iOS graphics are much better than Android

The new "Metal" tech inside iOS 8 bumped up the graphical capability of a 5S past the capability of a 360 and PS4, pushing close to XbOne and PS4 in fidelity.

It's quite amazing.

Metal is just a low level API, it doesn't allow hardware with ~50% the performance of the Xbox 360 and PS3, to offer graphical capabilities close to the Xbox One and PS4.

NVIDIA's Tegra K1 (32-bit variant) can offer performance slightly above an Xbox 360 or a Playstation 3. It offers ~2.5x the performance of Apple's A7 and ~1.8x the performance of Apple's A8 (based on GFXBench 3.0 Manhattan offscreen). It has the same API support and feature set as a desktop Kepler based graphics card (OpenGL 4.4, DirectX 12, tessellation, CUDA, etc.). The PowerVR 6/6XT GPU's inside of the A7/A8 only offers support up to OpenGL 3.x/DirectX 10, and equivalent features.

OpenGL NG is a low level API, and it will be an industry standard (even Apple's on board).

It's not quite at that level, we've got a ways to go before we have a package as thin as an iPhone 6 with the power of something like a PS4 or Xbox One. But quite honestly, games are like cameras. The best games we have are the ones we have with us and can play anytime. In that sense, the iPhone -- with the practically unlimited supply of high quality games via the App Store -- is second to none.

Metal will make them prettier and faster and give them longer legs. It's a very exciting time to have an iPhone.

Side note: With the smaller overall performance jump from the A7 to the A8 and given the fact that the 5s was the most popular, best selling iPhone before the iPhone 6 and 6+ (both of which I imagine will greatly outsell the 5s), expect that iOS devices sporting an A7 or A8 processor to have a long, long, long lifespan.

Certain games have updates in recent days to support Metal. The difference is definitely not negligible.

Asphalt 8 looks like a completely new game, adding visual effects that would push the envelope on a PS3, but is being processed on a 4S smoothly and with a consistent framerate. The gains are absolutely real.

All I did was update to iOS 8 and a few of my older games are already taking advantage of it. It's amazing to see the huge boost in visual gain just from a mere software update. Makes me even more excited to finally upgrade to an iPhone 6 with up-to-date specs and finally grab a new iPad Air when it comes out next year.

@mister-man Your exaggerations are ridiculous, these features would NOT push the envelope on a PS3.

Features added with Metal, such as the shadow effects seen in the images below, have been around since OpenGL ES 3.0 was introduced over a year ago (even longer on Tegra devices with unique features).

@NVIDIATI: Every year, the iPhone handily beats almost every Android handset in the graphics department in every benchmark. And with lesser specs. Android is such a horribly clumsy OS.

Behold. I bring to you official Anandtech benchmarks. Nevermind the iPhone 6, look at the 5S go! :lol:

@mister-man You fail to understand what those graphs mean and continue with your ridiculous claims. The screenshots you posted add nothing meaningful to this thread, you didn't even bother to explain why you posted them.

Those are on-screen benchmarks for GFXBench, meaning they only use the device's native resolution. Offscreen, which measures the SoC performance, the A8 isn't even close to Tegra K1 and it sits at par with Snapdragon 805 (an SoC from H2 2014).

The graphics component that you cherry-picked for 3DMark does not include physics, and even then, it's well behind Tegra K1 and only on par with Snapdragon 805.

GFXBench 3.0 Manhattan (Offscreen):

3DMark 1.2 Ice Storm Unlimited (Overall):

Android L will improve the performance even more.

@NVIDIATI: "Android will improve graphics soon.". Heard that every year. Every year the iPhone wins in benchmarks. Same as this year. One graph out of the rest where iPhone won out doesn't lay a dent.

It's a harsh reality but I'm sure you'll come to grips with an "under powered" device running laps around the rest

@mister-man

Stop ignoring evidence that says the exact OPPOSITE of your claims.

I don't even think System Wars would tolerate your troll-like behaviour, I certainly have no patience for it.

@NVIDIATI@mister-man

It's actually graphs like the one below that I think are most important and immediately relevant when examining performance because the immediate context is there.

It's actually graphs like this that convinced me to get an iPhone 6 instead of the iPhone 6+. Not only is the iPhone 6 incredibly well balanced in terms of power and specs, but it also has available to it an immense library of high quality games -- and that's only going to be increasing as time goes on.

These graphs are the ones I feel are most important. In context, the iPhone 6 is especially impressive, and it's that kind of context that actually matters from an end-user standpoint -- not useless raw performance analyses.

EDIT: It may be safe to say that games run better on iOS than they do on android, generally, for two reasons: 1. There are more of them, and 2. 99.9999% work on every phone that supports the latest operating system (which goes back as far as 2011). That's fairly convincing to me.

@musicalmac

Onscreen benchmarks do not reflect the real SoC performance as the devices are running at different resolutions. The iPhone 6 might offer a higher Fps onscreen than the iPhone 6 Plus, but the iPhone 6 Plus is pushing far more pixels. When compared at the same resolution, the iPhone 6 Plus offers slightly higher performance. A device can run an app at a different resolution and scale it to it's native resolution if need be.

This is why mobile benchmarks (Basemark, 3DMark, GFXBench) by default display the offscreen performance, this way all devices utilize the same resolution, regardless of their native resolution. Even desktop benchmarks follow that same standard.

But they do reflect the real SoC performance relative to the device that SoC is powering. That's absolutely the best way to test new devices because it's directly relevant to ones experience with that particular device.

I don't care too much about arbitrary benchmarks, I care about balance and how well a device handles itself given it's own unique properties. A benchmark for the sake of a benchmark is borderline meaningless in the greater context.

@musicalmac These aren't arbitrary. When a game is running at a 1920x1080 resolution, it will utilize high-res textures. If a boost in Fps is required, the application doesn't have to run at the device's native resolution. This is something that's familiar on laptop/desktop computers. Take for example a game like Riptide GP2, in the settings, you can adjust the game's native resolution and/or textures. Decreasing the resolution has a direct impact on the Fps.

Not sure how many FPS and benchmark comparisons its going to take for people to wake up from delusions.

Forget the fact that Android is "competing" in the graphics department. The fact that Android has to "compete" with specs that are half of what is found on Android specs is damn near telling, if not condemning.

Androids clumsy OS really shows its inefficiency in these benchmarks

In terms of looking at iOS, they are absolutely arbitrary. I suppose the same can't be said for android -- which is what makes comparisons difficult. Developing for android is quite a bit more complicated because of the fragmentation woes.

Has anyone seen the battery benchmarks?

iPhones last nearly double the time some of the Android handsets do, and with a battery nearly HALF as small.

The innovation lies in their software and hardware integration. Something no ither company has achieved thus far.

The efficiency.

@mister-man:

Just to clear things up, the Note 3 and the Galaxy S5 smashed iPhone 5S in the graphics department. The Note 4 and the Galaxy S6 have more impressive hardware than the iPhone 6 and will therefore be better. Its funny because I think the S5 still beats the iPhone 6 and this is last years model. Also, you have the Note 3 on there (which is last years model) and it is almost on the same level as "this" year's iPhone 6 model. Apple makes great phones, but from a hardware point of view, they have never been able to go toe-to-toe against Android's flagship devices, more specifically speaking Samsung's Galaxy line up.

So, how does the iPhone compare in a bending test? :P

your actually wrong, the 5S is either always on par or faster then the Snapdragon 801 found in the S5/M8/etc by about 5-9% and it came out in 2013.

And the A8 takes it even further, Note 4 GPU and iPhone 6 GPU are about on par with each other (Adreno 420/GX6450)

@xboxiphoneps3: @xboxiphoneps3:

Sorry, but that isn't true. In gaming capabilities, the S5 outperforms the 5S. The S5 is slightly less in terms of specs than the the new iPhone 6. The Note 3 was not as good in terms of processing, but it beats both the S5 and 5S in 3D mark scores. (Again in terms of video processing I have seen Note 3 and S5 outperform the 5S) I haven't seen benchmarks for the S6 or the Note 4, but in the graphics department they will most likely beat the iPhone 6 lineup. Plus, I don't see an 8 core processor being bested by a dual core processor, but that is just my opinion.

@Mighty-Lu-Bu:

But it is actually true, plenty of benchmarks both onscreen and offscreen tests in a few different benchmarking applications show this.

The Galaxy S5 (Adreno 330 upclocked)and 5S GPU are within 10% of each other, trading blows in certain titles . And since the 5S runs at a lower resolution then the S5, itll always usually run a game smoother then the S5.

The new Apple A8 has a new GX6450 GPU that is roughly neck to neck with the Adreno 420(found in the 805) and this too results show this aswell

The reason why the Apple A8 is superior because its architecture is very wide. It can decode, issue, and retire up to six instructions per clock cycle. For comparison, Swift and Krait (Qualcomm’s current mobile CPU core) can’t do more than three concurrent operations.

They are ARMv8 chips which is the newest ARM architecture, they do six ops per cycle , double of Snapdragon

Its basically the same comparison with Intel and AMD,

a Intel quad core beats out a AMD octo core thats clocked even higher, its because of stronger IPC and single core performance and single thread performance per core

Throwing more cores at a task doesnt magically make it better, it takes very solid multi threading and parrallelization support which isnt even that easy on desktops(alot of applications on desktop still use only 2 cores) the more advanced programs use multiple cores but you cant compare mobile to desktop at all

Mobile is constrained by heat, parallelization etc

(Basically majority of Android apps dont really use 4 cores and arent the best in multi core use efficiency and efficient parralleization in a mobile envelope)

Stronger single core performance and better single thread performance with is the way to go in mobile right now and allows for better efficiency and performance

-Galaxy Alpha just came out with a new 20nm Exynos octo core and it gets only 3 hours of battery usage time, even little less from what reviews are saying, what does that tell you, my iphone 6 gets much more then that and its also a 20nm SoC(latest manufacturing node for mobile right now)

Sorry for the large quote, from an Ice Storm Unlimited 3DMark benchmark the iPhone 5S scored a 13975. My phone (a Note 3) just scored a 16138. Now in terms of of mobile gaming, or gaming in general, sure the CPU plays a huge part, but it is usually more so heavily based on the GPU. The iPhone 5S could very well have an edge in the CPU department, but in terms of gaming the Note 3 wins especially with its 3GB of RAM.

For your Intel comment, the highest end Intel CPU is more powerful than the highest end AMD, however, the Intel has a terrible price to performance ratio. The highest end Intel CPU is maybe what a couple thousand dollars, where as the highest AMD is only about $300.

@Mighty-Lu-Bu: im talking about a i5 beating out the FX series CPU's from AMD, a i5/i7 is no where near thousands of dollars (although extreme i7's can run alot but im not even talking about that here)

Physics performance is a extreme rare task where multi core in mobile will shine, thats just one type of specific task that does have a little advantage on multiple cores but in essentially 99% of smartphone operation there is no physics related tasks , trying to push heavy multi threading usage and crazy parralleization on mobile is defitnetly not the best idea and focusing on single core performance and single thread performance is much more optimal for mobile

As for the Note having 3GB, thats quite solid ill admit, but 3GB of RAM does not make a huge difference and its mostly due to pushing 2k resolutions and running Touchwiz layer ontop of the not most resource efficient OS(That is Android) with dual windows on the screen, RAM isnt everything though

Just to inform everyone in this thread, @mister-manchanged the thread title to "iOS graphics are much better than Android", this was not the original title of this thread, nor was it the original focus point of discussion.

@xboxiphoneps3:

The FX 8350 beat every i5 line up in benchmarks and some i7s if I recall. The i7 4770K was, however, better than it. The newer FX 9590 successfully beat the i7 4770k in bench mark tests and costs around $100 less.

I'm sure there is a chart somewhere that proves this:

http://www.cpubenchmark.net/cpu.php?cpu=AMD+FX-9590+Eight-Core

The super high end "Extreme" i7s are at the top of the totem pole and I have seen them range from $950-$2000. The AMD FX 9590 is AMD's best cpu ever and I've seen it for less than $220. I'm not saying that Intel or Apple is bad, I am saying that there are alternatives that are just as good if not better.

@NVIDIATI: What the hell does it matter what kind of performance my device produces when rendering on a different screen with a higher resolution? Seriously. I play mobile games on my mobile device rendering at my mobile devices resolution.

You're really really grabbing for straws.

1. Your mobile device (1334x750) is producing a lower resolution image. A game will look better on a device running native at a higher resolution (1920x1080 or 2560x1440).

2. If the resolution on a device with a high resolution display (1920x1080 or 2560x1440) is causing performance issues, the performance can be boosted by lowering the resolution to that of your mobile device (1334x750).

This is why we compare hardware performance at the same resolution.

The pixels are already too small to see unless you stick your face so close to the screen that your eyes can't focus anyways. To imagine a game a 1920x1080 looking better on a 4.7" display than a game at 1334x750 is a special kind of ridiculous.

Mobile devices should be evaluated on balance and performance relative to the device, not relative to some standard benchmark that ignores the specific aspects of that device. Certainly you have to understand something so simple. Apple is undoubtedly the leader in this area -- look again at these benchmarks:

Make particularly sure you note the CPU clock, CPU cores, and memory.

If you were to give even an inch when a mile was appropriate, it's likely you'd find more meaning in every conversation in which you choose to engage.

@musicalmac said:

The pixels are already too small to see unless you stick your face so close to the screen that your eyes can't focus anyways. To imagine a game a 1920x1080 looking better on a 4.7" display than a game at 1334x750 is a special kind of ridiculous.

Mobile devices should be evaluated on balance and performance relative to the device, not relative to some standard benchmark that ignores the specific aspects of that device. Certainly you have to understand something so simple. Apple is undoubtedly the leader in this area -- look again at these benchmarks:

Make particularly sure you note the CPU clock, CPU cores, and memory.

If you were to give even an inch when a mile was appropriate, it's likely you'd find more meaning in every conversation in which you choose to engage.

Hence the notion of point number 2, if someone wants higher performance, they can just lower the settings.

This is why you don't bother using images or boasting arguments made by Apple's number one defender, Daniel Eran Dilger (a writer for Apple Insider). He couldn't even be bothered to get the labels on the SoCs correct (Nexus 5 uses Snapdragon 800, Moto X uses a variant of the Snapdragon S4 Pro). There is also the fact that those benchmarks compare Apple's 20 nm A8 to Qualcomm's 28 nm SoCs from H1 2013, H2 2013, and H1 2014. Qualcomm's H2 2014 SoC is the 28 nm Snapdragon 805 (Galaxy S5 LTE-A, Galaxy Note 4), and Google's H2 Android is version 5.0 (that uses a new runtime).

Aspects such as clock speed and core numbers are not directly relevant to the comparison (ie. you're comparing 4 mid cores to 2 large cores).

Take for example, the Xperia Z3 Compact (Snapdragon 801) with 4 cores @2.5 GHz. In Phone Arena's custom web-script battery test, it scores 10 hours 2min and has a GeekBench 3 score of 2856.

The iPhone 6 with 2 cores @1.4GHz scores 5 hours 22 min and has a GeekBench 3 score of 2891.

@NVIDIATI: Poor deflection. Just give a reasonable inch and restore for yourself an iota of conversational validity.

@NVIDIATI:

PhoneArenas testing is rubbish and they dont even disclose their testing methods, they also set the brightness at a % and not at same exact nits

I get noticeably longer wifi browsing time and usage on my 6 then 5 hours and 22 minutes. Something is defitnetly off there.

And also you forgot that Geekbench hasnt been recompiled for the new A8 and also to the new LLVM compiler in iOS 8, LLVM update in iOS 7 gave the 5S a 5-7% boost in Geekbench 3 scores after it was updated , boosting it on average of a single core score of 1325 to 1400. Once Geekbench 3 is updated for the new LLVM, iPhone 6 scores will go up even more in GB3

@xboxiphoneps3

The Z3 Compact's long battery life is no secret, look around at various reviews.

The Xperia Z3 Compact comes with Android 4.4.4 using Dalvik. Android 4.4 allows you to switch compiler to an ART pre-release. On the Nexus 5 (4.4.4) the Dalvik score is listed at 2451, when switched to the ART pre-release, it scores 2745. That's a 10%~11% increase on a pre-release compiler. So the Xperia Z3 Compact's score should go up just by switching compiler in 4.4.4.

The performance will only continue to go up when Android L with the official ART release comes in a month from now.

As well, Android L comes with Project Volta, which also improves battery life further.

@NVIDIATI:

yes ios graphics are mre better than android, because of the metal inside it.

@NVIDIATI said:

@xboxiphoneps3

The Z3 Compact's long battery life is no secret, look around at various reviews.

The Xperia Z3 Compact comes with Android 4.4.4 using Dalvik. Android 4.4 allows you to switch compiler to an ART pre-release. On the Nexus 5 (4.4.4) the Dalvik score is listed at 2451, when switched to the ART pre-release, it scores 2745. That's a 10%~11% increase on a pre-release compiler. So the Xperia Z3 Compact's score should go up just by switching compiler in 4.4.4.

The performance will only continue to go up when Android L with the official ART release comes in a month from now.

As well, Android L comes with Project Volta, which also improves battery life further.

Project Volta isn't going to help the Z3 compact get better battery life when it has its own power saving mode implemented, but Android L optimizations will help, plus those tests are pretty weak as they use the Stamina Mode in the devices, cutting out push email/any notifications and internet access.

I could put my iPhone 6 into grayscale mode and airplane mode and get just as good video playback and hours

also your showing one battery benchmark out of the other many that refute GSMarenas claims

http://www.cnet.com/products/sony-xperia-z3-compact/2/

CNET says here battery life is nothing too special, just a little better then average but nothing revolutionary, also other reviews from other websites say this too.

http://www.androidpit.com/sony-xperia-z3-compact-review

This review also says that its just a bit above average, the battery life.

@NVIDIATI:

How can you compare the Z3 compact to the IPhone 6? the Z3 is a $600 AUD device and the cheapest IPhone 6 is $1000 (that's outright just for comparison sake).

The Z3 Compact runs rings around the 6 whilst also being $400 AUD cheaper.

I love how non developers are comparing and talking like they know everything. Fact is iOS is supported by a number of device thus making app releases for it easier. You don't have to worry about different screens, memory allocation and a lot more that you probably have never heard of. Android can achieve what iOS can but developers won't push that far because they aim for money. (You can't get that much money of android by targeting only flagships) :) Everything is shifting towards android and you should really stop ignoring facts. I mean 90% probably don't know a single programming language, not to talk about game development. Listen to the numbers and NVIDIATI. He made some pretty solid arguments. You can't compare 1334x750 with 1920x1080. And the argument that you can't see the difference... really? Just really? Of course you won't see the difference in angry birds... Try something that pushes your device to the limits and compare it with it's android version on 1080p. The difference is huge. Even if you like iphones and apple in general you shouldn't ignore facts and use 3-4 year old arguments. Gratz for having an iphone enjoy using it but stay away from what you don't understand.

@NVIDIATI: I'm not ignoring evidence. I'm ignoring one questionable benchmark in favor of recognizing the insurmountable piling evidence in the form of twenty other benchmarks showing that you're the one in denial.

Android fans will never live this down. To be debating over the likelihood of an "incremental difference" in performance with a device that's equipped with half the spec numbers is so embarrassing.

To say "Don't worry. The next Octo-core 4GB ram with 2.4 ghz clockspeed will DEFINITELY outdo your dual-core 1GB ram iPhone" is embarassing all on its own, and is being viewed as an inherent fail.

every year it's the same thing.

However most android phones have the removable battery option, this means you can buy a battery with a bigger capacity. My Note 2 was packing a 6500 mAh battery which gave it unreal battery life. With that battery in it could game for at least twice as long as any Iphone could without charge

One thing I have noticed regarding games on my iPhone 6 is that games that support Metal are noticeably less stressful on my phone than games that haven't updated to support Metal yet. For example, Dungeon Hunter 4 supports Metal and the graphics are good, but it hardly has any more impact on my battery than browsing the internet with Safari. The same is true of Asphalt 8.

Pretty cool stuff. Looking forward to the next round of updates. Really hoping Capcom is either updating SF4 Volt or making a new SF4 game for iOS. It's still my go-to for quick games.

The iphone 6 has that screen that just wraps around you. lol.

How's the iphone 6 like the seahawks Defense? It bends but doesn't break.

What did the droid say to the iphone? "I like your curves"

Person 1: "I can't find my iPhone." Person 2: "It's just around the bend."

Apple has a new slogan: "IPhone 6. It's bending the rules".

I put this in the other thread, but it's worthy of being here, too. Just imagine how things improve when Metal is part of the picture. Good gravy.

Apple rules the gaming roost in the post-PC era. Look how far ahead the 6 is...

There's more to Project Volta than just a power saving mode (which was NOT utilized in the Nexus 5 numbers shown in my earlier post).

According to CNET, the iPhone 6 can't even last the entire day, you ignored to put that information in the context of their own testing standards.

It's enough to hang in for a chunk of the day, but it's not the all-day-plus battery life I wanted on a new iPhone.

Credit CNET iPhone 6 review

Meanwhile, the CNET review complains that the Xperia Z3 Compact can't last far into the second day.

With moderately heavy use -- a spot of Web browsing, sending and receiving emails, playing some games and and taking some photos -- I found the phone would make it most of the day, but you won't get far into the second if you're demanding of it.

Credit CNET Xperia Z3 Compact review

The Androidpit review didn't say "just a bit above average"...

the Z3 Compact's battery proved to be above average

Credit Androidpit Xperia Z3 Compact review

Battery life in most of these reviews are considered WITHOUT stamina mode or ultra stamina mode active.

Even without activating the Battery Stamina mode — which minimizes data transfers to improve endurance even further — I found the Xperia Z3 and Z3 Compact to be the longest-lasting smartphones I’ve tested.

Credit The Verge Xperia X3 Compact review

Android's AEP offers OpenGL 4.x extensions, GPU's such as the Adreno 420 inside Snapdragon 805 can utilize those extensions and offer features such as tessellation (not possible on a PowerVR GPU). NVIDIA's Tegra K1 goes a step further offering native OpenGL 4.x support which has allowed for 10+ PC games to be fully ported with OpenGL 4.4 running on Android.

GFXBench 3.0 Manhattan Offscreen (1080p):

• NVIDIA Shield Tablet (Tegra K1 32-bit) - 1920 frames (31.0 fps)
• Samsung Galaxy S5 LTE-A (Snapdragon 805) - 1197 frames (19.3 fps)
• Apple iPhone 6 Plus (A8) - 1166 frames (18.8 fps)
• Apple iPhone 6 (A8) - 1105 frames (17.8 fps)

3DMark Ice Storm Unlimited:

• NVIDIA Shield Tablet (Tegra K1 32-bit) - 30794
• Samsung Galaxy S5 LTE-A (Snapdragon 805) - 21250
• Apple iPhone 6 Plus (A8) - 17839
• Apple iPhone 6 (A8) - 17302

@NVIDIATI: You don't have to keep reminding us how much you like arbitrary, borderline-meaningless GPU benchmarks. If we didn't care about context we'd love them as much as you.

The meaningful tests are the ones that look at performance in the context of the device doing the work. But that's something you're struggling mightily to grasp. Mobile computing isn't about raw benchmarks, it's about balance. Not even a phonearena reference was enough to pull some reason out of you.

How many devices and games are available that take advantage of all those extra graphical goodies you reference time, and time, and time again? Please answer that question.

Edit: Removed obvious question.

@musicalmac You just quoted "onscreen" benchmarks, which doesn't hold relevance when comparing two mobile devices, as you're asking the devices to do two DIFFERENT tasks. I shouldn't have to keep repeating myself, if an application cannot run at the native resolution of the device, the resolution or high resolution texture settings can be LOWERED to MATCH the settings used on the iPhone.

These are two different numbers by a factor of ~2x:

• 1334*750 = 1,000,500 pixels
• 1920*1080 = 2,073,600 pixels

Just about any device with a Qualcomm SoC (and some others) since the Nexus 4 can utilize OpenGL ES 3 and some features of AEP. Devices with Adreno 420 or Mali-T760 can utilize DX11 equivalent features that can be used in AEP (which is planned to be released with Android L). Devices with Tegra K1 already run (desktop) OpenGL 4.x. The numbers you quoted are mainly comparing Apple's A8 with devices using Qualcomm's Snapdragon 801, which launched at the start of 2014. Qualcomm's current flagship SoC (launched this past summer in the Galaxy S5 LTE-A) is the Snapdragon 805. I can't provide a number, but just about every major game has been using OpenGL ES 3 since summer 2013. Titles for Tegra K1 devices use (desktop) OpenGL 4.x. When AEP launches with Android L later this month, then games on non-Tegra K1 devices will start to take advantage of those OpenGL 4.x extensions.

First of all, way to handpick benchmark results, the Apple A8 is on par/ faster then the Snapdragon 805, and this isn't even without Apple Metal being incorporated yet, why not link battery test reviews and great compliments from other reviewers saying the 6 has great battery life? My 6 easily last me all day with moderate to heavy usage, Apple Metal and apps being recompiled to iOS 8 and the Apple A8 will increase battery performance. Anandtech and others have gotten some pretty dam solid numbers from the 6, and even shows in Anandtechs testing the 6 Plus and 6 beating other flagships in wifi longevity

Also, the Plus and 6 have the least performance degradation over time, next to the SHIELD tablet(that is incredible that a mobile device like a smartphone in a smartphone envelope is able to not throttle back, and that it competes with tablets in a tablet envelope)

That is some serious performance degradation for the Android devices, it's all because of Imagination's new GPU architecture , where they are able to go for much longer without GPU throttling, it's some amazing stuff