Snapdragon 820 vs A9 vs Exynos 8890 vs Helio X20 benchmark chart scores one for Apple
5. roldefol (Posts: 4097; Member since: 28 Jan 2011)
Who cares? At least it's running neck and neck with the Exynos.
48. sgodsell (Posts: 3073; Member since: 16 Mar 2013)
All of the competition will most likely continue to crush Apple in the multi-core benchmarks. Considering most of the OS use multiple threads to run, then multiple cores is more real world, instead of a single core test.
78. Acer_Predator (unregistered)
exactly there is no use of single core all aplications auto uses multi cores
only stupid idion in geek bench made single core benchmark that naive ppl will see biiig high number :DDD
149. vincelongman (Posts: 3989; Member since: 10 Feb 2013)
Currently games still have poor multi core usage
Its due to OpenGL ES not being being a low level API, it doesn't scale well for multi core
But hopefully Android N brings Vulkan support, which would allow games to scale far better
IMO single and multi core are equally important on Android
150. Shocky (Posts: 2113; Member since: 16 May 2012)
Games on mobile devices don't need much in the way of CPU performance so it doesn't really matter yet, hopefully that will change.
It's not like they have complicated physics engines and character AI to process so they barely get used.
They're entirely GPU limited.
152. vincelongman (Posts: 3989; Member since: 10 Feb 2013)
That's true too
I was more trying to point out a specific area which doesn't scale well for multi cores
But game loading speeds do tend to be noticeably faster on e.g. the iPhone 6S+ than the S6 or 6P, whereas the S6/6P are usually equal in other app and web loading speeds, but then again the 6S+ does have a much newer and faster GPU
157. Shocky (Posts: 2113; Member since: 16 May 2012)
I understand what you were trying to say but how is it meant to scale when the games is are GPU limited?
You could put i7 4790k in there and it still wouldn't scale with that mobile GPU, game loading speeds are more to do with the speed of the flash storage than the CPU.
The iPhone 6S/+ has double the read and write speeds of the Galaxy S6 so nothing to do with the CPU.
165. vincelongman (Posts: 3989; Member since: 10 Feb 2013)
Loading a game would be read speeds
The iPhone 6S+'s real world read speed would vary because of the way its storage solution works
It is mainly TLC NAND, which is slower than the MLC NAND in most phones, as seen in the random read/write speeds where it can't predict what to do with the SLC NAND cache as well
But has a SLC NAND cache which is significantly faster than MLC NAND, as seen in sequential read/write speeds
So depending on the size of its SLC NAND cache, it may not even be in play if it's not big enough to hold the games files, so I wouldn't rule out the CPU
169. Emacs244 (Posts: 6; Member since: 29 Oct 2015)
> some websites
All websites. Websites are running by blink and V8 js engine typically on android. Both blink (chrome web engine) and V8 are single-threaded. Thus, the single core score reflects web pages execution performance directly.
> Currently games still have poor multi core usage
Parallel programming is extremely bug-prone and just complicated thing. The majority of algorithms just cannot be effectively parallelized and that's why the single core performance is still more valuable than multic core.
170. Emacs244 (Posts: 6; Member since: 29 Oct 2015)
Sorry, you hardly realize what are you talking about.
166. cheetah2k (Posts: 1499; Member since: 16 Jan 2011)
How does the A9 go on a 2k screen? Let me tell you not very well
167. strudelz100 (Posts: 644; Member since: 20 Aug 2014)
Could easily handle multiple 2K displays with the A9 GPU.
168. tedkord (Posts: 9942; Member since: 17 Jun 2009)
Yrs, they do. The multicore does are there to see.
14. Martin_Cooper (Posts: 1368; Member since: 30 Jul 2013)
No one who buys android phones cares if snapdragon can beat apple's cpu simply cause apple don't make android phones and no one else can make an iOS phone. But its good to know how they all compare and are progressing of course.
136. Pattyface (Posts: 1511; Member since: 20 Aug 2014)
This may be the first time I 100%agree with you.
21. jove39 (Posts: 1787; Member since: 18 Oct 2011)
Qualcomm is never going to make a cpu core that is more powerful than cpu core of apple ax series.
Nvidia is only one...capable/daring to make successor of denver core that'd be more powerful than twister core.
32. roldefol (Posts: 4097; Member since: 28 Jan 2011)
Nope, they'll just make four cores that are good as everyone else's 8 (or 10).
52. tedkord (Posts: 9942; Member since: 17 Jun 2009)
Maybe not individual core, but they're consistently more powerful overall.
90. bucky (Posts: 2448; Member since: 30 Sep 2009)
by overall you mean only in multi-core I'm assuming.
108. marorun (Posts: 2918; Member since: 30 Mar 2015)
And most games and app use multi-core so multi-core scoore is more important.
161. Shocky (Posts: 2113; Member since: 16 May 2012)
Got proof for that? since you're saying "most games and app" you must have a source that's tested a huge number of apps, yes?
109. Awalker (Posts: 1368; Member since: 15 Aug 2013)
Yes. It's been that way for years now. Mobile phone chips makers excluding Apple try to maximize multicore performance. Apple focuses more on single core performance.
134. engineer-1701d (Posts: 3214; Member since: 13 Mar 2014)
intel will make the move if not google remember google is looking into producing its own
26. Taters (banned) (Posts: 6474; Member since: 28 Jan 2013)
The title is misleading. It's typical iPhonearena reporting where they make the title favour Apple and show the only metric that the A9 wins and leaves out the other charts or put the other charts on the second page and in fine print.
137. Pattyface (Posts: 1511; Member since: 20 Aug 2014)
You act surprised by this haha.. PA is about as bad as verge with its bias
142. ya4dang1 (Posts: 3; Member since: 18 Nov 2013)
At the end of the day, only users' perception matters.
27. BLUEBLASTER (Posts: 256; Member since: 23 Feb 2014)
Apart from games. What else do we need all that power for though?
34. Ordinary (Posts: 2142; Member since: 23 Apr 2015)
Cracking Wi-Fi password from neighbours?
79. StanleyG88 (Posts: 230; Member since: 15 Mar 2012)
I'll use my i7 desktop for that. Why mess around with some mobile processor when you need REAL power.
80. Acer_Predator (unregistered)
u can't do that
60. VigneshRaja (Posts: 31; Member since: 17 Dec 2014)
I don't think even mobile games need that much raw power. Even 801 can do well.
111. marorun (Posts: 2918; Member since: 30 Mar 2015)
even SD800 is more than enuf to run everything smoothly.
145. Cela911 (Posts: 18; Member since: 15 Jan 2015)
My old LG G3 was struggling a little with some game's and its QHD screen paired with the 801 processor.
65. rd_nest (Posts: 1578; Member since: 06 Jun 2010)
You should look in a different way. Most of the time mobile SOCs stay in low power mode, and use such high performance states only for small time period. This allows them to complete the task faster and return to idle state faster. The more it idles, less energy it takes.
So, even though you may do the same activity, you are likely to conserve more enerygy with a higher performance SOC.
112. marorun (Posts: 2918; Member since: 30 Mar 2015)
Thats only part of the equation.
higher speed usually mean higher battery load..
This with the fact OEM keep on shrinking battery make current smartphone in general crappy for battery life.
110. marorun (Posts: 2918; Member since: 30 Mar 2015)
Even games dont use more than half of this.
138. jeroome86 (Posts: 800; Member since: 12 Apr 2012)
To make that super charge phone call and that blazing fast text.
49. Subie (Posts: 437; Member since: 01 Aug 2015)
Have to wait for the official release of SD and then benchmark it. Even the article gives Qualcomm and Samsung leeway on this ("Of course, the Snapdragon 820 benchmarks have likely been collected from tests on a Qualcomm reference platform, as there is still no device with it on the market, and ditto for the Exynos 8890")
51. TechieXP1969 (Posts: 7919; Member since: 25 Sep 2013)
It does. Single scores are stupid because today's mobile OS and applications never use just a single core---EVER!!!!!!!!!
57. AlikMalix (Posts: 4855; Member since: 16 Jul 2014)
You sure about that? I believe it's quite the opposite and apps rarely use more than one core.
75. uzimafioso (Posts: 195; Member since: 15 Jul 2014)
That would be true for Scorpion processors, not anymore (at least not for Android).
82. Acer_Predator (unregistered)
no my friend no.. they are all still in standby mode and if needed they just engage .. and they engaging very very often so you won't complain about lag or shuttering
83. Techist (Posts: 225; Member since: 27 Jan 2015)
Saying "I believe" doesn't make anything true. People believe all sorts of fictions, you being one of them. Here are the facts:
"When I started out this piece the goals I set out to reach was to either confirm or debunk on how useful homogeneous 8-core designs would be in the real world. The fact that Chrome and to a lesser extent Samsung's stock browser were able to consistently load up to 6-8 concurrent processes while loading a page suddenly gives a lot of credence to these 8-core designs that we would have otherwise not thought of being able to fully use their designed CPU configurations. In terms of pure computational load, web-page rendering remains as one of the heaviest tasks on a smartphone so it's very encouraging to see that today's web rendering engines are able to make good use of parallelization to spread the load between the available CPU cores."
"In the end what we should take away from this analysis is that Android devices can make much better use of multi-threading than initially expected. There's very solid evidence that not only are 4.4 big.LITTLE designs validated, but we also find practical benefits of using 8-core "little" designs over similar single-cluster 4-core SoCs.
"Clearly Android apps are able to take advantage of multi-core processors and big.LITTLE allows the scheduler to pick the best core combination for the current workload. If you still hear people saying things like “but a smartphone doesn’t need 8 cores” then just throw your hands-up in despair, as it means they don’t understand Heterogeneous Multi-Processing and they don’t understand that big.LITTLE is about power efficiency and not overall performance."
84. AlikMalix (Posts: 4855; Member since: 16 Jul 2014)
When people say "I believe" they're suggesting they're open for discussion on the matter because they may be unsure. You took it the wrong way - but I appreciate the breakdown and links - I'll look over that when I have more time.
115. roscuthiii (Posts: 2150; Member since: 18 Jul 2010)
It's arguing semantics, but "believe" does not mean open to discussion. Belief holds something to be true regardless of factual support.
Ideas can changed, beliefs are more set. People die over their beliefs. Historically, more so than anything else.
88. TechieXP1969 (Posts: 7919; Member since: 25 Sep 2013)
Yes I am sure. Do you own an Android phone? You can go into developer options and turn on the couch to show the cored on all screens. I have never opened any application and seen only one core...EVER!
When phones only had single cores, no one was even doing these types of tests. All of today's smartphones all use multiform operating systems and hardware. Thus single core tests are even real world.
For all you douchebag fanbois of racist apple, all who jump on Samsung saying they cheated because they white list benchmark apps and had a script which ran the clock at full speed through the whole test which is not real world usage, so how is this any less?
All mobile OS' are designed to use multiple cores even if the app you load wasn't originally designed this way. This had been proven for years even on Windows. Today's OSS all have the ability to split a thread of applications even if they were not specifically written for multiform platforms
This just proves my claim that PA cheery picks just like in reviews whereby hey cheery pick apps they know reload on Android and don't on its so their favorite can when in a situation where it would otherwise lose. That is fact.
Single core scores are not teal world usage. The fact is Apples new chip can't even beat the 7420 in multiform which is the focus of the OS and apls. Period.
You can't change that fact nor can PA.
113. marorun (Posts: 2918; Member since: 30 Mar 2015)
Maybe on iphone but not on android.
the way android work is even single treaded application will get its tread shared on all cores.
135. engineer-1701d (Posts: 3214; Member since: 13 Mar 2014)
i thought apple was moving to dual cores plus and moving to 3gb ram and up to start a trend, or is it to help its os from stuttering all the time like the 6s with 30fps in games and 60fps other times.
64. almostdone (Posts: 231; Member since: 25 Sep 2012)
The chart is very misleading as only the very last chart 2016 on the right is show casing all the latest prcoessors. This is only one of many synthetic benchmarks and we don't know if the Exynos 8890 or Snapdragon 820 is fully optimized yet. The difference is also not that great so in another benchmark the results may be different.
74. Neros (Posts: 1016; Member since: 19 Dec 2014)
Apple gets crushed when it comes to multicore score. Nothing new here. Apple will always have the best single core performance based on the differences architecture (fewer cores but bigger in size).
77. Acer_Predator (unregistered)
yeah and this stupid pillars for stupid ppl who thinks that higher score is better.. like you.. better for draining battery within 3 hours hahaha nothing else is good for .. so stupid.. best is mediatek efficient in saving battery
85. fyahking (Posts: 859; Member since: 28 Jan 2015)
But, Nexus 6p with the sd 810 is already faster than the iphone 6s plus. So, the benchmark don't mean crap when it's slower than certain phones.