The question is old, and in this regard, comments, discussions, debates were wasted: how realistic benchmark smartphones and processors that all manufacturers waved in four winds? Are the results recorded in these tests true?
To comment on doubts today comes the analysis carried out Anandtech and it is aimed atHelio p95, one of the flagship SoCs MediaTekDoubts arose when the results of his tests were compared with the results of Dimensity 1000L, later and intended for the same range of devices.
The source in question used a modified application for testing performance so as not to recognize an attempt to test the device. The results, you guessed it, are much lower than those advertised. The explanation is simple: Helio p95 maximizes performance (which therefore does not reflect “real” use) when it recognizes the application benchmark,
This could be an end in itself, also because a Taiwanese company quickly responded to allegations of “test fraud”: MediaTek he first talked about the performance of various processors based on the operations to be performed, including me benchmarkIn short, levels that are defined at the table, but which also include other possible areas of activity.
MediaTek he also spoke of the direct responsibility of manufacturers smartphone (which may disable “higher” levels of productivity) and practices that are also common among other competitors. Will there be enough of these answers to protect the company from criticism?