How Are Android Makers Cheating In Benchmarks
Whenever a new smartphone comes in sight, the benchmark tests are the best way to figure out where they stand in front of other smartphones already released. This comparison is meant to stress the smartphones ( or tablets ) CPU and GPU and offers a score based on how well they managed to deal with the volume of information. And because this can make a difference between a successful and a bad device even before its launch, the Android based makers are aware that the sales depend of these kinds of tests. So they found a way to cheat by including a few tweaks on how their devices are handling the test.
Based on the report made by our source, it states that devices from major Android makers are optimized to score unrealistic performances in the benchmark tests, and this way assuring a better percentage among competition. Until now, the suspicion was quite unfounded as nobody could really prove how they do it, but it seems that Android makers forgot to prepare their devices for all benchmark tests, therefore they have set maximum performance for well known tests and did nothing for influencing some other not so used benchmarks.
It is true that during the report, they have found some clean devices, who are presenting real data during all kind of tests and they worth mentioning their names: Motorola Razr i, Moto X, Nexus 4, Nexus 7 and Nvidia Shield. On the other side we have names like Asus, HTC and LG who optimized so “well” their devices so that they were able to influence the final results, making them score higher than it should.
But let’s cut to the chase and bring out real names, and we should start with the famous LG G2 and Asus PadFone Infinity who are optimized to score high in AnTuTu and Vellamo benchmarks. Snapdragon 800 is the CPU used in both cases, and when it detects one of these two benchmark tests, they go wild and run in the sixth gear, while some other 5 benchmarks in which they were tested, the interface didn’t bother to offer the same performance, scoring much less that initially. The list ends with “almost clean” Galaxy Note 3 who was optimized to boot in six out of ten benchmarks. They have done a pretty good thing but not until the end.
In the end we can clearly see how customers are fooled into a misleading performance of the product, and the Android manufacturers are investing huge amounts of money into cheating tests rather than investing them in real higher performance products. In order for you to get the big picture, we highly recommend you reading the full report and appreciate the work of made by Anand Lal Shimpi and Brian Klug who ran all those test and made public all results with detailed analyze for major products. You can acces it here and until next time, make sure that you choose the right product who scored well in all benchmark tests.