Originally Posted by Gigatel
Please name some grown up purposes. This is a very common Android talking point. Please give examples of whatever it is you're talking about.
If Apple was doing this, I'm quite certain we'd have heard about it by now. Anand said himself Apple and Motorola don't do this. But I guess he's bias and outright lying?
If there's no motive behind cheating the benchmarks, why would they cheat the benchmarks? Because they know several review sites will put the phone through its paces and post the results. Better results translate to better press, etc etc. It's pretty simple. Some users like to know the phone they're buying is reasonably better than their previous device. Phone reviews are in every major publication in the US, i.e. USA Today, NYT, WSJ, etc. People read these things and device OEM's are outright lying and cheating.
I'm not sure why OCN is okay with this. Oh right, the double standard. Just search OCN for mobile bnchmarks and it's pretty clear this community, which is sooooo focused on performance, is most definitely concerned with mobile performance. It is truly amazing how fast the collective brain trust changed its tune in light of 5s performance and Android OEM's cheating. Now all the sudden, the pursuit of performance doesn't matter.
Anandtech tested in an extremely inefficient manner, I did mention this and I mentioned how it can be improved. They also have no method of testing this on Apple phones. I'm inexperienced with Apple phones, but my understanding is that there is no open source tool to decompile the apps and view human readable code. Without that, there is no way of telling if Apple do this.
Anandtech simple used a hex editor to look at the .dex file, which is a rubbish method of looking inside. They saw the packagenames of the benchmarks and put 2+X together and got 4. IT's unclear that X = 2, but they've assume it was.
I think you misread me as well, there is clear motive for cheating the benchmarks, but they didn't falsely advertise (if we can be literal), nor did they lie. To lie you must be asked about a subject, and deny it. Cheating? That remains to be proved unequivocally.
I think, to me, it appears that you are targetting a select few on the forum with your comments about the benchmarks. I'm certainly new here, but I don't conform. I think benchmarks matter, but I don't care if X benches better than Y.