Apple iPhone 6 (Apple A8) performance review: CPU and GPU compared to the best Android phones out there

The Apple A8 is the first 20nm SoC on mobile, bringing a 25% boost in CPU performance and a 50% improvement in GPU compute…
This is a discussion for a news article. To read the whole news, click here


90. Taters

Posts: 6474; Member since: Jan 28, 2013

It's wrong because you are comparing cou/GPU power, not which phone uses the inferior screen by a far margin. Lol

111. TylerGrunter

Posts: 1544; Member since: Feb 16, 2012

Sure, but then why show only the test for the iphone 6 and not the iphone 6+? Not to talk taht we are talking about a SoC, not a phone. There the imporant one is off screen to know which oneis the better SoC. And that without taking into account that the ON SCREEN TEST ARE WRONG. They are rendering at the 5S resolutions and then upscaling. So pretty much giving false data. And what about not comenting somelthing on the line of: iphone GPU results are so high due the the LOWER resolution of the iphone. I have not seen anything like that in the review either. So it´s your choice, but they are not helping reader to make a choice, just giving the best face for the iphone.

83. TylerGrunter

Posts: 1544; Member since: Feb 16, 2012

Anandtech have updated their article after my feedback, the test in Basemark should not be an issue, as you only put the final score in the pictures which is calculated OFFSCREEN. The issue is happening in the low level when you do the on-screen test. Quoting from Anandtech: "Before I get into the results, I must caution that Basemark X will have inaccurate on-screen results as the benchmark was made using XCode 5.x in order to keep scores comparable between versions 1.1 and 1.1.1. This doesn't affect the overall score, which is solely calculated based upon off-screen performance." (no, I don´t constantly check their site, they just mailed me when they corrected the article based on my input ;)

107. Neo_Huang

Posts: 1067; Member since: Dec 06, 2013

What about one hand mode?

114. vincelongman

Posts: 5677; Member since: Feb 10, 2013

Why use browser benchmarks (Sunspider & Kraken) to compare CPU performance They mainly test Safari Vs Chrome Vs OEM's browser Unless if they are using the same browser, e.g. testing the iPhone 6 vs the iPhone 5S You should have used Geekbench (a CPU benchmark) for the comparisons (though ARM v8 has an advantage) ~4200 - Note 4 (Exynos 8433 ARM v8) ~3200 - NVIDIA SHIELD Tablet (K1 ARM v7) ~3200 - Samsung Galaxy Alpha (Exynos 8430 ARM v7) ~3200 - Samsung Galaxy S5 LTE-A (Snapdragon 805 ARM v7) ~2900 - Moto X (2014) (Snapdragon 801 ARM v7) Anyway, good work on the otherwise great article Any plans to do detailed articles like this when the 64-bit K1 or 64-bit Exynos 8433 are released?

94. strudelz100

Posts: 646; Member since: Aug 20, 2014

Onscreen is the only thing that will ever matter in daily use bro.

115. Taters

Posts: 6474; Member since: Jan 28, 2013

Not really because most daily use tasks only need like a 200 score on antutu. This is a dick waiving contest so if your screen is only 750p, your score should be irrelevant. Just look at the iphone 6 plus scores and ignore the iphobe 6, Alpha, and Z3 compact ridiculous 45000 and up on Antutu. If you just look at the iPhone 6 plus vs the GS5 LTE A. They are close as hell with the GS5 LTE A having two huge disadvantages. First, Chrome and IE should not be compared with Safari period. Safari is much lighter. And the GS5 LTE A has a QHD screen vs the 6 plus only 1080p.

73. roscuthiii

Posts: 2383; Member since: Jul 18, 2010

How many different A8 smartphones are out there Victor?

82. Victor.H

Posts: 1055; Member since: May 27, 2011

My point here was that Bay Trail is currently only shipping on tablets, not smartphones.

121. roscuthiii

Posts: 2383; Member since: Jul 18, 2010

My point is you got caught making a snarky comment before thinking it through.

126. bucky

Posts: 3776; Member since: Sep 30, 2009

his response was completely correct. Why was it snarky? The comment he replied to was snarky.

16. Muayyad

Posts: 240; Member since: Oct 05, 2012

Androids problem relies in cache. Now ART is coming, things will change a lot. I tried it in KK and you can tell it is going to improve the experience. It is still experimental in KK BTW.

19. DefinitiveKid

Posts: 261; Member since: May 15, 2013

"Naturally, using the 32-bit 805 translates into those flagships not being able to benefit from ART optimizations like the 20% performance and battery longevity increase in Android L." Didn't quite get that one. I thought the Nexus 5's battery life got longer with Android L? It's using a 32-bit chip so...

25. Victor.H

Posts: 1055; Member since: May 27, 2011

I'll edit this part out. You're correct about Project Volta bringing those particular benefits to 32-bit architectures as well, but the case remains for all else.

50. Muayyad

Posts: 240; Member since: Oct 05, 2012

Well, I switched to ART in my HTC one M8 and now apps don't use power as much as dalvik mode. Actually, I get at least 25% more juice by mid night also noticed scrolling got way better with ART.

26. twens

Posts: 1172; Member since: Feb 25, 2012

Enough with this bla bla 64 bit gimmick. How does the iPhone perform against the competition? That is the real question, from all the videos I've seen, the iPhone is still struggling to out perform the old 32bit chips on android. Even the notoriously heavy touchwiz on the s5 didn't give in to the iPhone 6. So to hell with this a8 hype and 64bit gimmick. The exynos 5433 is quietly killing the a8 and no one wants to even talk about it. The a8 is not that great so stop all this hype. No 64 bit will be great until paired with real 4gb of ram. We are not stupid, stop forcing what apple tells you on us.

38. DefinitiveKid

Posts: 261; Member since: May 15, 2013

I will never understand fanboyism. The iPhone 6 is extremely fast, get over it. TW didn't give in to it? Yeah right. The only phones that are up to par with the iPhone 6 speed are the new Moto X and probably the HTC One M8, because they have faster animations for their interface. GTFO with your TW comparison. When you're hating on something, at least try to be credible. Fanboys are the worst.

58. Taters

Posts: 6474; Member since: Jan 28, 2013

Um the Nexus 5 killed it and that's a $350 phone.

123. iushnt

Posts: 3085; Member since: Feb 06, 2013

did u miss the s5 vs 6 comparison?

152. CellPhonesExpert

Posts: 45; Member since: Oct 10, 2014

lol ...

31. tiara6918

Posts: 2263; Member since: Apr 26, 2012

We have to give credit when it's due. iPhone really is faster than any android smartphone as apple managed to show how smooth everything is on ios. Even my one m8 still lags even if I go to developer's option and max out the performance.

35. twens

Posts: 1172; Member since: Feb 25, 2012

Running smooth doesn't mean it's fast. That's where the confusion starts, IOS is more consistent in it's overall performance which is mainly due to the optimization and limited features of the OS. Android is generally fast but is not consistent in it's performance like IOS. That's what makes people think iPhones are fast. The iPad 2 is very smooth but is not fast compared to other android tablets. Consistency is the key android is yet to address and that's where I give IOS a thumps up.


Posts: 30; Member since: Aug 16, 2014

*eating popcorn

41. AnukulVcool unregistered

The fact that android runs on java "virtual" machine (JVM) obviously undercuts the performance of android when compared on same specs. iOS runs natively on its hardware. To all the the negligent android fans, this is a real fact, just digest it.

44. RebelwithoutaClue unregistered

Don't forget that Armv8 brings a lot of extra performance as well, even without the 64bit, just the v8 instructionset.

56. wilsong17 unregistered

I guess you don't realize that why we have squaq processor. . At the end of the day android is more powerful...

104. jbo5112

Posts: 11; Member since: Feb 24, 2012

Android software runs ion Dalvik virtual machines, which is much better optimized to the target platform, not Java, but if you're concerned about resource usage, Android has offered native apps in C++ for years. I'm pretty sure the actual OS is written in C/C++. An actual fact: The Nexus 4 phone (yes, the 2012 one) beats the iPhone 6 by around 14% in the 3Dmark physics benchmark, which uses a real-world physics engine. The developers have confirmed this isn't a software bug.

110. jbo5112

Posts: 11; Member since: Feb 24, 2012

I should also note that with the upcoming Android release, programs will be compiled into native code, no longer using the Dalvik VM.

43. caesantos

Posts: 90; Member since: Feb 27, 2014

Are we supposed to believe it? Good try, but you do not convince me at all.

45. fanboy1974

Posts: 1345; Member since: Nov 12, 2011

I'm not worried about the speed of iOS. How about fixing all the bugs and having the developers optimize apps for the new phones. No one has complained about the speed of iOS in years.

* Some comments have been hidden, because they don't meet the discussions rules.

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.