Last week, we detailed in a report the whole controversy about the differences that exist between the chips of the TSMC and the Samsung found on iPhones 6s and 6s Plus. While some tests have pointed out significant variations in battery consumption, Apple has come to the public to say that in real use this difference is only about 2-3%.
But is that what Apple talked about? O Ars Technica decided to check with new tests.
Four different tests were performed, all of them performed at least twice to avoid any discrepancies. All of them confirmed Apple's claim (one of them, that of WebGL, even gave a positive result to the Samsung chip), with the exception of only the Geekbench that had been used as a reference in other tests.
The explanation for this is that Geekbench leaves the CPU running at ~ 55-60% of its capacity during the entire run. That is, when iPhones are being used for heavy processor tasks it is possible that the difference between the chips becomes more noticeable.
Summing up the whole story, we are talking about very small practical differences that should not be a cause for concern for anyone. Still, the fact that it all came out and Apple's own public statement shows that this is yet another controversy that it could have avoided.