True only for the fact that, well, are there even any programs in iOS which would stress the CPU cores to anywhere near the extent of multi-threaded Cinebench? Because soon as you do give the A7 an actual workload you get throttling - http://www.anandtech.com/show/7519/apple-ipad-mini-with-retina-display-reviewed/2 And before anyone launches into the "well that's an unrealistic power virus" rant, from http://www.anandtech.com/show/7460/apple-ipad-air-review/3 we see that delta idle -> Kraken (single threaded benchmark) delta is 3W while idle -> 'unrealistic power virus' delta is 8W, or 4W per core. So that's a 33% increase between 'realistic' benchmark and power virus... which isn't all that far off from the increase seen between running Cinebench and AVX enabled linpack on a low voltage Haswell. (Playing with my i7-4770k running 3.5 GHz at 0.975V I saw a 59W delta idle -> load for Cinebench and 81W delta idle -> load for AVX linpack.)Depends on the platform. In general the A-series of chips from Apple don't throttle much over time. What you see is what you get.
Agreed, abuse in Android is far more prevalent since they attempt to circumvent their lack of proper hardware power management by specifically detecting benchmarks and maximizing their performance rather than allow their software DVFS to operate. But even without that 'cheat' they'd still have the issue of inflated benchmark performance when the device is cold/benchmark length isn't adequate to reach steady state.The benchmarketing shenanigans came more from the Android side, but only specific vendors, esp. Samsung.
http://arstechnica.com/gadgets/2013...rking-adjustments-inflate-scores-by-up-to-20/