Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
So only Google and apple are behaving when it comes to meaningless benchmarks?

Many an e-peen will be deflated with this news.
 
And this is why Android phones have been able to claim great performance via benchmarks and specs while iPhones have performed just as well or better in practice. I remember folks telling me that the Galaxy S2 performed better than my iPhone 3GS according to benchmarks, but in practice (side-by-side) my iPhone 3GS seemed to be twice as fast with no stutter on tasks that seemed to stutter and lag on the GS2.

As someone else stated, that's only part of the issue. Android is a bloated mess in comparison to iOS. I like my Nexus 4 and Android 4.3 but there's a reason why Android-based smartphones need more horsepower - Android (and thus most of the apps) is not optimized nearly as well as iOS.
 
So, all the Asian manufacturers are corrupt, and all the American ones are clean.

What about Nokia?

And are there any Russian or Middle-Eastern manufacturers?

I'm curious how this relates culturally.
 
Seriously - who cares how it benchmarks just as long as the phone is fast enough to use the apps I want.

The people choosing between 5 similar phones. All things being relatively equal why wouldn't you pick the faster one?
 
Asian market always takes competition to the next level, whereas American companies usually value integrity and honesty over sacrifice of principles. I am proud to be an American because all the American companies here (Google, Moto, and Apple) have class and present their phones as they are. I could imagine the decision in the engineering department, a VP thought of the good old days of a budding Silicon Valley where a young boy could get a job at a company by walking up to its CEO, where innovation thrived, driven by a group of idealistic young men, so he scoffed at the proposal to compete with these Asian newcomers at their level. Samsung might be the biggest smartphone maker, and HTC and Asus are poster childs for successful Taiwanese companies, but they have nothing of the tradition and history of people like Robert Noyce, Gordon Moore, Sherman Fairchild, and Andy Grove. This is why Silicon Valley continues to be the birthplace of most cutting edge technologies and inventions while their Asian counterparts master their arts of copying and expanding. I think what it comes down to is risk. What makes Silicon Valley special is the risk-taking courage. Samsung could come up 50 useless whiz features because as long some of them work, some generate buzz, then they accomplish a goal. It is 1,000 harder for companies like Google and Apple to say, create an entire new platform, or embed a finger print sensor, or use a new screen technology (retina), a new casing material (metal and glass); these are much more risky because they require huge capital investment and a commitment of your product line and company identity where a failure cannot be an option. Then once they pioneered it, it is much easier for Asian companies to follow suit and refine. "You got a retina? I will out 1080p in my phone." Well you see, it is not hard to put 1080p in a phone, it is hard to be the first company to take the huge risk to explore a new market. Once proven, of course it is easier for other companies to learn and follow. People of the world need to give more credits to this tradition entrepreneurial culture that differentiates American companies on this list. Without Google and Apple recently, or Intel, Fairchild, Bell Lab, Microsoft, etc, Samsung and LG will be still copying Sony and make DVD players.

----------

But, what I can't figure out is why fake these scores. The galaxy s4, htc one and most of the phones and tablets on this list are pretty fast. Of course the galaxy line has its stutters because the os isn't fully optimized, but why cheat if you've already passed the test lol. But I use an iPhone so doesn't matter to much to me

Think about it from the engineers perspective. They always want to build and code their products to be better, not meeting a standard. Now if they know how to ace a test, why wouldn't they tweak their systems to do it? Well some engineers won't, those with class and confidence that their PhD degrees probably from a better school taught them to use their time well, on work and research of actual substance and impact, instead of wasting time on inflating their works.
 
So the Samsung's Quad-Core phone cheated benchmarks and still get lower benchmark score than Apple's straightforward Dual-Core CPU?


Hmm interesting.
 
i don't really know many (if any) people who actually use benchmarks to decide which phone they are going to purchase. especially when trying to decide on a platform (i.e. ios, wm, android)

i guess this is mostly used on the android side since there are so many different sets put out?

I think they're looking for the customer that already is committed to Android but is trying to decide which handset to upgrade to.
 
Why am I not surprised? If the phone does what you require of it, do the benchmarks really matter? Besides, how many people pick a phone because of benchmarks. I would think it to be a very low percentage. The average person would not care.
 
I dont like cheaters.I also dont care about benchmarks of mobile devices as long as they run the apps I want and that until now always worked.
 
All they do in Korea is copy everything. I love Koreans but they don't innovate in any business. When you go there all you see is fake Nikes, Adidas, cheap suits, faux leather, faux wood, faux gold, faux everything. It should be expected that they cheat in benchmark tests.

I didn't knew that Koreans write with Chinese symbols.
 
I also believe benchmarks are pretty meaningless. It's still super embarrassing when you cheat to inflate your score.
 
Despite all of the effort that OEMs put into benchmark optimizations, the gains are negligible. The impact on CPU tests revealed a 0 to 5 percent performance increase, and a less than 10 percent increase on GPU benchmarks.

They might be doing it for a different reasons then?

I see lots of talking about the benchmarks in the article. But I do not see any attempt to see whether it is not the benchmarks which are targeted, but particular types of the workloads. I know for a fact that GPU drivers (from both nVidia and AMD/ATI) can adapt some settings in run-time based on how the game has initialized and uses the 3D subsystem. But then, with PC games/benchmarks/drivers the cheating was confirmed by simply renaming the executable. In the article, nothing similar was done, and neither tests were done with real actual workloads to see whether they trigger the same behavior.

All in all, if there is no change in performance, why they then consider it to be a cheating at all?
 
Glad to see the Nexus 4 being well behaved.

And here's hoping that the Nexus 5 is based off the G2 (with no cheating!)

The only reason the nexus 4 is behaving is because its an android pure device. Its once the OEM's get their hands on those cheap plastic phones that thematic starts to happen.
 
I didn't knew that Koreans write with Chinese symbols.

adadis is english. If your talking about the REAL iPhone vs. the FAKe iPhone. They both have chinese characters. Does that make the REAL iPhone chinese? Well its made in china but its an American company.
 
The only reason the nexus 4 is behaving is because its an android pure device. Its once the OEM's get their hands on those cheap plastic phones that thematic starts to happen.

Nexus 4 cheap plastic phone?

adadis is english. If your talking about the REAL iPhone vs. the FAKe iPhone. They both have chinese characters. Does that make the REAL iPhone chinese? Well its made in china but its an American company.

First, Adadis is english?

And second, you have show Chinese knockofss, they don't have nothing to do with Korea
 
They might be doing it for a different reasons then?

I see lots of talking about the benchmarks in the article. But I do not see any attempt to see whether it is not the benchmarks which are targeted, but particular types of the workloads.

ArsTechnica looked into that on the Note3, there is a file with a list of benchmark applications to trigger the behaviour based on the name of the application. If you rename your benchmark application then it doesn't boost it.
 
Of course it's legal. Why would it be illegal? Is it ethical? Probably not.

In Germany, it would fall under "unfair competition" and would be illegal. If company A spends a million dollars to make the processor in their phone 20% faster, while company B spends 1000 dollars to hack the operating system to make their phone appear 20% faster by cheating on benchmarks, that's clearly unfair competition.
 
In Germany, it would fall under "unfair competition" and would be illegal. If company A spends a million dollars to make the processor in their phone 20% faster, while company B spends 1000 dollars to hack the operating system to make their phone appear 20% faster by cheating on benchmarks, that's clearly unfair competition.

Apart of the moral question, they are not making the processor appear 20% faster, it is its nominal speed, they are not overclocking it.

It is illegal optimizing javascript engines for benchmarks?
 
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.