[HeyCori]

Exposed

Note 3's benchmarking "adjustments" inflate scores by up to 20%

[Jazmac]

Wow. If this is faked, what else did google/samsung fake?

[HeyCori]

Cheater's gon' cheat

Samsung reportedly not alone in cheating Android benchmarks

[spazzmeister]

This is true, the benchmarks when rerun without overclocking are still impressive and the fastest at the moment.

[xandros9]

I'm actually on the fence on this.

Because the point of benchmarking is knowing what the upper limits of the hardware are, and pegging it up to max would make sense in that context. I think it would've looked best for them to give an option, "conventional use" with scaling and powersaving measures enabled and a "performance" mode that goes all out no matter what.

[tissotti]

I'm actually on the fence on this.

Because the point of benchmarking is knowing what the upper limits of the hardware are, and pegging it up to max would make sense in that context. I think it would've looked best for them to give an option, "conventional use" with scaling and powersaving measures enabled and a "performance" mode that goes all out no matter what.

Benchmarking is flawed from the start. You can get some useful information out of it if you actually know the parameters and how everything actually affects everything. Problem is that people usually looking at benchmarking results are just those who don't get it. That is if the benchmarking program is open about what it exactly does.
GPU world is especially famous about this and has been going since the start.

To me it just amazes Samsung would actually do this on their position. I would love not to hate Samsung as they do make true hardware innovation, but they have always seemed like such rats, especially when they grew up (similar to Huawei blatantly stealing tech from nsn and Ericsson, going so far that they copy closed B2B site straight from nsn).

[BIGPADDY]

Dont trust any Android device benchmarks!!!