Wednesday, 09 October 2013 20:45

AnTuTu speaks out against cheating on benchmark tests

Written by

Reading time is around minutes.

Popularity of benchmark application was exploited by manufacturers of mobile devices that managed to present their device better than it really was with little tricks like short-period overclocking and similar. Thanks to these practices over time, benchmark tools have become quite unreliable. This was also due to fact that different results were achieved when testing the same device with different benchmark tools.

Research that showed that Samsung and other manufacturers of smart device were adjusts settings of their devices when they detected that one of the most popular benchmarks, AnTuTu, is running. Now AnTuTu decided to put an end to it in order to protect their name ,the relevance of the results, and of course their business.

AnTuTu has released a new version of their benchmark application, which looks identical to the old one, but is enhanced by additional tricks that should prevent producers in fixing benchmark tests. What exactly those tricks are, AnTuTu has not specified, which is understandable since it would be stupid for them to expose their weapon in the fight against cheaters.

[Ed - We are glad that AnTuTu is taking this stance, but there is very little that they can do to stop manufacturers from putting in cheats and optimizations for their benchmarks. The same thing is still happening in the PC world and will continue to happen. The best thing that can be done is to make people aware of the issue and remind them that synthetic benchmarks are (and always will be) only part of the performance story. They represent a small portion of the actual abilities of a device and too many web sites rely on these tests to show case the performance of mobile devices. Very little time is spent these days on actually testing the product to see how it feels to the end user or how easy it is to use the product. Maybe this revelation will move the industry to start really testing products instead of just running canned benchmarks.]

Tell us in our Forum

Read 2244 times Last modified on Wednesday, 09 October 2013 20:48

Leave a comment

Make sure you enter all the required information, indicated by an asterisk (*). HTML code is not allowed.