Search Search

    Android smartphone makers “cheat” in benchmarks

    One more piece of news to unnecessarily tarnish the reputation of Android smartphones. Android smartphone manufacturers, namely: HTC, Samsung, Asus e LG, seeking to stand out, “cheat” in benchmark tests. As? Putting instructions for the smartphone to work beyond its capacity, only during benchmark tests. The practice would even be cool, but as this doesn't happen in any other application, there's a sense of deception.

    The trick is so “in the face”, that the Arstechnica website found a file, with a readable code, specifying in which apps the smartphone is allowed, in this case Samsung Galaxy Note 3, to set the clock “topado” of the processor cores at 2.3 GHZ and the GPU at 533MHZ. So when renaming the benchmark apps, the device shows different results. Have you thought? Renaming a file in Windows and making the computer run faster or slower?

    It may seem like a small thing, but it leads to incorrect data. Imagine if you buy a 2.0-liter car, you test-drive it and the car surprises you in terms of performance, but when you get home you notice that it is the same as the other 2.0-liters?

    In my opinion this is just one way that companies have sought to try to stand out. It turns out that the manufacturers of smartphones and tablets with Android are betting on very similar models, with almost identical configuration, always being in the “war” of those who launch the latest processor version that most of the time are not even manufactured by them, but by third parties. .

    As there is no manipulation in the kernel in Android ROMs, only a “makeup” of the interface, the natural tendency is for the devices to have very close results in the benchmarks. When it doesn't, it may raise an eyebrow, as someone is using tricks to look good in the picture.

    But surprising is to realize that almost all manufacturers do this. An article posted yesterday by Anandtech, made an overview of the main devices on the market and showed that only Motorola and LG smartphones from the Nexus line do not have any kind of “tweak/cheat” for benchmarks.




    Maybe many fanboys, of one brand or another, don't see any problem, but the word to use is HONESTY. It would be nice and seen with good eyes, if the manufacturers informed about the “tweak” to be carried out during the benchmarks. But anyway, it's a dishonest competition with those who don't use such “tweaks”.

    Why do such instructions exist in these files and because they are exclusive to benchmarks? Why do they exist on some devices and not on others? It would be fairer for everyone to have it, wouldn't it? The problem is really just that, such instructions exist on some smartphones and not on others, it would be legal and more transparent for the consumer if everyone used it or no one used it. This fact can lead the consumer to think that a smartphone is much more powerful than another, when in fact it is not!

    Note children, the question seems to be about specification and performance, but IT IS NOT! The issue is about honesty and that's what these companies above lacked. You can spend the whole day trying to fool me with voltages, GPU, Clocks and the devil at four, which won't explain why there is a specific code for Benchmark that was not placed by Google, but by the manufacturers, aiming, of course, to do well in the benchmarks .


    add a comment of Android smartphone makers “cheat” in benchmarks
    Comment sent successfully! We will review it in the next few hours.