Anonymous, 11 Feb 2018Of those running same machine with both OS I have seen a difference closer to 5%. Which sites ... morean i3-6100 scores ~3100 on Windows 10, whole scores ~3500 on Ubuntu 17.04 LTS. that's quite a difference.
well, maybe you find reputable reviewer not using GB as benchmarking tools as fine. but you should think about it.
yes, A11 is the fastest smartphone chip on the market, that's a fact. but originally, my comment was to not compare the score with SD845 score directly, as it makes no sense and doesn't translate to real world performance. synthetic benchmark =! realwork usage. besides, what is scores when in real life an android is as fast as an iphone even though it's 75% slower in synthetic single core benchmark?
Fulljack, 11 Feb 2018No professional desktop hardware reviewer ever use Geekbench. Not Ars, not AnandTech, not PC P... moreOf those running same machine with both OS I have seen a difference closer to 5%. Which sites you find "professional" is up to you. It's still being used and again there are tons of (better) options on desktop, on mobile there is not.
Adding a 5%, 10% or even 15% difference would not make a difference. The A11 is so far ahead (75% faster than these scores in Single score) you can not say this difference i just because of OS.
And you did not answer my question. Do you not believe the CPU in A11 currently is the most power CPU on the mobile platform?
Anonymous, 10 Feb 2018Geekbench 4 is being used as a benchmark tool in reviews for desktop hardware. The difference ... moreNo professional desktop hardware reviewer ever use Geekbench. Not Ars, not AnandTech, not PC Perspective, no one has ever use it.
10% difference aren't margin of error, it's an evident of an underlying flaw.
I despise Geekbench because it's not a tool that you can use to compare two different SoCs that running on two different platform. If you compare SD 660 with SD 835 with Geekbench, I don't really mind that. But A11 against SD 835 based on Geekbench alone? yeah, the test was flawed from the beginning.
Hip hip hooray,
HB266 (Xperia XZ1)
It's Sony... Sony.. Sony...
Fulljack, 10 Feb 2018yes it is. just because both are Linux-based, doesn't mean the compiler are the same. optimiza... moreGeekbench 4 is being used as a benchmark tool in reviews for desktop hardware. The difference is that for desktop we have a wide range of benchmark tools to use where as with mobile platforms we don't. Geekbench 4 is not perfect but its the best we have.
Comparing a desktop CPU running on Windows vs Linux will give you single/multi scores within the margin of error. Nothing points for it not being the same on mobile.
Why do you oppose this so much? Do you not agree that A11 is the most powerful ARM CPU out there at the moment (and probably for quite som time more)?
Anonymous, 09 Feb 2018You very much can. CPU performance has *nothing* to do with underlying OS (unless there are pr... moreyes it is. just because both are Linux-based, doesn't mean the compiler are the same. optimization are bound to happen. I'll ask you this, why do you think in professional PC segment no one has ever use Geekbench for comparison tool?
20% IPC increase ain't shit. It's not easy to create an architecture from scratch, AMD even need half a decade to do that
wow good specs!
Hmm we'll see. Interesting...
I hope this one is not a sony phone
20% faster. Really, only 20% faster. Apple is like 30-40% year over year. At this rate SD 855 will finally catch up to A11
where are you stupid intel processor
Vegetaholic, 09 Feb 2018Of course you can, they both based on Linux and are similar in core and I already mentioned 20... moreYeah and it can draw more power and throttle more too.
Especially since the batteries are already smaller than average
Anonymous, 09 Feb 2018You very much can. CPU performance has *nothing* to do with underlying OS (unless there are pr... moreCongratulations on a blind view on everything. When there are variables a-z you cant just look at a and b and go welp this means this. End of story. ...
But sadly that's how you and 90% of the world view everything and why it continues to be a place of destruction and hatred
I hope that isn't near final, cause that is a small improvement compared to exynos 8895. When oc'ed to 2,65 GHz, it reaches ~2240 single and ~6850 multi. If oc'ed to 2,8, it prob reaches 2300+ and 7000+. Exynos 9810 will prob have a much higher score, they build better big cores then qcom atm.
Fulljack, 08 Feb 2018before you guys comparing it to iPhone's Geekbench scores, I suggest you DON'T. You can't comp... moreOf course you can, they both based on Linux and are similar in core and I already mentioned 20 times that benchmark gives same task to processors regarding OS or chipset, it's damne very accurate what geekbench 4 offers between iOS and android and everybody knows that a11 destroys any Snapdragon or exynos, cause it's much larger in die so numbers are completely true
Very disappointing results as I understand correctly, exynos in S9 should reach single core scores close to A11? So basically 845 will be canon meat this year or Samsung exegurated very badly about their exynos chipset