Michal Ždanský Server & Hosting AnandTech.com made a scandalous revelation that caught many Android phone manufacturers cheating on benchmarks by purposefully overclocking their chipsets during testing:
With the exception of Apple and Motorola, literally every OEM we've worked with sells (or sold) at least one device running this silly optimization. It's possible that older Motorola devices may have done the same thing, but none of the newer devices we've had with us have shown this behavior. It's a systematic problem that has apparently surfaced over the past two years, and it's far from just Samsung.
This revealing article was preceded by several other convictions, on the one hand in the case Samsung Galaxy S4 and the latest Galaxy Note 3:
The difference is respectable. In Geekbech's multi-core test, the Note 3 benchmark scored 20% better than it would have under "natural" conditions. If the possibility of a performance increase in the benchmarks is bypassed, the Note 3 will fall below the level of the LG G2, which we originally expected due to the identical chipset. Such a large increase just means that the Note 3 is messing with the CPU at idle; much more performance is made available when benchmarked on this device.
Samsung, HTC, LG, ASUS, all these manufacturers deliberately cheat in the benchmarks by purposefully overclocking the CPU and GPU to achieve higher results on paper. However, this increase only works for benchmarks included in the list inside the system, which is not easy to work towards. There is apparently a belief among manufacturers that “if he cheats others, we must too. After all, we will not be behind in the benchmarks".
Apple has never boasted about CPU clocks or benchmark results (with the exception of web browser benchmarks) on its iOS devices, it didn't need to. If the device works perfectly smoothly, the customer doesn't care about test scores whose names he can't even pronounce, let alone remember.
In the world of Android, everything is different, manufacturers are fighting with the same (or similar) weapons, and benchmarks are one of the few places where they can show that their device is better than others. However, this disclosure renders most benchmarks irrelevant, as reviewers and readers can no longer be sure who is cheating and who is not. A popular technical thing that is only used by reviewers to prove that they have really tested the device thoroughly, and to geeks for whom these numbers really mean something, maybe it will completely disappear from the mobile sphere and everyone will instead start looking at whether the system is smooth, as well as the application inside it . After all, it's always been that way with the iPhone.
It may not surprise anyone these days that Samsung and other manufacturers cheat to make themselves look better. But it's sad and embarrassing at the same time. Great admiration, on the other hand, goes to the server AnandTech i ArsTechnica, which proved specific lists of "supported" benchmarks parse from the code.
Have you seen the "sales graphs" of "Czech cars from Nošovice" compared to Skoda, which the Korean is literally obsessed with??? It is enough to move this or that model to another group, in which it is not normally, and you are the winner in terms of sales right away. :-D Koreans… North or South…
It's not quite the same thing. It's a different type of ochcavka, and on the other hand, the price section is the answer :-).
Oh yeah. :-) Koreans just have to defend their nationalistic lathe no matter what… :-)
For me as a customer, it is better to compare cars at the same price level. That is it's good for me
If someone outright deceives me by claiming to have performance that they don't have, that's another matter ;-).
With all due respect, this is myopic demagoguery... Either you're working in HJUNDÉÉÉ or you don't read it yourself... Reassigning a vehicle to a different class than the one it's in is a simple fraud. The classes primarily determine the size of the car... If we rely on this simple and probably immutable fact for Koreans (even though...), then we have to think like this: If you need the size of a middle class, you have to buy a middle class. According to your logic, it wouldn't matter whether I buy a BMW 1 series or a new Škoda Superb, because they basically cost comparable money... Of course... They are completely different cars, which, with certain space requirements, can at least "surprise" you with your approach. :-)
I don't work in the automotive industry ;-).
Everyone, I meant a comparison file between Combi and Combi + depending on where the price is.
If it's a 2-door and some kind of combi car, that's really nonsense, but I probably haven't seen that yet :-).
Well, that's what Hjundééé did... They moved the Rapid, for example, because (this is the official press statement, please) "they put it in a different class of cars themselves". :-D This is a classic Korean scam…
But still, the Koreans are right that a Škoda is just a rolling trunk :) and the English wouldn't lean a wheel on a Škoda since time immemorial.
Clear. Almost everyone cheats, only Apple doesn't :D
Apple doesn't need it, its 5S is the best :D
I think that the most important thing these days is rough performance. It's like when years ago everyone bought cameras that had the most MPx. At first it was clear that it could produce a decent photo at 15×10, but a year later it was the least important factor that people still cling to today. They don't care that they have plastic optics and that the camera can't focus ;-).
There's no point in cheating. What will it do me if the benchmark shows me some number? I won't be happier for some number :-).
Either the phone does what it's supposed to and quickly and I'm satisfied, or not, and then I'm completely on my ass what number he got.
So far, every Android I've had in my hands has been limited to basic functionality, so it's dead (not optimized for HW).
In addition, Apple has no reason to cheat because it only produces the iPhone
he. If it was made by multiple manufacturers, like Android, there might be a reason ;-).
Apple has no need to cheat because only it makes the iPhone. That's a goddamn argument. You are probably missing the biggest rivalry with Samsung. So the fact that it is the only one to make an iPhone has nothing to do with it. I understand people who don't allow Apple to be allowed, I was like that myself when I still had reasons, but people are out of fashion now. Kor after iOs7. Where is the place for the text button and the ubiquitous white color. Terrible bastard. I wouldn't even want to test this, but I'm forced to.
So iOS7 is definitely the worst thing that happened to me during the last cleanings, I agree. However, it still doesn't reach a completely premium Android (which I own, so I know what I'm talking about).
Otherwise, I am personally at such a level that I would definitely not choose a phone based on the score it gets in embarrassing benchmarks (especially when big companies can afford to pay for optimization for such SW :-) ).
Another thing I think is that a person who knows a little about it will not decide between operating systems until the end, so no performance comparison is important. The right OS is much more important, and you can go for any.
Then I don't care if there will be an embarrassing operak on a phone that has the power to calculate nuclear reactions, if there won't be a single application on it, or if the OS will be unstable, or if it will give away my personal data.
So the user does not have to pay any attention to any media rivalry and should think about what the mobile phone is actually for. For me, the performance of the iPhone 5S is absolutely useless, because I don't need to play the latest 3D games and video editing on my mobile. Of course, if the basic operations get stiff or fall and I don't function otherwise, that's essential for me.
:) hehe, sorry, but the output of the iPhone is also a big "Bullshit" (I have a 4Sko, and sometimes I have to recharge it daily)
Well, you can't beat him like that... ;)