Alberto (git.delete@this.git.it) on February 7, 2015 11:22 am wrote:
> Umm yes, because they not say if the IPC increase claim is "mostly 10%" or "mostly 50%"
> :), neither they say if the sustained increase in performance at the same power is "up to"
> or "in average" or "in some lucky cases" or "in most cases" etc.....they are vague.
> And yes some informations on test are welcomed.

Peter's descriptions were clear and unambiguous, but obviously he is operating outside of official releases and there's a limit to what he can say - like giving large data sets of benchmark results. But I doubt anything would really be good enough for you.

> Still i think the gain is mainly in the power management, not so strong in both A15 and A57.

How exactly does a change in power management result in an increase in IPC by 10-50%? If all of the performance was gained by an increase in clock speed enabled by a decrease in power consumption you may have an argument, but that is clearly not the case.

> This time they have worked like Qualcomm have done with its very power efficent Krait core line???

Krait gets all this praise for being so power efficient but I have yet to see a real comparison with normalized performance on the same process and all that. All I've seen is comparisons where A15 uses more power (sometimes a lot more) while also performing better (sometimes a lot better). I repeat it constantly, but perf/W is not linear and these comparisons are meaningless. Reduce the clock speed on the A15 until it matches the performance on the Krait core and then do a fair measurement - of course, this will vary tremendously from benchmark to benchmark, especially since Krait appears to have some big glass jaws.

What we do know is this: despite your remarks that power management is "not so strong" in A57 Qualcomm still preferred to use it instead of a version of Krait that only added ARMv8 support, something I'm sure they could have done in time if it would have given them such an advantage.