoothly on their phones in my experience.

Apple is another story however, they do brag a lot, but most of the time it seems justified based on what various sites have concluded. Of course they do cherry-pick their scores to show the best diagrams when comparing to the competition, but these scores appear to be thruthful. These guys don't like bad PR at all, I think they'll avoid giving the other OEMs the joy of debunking and shaming them on that. (They can sometimes screw up enough by themselves on other subjects to add that kind of insult to injury xD)

Arkos Reed said,
Apparently so, but honestly Nokia doesn't have much to brag about when it comes to performance, they concentrate more on features. The rest of the OS runs quite smoothly on their phones in my experience.

Apple is another story however, they do brag a lot, but most of the time it seems justified based on what various sites have concluded. Of course they do cherry-pick their scores to show the best diagrams when comparing to the competition, but these scores appear to be thruthful. These guys don't like bad PR at all, I think they'll avoid giving the other OEMs the joy of debunking and shaming them on that. (They can sometimes screw up enough by themselves on other subjects to add that kind of insult to injury xD)

Both Nokia and Apple have a similar strategy - marketing the device on the whole experience rather than narrowing in and focusing solely on CPU or GPU speed. End of the day the customer cares about the whole experience; if the UI is a nightmare and difficult to use then what use is it if the process is "super duper fast"?

One thing I don't understand is why even do this. They know that this will only matter to tech people, but for consumers, not at all. Worse part of these, consumers will not know about this, and they wouldn't care.

No kidding. They're going out of their way to improve benchmarking scores it looks like. However, while many people may not go by these benchmarks (as they truly shouldn't), we all have that one "tech" friend who thinks he knows it all, and spreads misinformation (or even disinformation in some cases, depending on bias) which leads people walking into my store telling me they want or don't want X because of [stupid reason].

Things like that are how false stories get spread around, much like how some customers hate Windows 8 not because they tried it, but because their buddy at work said it sucked.

dead.cell said,

Things like that are how false stories get spread around, much like how some customers hate Windows 8 not because they tried it, but because their buddy at work said it sucked.

Hey Dead.Cell. you got that right. However, for some reason, it is different when it comes to PC and mobile devices. With the mobile devices, as long that they can access their email, tweet, facebook, and such, it really wouldn't. When it comes to Windows, well, we all know that is different story.

RommelS said,
One thing I don't understand is why even do this. They know that this will only matter to tech people, but for consumers, not at all. Worse part of these, consumers will not know about this, and they wouldn't care.

i think the info does pass down to consumers in the form of biased advise from their 'techy' friends or showroom staff, when the decision gets to a point of 'ok its between these two, which would you recommend?' often the misinformed showroom agent or friend will subconsciously refer back to some stats or benchmarks which 'clearly showed ' that x was faster than y.

dead.cell said,
No kidding. They're going out of their way to improve benchmarking scores it looks like. However, while many people may not go by these benchmarks (as they truly shouldn't), we all have that one "tech" friend who thinks he knows it all, and spreads misinformation (or even disinformation in some cases, depending on bias) which leads people walking into my store telling me they want or don't want X because of [stupid reason].

Things like that are how false stories get spread around, much like how some customers hate Windows 8 not because they tried it, but because their buddy at work said it sucked.

Also, this affects a lot reviews of other tech sites by having some conclusions to some non-tech people by stating that one is "faster than the other".

At least great reviews do compare them via games, and battery life.

You may not, and are wise in that respect, but many people don't have the technical know-how to differentiate bogus results for legitimate ones, or just flat out rigged scores like contrast ratios on TVs / Monitors

Jockulation said,
Can't say I've ever made a smartphone purchasing decision based on benchmarks... ever!
Many people here live and die by benchmark scores.

and one of the biggest ones is chatting in this article thread. We all know who he is too

JHBrown said,
Many people here live and die by benchmark scores.

Jockulation said,
Can't say I've ever made a smartphone purchasing decision based on benchmarks... ever!

Same.. PC hardware sure. Phones? I go by the "fast enough" scale. Not going to quibble or even care about benchmarks, fudged or otherwise. If it feels fast, it's fast enough. If it feels laggy, it gets tossed back into the pile.

Arkos Reed said,
You may not, and are wise in that respect, but many people don't have the technical know-how to differentiate bogus results for legitimate ones, or just flat out rigged scores like contrast ratios on TVs / Monitors

True, when some customers are faced with two very similar phones what I hear (as a customer browsing in the background) is "well, which one is faster". I'd love to live in a world where people evaluate something on the overall package rather than a specific single point but alas we've been here done that before; gigahertz wars, the storage wars where massive storage is promoted by the rotation speed is pathetically slow resulting in slow seek times as a consequence laggy responsiveness etc. the GPU war where the prices are so similar the only thing left is performance. End of the day it kind of goes back to what consumer protection should cover and that is unsubstantiated claims that can't be backed up with a real world independent test.

Freaking unbelievable! I will light fireworks as soon as I hear that FTC or at least EU sends an ungodly firestorm up the basses of these cheating bassclowns @#*#%(%&&(#$

isn't that the whole point of these benchmarks? I mean benchmark are synthetic measurements anyway so what is wrong with running the processor at highest speed possible? For example, if you want to benchmark your GPU, you run it at the highest speed possible or even overclock it and not run it in some power saving mode.

do not confuse commercial benchmarks, with PR benchmarks

Commercial ones are typically meant to compare platforms one to another to determine the most appropriate one for purchase in a typical usage scenario, on a level playing field (that's what's being rigged here)

PR Benchmarks, or ePeen as I like to call them, are just for bragging rights to say hey I have the highest score when I modify my system this way or that way.

The term is the same (just like there are enterprise process benchmarks to see how well tuned your enterprise ERP is), but the destination isn't.

Typically a benchmark will show you a reasonable peak that you can realistically get out of the hardware. So for instance on a PC, some benchmarks will show FPS you can expect for specific games at certain resolutions and settings. The key being that the performance is actually achievable. These phone benchmarks will never be achievable in anything except the benchmarking app.

It's just a shady way to claim performance that is technically not attainable.

Arkos Reed said,
do not confuse commercial benchmarks, with PR benchmarks

Commercial ones are typically meant to compare platforms one to another to determine the most appropriate one for purchase in a typical usage scenario, on a level playing field (that's what's being rigged here)

PR Benchmarks, or ePeen as I like to call them, are just for bragging rights to say hey I have the highest score when I modify my system this way or that way.

The term is the same (just like there are enterprise process benchmarks to see how well tuned your enterprise ERP is), but the destination isn't.

Don't confuse a knowledgeable person with one that is not.

Oh Teh funny, even Anand Shimpi calls them all out as liars and cheaters with a special mention for Samsung rigging the GPU scores as well by raising the allowed thermal limits and frequencies beyond specs.

I wonder how our resident Samsung nutcase will wiggle his way out of that one (I bet some **** along the lines of *they all do it, it's fine!* or *hey it's benchmarks I want the best epeen!*), not that I'll know or that I care, I've blacklisted him

Arkos Reed said,
Oh Teh funny, even Anand Shimpi calls them all out as liars and cheaters with a special mention for Samsung rigging the GPU scores as well by raising the allowed thermal limits and frequencies beyond specs.
I wonder how our resident Samsung nutcase will wiggle his way out of that one (I bet some **** along the lines of *they all do it, it's fine!* or *hey it's benchmarks I want the best epeen!*), not that I'll know or that I care, I've blacklisted him

An ARM processor hardly overheats, they still don't have enough power for that.

Yup, indeed, they are power sippers, but it's not the same about the (on-die or separate) GPU components of these SOCs, they are heavily threaded and tend to overheat a lot, that's the specific part I was mentioning.

Sure, as much as Exynos Octa is concerned, 50 degrees Celsius is not overheating. It's such a nice hand-warmer, it's not a bug, it's a feature.

Its pretty apparent now that quite a few OEMs are doing this so...

Business as usual? Don't be an idiot and buy something according to benchmark data?

I'd say, better ignore these benchmarks for the time being until things are sorted out, like the ATI/nVidia wars in the past, if there's enough consumer backlash and bad publicity they'll back out of that nonsense.
And for now decide based on your experience/affinity with the devices.

Arkos Reed said,
Oh Teh funny, even Anand Shimpi calls them all out as liars and cheaters with a special mention for Samsung rigging the GPU scores as well by raising the allowed thermal limits and frequencies beyond specs.

I wonder how our resident Samsung nutcase will wiggle his way out of that one (I bet some **** along the lines of *they all do it, it's fine!* or *hey it's benchmarks I want the best epeen!*), not that I'll know or that I care, I've blacklisted him

What you said was blatantly false. The CPU were not over clocked, On the Samsung device they were simply all running at 2.3Ghz which is the native clock speed of the CPU.

Arkos Reed said,
Oh Teh funny, even Anand Shimpi calls them all out as liars and cheaters with a special mention for Samsung rigging the GPU scores as well by raising the allowed thermal limits and frequencies beyond specs.

I wonder how our resident Samsung nutcase will wiggle his way out of that one (I bet some **** along the lines of *they all do it, it's fine!* or *hey it's benchmarks I want the best epeen!*), not that I'll know or that I care, I've blacklisted him

Here is the actual tests with pictures...show me where the CPU was clocked past 2.3Ghz which is its native ratil released speed? Please show us you can count!

http://www.slashgear.com/galax...igans-accusations-01299884/

Repeating a lie doesn't make it the truth...I hope you know that.

HipHopSinceFriday said,
Here is the actual tests with pictures...show me where the CPU was clocked past 2.3Ghz which is its native ratil released speed? Please show us you can count
Kudos on the strawman...attack his use of the word "overclocked" instead of trying to defend the fact that during benchmarks the CPU / GPU are pushed beyond thermal limits and what any other application would get.

the link goes to anandtech regardless, just a minor editing error, typical John, no worries he often mixes things up when copy/pasting