Apple A9 chipset scores high in GeekBench tests; huge improvement over A8

Apple A9 chipset scores high in GeekBench tests; huge improvement over A8
According to a series of benchmark tests conducted on GeekBench, it looks like the Apple A9 chipset is quite a powerhouse. Carrying a dual-core 1.8GHz CPU, the A9 scored an impressive single-core score of 2522 on both the Apple iPhone 6s and Apple iPhone 6s Plus. That tops the 1316 single-core score tallied by the Exynos 7420 on the Samsung Galaxy S6 edge, the 1061 produced by the Snapdragon 805 powering the Nexus 6, and the 1953 score belonging to the Tegra K1 inside the Nexus 9 tablet.

The single core score of the A9 is 57% better than the 1605 produced by the A8 chipset that is found inside the Apple iPhone 6 Plus. The A9's multi-core score of 4423 on the iPhone 6s Plus is a 54% improvement on the A8's multi-core score on the iPhone 6Plus. The only multi-core score that beat out the A9 was the 4404 scored by the Exynos 7420. That topped the multi-core score of the A9 powering the iPhone 6s, by a single point.

The Apple A9 chipset produced by Samsung uses the 14nm FinFET process, and features 2GB of LPDDR4 RAM. That is twice the amount of RAM that Apple employed with the A8.

source:  AppleInsider



1. carlemillward unregistered

No surprises here.

35. vincelongman

Posts: 5745; Member since: Feb 10, 2013

Agreed Apple's SoCs have always matched the Qualcomm/Samsung equivalents in performance So it shouldn't be surprising that the A9 matches the 810/7420 Especially considering those are 6 months old now Also everyone needs to stop to whole dual core vs octa core thing Apple designs for iOS, so they want max single core performance, hence why they use dual cores Qualcomm/Samsung designs for Android, so they want max multi core performance, hence why they use quad cores or octa/dual-quad cores Also these benchmarks are misleading The Exynos 7420 usually gets over 5100

37. Diezparda

Posts: 941; Member since: Oct 23, 2013

I think am the only one who sees S6/edge scored so low at phonearena table comparison. It usually scored 1500 single-core and above 5100 in multi-core... Great benchmark score by A9, kudos to Apple... I really want to see real-life performance between 6S & S6

46. vincelongman

Posts: 5745; Member since: Feb 10, 2013

The 6S/6S Plus will be faster in games (due to faster GPU, lower res and Metal API) And 6S/6S Plus will be faster for web browsing (due to content blocking, Safari's great JavaScript engine and websites tend to be more single threaded) I reckon they will be similar in terms of app opening (due to similar mutli core scores) But Apple wins overall because TouchWiz has horrible ram management

84. Diezparda

Posts: 941; Member since: Oct 23, 2013

Probably as you'd expect, i'm curious to how developer will implement force touch to new gaming experience. Again i'm not Apple fans and almost never play mobile games LOL... I'm touchwiz fans too, what people claim as bloat sometimes i find it as nice feature (always used multi window on my Note 3). Quite impressed by how Samsung at least try to simplify their ui. Some says their horrible ram management was purposely - to keep the phone performance, is that true?

74. Nathillien unregistered

No, you are not - phonearena is shilling for whoever pays the most because:

106. mitoita

Posts: 77; Member since: Apr 03, 2014


111. j2001m

Posts: 3061; Member since: Apr 28, 2014

The above scores are crap for the edge plus, it's also all down to software and things running on a vm, if Samsung run tizon on a 7420 you see way bigger benchmarks, oh but then geakbench will have to make an app and we have no idea how the app changes from Apple to android phones Here are my scores for my sim free just updated edge plus And

124. astroboy_888

Posts: 6; Member since: Jun 12, 2015

That is because benchmark is being ran by a 3rd party. And that is the true performance number. Samsung has been caught building in benchmark accelerator (like VW emission cheater). When the device detects a benchmark program, it kicks the CPU into over-clocks mode, so that it can give a really high number. So no one should believe any number put out the company. Only trust independent 3rd party benchmark numbers on Samsung products.

41. ThePython

Posts: 902; Member since: May 08, 2013

Excellent comment. +1

53. arch_angel

Posts: 1651; Member since: Feb 20, 2015

IKR And (dual-quad cores) I Think You Meant Hex Core LOL.

55. vincelongman

Posts: 5745; Member since: Feb 10, 2013

By dual-quad cores I meant big.LITTLE octa cores (4 big + 4 LITTLE) Since there's 4 big cores are for performance and 4 LITTLE cores are for efficiency E.g. the Tegra 3 had 5 cores, but was called a quad core because 1 core was for efficiency Although that's not quite the same since the A53s do have far higher performance than the Tegra 3's power saving core

58. arch_angel

Posts: 1651; Member since: Feb 20, 2015

LOL I Thought You Meant Two Plus Four.

103. Shocky unregistered

It's strange, all the tech sites are quoting results for the Galaxy S6 which are lower than the devices actually score. I could run Geekbench 3 all day long but I'l never reproduce those results, I can think of conditions where it would score that low but the user would have be an idiot to run benchmarks in that situation.

107. j2001m

Posts: 3061; Member since: Apr 28, 2014

They used a very low score for the edge plus, my edge as always scored way better than the score

45. DarkStar286

Posts: 229; Member since: Mar 18, 2014

Actually I do find it a little surprising. Apple are usually clear at the top of the table until the next round of flagships in the new year, but this time they're only just beating the best of the rest. I hadn't realised just how much Samsung's SoC had improved over the last year's version; I thought it was more a case of Qualcomm dropping the ball, but Samsung really did nail it with the 7420.

72. Nathillien unregistered

78. nodes

Posts: 1162; Member since: Mar 06, 2014

A9 has almost 2x single-core performance than Exynos 7420, while retaining the same multi-core. and not to mention, Apple's SoCs are the most stable throughout higher temps, while Exynos and Snapdragons throttle a lot.

109. xondk

Posts: 1904; Member since: Mar 25, 2014

One of the problems I have with these tests is that Apple's score almost always seems 'better' in initial tests, but those I know that have (though not iphone 6s yet) don't see that performance increase in actual realworld scenario's, and don't misunderstand. I don't mean comparing iOS interface and such how it feels 'faster' then previous, because of course it does, Apple can specifically tweak and do whatever they want to make it do so, the benefit of so few products. Personally I would love to see people playing actual games on previous and new phones and such, real world tests of various different software rather then these synthetic benchmarks, because lets face it cheating has been seen and it wouldn't surprise me if it still happens..

115. nodes

Posts: 1162; Member since: Mar 06, 2014

because specs on modern smartphones are good enough for simple tasks. just like other flagships too, it won't differ much in UI speed, but when it comes to heavy gaming, you will feel the difference.

125. xondk

Posts: 1904; Member since: Mar 25, 2014

I can't speak for 6s naturally but that's not the impression I've gotten from those I know that own iphones and upgraded from 5->6 , though yes, granted higher resolution, it didn't feel 'better' it felt the same in terms of responsiveness and fps in games. I'm not saying it isn't as powerful as it is, i'm just saying numbers of benchmarks really don't say a lot.

62. AlikMalix unregistered

I'm really interested where Note5 is on these comparisons. Am I on the wrong planet today? I know I'm an iPhone fan, but wouldnt Note 5 bench higher than the androids on the list above?

89. YourKnowledge

Posts: 60; Member since: Aug 17, 2015

Note 5 has the same Exynos 7420 used in the S6 so no it wouldn't bench higher.

94. AlikMalix unregistered

Oh, so the beast got beaten by 2 core 2012 device?

117. alex3run

Posts: 715; Member since: May 18, 2014

Where is it beaten?

122. AlikMalix unregistered

It's as fast with multicore score (but only with two cores vs 8). But did you see single-core score? iPhone imbarrased Samsung (and any other device) there!

98. nodes

Posts: 1162; Member since: Mar 06, 2014

Note 5 has the exact same chipset and frequency as the Galaxy S6. Note 5's advantage is only the amount of RAM, which is unrelevant here.

108. j2001m

Posts: 3061; Member since: Apr 28, 2014

The note 5 also does not get hot at all, like the edge plus unlike the s6 or s6 edge and will go longer without underclock, also the above scores are crap for the new Samsung phone they are bigger, I have never got the above scores

Latest Stories

This copy is for your personal, non-commercial use only. You can order presentation-ready copies for distribution to your colleagues, clients or customers at or use the Reprints & Permissions tool that appears at the bottom of each web page. Visit for samples and additional information.