It’s a headline we’ve been seeing pop up a lot lately: “__ manufacture cheats on benchmarks, Android community appalled”. While that may not sound like a very big deal to some of you, keep in mind companies like AnTuTu who specialize in benchmarking software are taking this very seriously. Why? Well, if everyone begins “cheating” on benchmarks and skews results in hopes of presenting their devices in the best light, benchmarks will lose credibility and user demand for them will decrease. This is something AnTuTu takes very seriously.
With their livelihood at stake, AnTuTu announced via their blog today that they’ve created a new app called “AnTuTu X Benchmark”. While essentially the same in UI and function, X Benchmark uses AnTuTu’s new “camouflage tricks” to sidestep OEMs from targeting their app. Exactly how that’s all done is a mystery, but it could be enough to make AnTuTu X Benchmark a reliable option for checking out benchmarks on your favorite Android devices.
We have a feeling the cat n’ mouse game between OEM’s and benchmarking apps will continue, but let’s hope AnTuTu can keep Android manufacturers honest. At least for now.
[AnTuTu | Google Play]
Don’t care. Never relied on benchmarks for making my next phone decision, which Turner’s out to have been a very good thing! My buddies HTC One spanks my GS3 pretty good, but my GS3 still runs nice and smooth, so who cares? It’s just a number.
it is just a number, but its nice to have those numbers when comparing processors from one generation to another… or when new chip makers come to the scene, or if you are buying a phone that you cannot test out in person (like an international variant or phone), knowing what ball park it might fall into.
Good on ya AnTu TU
One of the most unreliable benchmarks, different scores with every update.
I wouldnt be surpised if the iPhone gets knocked down a peg with that new app lol
You have no idea what you’re talking about
I do , i am joking that the iphone is cheating on its benchmarks
Actually, there was a study done recently that showed that Apple and Motorola were the only manufacturers (at least, recently) to not implement any benchmark “cheating” on any of their phones. http://www.anandtech.com/show/7384/state-of-cheating-in-android-benchmarks
I didnt know motorola wasnt spoofing
And Google, no Nexus devices were caught ‘cheating’. Motorola isn’t Google. Google owns Motorola. If your only going to mention one company it would be more accurate to mention Google and not Motorola, otherwise mention both.
and Nvidia! everyone forgets Nvidia!
I don’t see why you would HAVE to mention both every time. Google has apparently given Motorola some autonomy anyway.
That said it should certainly be noted that the Nexus devices (notably made by LG and Asus) don’t seem to be playing this game and is important to bring into the conversation.
Google is not a hardware manufacturer. Yes, Google owns Motorola, but that doesn’t make Google a manufacturer, just an owner of a manufacturer. Further, as I understand it, Motorola is run almost completely independent of Google, and vice versa.
Google sells the Nexus devices with their name and the hardware manufacturers name on it. The hardware manufacturers (LG, Samsung, HTC, e.t.c.) have been caught but Googles Nexus devices haven’t. Personally I feel this needs mentioning and to not mentionmention Googles clean track record is an injustice despite your attempt at a justification.
Apple doesn’t manufacture their hardware yet most people would be happy describing AppleApple as a mufacturer as you yourself did in your initial comment. You can’t have it both ways!
they were not found to be currently cheating based on current checks.
and nexus devices…
the 5s has good benchmarks because they updated the processor but they kept the same screen red so it gets boosted a bit. and the ui is really optimized because of the very few devices they have to make it for
that i know
The 5S si above the crowd, even in resolution independent tests (offscreen), and even in OS independent tests (GeekBench). It’s a beast, that’s all.
Cafreful.. most Iphone users are emotionally unhinged hipsters…
Ever since that big stroy about how Samsung was gaming the system broke a while back I’ve been thinking about how pretty much everyone probably games the system to a degree. I noticed looking at graphics card benchmarks recently that some benchmarks had varying results, and some placed some cards higher than others, but the reverse was true on a different benchmark. Also real world gaming benchmarks were notably different than the controlled environment in the benchmark. I wondered just how much this was going on with every kind of chipset maker, or other OEM.
But hey, it’s like Chris said, this stuff will go on till the end of time. The only thing that will vary is who has the upper hand at any given time…somedays it will be the OEMs, others it will be the benchmark guys.
I did the AnTuTu benchmark test on my Samsung s4 it gave my device top marks the high 876257 what is it measuring and also what does rooted meen in a simple explanation
I noticed you guys never covered this one, despite me submitting it… I guess because it doesn’t make Samsung look like the sole villain:
funny, one person says a site is run by samsung fanboys and another says the same site is run by samsung haters.
The only person I imagine would think the former would be an iPhone user who doesn’t actually read the articles. Their first “top 5” article including the two at least tied the HTC One w/ the S4… every one since has given the HTC One the edge despite lack of fundamental features once touted as strong reasons why Android devices are superior to iDevices (namely removal battery/SD)… They’re both great devices with their own strengths and weaknesses, but calling one definitively better than the other is purely subjective based on user preferences. Not acknowledging this displays clear bias.
I’ve also seen at least 3 articles on here about Samsung cheating benchmarks written with a condescending “shame on you!” tone. Of course when it comes to light that all major manufacturers are doing this, it is conveniently ignored. We get it, the HTC One has a beautiful chassis… but I stopped buying a phone for its chassis when the Nexus One came out and I gave the iPhone to my wife.
Removable battery/sd are cool features, but their convenience/usefulness has much less of an impact today is they did in 2010. 3 years ago when your phone would only last 8 hours to a charge, and it came virtually no internal memory and nothing but an 8gb sd card, these things were a must. Now with some phones lasting 16 hours up to a day and a half on one charge and loaded with 32gb of internal memory for base models, those features for the most people are no longer necessities, but a pro bullet point on the pro/con list.
Also they have said that those devices are pretty much a tie, n they even listed it as such until i think the 2nd or 3rd unofficial smartphone ranking (which is based purely on personal preference of the editor) where they said “i finally decided to give the HTC one the edge.
It’s also not ONLY about aesthetics and feel, many people prefer the new Sense to be > than the new touchwhiz. But i wouldnt get upset because your favorite phone only ranked 2nd on the best android phone list by someone who personally decided to buy an HTC one. (which by the way was lauded on many sites as well to have a slight edge if they were forced to choose)
This is what an objective unbiased comparison of the two looks like:
But reviews aside, ignoring the fact that HTC is cheating benchmarks as well is a little more than a tell. I’ve seen far less android-relevant articles here.
yes that a nice comparison… however when this site posts this: “In the end, you can’t go wrong when choosing between the HTC One and Samsung Galaxy S4. Any number of the above categories could have resulted in a tie, and many will come down to personal preference.” – Phandroid I don’t see what the big deal is.
I understand they didn’t post that article you wanted them to post, but with the gs4 being tied for 1st on their smartphone list of top 5 smartphones to buy, and the gnote 2 being on the list for an entire year, and the gnote 3 being number 1 and gs4 being number 3… i don’t really get the same feeling that you get about being anti-samsung.
Was pointing at a bias, not a dogmatic hate. Putting any HTC device ahead of the Note 3 would result in a complete credibility loss. Citing the fact that HTC is cheating bencharks too would have also invalidated all their prior finger wagging at Samsung.
I just don’t get how they are “cheating”. Everything seems to imply that the phones boost their performance during benchmarking by actually boosting performance…. how is that cheating?
They are “cheating” because when a benchmark app is run, the code will boost the performance of the CPU/GPU during that time, and during that time only. You don’t get the same performance when doing other things, no matter how graphic/processor intensive it is. Essentially, as I understand it, this is no different than running the test with an under-clocked processor, and then running it with a normal or over-clocked one.
So what you are saying is that my games need to have this “benchmark” flag in then and then they will run faster?
just checked and my phone cranks the cpu during Must games. Doesn’t this imply that the benchmark is at fault if it doesn’t crank up the cpu?
It cranks up the CPU during games, but not as high as it does during benchmarking. Benchmarking apps get special treatment that NO OTHER APP does.
So you just need to rename whatever game you’re playing to ‘Geekbench’, then you’re good to go :)
While I’m not a developer, I imagine any bit of code that cranks up the CPU more than usual will have drastic effects on battery life (negatively), which will more than likely piss off a ton of the users of your app(s).
And to be blunt, if you need to resort to these kinds of methods to make your app/game run faster, you’re probably either not coding correctly, or you like taking the easy way out. Either way, in my opinion, you’d be no better than the manufacturers who implement this code. Just my 2 cents.
I’m not a developer but if my game might run better and I’m already plugged in I want the maximum performance. I didn’t pay for a note 2 without wanting the full power it can provide me.