Certainly left the other devices smoking.... Thing is, the iPhone 5 -excels at what the PS VITA should have been good at, but fails at what a smartphone should be capable of, thanks to the unforgivably horrible Maps app.
It's certainly impressive to see what they've done with the CPU, GPU and camera... but the Maps app alone makes it a no buy.
Ditto. It works very smoothly for me, it's definitely replaced my old standalone GPS (which can take several minutes to lock, and a minute to calculate/recalculate the route!). I've also enjoyed being able to tell Siri to get me directions to <name of a place> and have it just work.
The only thing I've found any issue with in iOS 6 is the new App Store. I definitely like the new functionality/interface, except that there's lots of areas it's not smooth (i.e. often if I click to the updates tab it just sits for a few seconds before it switches tabs and starts loading, instead of instantly switching and showing it's loading). Nothing serious, but it definitely could have used some more testing (I don't believe it was included in the betas, was it?).
Because, unlike the Android fanboys posting here, they've actually used it! And the tumblr post doesn't take into account how many issues there are unlike a real world test.
The new maps app is atrocious. I can't believe you are even trying to defend it. I don't know where you live but where I am, in southern CA, is not exactly remote and the level of detail for the local maps is barren. You would be able to navigate to a written address if you had one, but if you were exploring an area you would have no idea what is actually in the places you were looking at. Actually the better question is how you actually use the maps app. If you are finding driving directions to addresses with no regard to anything else then yes, the app is serviceable. If you need the app to provide you with details for what is actually in the areas you are viewing, it is completely useless.
Comments like this remind me of the whole antennagate non-issue. Maps working fine for me so far in London. As a backup I have Nokia and GMaps as one tap web apps if i need them so its not a question of having a useless phone as the above comment and a few others imply. This whole maps argument is part of a juvenile religious war IMO, an attempt to blow up a minor problem to score points.
Each mobile OS has strengths and weaknesses. Users should use the same criteria as when buying toasters. Which better suits your needs and fits with your taste and requirements.
since when was antennagate a non-issue? The iphone 4 and 4S have the worst reception of any modern smartphone. That's not a non-issue, it's a fact. The 5 has apparently fixed the issue though.
You know they fixed the problem on the 4s right? And if i remember correctly, anandtech actually tested both the 4 and 4s as having some of the best reception of the smartphones available at the time.. So you might wanna look over your "facts" again..
That is utter nonsense! The iPhone 4S had extremely strong reception compared to the iPhone 4 which, granted, did have some slight issues upon launch. They redesigned the antennae on the 4S and the 4 for Verizon which meant it did not suffer from the recption issues.
The 4S reception is leagues better than the Samsung Galaxy S3, which in the same location on the same network for me would actually drop virtually every call. The 5 is once again much better than any previous iPhone for voice call quality and reception, as all the reviews have corroborated and I have experienced. Stop spreading incorrect information which you know nothing about.
You should probably read anandtech's analysis on it. Certainly not the "worst" of any smartphone, and certainly not a fact. Their tests showed both pros and cons, and the con was at most a 29 dB drop in signal.
Not to mention the antenna was altered for the Verizon 4 and 4s's in general.
I can't speak for the 4, but I can tell you from firsthand experience that the 4S has spectacular reception. Around town there's not much different, but my 4S grabs signals quicker and hangs onto them longer in mountainous terrain than any of my friends' Android phones.
So what? Eye candy needs work. Vector based mapping and nav routing seems solid. As does the voice activated turn by turn. Its not like the maps app is routing you into oblivion. Nit picky bs. No public trans maps is a stupid miss though and I dont even use PT.
Yes, not everyone has the same experience. For the large majority of people, life is fine but for some (mainly outside the US), you will have problems. This is their first time with public maps. Its their beta.
My direct experience of the maps app on iOS6 has been fine, I think people are just looking for something to bash Apple over the head with now the benchmarks show it leaving the S3 and the others even more in the dust, and it has somehow become fashionable to bash Apple. Every year, with every successful product launch, they find something minor to complain about, it seems.
The thing is also that the graphics prowess of the iPhones are actually used by Developers far more, vs those found in Android phones. They are mainly used to try and make the interface smooth, but try and find actual games which will use the power contained in the handset and what do you get? Angry Birds Space. Yes it's smooth, but there seems to be very little dedication on the Android front to making Apps which run decently and take advantage of the full power of the graphics in higher end handsets like the S3 a la Infinity Blade, and all the other games which constantly push what can be achieved on a mobile device.
Without them, you are simply left with benchmarks and bragging rights, and what's the point of that? Yes I know that you can do things like moving video previews and floating video windows but everybody knows that it is games which push graphics. I felt pretty demoralised after getting an S3 and then realising that there was really nothing decent gaming wise available.
I don't really see what Google can do to fix this fragmentation other than adopt a platform mentality kind of like it does with the Nexus line, but really, they should just have power profiles which a device can hit which games target, however this doesn't appear to be working very well at the moment, and overall for performance reasons, I'm much happier with my new iPhone 5.
Better watch your data usage in Apple3D maps. Especially with LTE speed and Apple's new A6 speedy CPU. Can readily guzzle ~ 30MBytes/minute of urban 3D viewing. Owners of the iPhone5 on a LTE network and with capped data plans will now burn through their data allotment on apparently trivial functions, where otherwise the data usage would have been throttled by either the network bandwidth or the slow processing speed. Expect a huge chorus of iPhone5 owners mad as hell about data-caps and overage costs within the next month.
Just a follow-up. Seems to me that the iPhone5 would benefit from a built-in continuous tracker of data usage with its output user-selected to be optionally visible in a corner of the "current" screen.
Android not only lets you set a warning and force a limit, it also breaks down data usage by app. I'm sure they'll build that into iOS 7... You can find a lot of similarly minor but very useful features that either OS has/lacks, stuff that mostly doesn't get much press but really completes the user experience.
People can only give opinions on the functionality of the maps-app based by the area where they use it.
Considering how appalled international users have been in some cases, I'd say that the new maps is not really "fine", but instead needs tremendous amount of work to catch up with Google.
Not that it necessarily is such a bad thing. Google maps is not perfect either for people who occasionally stroll off the beaten path, so competition should benefit us all in the long run.
Same here. I've had nothing but very positive experience with Apple's maps so far. While I don't doubt some issues exist in some places, I'm quite certain that it's nowhere near as bad as what's being reported by the tech press.
Show us one example where maps actually screwed up for you that wasn't posted on the net. I agree some of the stuff posted the last few days is laughable, but I've yet to find something not working in my area and I would imagine that holds true for at least 90% of us. It's ridiculous to call it unforgivable, give me a break.
There are millions of iphone users that couldn't give two shits about apple vs everyone else. People buy them simply because they are the standard every other manufacturer tries to imitate. They work extremely well minus a few overblown issues that are bound to crop up in any product given so much scrutiny.
Do you live in the US tnicks? Because if you live outside of the states and don't live in a major city then yes, it is pretty poor from all of the reports I've had (from friends, not just random net users). The town where I live in the UK, the population is around 100,000 people, and it's just a giant green patch on maps. I'm not an iphone user, but this is my friend showing me on his updated 4S at work on Friday, it's a bit of a joke. The 5 is clearly a very good phone (benchmarks alone show what a good job Apple have done with the A6), but this had been a PR disaster.
And like you said, there are millions of iphone users. Release a new bit of software that doesn't fully work as people intended and of course it's going to blow up!
"People buy them simply because they are the standard every other manufacturer tries to imitate."
No, they buy them because they simply aren't paying attention and/or are locked into the Apple ecosystem. Apple's competitors have long since passed them up in terms of both OS and hardware design.
Keep telling yourself that! I had an S3 back in June and sold it after a month and went back to Apple, because I found life is still much worse on the Android side, apart from the OS customisability aspect and widgets, despite how much things have improved since my last Android phone. Those last two aspects are not as important if a competing phone does everything you want without customisation though.
-There is no iTunes equivalent, you have to drag and drop everything, which is simply not practical for most people with complex playlists. You might think that you won't mis iTunes, but IME, there is nothing which does so much an offers the 'one-stop' convenience.
-Simple stuff like music playback is mind bogglingly complex to use when the phone is locked becasue you can't put a music playback control on the lock screen using the stock rom. A lot of the software both on the S3 and the supporting software like Kies feels under developed and unreliable.
-Tthe Apps available remain a poor shadow of the Apple ecosystem. Even the apps which do exist on both platforms are usually feature incomplete on Android. The only apps which are leagues better are the Google ones, which I miss slightly.
-The hardware and design of the iPhones blows away anything Samsung or anybody else for that matter have come up with to date. It is refined and feels like quality, not the cheap, thin plastic of the Galaxy, which is supposed to be a flagship device. The screen, whilst bigger on the G3 has a cold tint to the colours on the display and it not retina quality (it looks fuzzy round the sides of text)
-If you want to talk about lock in, there are features on the Galaxy 3 which will only work with other people who own the Galaxy 3 like the photoshare thing, which is hardly useful, and a great deal of what Samsung added to Android like 'S Assist' and 'Smart-stay' just don't work very well.
I'm a power user and know exactly what i'm buying and the reasons for this. Stop patronising people who have weighed up all the pros and cons of a device's hardware and software integration, plus their own needs and opted for a choice other than yours. I would happily buy whatever device I find to be the best, the cost is only one consideration, and for now that device is the iPhone 5.
lol, you listed iTunes as a positive point? NO ONE I come across in real life defends that POS, not even the most die hard Apple lovers. Apparently, my magical experience did not match up to yours, every time I sync with my old iPhone, something goes wrong, in one of the worst case scenario, my prayer didn't go right with Lord Steve J, and all my photos were gone!!!
It's got to be the worst piece of shit in the history of all garbage software. I ran iTunes in windows 7 and 10.7 OSX before. Biggest reason why I'm left iPhone.
Someone also mentioned iOS being smooth, sigh... my MOTO Razor flip phone was smoother than any iPhone. But that didn't make my stick to it, I wonder why?
Sure, software if you're willing to buy every year, since it seems most androids get left out in the cold when it comes to OS upgrades. And if you're ok with a few glaring bugs. My friend's Android often won't show the full keypad after she locks it, which means she can't unlock the phone without rebooting. Not saying Apple software is bug free, but that's pretty a fundamental bug.
Hardware design, how? NFC? That's currently a gimmick/non-standard. I have never seen anyone use their android for that; heck, people barely ever use the wireless feature on their CCs. More RAM? Well the tests historically have shown that hasn't seemed to matter, I suspect partly because Android needs the Dalvik virtual machine to be running full time.
There is more to a smartphone than a map app. There are many map apps for the iPhone. The built in one is guaranteed to get better with time particularly with the help of third parties who specialize in maps.
Maps works well for me. Routes and locations are fine.
I've had some major screw ups with Google maps over the last few years. I think a lot of this is overblown. You can also download other map apps, such as MapQuest, which gives you turn by turn navigation.
I do believe Apple's data is a little behind Google's but that will get resolved in a matter of months. I don't think it is even close to a major issue at this point. Might be a little worse initially when it starts selling in more countries since I assume international data will be weaker.
That being said, it is based on OpenStreetMaps with an open data set anyone can contribute to, and a bunch of companies are supporting it over the proprietary Google maps data. I wouldnt be surprised at all to see it far exceed Google Maps's data set within this calendar year itself, especially after Apple starts adding to it based on the data they get from iPhone users.
Personally, i haven't had any issues at all. Initially I was annoyed by having to download a new app for transit directions, but after a couple of days I love it, because the app gives me advisories and offline maps, which the older Maps could t do. And the usage of vector base maps, as opposed to bitmaps has made the map browsing experience much better.
Lame excuse. You do know that there are plenty of alternative Navigation apps available, right? Some are free like MapQuest and Waze that are pretty damn good. My only issue is I wish they had made the screen a little bigger and wider.
That's like not buying a Windows computer because it comes with Internet Explorer! Some haters just need something to rage at, because it's incomprehensible that Apple maybe has the superior product.
Apple Maps has worked fine for me as well. I used it for both searching and also for turn-by-turn directiions. It is also much faster than Google maps in my 4S. I am pretty satisfied with it. I am sure that it can be improved and some mistakes will appear and be corrected. It is really a non-issue and certainly not a deal breaker, as MILLIONS of people have already proved by voting with their money.
In my area in France I don't encounter problems using the new Maps. I live in a small town and the satellite view is actually better than Google Maps. The precision is better. Also turns by turns direction is works well. But TomTom turns by turns work well here also (could be related).
As for the performances of the A6, what I would like to see is a more real world application performance. Geekbench use C language for its tests (Android NDK) but most applications are written in Java. So there a huge performance penalty here.
What would be interesting to see is a algorithm written both in Objective C/C and in Java running on say a S3 and the iPhone 5. From my point of view a quad core processor brings not that much in real world tests (because most applications rely on one or two threads not four). And then there is the penalty of converting pseudo code to machine code with the Java Dalvik VM. How much does it cost today on modern phone. It is always a 5-10x penalty like on Java on the Desktop? Note that I really like Java as a language but still it cannot compare to native code and on a small device it could be impacting. As a developer I see huge differences in things like parsing an XML file (a common thing in mobile applications).
Using for development purpose both a Nexus 7 (not the most sluggish Android device) and iPad/iPhone I can quite tell you that iOS fly in comparison. Everything is smoother. And more that that everything seems to closely follow the fingers of the user. It's a really important point in the user experience. You don't have the feeling that is a computer but more a real objet dans is represent in the user interface. It follows the physics of finger movements.
I found the UI of Chrome better (the number of tabs is not limited) but scrolling in a page is slower than in Safari. Safari makes also a more profound usage of GPU acceleration (and iOS use GPU acceleration everywhere from the beginning).
The main thing has see as an advantage on Android is the ability to share more easily file between applications. You can also do it on iOS but is more limited (It should improves over times I hope with iCould).
Wow, he provided legit, unbiased info about why Android handsets are running slower and he's full of sh**. So now anybody who talks against Android is full of sh**? Is it a religion now?
He provides his own view of the two environments. I myself have found that unless I start up a ton of background tasks, my SGS2 and my Chinese tablet are smoother and load apps faster than my friends iPhone4s. And because of his lack of widgets, he needs more inputs to get basic information than I do.
I call BS, I have a Nexus 7 running Jelly Bean and it is in NO WAY smoother than iOS even going back to version 1. It's an alright device, I guess, build for a price point and you get a lot for your money, but don't kid yourself that it competes with any iOS device such as the ipad 3 or iPhones in this respect.
Wait until Google releases an update to Chrome. Once they compile it with the iOS 6 SDK, scrolling will get a bit better. Apple enabled the background rendering in UIWebView which will make third party web apps a bit smoother to scroll (not as much as Safari, but still a big improvement).
Well, you make some...interesting...claims. First, if you think Dalvik==JVM (which I assume is your point since you are making sweeping, and inaccurate, claims about Java and seem to be confusing the language with the runtime) you are mistaken. The Sun/Oracle JVM (Hotspot) is extremely fast, and very competitive for many algorithms, excepting those with very short run times (http://shootout.alioth.debian.org/u64q/which-progr... Dalvik isn't as optimized as Hotspot (but it has already improved greatly with their new GC in 3.0). So, you can find benchmarks with 5-10x differences but to say that is the norm is incorrect (though, I suppose, it depends on what you are working on from day to day, but hopefully you understand my point). Second, Geekbench uses native code b/c in order to measure low level performance (like the mem speeds), but since it isn't open source, we don't really know how they are doing things (seriously, people, why are we relying on non-open source benchmarks?). Regardless of GB, though, I would imagine most, if not all, of the Google apps on android are native. Third, can you post a video showing the difference in latency between nexus 7 and ios device? I ask because I've used both (though own neither) and each time I am surprised with both. One, I keep thinking iphone/ipad is more responsive than it is (I use vertical text scrolling in order to get an idea of input lag), and two, I think Android is less responsive than it is (this, at least, is reasonable since I own a Nexus S which simply isn't fast enough for the software, but the Nexus 7 doesn't seem to have any of the same problems). Frankly I've looked for a video of this type (Tom's Hardware used to use a 1000 frames/sec camera to show input lag but they didn't use it with the Nexus 7) but have found nothing. Lastly, can you please tell me what you think ios "accelerates" that android doesn't (at least since 3.0)? The browser, in particular, is something that has been mentioned in Android developer blogs as something where they "recently" changed the rendering model (though, still, it has been "accelerated" for years) to a tile based, on demand approach (like Safari, they mention).
Fine. One example is Chrome on my Nexus 7 running JB, scrolling is not at 60fps. On iOS it is. Having said that, on iOS6 the App store has gotten slightly laggy to scroll down slowly, though if you scroll and let go, it goes back to being smooth, which is a little odd.
I'm not sure why you are giving me this example. I never said that android never lags. I was asking the above poster what parts of the drawing pipeline he thinks ios accelerates that android doesn't. Since you mentioned this, however, let me ask: does chrome always lag when scrolling? Does it lag when the page is being assembled and drawn, or afterwards as well? Does it depend on page content (do text pages lag as much as media heavy pages)? Does it lag mostly when memory is low or does that not matter? That is, try restarting the tablet, then open only chrome, and then only open a single tab to a relatively simple page. If it lags regardless then chrome has an issue. If that is the case, try downloading firefox (i stay on the nightlies since they've been so stable), or opera. If those lag regardless of circumstance as well, then either android is doing something wrong or you have problem with your tablet (either mechanical or software issues due to mechanical issues).
Apple created OpenCL (Open Computing Language) that is specifically designed to shift general purpose work away from the CPU and onto the GPU in order to take advantage of its power. Using C/Objective-C, you can get the GPU to do tasks that were normally done by the CPU, such as calculating scroll rates in user windows based on touch inputs. GPUs are not solely used for 3D gaming. Apple uses OpenCL to get the GPUs to accelerate every aspect of the UI, hence a more than 2 year old iPhone 4 running iOS 6 is smoother than any equivalent Android device of the same era (that I know of and have seen testing of).
Since Apple is the only manufacturer that designs both the hardware (right down to the SoC) and writes its own OS completely, it designs both in tandem. That's a major advantage over all the competition when it comes to performance and this article merely demonstrates that.
And as for maps, I've been using it for a few days now in Sydney, Australia and the only issue I have seen is that some of the restaurant data it is getting from Yelp is not up to date. I plan on using Yelp to update my favourite restaurants now on the map. Otherwise, I really like both the look and the smoothness of operation. Also, it has a big feature that the previous maps lacked, turn by turn (although we in Aus don't get this turned on till next month). So even if it has bugs, its a feature that wasn't present previously.
Wait, spec whores on the Android side are spec whores till... now? Now it's all about the apps! Something iOS has in spades, unquestionably the champ. I use a terrible phone (Android HTC Tbolt) and am replacing it with the iPhone5. I'd rather have a solid QA'd, engineered phone that's fast and has great apps than whatever it is Android is trying to sell itself on (specs? Maps? hit and miss phones?).
LG G2X good bye. I tried to like the Android echosystem, but fragmentation and support is just pathetic. It, like the Tbold, was a *freaking Flagship device* 1 year ago, and have been essentially 'off support' for their eintire product life.
The HTC "ONE" and Galaxy SIII don't even use the same processors/gpu/memory from carrier to carrier from what I can tell.
How the hell can you even tell what you're buying? How much research should you have to do to buy a flagship product and expect it to work for the duration of your contract?
I bought the G2X on the basis of a ton of research for a consumer purchase, including very favorable reviews on this site - and I believe Anandtech is hands-down the best technical review site on the planet.
But there are too many variable that go into a phone purchase that go beyond the specs and hardware. I have learned the hard way that manufacturer support, carrier support, and the carrier-manufacturer partnership are AS IMPORTANT AS THE HARDWARE in picking a phone that will have value for the term of the contract.
And if it's an Android, very close behing the manufacturer of the phone itself, you have to evaluate the relationship of the component manufacturers to the open source community. NVIDIA - fuck you.
Because if you think 'community support' somehow makes the carrier-manufacturer relationship irrevelant, you beter think again. Open source developers can only hack without drivers, and when it gets hard, they lose interest, sell their phones on craigslist, and develop for other platforms.
Couple that with the frequent releace cycle of Andrioid, and it's really screwed. Because the community gets ICS support 60% complete for the G2X, well here comes Jelly Bean. And the 3 guys that still had interest? They say 'screw it' and go get a Nexus.
We need review sites to do a better job grading manufactures, carriers, and component suppliers with respect to smart phones... because those things matter as much as the hardware.
And from what I see, Apple benchmarks so far ahead of the rest of the industry on those measures, it isn't even funny.
That Optimus G hardware looks compelling? Run screaming from the building. Because you will never, ever be supported on that phone.
TL;DR: Andriod is a fragmented mess. The review community must look beyond hardware and software. Fuck NVIDIA.
Wow, it's actually rare that I find someone who experienced the same problems with Android that I did with the Galaxy 3. I think they would seem really good if you've only ever used Android, but compared to the very mature iOS with its App library and the more polished iTunes, I would take an iPhone any day.
LG are known to be pretty bad for support generally, but then you only usually find this out after you buy it, and by then a new version of Android has come out that can't easily upgraded to, your phone is no longer the top of the benchmarks. Android phones seem to depreciate like crazy compared to iPhones too, which I found rather alarming. I think the support which you are effectively paying for through the higher cost of the iPhone plus the fully QA'd software like iTunes and iOS is worth paying for, especially if you are tied to a 2 year contract. Hope you are happy with your choice.
The simple answer is buy whatever the current Nexus is OFF CONTRACT (the Google model, not the verizon model). The more thorough answer is to research popular models and buy those off contract (in case you aren't getting the message, buy your phones off contract--- it costs a bit more upfront, but less in the long run and you get much more flexibility but you really need to stick with GSM). Then you can either rely on kies (which I"ve never used but have heard is pretty good), or run the latest AOKP/AOSP ROM of your choice. Frankly, the problem is the expectations of smartphones. Although Apple's model insulates people from this (on all sides, not just the mobile area), essentially, the problem is that which Window's faces. Smartphones ARE COMPUTERS. Just as Windows runs on radically different hardware so does Android. This means that the carriers need to get out of the way of the upgrade process (why are they involved so much anways? I thought the telephony subsystems on these SoC ran their own RTOS?). I think this will happen in time b/c you can currently upgrade your phone outside the carrier approval process. Android, being linux based, should be able to run on any of these SoC as long as the drivers are provdied (which the OEM's can handle).
Yea but I mean I'm buying a 200$ smartPHONE for only the maps, maybe ill get two and use the other one as a phone, music player, video player, web browser, video caller, wireless hotspot, address book, document editor, gaming device, social networking device, quick camera fix, video recorder, voice recorder, personal reminder, alarm clock, voice assistant, send SMS, send email etc... But the maps suck so all that other stuff doesn't matter. Logic at its best.
I'm starting to wonder how custom it is, since we now know its max clock is 1.2GHz vs 800MHz in the 4S, that's 1.5x the speed right there, fix the memory controllers (which were a sore point in A9) and you might have 2x without that much customizing.
Of course, as we've all learned by now browser benchmark performance is largely a measure of how well you cheat (see: IE9 SunSpider scores), so I take any comparison to Android handsets with a grain of salt, especially in light of this: http://arstechnica.com/apple/2012/09/developers-cl...
But those are still some darn impressive numbers compared to the outgoing 4S. I'm looking forward to TI's and Samsung's Cortex-A15 CPUs even more now, since at the end of the day the iPhone "5" still has too tiny a screen, no simultaneous voice and LTE, and runs iOS.
I wouldn't expect the Cortex A15s to be radically better than krait in pure computation, though they have other advantages (the biggest, IMHO, are the virt extensions).
@AT: Thanks for including other benchmarks than just sunspider! I'd still like to see the Kraken benchmarks and Firefox browser included, but this is certainly better.
"Hmmm, yeah the camera, the CPU, the GPU, the build quality, the look, the screen but... on the maps app there is cloud cover in Scotland, so yup its trash!"
"Specs are everything! - Well.... you can cheat these benchmarks....."
"Yeah but can you flash the boot loader and change the power on graphic? Ha iOS is a closed crappy OS!"
(For the record, maps has been working like a charm since the first beta.)
I was giving the benefit of the doubt because he said "I'm not aware of ANY benchmarks where the GS3 bests the iPhone 5.", and right there on that that site, the GS3 clearly wins the Geekbench Floating Point benchmark 2315 to 2099 and then the simplest of googling reveals even more tests.
I think part of the confusion is the quad-core vs dual-core GS3
> I think part of the confusion is the quad-core vs dual-core GS3
Maybe Samsung should stop giving different regions drastically different phones, all branded similarly? They seem to have issues trying to support so many different variants (as I've had the personal displeasure of experiencing).
How much research should a consumer be expected to do the buy a "Flagship" product? That HTC and Samsung GSIII products have different freaking displays, processors, memory, battery, radios, etc. from carrier-to-carrier.
Seriously, "part of the confusion is the quad-core vs dual-core GS3"?
How can that even be possible?
If I'm watching a Galaxy S3 commercial and then go buy the dual-core version.... that can happen.
Really? Tho, seriously, tho. Really? It's like the Andrioid device makers are trying to make it impossible for the consumer to be informed.
As a consumer shopping for a *flagship* Android phone, you can research for hours, for days! And when you make your purchse, without the most careful attention to detail, you could have no clue if the product bought has anything to do with the products you researched.
I get what you're saying, but you're either stretching it too far or you downright suck at research. But yeah, with Android it's way easier to get burned by buying a phone that falls out of favor with the manufacturer or even users/developers...
It's not too hard to pick out which are gonna be the most popular models with the most support though. The Tbolt was overhyped and delayed, the G2X was never gonna be such a thing (LG's known for flaky software builds and it came out on the smallest carrier only).
American SGS3/One X have a different processor in order to have built in support for LTE, but performance between Krait/S4 and Tegra 3 is very very comparable so I don't see what the big deal is. S4 is slightly more efficient but that's about it.
You'd certainly never be able to tell both versions apart unless you started benchmarking them. Same displays, same batteries. If you don't wanna do any research and you don't care for having different options then clearly Apple's the perfect option for you.
There arent really any games for either which take full advantage of their graphical capabilities, so you would probably never know the difference unles you were running benchmarks lol
a) It lost on every single other benchmark (and there's loads) b) Anandtech already pointed out that Geekbench can't be used to compare cross platform.
Which version of the S3? The international Quad core with good graphics or the American Dual Core with middling graphics? What do you use the power for, Angry birds? If you think the S3 beats the iPhone 5 in anything in the real world, jog on, you are the apologist here. I have used the new maps and by the way they work fine.
Can we get the date of those benchmarks, and the OS version used?
I ask because my VZW Galaxy Nexus running AOKP Jellybean (stock clocks and voltages) scored way better in Sunspider than the one listed here. Haven't tried the rest of the benchmarks, but I'd imagine all of these are old, taken when the phones were first released.
If the benchmarks are not going to be updated to show the actual current performance of the devices, then at least let people know that.
I'm also noticing a lack of Geekbench scores comparing the iPhone 5 to any other device, in contrast to all of the other benchmarks.
(eg: I see there are included benchmarks for the iPhone 4S running iOS 6.0, which is marginally worse than my own GNex, but much better than the one listed in the chart)
Also, just clocked my GNex to 1.4 GHz and scored 1525 on Octane, basically doubling the 4S's perfs and about 25% higher than the Krait listed.
Since there's no way the GNex is faster that the S4 when clocked 100 MHz lower, my best guess would be that the difference is in the OS version (JB vs ICS).
It's an OTA update that started rolling out today. Much the same as the update for the iPhone 4S to iOS 6.0
Also, minor updates with performance enhancements have been issued over the past year the phone has been available. If these benchmarks were run in December of 2011, when the phone was first released, they would not be applicable to the phone today, just as an iOS 5.0 benchmark would not be applicable to the iPhone 4S today.
Not to mention, said Jelly Bean update has been available to GSM Galaxy Nexus users for months, as well as to more enterprising users of toro[plus].
I agree that dating these benchmarks should be a requirement, and re-running them (at least for flagship devices/major updates, like JB on the Nexus) would be hugely valuable. Especially for benchmarks like Sunspider, where software optimization is just as relevant as the SoC in question.
All that said, I'm pretty jealous of A6 users. Those of us who don't want everything that goes with buying into Apple are stuck waiting for S4 Pro (hopefully soon) or TI/Exynos A15s (god only knows when).
So, given the lack of a direct response to my question, am I to assume a direct response would be something along the lines of, "No, they don't sell it with Jellybean, and while I now understand why only certain benchmarks are updated, I still don't like it"?
In a couple of weeks or a month at worst (the update came out recently) I'm sure VZ will be selling them with the update already on board. Then again, a month or two from now they might be selling the next Nexus. :p
I was surprised by many things: --how good the A6 cores are --how good the PowerVR SGX543MP3 graphics are --how good the Adreno 320 and 4 Krait Cores performed (in Brian's LG Optimus G)
How much of the iphone 5's outstanding performance was due to its improved operating system and middleware versus hardware? For the results to be this uniformly good, didn't at least some of the improvement have to have come from improved software?
Anand Shimpi, thank you very much for the Mobile SoC GPU Comparison chart and the easy to understand explanation regarding how the SGX543MP3 GPU works.
Some questions for you if I could: " Imagination Technologies' PowerVR SGX543 GPU core features four USSE2 pipes. Each pipe has a 4-way vector ALU that can crank out 4 multiply-adds per clock, which works out to be 16 MADs per clock or 32 FLOPS." Please explain to me why 16 MADs per clock (I understand where that comes from) results in 32 Flops per clock? How can 1 MAD per cycle result in 2 flops per cycle? What am I not understanding?
"Imagination lets the customer stick multiple 543 cores together, which scales compute performance linearly." How close to linear is it? Is the performance of three 543 cores = 2.8 times the performance of one 543 core? 2.9 times? Or some other number?
"The A5 featured a two core design, running at approximately 200MHz based on our latest news. The A5X in the 3rd generation iPad featured a four core design, running at the same 200MHz clock speed. The A6 on the other hand features a three core PowerVR SGX 543MP3, running at higher clock speeds to deliver a good balance of die size while still delivering on Apple's 2x GPU performance claim. The raw specs are below:"
Is the clock speed of the 543MP3 = [25.5 Gigahertz/19.2 Gigahertz]*200 Megahertz = 266 Megahertz?
From the Mobile SoC GPU Comparison chart, the 543MP3 has 12 SIMDs (3 graphics cores multiplied by 4 SIMDs per core). Each SIMD has 4 MADs. Therefore the 543MP3 has 48 MADs = (3 graphics cores multiplied by 4 SIMDs per core multiplied by 4 MADs per SIMD). Since each MAD has 2 flops per clock, the total number of flops is 96 per clock for the 543MP3. Multiply 266 Megahertz by 96 flops per clock and we get 25.5 gigaflops. This math computes.
Are the 48 MADs all 64 bit? Which results in two flops per MAD? Or is there some other explanation?
Many thanks in advance for your help and the help of the many commentators on this blog.
I presume that 1 MAD equals 2 FLOPS because a MAD is algorithmically decomposed into a multiply and an add in terms of raw computation. In hardware however the multiply and the addition/accumulation would be performed in a single cycle.
Glad to see you're finally including other JS tests, as Sunspider was really getting long in the tooth. It reminds me of when the A5 scored like 4-5x in graphics over other phones in other simple graphics tests before. But those are kind of irrelevant. We need more complex tests that show a more realistic difference in performance between different devices and chips.
So I'm glad you've included the V8 test, which was specifically made by Google as an alternative to Sunspider and as a much more complex and realistic test of Javascript performance. Google has been saying Sunspider is irrelevant and they won't optimize for it anymore for about a year now. That might be why Chrome tests even lower than the stock browser in some Sunspider tests. Although it's still strange Chrome score lower in V8 as well compared to the stock browser. I think Google is not where they want to be with Chrome for Android yet. Hopefully the next version of Android that will be unveiled this November and the next version of Chrome will fix that.
What's your conclusion on why the A6 CPU is so fast in these tests? Does it really have much higher IPC than Krait? Or is it because memory was a bottleneck for Krait, and Apple fixed that in the new iPhone 5? Should we expect higher performance from Krait as well when paired with better memory?
I'm also looking forward for you to test Samsung's Exynos 5 Dual with Cortex A15 and Mali T604. No idea why Samsung is dragging its feet and not releasing it with a device already.
We now know it's clocked up to 1.2GHz, vs 800MHz in the A5, so that alone accounts for 1.5x change. The IPC doesn't need to be hugely higher to account for the rest, if they fixed the memory controllers from the A9 that might just account for the rest.
Very good gpu, cpu too. I become a android user for navigation and share trading. Apple would of possibly got me if they had the maps earlier. Now I will stay with android for Swype keyboard and widgets and being able to attach files to email easy and the ease of sharing files with other apps. Still can't recommend Apple to anyone on moral stand.
The problem with moral stand is that it's usually filled with personal stance and when you put them all together you'll always end up with a lot of holes.
I like competition I don't like a company as large as Apple trying to squash competition, by claiming silly patents. If Microsoft did this they would get done over (and did)
Rather than copying what already exists... are you talking about copying a dock, icon based os, copied clock, notifications (al la ICS)... or are you talking about copying 'rounded cornered rectangles' kind of copying.
Every company 'copies' each other (what already exists) to an extent. Sometimes 'copied' things are no-brainers... phones have never been circular, have always had buttons on them, and can all make <gasp> calls.
I would have no problem with apple's design patent if it clearly specified a specific aspect ratio or specific dimensions of a phone ie: 3.76" x 5.15" with 15 degree rounded corners etc, but that isn't what has happened here.
There's plenty of rounded rectangles in technology from computer monitors to pc speakers, card readers, digital picture frames, etc. that pre-date the iPhone.
It's hard to find good explainations about 'Trade Dress' and the meaning of Apple's patent. I had a hard time wrapping my mind around it until it was explained to me like this:
Coke has a registered trade dress on a glass bottle with curved sides. When you see it, you know it's a 'coke bottle'. It's an iconic shape. Nobody else can sell a beverage in a similarly shaped bottle in the United States. It doesn't matter if it's diet coke with a white lable, or regular with a red, or no lable at all. When you see that shape, it's a coke bottle. There are 2 liter, 1 liter, 20 oz, 12 oz, 8 oz sizes. The ratios of hight to width are not constant. Some are glass, some are plastic. But's it's all a 'coke bottle' and you know it when you see it.
That same form of recognition is what Apple set out to establish with the Iphone: A single piece of glass, surrounded by a thin bezel, a speaker centered above the screen, a single button beneath the screen. It's an iconic piece of design, and they registered the essential elements as 'trade dress'.
I think the thing that got Samsung in court was the single button and the placement of it in relationship to the other elements of the smartphone design.
When you see a coke bottle on a shelf, you know instantly it's 'Coca-Cola'. There is no doubt, no question. Apple set out to establish that same level of iconic status. They designed that concept, and registered it with the PTO.
What Samsung did with the Galaxy S was the equivalent of Pepsi putting their beverage in a coke bottle with a red label that says 'Pepsi'. You see the product, and it first envokes the brand 'Coke', even tho the lable says 'Pepsi'. The fact that Samsung has rounded corners on a square button or that 'Samsung' is printed between the speaker and screen were trivial - like the difference between glass and plastic coke bottles.
The Galaxy S too closely resembled Apple's design, as did subsequent Samsung products with single buttons centered on glass faces with narrow bezels.
There were other glass bottles before the coke bottle, and there have been many after. "Rectangles with Rounded Corners" is too much of a generalization to describe the iPhone's trade dress. There is more too it, and Samsung hit all the essential elements.
I agree the button is iconic of Apple, but the speaker placement is set by where your ear would be lol. Round corners are not Apple. And pinch to zoom is a fundamental was to use your phone I don't think that should be patterned and if it is should be cheap and available to everyone, like notifications that Apple copied from android. When I look at a Samsung phone I don't think that looks like a iPhone at all. And my local super market Woolworths logo does not look like apples logo.
Dock - not patentable Icon based OS - not patentable 'Copied Clock' - do I even need to answer this one. Notifications - remains to be seen if they infringed Google's patent in implementation their own version. Google hasn't sued them over it yet which gives you part of your answer.
Copying things is legal, when it's legal. It's illegal, when you infringe patents / trade dress.
I don't know why you state phones have never been circular (shape wasn't anything to do with why Samsung lost), nor buttons (again, not why Samsung lost) nor making calls (again, not why Samsung lost). Ironically, the FRAND patents which Samsung is suing Apple over are for both 3G (which they tried and failed to ban the 4 over) and LTE (which they claim they will try to ban the 5 over).
The main items Samsung lost here were on the Utility patents, which has nothing to do with the shame. The decisions they did win on the design patents and trade dress were nothing to do with the shape claims. Making your final paragraph irrelevant.
To quote an article from a clearly more educated individual:
'Samsung contributed greatly to this with a post-trial statement that said: ““It is unfortunate that patent law can be manipulated to give one company a monopoly over rectangles with rounded corners, or technology that is being improved every day by Samsung and other companies.” It’s more unfortunate that the claim was a gross exaggeration that was swallowed whole by many writers.
Apple claimed that Samsung infringed on four design patents. The D’677 patent covers the overall design of the iPhone while D’305 covers the layout of icons. These claims were upheld. But the jury rejected infringement claims based on patent D’889, which covers the iPad, and rejected eight of 13 claims under D’087 and which deals specifically with the rectangles-with-rounded-corners design of the iPhone (see the relevant sections of the jury verdict form below.) '
It's only very simplistic, uninformed people who claim this case was about rounded corners, etc - exactly the type of person I wouldn't expect to find on Anandtech.
Most companies don't sue other company's over minor infringements. This has been a core part of apples strategy And this is the reason I dont own a apple product. Its not that I dont think they are good for people that want a fun phone with good games thats easy to use. And the Ipad is really good. Apple still sucks a55 as a company. And even Woz said this last win over samsung was BS. Your link is bias and there are plenty with the opposite bios. Jobs wanted to go "Thermo Nuclear" on android, they have proven this to be the case. And if this was just about Samsung I wouldnt have a problem, but its been every company thats produced a competing phone. They just dont want to have competition. I like competition hence I dont like a company that tries to remove it.
Note: The original version of this post said the jury had rejected all claims regarding the rounded-corner design. The jury in fact rejected all claims only regarding willful infringement. On the simple question of infringement, the jury rejected a majority of claims, but did accept five regarding the iPhone. The corrected version appears above.
I quoted the amended article, it needs no further qualification.
That your only argument to the article is that it is 'bias', is frankly only representative of your own bias.
The 5 claims which were accepted did not related to rounded corners, as the quote I already pasted proves.
And to claim other companies aren't suing over trivial things - they've gone a step further. Samsung tried to ban the 4S over 3G - with a zero hit rate, and claims it will try to ban the 5 over LTE - both FRAND patents and Samsung is already under investigation for that.
They don't sue over 'trivial' (read: utility, design & trade dress) patents because Apple hasn't infringed any.
Apple copied as much, if not more, from other companies' products as anyone copied from Apple. Quit pretending that Apple's legal shenanigans are anything other than anti-competitive trolling, please.
There is a difference between copying things which are in the public domain, not patented/patentable - and doing them in your own way, to outlining a 132 page document on how you plan to replicate technology which has been patented whilst offering no value or originality of your own.
If Microsoft did this they would get squashed? Heh, Microsoft would never do that, they're smarter than that. Instead they forced every Android OEM into licensing deals rather than trying to stomp them in court. Microsoft makes money off every Android phone sold except Motorola's.
I agree moral stands are silly, but Apple's crusade is just as silly. Start milking the patents and stop trying to run competitors off the road. That doesn't affect my buying decisions in the least tho.
Its about patents not just market share. They have a monopoly of patents as they were the 2nd in the market with a touch screen phone and they built an arsenal of patents ready to swash any competition around. Let me say this touch phones was about to happen anyway (LG Prada). The technology to make them viable was just happening. Multi touch was shown years earlier at TED's conference and as the guy said there "this is not new" but they still had troubles getting it in a small form factor with high power consuming components. PS When S Jobs showed multi touch he said "we invented" lmao
Microsoft does do this. All of the Android OEM's (except for Motorola) are paying Microsoft a fee for every Android phone they sell. Google and Samsung combined made almost $50billion in 2011. Neither is small nor in danger of being squashed.
And you wouldn't think patents are silly if you were patent holder.
I admit, its easily nicer and stronger than the iPhone4. It does have a nice feel... and its size is pretty good.
But it really functions mostly the same as the iPhone4s... so I don't quite get the mad-rush to buy the iPhone5. The UI is the same (mostly) changed apps (bye bye google) and longated screen.
I still think most people just use their phones until contract is up or the unit dies... and buy the best they like for the price.
Going from Android 2.x to 4.0 has been very nice. My iPad1 (iOS 5) is so... old.
If you ask KoolAidMan and FrenchMAC, yes you should always buy the iphone, whatever iteration, because it is the best; or so they say between the words.
And if you ask one of the millions of Android fanboys, they will tell you they would never consider an iPhone regardless of what it can do or if it has the best performance in the thinnest package with superior battery to the SG3.
If anything I'm seeing more extremism on the anti-Apple side of late.
Amen. Not sure where they are all coming from. Almost like a paid herd of a thousand or so Samsung/Android kids hanging out in tech forums JUST to remind us of Apple's inferiority. It's actually kinda bizarre. Why not just go hang out with other fans of Android. Ya know, in Android forums, chats, and group discussions? If anything, this forum is a mature group of technology enthusiasts. Regardless of OS. No reason to bag on one OR the other
Lol, yeah. Totally objective, no bias and emotional attachment to your preferred brand at all.
Please. I have no attachment to Android (I don't even use an Android phone), but your examples of your "objective" pro-Apple arguments vs. "emotional" pro-Android arguments are just laughable.
Given the A6 CPU clock speed varies between 800Mhz and 1.2GHz, do we know if the 1.2GHz is sustainable or is it just a temporary Turbo mechanism? I'd be interesting to find out if the CPU can hit 1.2GHz when the GPU is running full tilt given they share the same die. Any idea if there is any type of single core Turbo?
Are you sure that both the SGX543MP2 and SGX543MP4 are clocked at 200MHz? If I'm not mistaken, the SGX543MP2 in the A5 is clocked differently for the iPhone 4S and the iPad 2 since the iPad 2 has stronger graphics performance. Something like 200MHz for the iPhone 4S and 250Mhz for the iPad 2, one quarter for their respective CPU clock speeds, would match the performance difference. If the SGX543MP2 in the iPad 2 is clocked at 250MHz, then the SGX543MP4 in the 2012 iPad should also be clocked at 250MHz and not 200MHz since Apple said the 2012 iPad GPU is 2x faster than the iPad 2.
However can the article quote gflops for the iphone5 gpu without ever quoting its clock ? Earlier anandtech articles cited 266mhz, but this is on the low side to me. To get similar performance tp ipad3 you would need >300mhz on mp3 against 250mhz on mp4. iPhone5 offscreen is showing slightly better performance than ipad3. So 325mhz seems right to me.
Please explain how Apple can't possibly have a team of 10 engineers who's job it is to tweak how iOS behaves so that it'll display higher numbers on the benchmarks for which they have full access to source code, so they'd know exactly how to cheat.
Please explain why benchmark numbers jump for devices that get upgraded to Jelly Bean from ICS (SGS3, HTC One X, Motorola M/HD, Optimus G all around October).
Please explain why benchmark numbers jump for rooted Android devices that install certain cyanogenmod or other mod.
I think these benchmarks cannot be used to compare the hardware and software performance of Android vs iPhone.
Real performance comparisons should be done using carefully designed test suites of for example loading certain web pages, loading certain comparable app functionalities, rendering certain verifiably comparable graphics. The way to measure would be to actually verify the graphics rendered on the screen, verify the actual bandwidth usage chronology, you could measure that using high-speed cameras, for example use one of those 240fps cameras, place the devices carefully in a system that coordinates capacitive inputs to trigger and that measures timing to the fraction of the millisecond. But even then, you have to measure through the camera that rendered data/graphics is exactly same.
It sounds complicated, but I think this is how performance comparisons between different platforms must be measured if you want to do it in a lab. You can also do some more general performance tests based on the speed of usage of the user interface by actual users. There, you get users of different levels of experience to be given some tasks to do on each smartphone platform. For example "go to the new york times and look for the latest article about X" and measure time. You'd need to do this test on several different people and have them each do the same tests on each platform but in random order. Come up with as many different relevant tests as you can.
You'll find Android has been much faster than iOS for years. And that the Exynos 4412 Quad-core 1.5Ghz with Mali-400, S4 Pro Quad-core 1.5Ghz with Adreno 320, OMAP4470 1.5Ghz Dual-core with SGX544, Huawei K3V2 and Tegra3+ are all faster than Apple's A6 and A5X etc.
Android has not been faster years. It s quite the opposite in fact. You cannot beat a platform where everything is built from Llvm compiler with code coming from a JIT java VM. Add to that the general lack of GPU optimizations of the UI and the lower quality of applications on Google Plays compare to those of the AppStore.
Android has it s plus but it not in the raw performance area. That what I see for using both platforms at work.
Android uses virtualization and compatible processes to reach as broad a range of hardware as possible. They can progressively optimize how the virtualization is done, hooking in more and more hardware acceleration mainly because all the different hardware platforms use more and more compatible hardware acceleration architecture. The point is Android has FOR YEARS been seconds faster than iOS at MOST actual real use cases, for example ask people of different experience levels to accomplish the top-20 most important smartphone tasks, Android has been faster than iOS at these real performance tests since Android 1.6 in 2009.
"The point is Android has FOR YEARS been seconds faster than iOS at MOST actual real use cases, for example ask people of different experience levels to accomplish the top-20 most important smartphone tasks, Android has been faster than iOS at these real performance tests since Android 1.6 in 2009."
Any chance you could point is to this 'study'? I'm curious as I've been developing now for Android for 4 years, iOS for 3. In my experience, as well as the dozen or so full time colleagues I regularly develop with--we've found the inverse to be true. From the top down. Whether we are talking UI fluidity, app launching, browser speed/loading, email fetching and all around application performance to memory management...iOS is clearly the victor. Whether benchmarking (artificial) or real world, perceived performance (subjective)--in an apples to apples comparison (no pun, talking about current hardware vs current hardware) Apple's devices are clearly 'faster' I wish...I really do that Google would actually close down it's open source idea a bit. Take more control over both manufacturing AND carriers/operators. The integration between soft, middle and hardware that Apple imposes clearly benefits end user experience. Google has the ability to do the same with their Nexus devices...but when it comes to other manufacturers abiding by a simple set of rules set by Google, I think drastic changes could be made is some areas of Android. Namely fluidity, perceived speed....and software development.
As a developer on both platforms, I notice the same thing. Everything in iOS feel faster. It show that android has been created before the multi touch era. Seriously, android 1.6 was soooo slow...
Even the 2.x serie has't much optimization in the 3d department for the ui.
You’re just sounding butthurt here. How many people will have a rooted Android device? "Real world" performance tests should be done with what most people will be using.
If you had one bit of real world experience developing either software or hardware you’d understand that putting aside 10 engineers to optimize a platform for a single benchmark suite is retarded, and double retarded if you are on a tight, 12-month release cycle schedule.
And, no, Android hasn’t been "much faster" than iOS. It’s been faster in some areas, perhaps, and slower in others (UI performance being the most obvious one—you can’t "fake" smooth animations vs jerky ones). It would actually take holding more than one device in your hands to understand that, though, and implying that AnandTech is too stupid to catch any kind of "benchmark tweaking" from either Apple or any other hardware manufacturer is insulting.
All in all, you can’t handle the fact that Apple is capable of producing a solid CPU/GPU combination. Which is surprising, seeing how they have shit loads of money to throw at the problem if they wanted to.
You're retarded if you think a $600 Billion company is going to choose not to use 10 benchmark cheating engineers if that can make some people online believe they are faster.
The point is rooted Android is not necessary faster at everything, it's faster at BENCHMARKS ONLY. The point is how OS is made influences how benchmarks score. And the OS does not need to be fast everywhere, it just needs to render the benchmarks faster!!! There is no proof that all elements of benchmarks are fully rendered, they can easily code OS to skip rendering aspects, skip frames even, just show a faster scrore.
I'm not saying Apple's ARM processors are crap, I'm saying Android ARM Processors are FASTER or at the least comparable. It's just moronic to suggest that the whole world of processor designers can't match or outcompete one little Apple process design team.
LOL... I mean really. Do you even read what you write?
You called another poster "retarded" because, of *course* Apple hired "benchmark cheating engineers" to rig the benchmarks? :) That's really just so obvious that you'd have to be "retarded" to not see it?
I'm sure it's possible though! So, show us some real world applications where the iPhone 5 can't keep up with these FASTER (just slower in benchmarks) "Android" ARM processors.
"You're retarded if you think a $600 Billion company is going to choose not to use 10 benchmark cheating engineers if that can make some people online believe they are faster."
Well said. I thought everyone knew the only reason Android appeared to come close to iOS levels of performance is because Google hired 10 guys to doctor the benchmark. Hell, they even went so far as to create their own benchmark that was "more realistic." I mean, how transparent is that??
I just can't believe how many sheep continue to buy Android phones, just because someone told them "it's almost as good as an iPhone." Don't these people see that Google are just a bunch of anti-competitive patent bullies? They're trying to ban Apple products because they claim they have a patent on messaging FFS. Why can't they just put out a competitive product instead? Are they afraid of innovation and competition?
Notice that Apple hasn't posted any benchmarks, nor have they really gone on and on about the performance of their new SoC. All they really say is that it has 2X the CPU and GPU performance of the A5. If anything, they're seriously downplayed just how well this chip performs compared to the competition, even when the competition is using devices nearly twice the size to house their respective chips and huge batteries.
iOS has always "felt" much faster than Android even on much more modest CPUs. Now that Apple has released a chip that has CPU performance on par or better than what's being used in current Android devices, that gap is only going to be wider.
I would like to see some real world apps that actually "do" something to compare performance. Thing is, it seems most of those types of apps exist only on iOS (advanced audio apps that are CPU limited on the #s of real time effects and simultaneous tracks for example).
There are plenty of apps to compare actual performance. Web browsing, UI, apps, games, those run faster on Android over iOS since 2009. You measure that not by bogus faked benchmarks, you measure that by measuring newbie/expert users actual usage of the device! Perceived smoothness of UI is a completely bogus way to measure things. Andoid is real multi-tasking, Apple cannot do real multi-tasking. Android uses virtualization throughout, Apple cannot. It's just ridiculous to dismiss Android's huge advantage in being based on multi-tasking and virtualization, it's pure technological ignorance.
Web browsing? Yeah that will really show how much faster these Android devices are... What with how laggy and choppy it is scrolling and tapping to zoom even on dual and quad 1.5ghz devices. Meanwhile a 1 ghz A8 iPad 1 has no problems scrolling and zooming like a champ (it has other problems due to its limited RAM, but that's another thing entirely).
"Perceived smoothness of UI" is quite a valid way to measure things. After all, I'm assuming Google and the OEMs want it to be smooth. In fact they made and named a "project" after the idea of making things smooth.
"Apple cannot do real multi-tasking" - iOS does do real multitasking, it's a UNIX-like system, as is Android. It runs a string of background tasks like Android. The primary difference is that user facing apps must use iOS's more strict APIs for multitasking- which has its advantages (difficult to have an app spawn some kind of persistent spyware or drain the battery) as well as the obvious disadvantages. Still, most tasks that most people do are covered by it. Saying "it can't multitask" is lazy unless you really aren't familiar with iOS.
"Android uses virtualization throughout" do you mean the Java-like virtual machine? That's one of the reasons why code runs more *slowly* on Android. There's a tradeoff to that too. And no one is saying those are bad things to have or dismissing them, they're just saying iOS has overall better *performance*.
So you just made all the up because you know its true. My technological knowledge and first hand use of Android phones since the G1 tells me your full of s**t. A single processor phone running a bloated OS with no hardware acceleration, real multi-tasking and virtualization is going to be slow. Android OS did get close to fast or smooth until the dual core phones were released in 2010.
Has it seriously become Android fanboys policy now to simply deny the results of objective and multiple benchmarks?
'You'll find Android has been much faster than iOS for years. And that the Exynos 4412 Quad-core 1.5Ghz with Mali-400, S4 Pro Quad-core 1.5Ghz with Adreno 320, OMAP4470 1.5Ghz Dual-core with SGX544, Huawei K3V2 and Tegra3+ are all faster than Apple's A6 and A5X etc.'
What we'll find is that no phone out right now can match the performance of the iPhone 5 - why can't you just accept that reality?
Because they don't live in the same reality you do.. there is obviously some crossover some where on the web to a reality where Android is faster and has been for years....
Or... he is a 5 year old that still has imaginary friends
This is just pure fakeness. Go do a search for benchmark bogus, benchmark cheat etc. It's a fact for decades that software benchmarks ARE MOSTLY BOGUS since hardware/software makers can always cheat on a benchmark. Apple has access to benchmark source code, it's just pure bogus to think that Apple doesn't look at it and make sure numbers are fake when those specific benchmarks are run on the device. SGS3, LGG, HDQ, A101XS, and upcoming Tegra3+ devices are all faster and much better value for money overall compared to iPhone5.
Let me guess, in your tinfoil hat conspiracy fantasy world, only Apple could possibly "cheat" on these benchmarks. Surely Samsung or nVidia or Qualcomm would never do such a thing in your world, right? :)
They all cheat. Just wait for SGS3, LG G, OMAP470 JB updated stuff, it'll all show Android way above these iPhone5 numbers, but regardless the numbers are bogus by design, a benchmark is a bogus way to measure performance. Measuring the performance of an ARM platform is NOT as easy as installing one apk or ios app and click run and note a number.
we all know the s3 at&t and T-mobile versions are running a dual core chip, thats why it has lower score than the iphone5. anandtech didn't include the standard version of s3 which is using a quad core chip. this may seems unfair....
How is it unfair for a US site to review phones for sell in the US? The versions of the S3 and One X your referring to are not or sale in the US. I'm sure there's a European website with the benchmarks your looking for.
In my area in France I don't encounter problems using the new Maps. I live in a small town and the satellite view is actually better than Google Maps. The precision is better. Also turns by turns direction is works well. But TomTom turns by turns works well here also (could be related).
As for the performances of the A6, what I would like to see is a more real world application performance. Geekbench use C language for its tests (Android NDK) but most applications are written in Java. So there a huge performance penalty here.
What would be interesting to see is a algorithm written both in Objective C/C and in Java running on say a S3 and the iPhone 5. From my point of view a quad core processor brings not that much in real world tests (because most applications rely on one or two threads not four). And then there is the penalty of converting pseudo code to machine code with the Java Dalvik VM. How much does it cost today on modern phone. It is always a 5-10x penalty like on Java on the Desktop? Note that I really like Java as a language but still it cannot compare to native code and on a small device it could be impacting. As a developer I see huge differences in things like parsing an XML file (a common thing in mobile applications).
Using for development purpose both a Nexus 7 (not the most sluggish Android device) and iPad/iPhone I can quite tell you that iOS fly in comparison. Everything is smoother. And more that that everything seems to closely follow the fingers of the user. It's a really important point in the user experience. You don't have the feeling that is a computer but more a real objet dans is represent in the user interface. It follows the physics of finger movements.
I found the UI of Chrome better (the number of tabs is not limited) but scrolling in a page is slower than in Safari. Safari makes also a more profound usage of GPU acceleration (and iOS use GPU acceleration everywhere from the beginning).
The main thing has see as an advantage on Android is the ability to share more easily file between applications. You can also do it on iOS but is more limited (It should improves over times I hope with iCould).
I doubt iCloud file sharing between applications will improve, as each application gets its own sandbox to play around in and has no access outside of that. Perhaps if Apple one day implements a "shared iCloud" sandbox, but that might take a while to get right.
iOS Chrome will become more smooth when Google releases an update for it. UIWebViews in iOS 6 will now render in a background thread which makes scrolling and resizing much smoother. The apps just have to be recompiled with the new SDK.
Yes, I know about no JIT in UIWebView, but at least in iOS 6 the page rendering will work in a background thread like it does in Safari, so scrolling isn't as "choppy". It's still a marked improvement over how Apple had you use it before (unless you used a private API call that could get your app rejected).
Adreno 320 on a phone is looking great. I'm also wondering about the power profile of those chips. I also can't wait for the tests to come to compare these custom chips to the coming Exynos 5 Dual SoC, I really want to know how they stack up to good ole A15 and Mali t600 series.
Yeah, I was quite surprised with how well the 320 performed. It's nice to see another phone actually competitive in the GPU race at launch. The other high end Android phones seem to be still working on beating the 4S, rather than being a threat to the 5!
So it's roughly double the A9 800MHz (A5) performance with it's 1.2GHz processor. Seems to be pretty inline with 1.5GHz Krait. Nice to see that they do match 3D-performance of the A5X. To bad they don't use the A6 on the expensive iPod Touch though.
Well think of how it would match the PS Vita if it had the A6, as it does have their great (LG/Sharp I guess) IPS-screen here not TN-panels like in earlier iPod Touches. You can still get a full power smart phone or gaming friendly tablets for same or less then the new iPod Touch. Galaxy SII matches the iPod Touch 5-gen here in Sweden. Sony Xperia P costs less. All the Tegra3 or Exynos-tablets with less then 8.9"displays typically costs less too. A Galaxy Tab 2 10.1 ends up at the same price. Also consider that the still selling 4-gen iPod Touch is a old device with an A4 processor. Plus the iPod Touch 4G gets beaten by even MSM8255-products like HTC One V in GLBenchmark. A5 might be decent, but still off-puting as a second device, or gaming console. It's also a steep price to pay to get an A5. Overall a high-price to pay to get into the ecosystem and use it as an entertainment device when comparable stuff with all the telephone-bits or larger screens ends up virtually at the same price.
But the big difference is that there is tens of milion of each version of the iPhone in the hand of users. So developer can spend more time porting games or devoloping specific games. People tends also to buy a lot of software. And so there are a lot of quality game in the AppStore.
Big difference? They could have made it equal to porting games to PS Vita. I.e. not just traditional mobile games. If a phone costs as much as the Touch then the appeal of it is drastically reduced. If you can get a SGS2 for the same you get games, software / apps and even more features and a much more flexible device. It's more then (i.e. higher price) popular gaming friendly Android tablets. If you don't need a telephone it's still hard to justify the price. As a second device, or gaming device, portable computing device it gets almost impossible to justify. Despite that it should have many uses in a household. You can control stuff remotely, stream music to stereos and tvs, stream tv-shows and movies (at least to Airplay devices) and much more. When that is not defensible it becomes take it or leave it with the iPhone 5 instead while in reality people have many different devices and generation of devices in a household.
Going from 800MHz to 1200 accounts for a 1.5x change in speed, we know that the Cortex A9 was penalized from bad memory controllers, I'm wondering if getting to 2x was done just by changing those two things, and not so much the actual processor architecture. Does that seem feasible?
One minor nitpick - you have the iPhone 4S's screen resolution as 540 width, when it's 640 across in the legend on some of the comparison graphs (GL Benchmark, etc.)
Damn, I know Sunspider is generally crap, but my old Dell D420, with a low voltage Core Duo (U2500) clocked at the same max GHz as as the A6 (1.2 GHz) scores only slightly faster 820ms vs 908ms for the A6. In terms of GPU performance the A6 creams the integrated Intel part.
Given that this CPU was not cheap at launch, and sucks down 9W for the CPU alone, it is amazing to see how far Smartphones have progressed since the first iPhone / Android devices.
However, I can imagine by 2015 Smartphones running quad-core Arm V8 64-bit processors at ~ 3 GHz with an exponentially faster GPUs, combined with a wireless dock, it truly will be the post PC era.
Question is why would 95% of population need anything faster?
That sounds not right. My parents Pentium 4 2.9GHz, which is way slower than that Core Duo, gets 750 on Chrome, and an old Core Duo laptop I have gets ~400. What are you using, IE?
Yep the IPC of the ULV Yonah (2006) chips is far low than the 2011 i5. However if you compare the per watt performance of the A6 / A15 class chips, they are outstanding.
My main point was that my old Dell U420 equipped with a slow zif SSD, is still a sufficiently swift web / youtube browsing machine for many people.
Given another 3 years software / Hardware development and the 64-bit architecture of the Arm's V8 design to exploit 4 GB + worth of RAM, I can't see most people needing anything more than a tablet / smartphone, especially with technologies such as Intel WiDi improving rapidly. I just can't see how companies like HP will stay in business, as brand is so important in the mobile computing market.
You described the A6 as "very good". I think a more appropriate description would have been "Outstanding". It blew away everything on almost every benchmark. How is that not outstanding?
Also, it's running on two CPU cores! There is tremendous headroom as Apple can increase the clock speed (e.g. run the baseline at 1 GHz) AND increase to 4 cores. These are the obvious steps to increasing performance another 50-100%
As someone who loves tech and innovation, I frankly thought you'd be more wowed by what Apple has done here than indicated by this review
I can guarantee you both anand and his entire staff tha did these prelims are extremely wowed by the A6. But being an unbiased review website they need to keep those feelings bottled in otherwise people will call them one sided fanboys and ruin the validity of the site. By always politely embracing both sides of the fence anand will always have a peaceful and very informative site.
Ps- the results speak for themselves. I think that in itself will wow both anand and the consumer.
Maps? Who cares. Works ok for me, I've also gone to maps.google.com.au and added it as a button to my home screen for when I want google maps. Works fine also.
This post is about performance, and the numbers certainly match what I've been seeing with my iPhone 5. Really quick. And while I've not done any formal testing, the battery life seems far better than the 4S. We shall see how that goes when Anand tests that later.
Quite impressive combination of performance and battery life considering the svelte form factor.
The standard Fandroid fallback of "Yeah, so you beat a phone that's 12 weeks old, what'dya expect?" falls a bit short when last years iPhone is hanging with or BEATING THE CURRENT CROP OF PHONES. I'm amazed that the year and a half old A5 beats the One X and GS3 in many of the graphics tests.
There is still a retail season to go, which includes new SOCs and new designs. The iPhone 5 is just an indicator of things to come.
The fight I'm looking at is Intel verses all the ARM SOC, and power consumption in general.
I think we need a better method to quantify software power consumption for consumer devices. An industry standard set of metrics applicable to any device, with a selection of consumer realistic use cases. Making it easier for a consumer to select a device based on their real world use case(s).
Caching Ajax calls has nothing whatsoever to do with benchmarks.
Ajax call caching cuts down on internet bandwidth usage and load speeds, but it's associated with retrieving data from an external web address, which is also limited by bandwidth.
Thus, none of the benchmarks use any ajax requests as part of their procedure.
It's amazing that so many Android folks come in here just to be defensive and attacking the new iPhone 5 A6 whipping the industry heavyweights.
It's so sad to see the legions of Android phone nerds that won't admit that the Android design by nature can't come close to Apple's tightly screwed design.
We've seen plenty of fake quad-cores non A15s running as high as 1.5GHZ only to score abysmally against the A6 running lower clocks.
This is as bad as AMD throwing out 8 cores and 3GHZ CPU getting trounced by Intel running 50% slower clocks.
Battery life on Android phones are terrible right now and I can't see how innovative Android gets beyond constantly stretching screen sizes and increasing GHZ.
Samsung is good at manufacturing but in terms of design they suck hard. They make such a good silicone with Apple's design.
By taking his quote out of context you missed the point.
"Battery life on Android phones are terrible right now and I can't see how innovative Android gets beyond constantly stretching screen sizes and increasing GHZ.
Samsung is good at manufacturing but in terms of design they suck hard."
The point is that Android phones did these things at the expense of others, either making the phones massive or having bad battery life. The iPhone 5 increased size and leapfrogged everything else in performance without compromising on either size (it actually got smaller) or battery life.
The iPhone 5 is faster than devices twice its size while maintaining excellent battery life. That is good design.
GoodReader (among other apps) connects to Samba shares. I actually wish there were an all in one army knife file manager app as good as GoodReader for Android.
Airplay.
Yes i sure would love to play GTA3 with a controller much larger than the iPhone itself though!
(Nevermind, I read your Flash comment, I didn't think you were joking at first :)
Am I reading some of these posts properly? some people are accusing Anand and Apple of cheating these results????? LMAO. I cant do anything but laugh to those responses. Come on guys. Anandtech remains one of the few unbiased hardware/software review sites left on the net. I highly doubt anandtech would EVER post faulty and falsified results or even if he had knowledge that the results might be skewed.
My best buddy has the GS3 and I just purchased the IP5 so I have experience with both phones. Both phones are extremely smooth, extremely fluid, and internet browsing/app loading has increased dramatically. But IMO, the iphone 5 for me, coming from my iphone 4s was a night and day difference. It is dramatically faster than both my IP4 and my Ipad 3, the A6 was an astonishing achievement IMO. When I say dramatically faster I refer to web browsing, app loading, fluidity of screen swiping and browsing, 1080p movies, 1080p gaming, using the camera for stills and video, and internet download speed/response time.
I am in no way an apple fanboy, I own a 24/7 phase change cooled i7-3770k with windows 7 as my main PC. I also own a GS2 as my business phone which I chose over the Iphone 4S last year so I am what you can call a hardware fanatic, not a fanboy one way or another.
Even though the tests compare to the other processors ie games, but is it fair to say that the benchmark don't account for optimization eg Intel ohttp://www.anandtech.com/show/6324/the-iphone-5-pe... code for Intel processors but codes run slowly on AMD processors; Nvidia code runs better on Tegra processors but less on Tegra processors, etc... How much does java handicap android vs iOS?
Just wondering if the Adreno 225 in the Galaxy S III runs at 400 Mhz because if it does, that would put the GLOPS at 25.6 which is pretty close to the new iPhone 5. I know there is a lot more involved than just GFLOPS for performance but I would think it would be a bit closer in the benchmarks. Unless of course the GPU isn't running at 400 Mhz. Or maybe GFLOPS don't scale per Mhz like I think they do.
It appears that Apples 6th IPhone has had some significant internal development. Too bad all buyers are restricted to use Apples software and market place.
I feel the DOJ is overdue in addressing Apples closed ecosystem and allow users an option to choose (like Android does) if they wish shop for content from other marketplaces. Imagine if you were only allowed to buy food, gas, clothes, etc. from only one store. Competitive pricing is non-existent and price fixing occurs (example: e-book price controls/fixing....sound familiar?).
With the exception of apps, you can pretty much load the iPhone with videos (movies/TV shows), music, books etc. that you buy elsewhere provided they're the same format. All you need to do is basically import the file onto the iTunes and sync the phone.
Android apps don't work on Apple's iOS even when jailbroken. iOS apps won't work on Android at all.
The DOJ won't address Apple's ecosystem because it is NOT a monopoly. Android's ecosystem likewise is not a monopoly.
The eBook prices are set by the publishers on Apple's platform. Each publisher sets its own price. Moreover, each author can set his or her own price if he self-publishes.
Amazon itself is a monopoly on eBooks. Realize that it is a publisher itself. As a publisher, it is competing with other publishers, and undercutting their prices to kill competition - just like Microsoft does.
One thing that needs to be noted for this test is that the Galaxy SIII does not have Android's latest OS available to it. That being said there are significant performance differences between the two. If you don't think there is than run iOS4 instead of iOS5 and compare the same iPhone5.
Why does it need to be noted? Jelly Bean is barely available for the GS3 at the time of me writing this (by barely, it's been released in only a few countries).
Why is it when comparing with Android devices we need to compare with: latest OS that isn't really available other than hackery, custom OS, custom kernel, etc?
Following your logic the only Android phone that should be reviewed is the Galaxy Nexus since its the only phone using the latest version of the OS. Lol, your grasping at screws with that one
It is amazing the way people can blindly support Apple. Makes me wonder who is getting paid to do so. The maps program sucks. Period. The phone and Apple are examples of excellent engineering, with a heavy dash of arrogance.
The guys from Prime Labs (Geekbench) now claim that A6 is actually working at 1.3Ghz (and that their software didn't measure the speed correctly before) ... I think we all knew that a 1Ghz A9/A15/somethingLikeThis processor wouldn't score as high as the A6 was scoring ... and even at 1.2Ghz it was beyond belief.
Now, at 1.3Ghz, with a very good memory interface, I can imagine even an A9 type of architecture getting most of those scores. It all makes sense now, I can go on with my life :)
OMG! LG Optimus G for the first time I see Android device could match iPhone performance. Maybe if not because of higher resolution, Optimus G will beat iPhone 5.
I think it's about time someone at Kishonti or Anandtech comments on the fill rate numbers that the Apple devices achieve. It's obvious that the result does not reflect real fill rate, question is why? MTexels/s implies we are measuring how fast the texture units can read texture data. At some point it's also resonable to assume this date is written to the render buffer. I beleive we are looking at 32bpp color and uncompressed textures. So, iPhone5 1650 MT/s x 4bytes x 1 read + 1 write = 13,2 GB/s. That is for the GPU only. In addition the display drive and other masters on the SoC use the DRAM as well. If we assume they do it in a diffrent way and use several reads / write and blend the pixels in the shader we get (per pas) 8 reads/1 write = 7.4GB/s or 4 reads / 1 write = 8.2GB/s. Obviously it's clear that the device has nowhere near that memory bandwidth for the GPU, especially since the calculation assumes perfectly alligned acceses with a 100% texture cache hit.
What is surprising is that I don't think Apple needs these bogus numbers to be sucessful, the devices are awesome in so many ways that are relevant and I don't think the fill rate result is one of them. So why not come clean and maybe reduce some of the douchebag points that have been growing.
The iPhone 5 is certainly king of the hill when it comes to the GPU, but there are tests where the SGS III International (the "true" SGS III) soundly beats it, for example: Google V8 SGS III: 1960 (stock browser) iPhone 5: 1672
Also on Octane my SGS III gets the same numbers as the iPhone 5 BUT running one test less (this test fails due to compatibility reason), i.e. once this is fixed the score on the SGS III will be higher.
I downloaded the new operating system onto my 4s and am waiting for my 5. I used the maps and compared it to my dedicated Tom Tom GPS. It performed as well as my dedicated GPS and I actually liked the look of it at least as well as my Tom Tom. My GPS is not a low end GPS and is less than a year old and I had NO problems with the IPhone maps. It was just as accurate. Maybe it is due to the fact that I live in Houston which may be mapped a little bit better than some small town in the middle of nowhere. I have been hearing about the map problems and have yet to see what everyone is talking about. Probably my location I guess.
We’ve updated our terms. By continuing to use the site and/or by logging into your account, you agree to the Site’s updated Terms of Use and Privacy Policy.
Facebook
Twitter
RSS
237 Comments
Back to Article
TareX - Friday, September 21, 2012 - link
Certainly left the other devices smoking.... Thing is, the iPhone 5 -excels at what the PS VITA should have been good at, but fails at what a smartphone should be capable of, thanks to the unforgivably horrible Maps app.It's certainly impressive to see what they've done with the CPU, GPU and camera... but the Maps app alone makes it a no buy.
semiconshawn - Friday, September 21, 2012 - link
Maps work fine for me. 3d eye candy needs work but the routes have been spot on.Tegeril - Saturday, September 22, 2012 - link
Same for me, was using it for routing since iOS6 GM hit and it's not yet made one mistake.KitsuneKnight - Saturday, September 22, 2012 - link
Ditto. It works very smoothly for me, it's definitely replaced my old standalone GPS (which can take several minutes to lock, and a minute to calculate/recalculate the route!). I've also enjoyed being able to tell Siri to get me directions to <name of a place> and have it just work.The only thing I've found any issue with in iOS 6 is the new App Store. I definitely like the new functionality/interface, except that there's lots of areas it's not smooth (i.e. often if I click to the updates tab it just sits for a few seconds before it switches tabs and starts loading, instead of instantly switching and showing it's loading). Nothing serious, but it definitely could have used some more testing (I don't believe it was included in the betas, was it?).
B3an - Saturday, September 22, 2012 - link
http://theamazingios6maps.tumblr.com/Yeah REALLY great Maps app.
/s
Formul - Saturday, September 22, 2012 - link
your point being? hearsay > actual user experience? or that the google maps have no problems what so ever?KoolAidMan1 - Saturday, September 22, 2012 - link
Of course B3an is going to rely on hearsay and anecdotal that backs up his extreme bias, what else would he do?semiconshawn - Saturday, September 22, 2012 - link
The 3d flyover stuff needs work. Its cool but its not great. Biased or not he is right.MadMan007 - Saturday, September 22, 2012 - link
How are some posts by users on a forum any more or less hearsay or user experience than results posted to a tumblr account?semiconshawn - Saturday, September 22, 2012 - link
cmon thats not fair your argument uses up all the common sense.Tabular - Saturday, September 22, 2012 - link
Because, unlike the Android fanboys posting here, they've actually used it! And the tumblr post doesn't take into account how many issues there are unlike a real world test.tuxRoller - Sunday, September 23, 2012 - link
I don't think hearsay means what you think it means...NA1NSXR - Tuesday, October 2, 2012 - link
The new maps app is atrocious. I can't believe you are even trying to defend it. I don't know where you live but where I am, in southern CA, is not exactly remote and the level of detail for the local maps is barren. You would be able to navigate to a written address if you had one, but if you were exploring an area you would have no idea what is actually in the places you were looking at. Actually the better question is how you actually use the maps app. If you are finding driving directions to addresses with no regard to anything else then yes, the app is serviceable. If you need the app to provide you with details for what is actually in the areas you are viewing, it is completely useless.relentlessfocus - Saturday, September 22, 2012 - link
Comments like this remind me of the whole antennagate non-issue. Maps working fine for me so far in London. As a backup I have Nokia and GMaps as one tap web apps if i need them so its not a question of having a useless phone as the above comment and a few others imply. This whole maps argument is part of a juvenile religious war IMO, an attempt to blow up a minor problem to score points.Each mobile OS has strengths and weaknesses. Users should use the same criteria as when buying toasters. Which better suits your needs and fits with your taste and requirements.
abhaxus - Sunday, September 23, 2012 - link
since when was antennagate a non-issue? The iphone 4 and 4S have the worst reception of any modern smartphone. That's not a non-issue, it's a fact. The 5 has apparently fixed the issue though.LasseGrr - Sunday, September 23, 2012 - link
You know they fixed the problem on the 4s right? And if i remember correctly, anandtech actually tested both the 4 and 4s as having some of the best reception of the smartphones available at the time..So you might wanna look over your "facts" again..
robinthakur - Monday, September 24, 2012 - link
That is utter nonsense! The iPhone 4S had extremely strong reception compared to the iPhone 4 which, granted, did have some slight issues upon launch. They redesigned the antennae on the 4S and the 4 for Verizon which meant it did not suffer from the recption issues.The 4S reception is leagues better than the Samsung Galaxy S3, which in the same location on the same network for me would actually drop virtually every call. The 5 is once again much better than any previous iPhone for voice call quality and reception, as all the reviews have corroborated and I have experienced. Stop spreading incorrect information which you know nothing about.
moogleii - Monday, September 24, 2012 - link
You should probably read anandtech's analysis on it. Certainly not the "worst" of any smartphone, and certainly not a fact. Their tests showed both pros and cons, and the con was at most a 29 dB drop in signal.Not to mention the antenna was altered for the Verizon 4 and 4s's in general.
Tiny_elvis - Sunday, September 30, 2012 - link
I can't speak for the 4, but I can tell you from firsthand experience that the 4S has spectacular reception. Around town there's not much different, but my 4S grabs signals quicker and hangs onto them longer in mountainous terrain than any of my friends' Android phones.Blookie Spookwell - Tuesday, October 9, 2012 - link
I have a 4S and have never had a reception problem. You must live in the middle of nowheresemiconshawn - Saturday, September 22, 2012 - link
So what? Eye candy needs work. Vector based mapping and nav routing seems solid. As does the voice activated turn by turn. Its not like the maps app is routing you into oblivion. Nit picky bs. No public trans maps is a stupid miss though and I dont even use PT.nicmonson - Saturday, September 22, 2012 - link
Yes, not everyone has the same experience. For the large majority of people, life is fine but for some (mainly outside the US), you will have problems. This is their first time with public maps. Its their beta.Responsive - Sunday, September 23, 2012 - link
Web centric designs seems to have customization with the designs they make with their websiteshttp://bit.ly/wpadaptly
robinthakur - Monday, September 24, 2012 - link
My direct experience of the maps app on iOS6 has been fine, I think people are just looking for something to bash Apple over the head with now the benchmarks show it leaving the S3 and the others even more in the dust, and it has somehow become fashionable to bash Apple. Every year, with every successful product launch, they find something minor to complain about, it seems.The thing is also that the graphics prowess of the iPhones are actually used by Developers far more, vs those found in Android phones. They are mainly used to try and make the interface smooth, but try and find actual games which will use the power contained in the handset and what do you get? Angry Birds Space. Yes it's smooth, but there seems to be very little dedication on the Android front to making Apps which run decently and take advantage of the full power of the graphics in higher end handsets like the S3 a la Infinity Blade, and all the other games which constantly push what can be achieved on a mobile device.
Without them, you are simply left with benchmarks and bragging rights, and what's the point of that? Yes I know that you can do things like moving video previews and floating video windows but everybody knows that it is games which push graphics. I felt pretty demoralised after getting an S3 and then realising that there was really nothing decent gaming wise available.
I don't really see what Google can do to fix this fragmentation other than adopt a platform mentality kind of like it does with the Nexus line, but really, they should just have power profiles which a device can hit which games target, however this doesn't appear to be working very well at the moment, and overall for performance reasons, I'm much happier with my new iPhone 5.
kilkennycat - Saturday, September 22, 2012 - link
Better watch your data usage in Apple3D maps. Especially with LTE speed and Apple's new A6 speedy CPU. Can readily guzzle ~ 30MBytes/minute of urban 3D viewing. Owners of the iPhone5 on a LTE network and with capped data plans will now burn through their data allotment on apparently trivial functions, where otherwise the data usage would have been throttled by either the network bandwidth or the slow processing speed. Expect a huge chorus of iPhone5 owners mad as hell about data-caps and overage costs within the next month.kilkennycat - Saturday, September 22, 2012 - link
Just a follow-up. Seems to me that the iPhone5 would benefit from a built-in continuous tracker of data usage with its output user-selected to be optionally visible in a corner of the "current" screen.pukemon1976 - Saturday, September 22, 2012 - link
welcome to android. 4.0 or higher.Impulses - Sunday, September 23, 2012 - link
Android not only lets you set a warning and force a limit, it also breaks down data usage by app. I'm sure they'll build that into iOS 7... You can find a lot of similarly minor but very useful features that either OS has/lacks, stuff that mostly doesn't get much press but really completes the user experience.moogleii - Monday, September 24, 2012 - link
Which, to be fair, doesn't work on many not-that-old androids.zepi - Saturday, September 22, 2012 - link
People can only give opinions on the functionality of the maps-app based by the area where they use it.Considering how appalled international users have been in some cases, I'd say that the new maps is not really "fine", but instead needs tremendous amount of work to catch up with Google.
Not that it necessarily is such a bad thing. Google maps is not perfect either for people who occasionally stroll off the beaten path, so competition should benefit us all in the long run.
techconc - Friday, September 28, 2012 - link
Same here. I've had nothing but very positive experience with Apple's maps so far. While I don't doubt some issues exist in some places, I'm quite certain that it's nowhere near as bad as what's being reported by the tech press.tnicks - Saturday, September 22, 2012 - link
Show us one example where maps actually screwed up for you that wasn't posted on the net. I agree some of the stuff posted the last few days is laughable, but I've yet to find something not working in my area and I would imagine that holds true for at least 90% of us. It's ridiculous to call it unforgivable, give me a break.There are millions of iphone users that couldn't give two shits about apple vs everyone else. People buy them simply because they are the standard every other manufacturer tries to imitate. They work extremely well minus a few overblown issues that are bound to crop up in any product given so much scrutiny.
L-Set - Saturday, September 22, 2012 - link
Do you live in the US tnicks? Because if you live outside of the states and don't live in a major city then yes, it is pretty poor from all of the reports I've had (from friends, not just random net users). The town where I live in the UK, the population is around 100,000 people, and it's just a giant green patch on maps. I'm not an iphone user, but this is my friend showing me on his updated 4S at work on Friday, it's a bit of a joke. The 5 is clearly a very good phone (benchmarks alone show what a good job Apple have done with the A6), but this had been a PR disaster.And like you said, there are millions of iphone users. Release a new bit of software that doesn't fully work as people intended and of course it's going to blow up!
kyuu - Sunday, September 23, 2012 - link
"People buy them simply because they are the standard every other manufacturer tries to imitate."No, they buy them because they simply aren't paying attention and/or are locked into the Apple ecosystem. Apple's competitors have long since passed them up in terms of both OS and hardware design.
robinthakur - Monday, September 24, 2012 - link
Keep telling yourself that! I had an S3 back in June and sold it after a month and went back to Apple, because I found life is still much worse on the Android side, apart from the OS customisability aspect and widgets, despite how much things have improved since my last Android phone. Those last two aspects are not as important if a competing phone does everything you want without customisation though.-There is no iTunes equivalent, you have to drag and drop everything, which is simply not practical for most people with complex playlists. You might think that you won't mis iTunes, but IME, there is nothing which does so much an offers the 'one-stop' convenience.
-Simple stuff like music playback is mind bogglingly complex to use when the phone is locked becasue you can't put a music playback control on the lock screen using the stock rom. A lot of the software both on the S3 and the supporting software like Kies feels under developed and unreliable.
-Tthe Apps available remain a poor shadow of the Apple ecosystem. Even the apps which do exist on both platforms are usually feature incomplete on Android. The only apps which are leagues better are the Google ones, which I miss slightly.
-The hardware and design of the iPhones blows away anything Samsung or anybody else for that matter have come up with to date. It is refined and feels like quality, not the cheap, thin plastic of the Galaxy, which is supposed to be a flagship device. The screen, whilst bigger on the G3 has a cold tint to the colours on the display and it not retina quality (it looks fuzzy round the sides of text)
-If you want to talk about lock in, there are features on the Galaxy 3 which will only work with other people who own the Galaxy 3 like the photoshare thing, which is hardly useful, and a great deal of what Samsung added to Android like 'S Assist' and 'Smart-stay' just don't work very well.
I'm a power user and know exactly what i'm buying and the reasons for this. Stop patronising people who have weighed up all the pros and cons of a device's hardware and software integration, plus their own needs and opted for a choice other than yours. I would happily buy whatever device I find to be the best, the cost is only one consideration, and for now that device is the iPhone 5.
amdwilliam1985 - Tuesday, September 25, 2012 - link
lol, you listed iTunes as a positive point?NO ONE I come across in real life defends that POS, not even the most die hard Apple lovers. Apparently, my magical experience did not match up to yours, every time I sync with my old iPhone, something goes wrong, in one of the worst case scenario, my prayer didn't go right with Lord Steve J, and all my photos were gone!!!
It's got to be the worst piece of shit in the history of all garbage software. I ran iTunes in windows 7 and 10.7 OSX before.
Biggest reason why I'm left iPhone.
Someone also mentioned iOS being smooth, sigh... my MOTO Razor flip phone was smoother than any iPhone. But that didn't make my stick to it, I wonder why?
moogleii - Monday, September 24, 2012 - link
Sure, software if you're willing to buy every year, since it seems most androids get left out in the cold when it comes to OS upgrades. And if you're ok with a few glaring bugs. My friend's Android often won't show the full keypad after she locks it, which means she can't unlock the phone without rebooting. Not saying Apple software is bug free, but that's pretty a fundamental bug.Hardware design, how? NFC? That's currently a gimmick/non-standard. I have never seen anyone use their android for that; heck, people barely ever use the wireless feature on their CCs. More RAM? Well the tests historically have shown that hasn't seemed to matter, I suspect partly because Android needs the Dalvik virtual machine to be running full time.
Yes, vastly superior indeed.
jameskatt - Saturday, September 22, 2012 - link
There is more to a smartphone than a map app.There are many map apps for the iPhone.
The built in one is guaranteed to get better with time particularly with the help of third parties who specialize in maps.
Maps works well for me. Routes and locations are fine.
The iPhone 5 is a whole lot of FUN!
kmmatney - Saturday, September 22, 2012 - link
I've had some major screw ups with Google maps over the last few years. I think a lot of this is overblown. You can also download other map apps, such as MapQuest, which gives you turn by turn navigation.addicted44 - Saturday, September 22, 2012 - link
I do believe Apple's data is a little behind Google's but that will get resolved in a matter of months. I don't think it is even close to a major issue at this point. Might be a little worse initially when it starts selling in more countries since I assume international data will be weaker.That being said, it is based on OpenStreetMaps with an open data set anyone can contribute to, and a bunch of companies are supporting it over the proprietary Google maps data. I wouldnt be surprised at all to see it far exceed Google Maps's data set within this calendar year itself, especially after Apple starts adding to it based on the data they get from iPhone users.
Personally, i haven't had any issues at all. Initially I was annoyed by having to download a new app for transit directions, but after a couple of days I love it, because the app gives me advisories and offline maps, which the older Maps could t do. And the usage of vector base maps, as opposed to bitmaps has made the map browsing experience much better.
Capt Caveman - Saturday, September 22, 2012 - link
Lame excuse. You do know that there are plenty of alternative Navigation apps available, right? Some are free like MapQuest and Waze that are pretty damn good. My only issue is I wish they had made the screen a little bigger and wider.EnzoFX - Saturday, September 22, 2012 - link
That's like not buying a Windows computer because it comes with Internet Explorer! Some haters just need something to rage at, because it's incomprehensible that Apple maybe has the superior product.kyuu - Sunday, September 23, 2012 - link
Yeah, that's it. People get upset at Apple's shortcomings because Apple has the superior product. Wait what?Mark678 - Saturday, September 22, 2012 - link
Apple Maps has worked fine for me as well.I used it for both searching and also for turn-by-turn directiions.
It is also much faster than Google maps in my 4S.
I am pretty satisfied with it. I am sure that it can be improved and some mistakes will appear and be corrected.
It is really a non-issue and certainly not a deal breaker, as MILLIONS of people have already proved by voting with their money.
FrenchMac - Saturday, September 22, 2012 - link
In my area in France I don't encounter problems using the new Maps. I live in a small town and the satellite view is actually better than Google Maps. The precision is better. Also turns by turns direction is works well. But TomTom turns by turns work well here also (could be related).As for the performances of the A6, what I would like to see is a more real world application performance. Geekbench use C language for its tests (Android NDK) but most applications are written in Java. So there a huge performance penalty here.
What would be interesting to see is a algorithm written both in Objective C/C and in Java running on say a S3 and the iPhone 5. From my point of view a quad core processor brings not that much in real world tests (because most applications rely on one or two threads not four). And then there is the penalty of converting pseudo code to machine code with the Java Dalvik VM. How much does it cost today on modern phone. It is always a 5-10x penalty like on Java on the Desktop? Note that I really like Java as a language but still it cannot compare to native code and on a small device it could be impacting. As a developer I see huge differences in things like parsing an XML file (a common thing in mobile applications).
Using for development purpose both a Nexus 7 (not the most sluggish Android device) and iPad/iPhone I can quite tell you that iOS fly in comparison. Everything is smoother. And more that that everything seems to closely follow the fingers of the user. It's a really important point in the user experience. You don't have the feeling that is a computer but more a real objet dans is represent in the user interface. It follows the physics of finger movements.
I found the UI of Chrome better (the number of tabs is not limited) but scrolling in a page is slower than in Safari. Safari makes also a more profound usage of GPU acceleration (and iOS use GPU acceleration everywhere from the beginning).
The main thing has see as an advantage on Android is the ability to share more easily file between applications. You can also do it on iOS but is more limited (It should improves over times I hope with iCould).
bplewis24 - Saturday, September 22, 2012 - link
Way to be full of sh**.Piffington - Saturday, September 22, 2012 - link
Wow, he provided legit, unbiased info about why Android handsets are running slower and he's full of sh**. So now anybody who talks against Android is full of sh**? Is it a religion now?Death666Angel - Saturday, September 22, 2012 - link
He provides his own view of the two environments. I myself have found that unless I start up a ton of background tasks, my SGS2 and my Chinese tablet are smoother and load apps faster than my friends iPhone4s. And because of his lack of widgets, he needs more inputs to get basic information than I do.robinthakur - Monday, September 24, 2012 - link
I call BS, I have a Nexus 7 running Jelly Bean and it is in NO WAY smoother than iOS even going back to version 1. It's an alright device, I guess, build for a price point and you get a lot for your money, but don't kid yourself that it competes with any iOS device such as the ipad 3 or iPhones in this respect.darkcrayon - Saturday, September 22, 2012 - link
Wait until Google releases an update to Chrome. Once they compile it with the iOS 6 SDK, scrolling will get a bit better. Apple enabled the background rendering in UIWebView which will make third party web apps a bit smoother to scroll (not as much as Safari, but still a big improvement).tuxRoller - Sunday, September 23, 2012 - link
Well, you make some...interesting...claims.First, if you think Dalvik==JVM (which I assume is your point since you are making sweeping, and inaccurate, claims about Java and seem to be confusing the language with the runtime) you are mistaken. The Sun/Oracle JVM (Hotspot) is extremely fast, and very competitive for many algorithms, excepting those with very short run times (http://shootout.alioth.debian.org/u64q/which-progr... Dalvik isn't as optimized as Hotspot (but it has already improved greatly with their new GC in 3.0). So, you can find benchmarks with 5-10x differences but to say that is the norm is incorrect (though, I suppose, it depends on what you are working on from day to day, but hopefully you understand my point).
Second, Geekbench uses native code b/c in order to measure low level performance (like the mem speeds), but since it isn't open source, we don't really know how they are doing things (seriously, people, why are we relying on non-open source benchmarks?). Regardless of GB, though, I would imagine most, if not all, of the Google apps on android are native.
Third, can you post a video showing the difference in latency between nexus 7 and ios device? I ask because I've used both (though own neither) and each time I am surprised with both. One, I keep thinking iphone/ipad is more responsive than it is (I use vertical text scrolling in order to get an idea of input lag), and two, I think Android is less responsive than it is (this, at least, is reasonable since I own a Nexus S which simply isn't fast enough for the software, but the Nexus 7 doesn't seem to have any of the same problems). Frankly I've looked for a video of this type (Tom's Hardware used to use a 1000 frames/sec camera to show input lag but they didn't use it with the Nexus 7) but have found nothing.
Lastly, can you please tell me what you think ios "accelerates" that android doesn't (at least since 3.0)? The browser, in particular, is something that has been mentioned in Android developer blogs as something where they "recently" changed the rendering model (though, still, it has been "accelerated" for years) to a tile based, on demand approach (like Safari, they mention).
robinthakur - Monday, September 24, 2012 - link
Fine. One example is Chrome on my Nexus 7 running JB, scrolling is not at 60fps. On iOS it is. Having said that, on iOS6 the App store has gotten slightly laggy to scroll down slowly, though if you scroll and let go, it goes back to being smooth, which is a little odd.tuxRoller - Monday, September 24, 2012 - link
I'm not sure why you are giving me this example. I never said that android never lags. I was asking the above poster what parts of the drawing pipeline he thinks ios accelerates that android doesn't.Since you mentioned this, however, let me ask: does chrome always lag when scrolling? Does it lag when the page is being assembled and drawn, or afterwards as well? Does it depend on page content (do text pages lag as much as media heavy pages)? Does it lag mostly when memory is low or does that not matter? That is, try restarting the tablet, then open only chrome, and then only open a single tab to a relatively simple page.
If it lags regardless then chrome has an issue. If that is the case, try downloading firefox (i stay on the nightlies since they've been so stable), or opera. If those lag regardless of circumstance as well, then either android is doing something wrong or you have problem with your tablet (either mechanical or software issues due to mechanical issues).
Mohjoe - Monday, September 24, 2012 - link
Apple created OpenCL (Open Computing Language) that is specifically designed to shift general purpose work away from the CPU and onto the GPU in order to take advantage of its power. Using C/Objective-C, you can get the GPU to do tasks that were normally done by the CPU, such as calculating scroll rates in user windows based on touch inputs. GPUs are not solely used for 3D gaming. Apple uses OpenCL to get the GPUs to accelerate every aspect of the UI, hence a more than 2 year old iPhone 4 running iOS 6 is smoother than any equivalent Android device of the same era (that I know of and have seen testing of).Since Apple is the only manufacturer that designs both the hardware (right down to the SoC) and writes its own OS completely, it designs both in tandem. That's a major advantage over all the competition when it comes to performance and this article merely demonstrates that.
And as for maps, I've been using it for a few days now in Sydney, Australia and the only issue I have seen is that some of the restaurant data it is getting from Yelp is not up to date. I plan on using Yelp to update my favourite restaurants now on the map. Otherwise, I really like both the look and the smoothness of operation. Also, it has a big feature that the previous maps lacked, turn by turn (although we in Aus don't get this turned on till next month). So even if it has bugs, its a feature that wasn't present previously.
tuxRoller - Tuesday, September 25, 2012 - link
Was that advertisement a response to me?SanX - Saturday, September 22, 2012 - link
Get out with your agenda, moronSanX - Saturday, September 22, 2012 - link
That was slipped from thread respond to TareX who inflated Maps issueObsoleet - Saturday, September 22, 2012 - link
Wait, spec whores on the Android side are spec whores till... now? Now it's all about the apps! Something iOS has in spades, unquestionably the champ. I use a terrible phone (Android HTC Tbolt) and am replacing it with the iPhone5. I'd rather have a solid QA'd, engineered phone that's fast and has great apps than whatever it is Android is trying to sell itself on (specs? Maps? hit and miss phones?).cserwin - Saturday, September 22, 2012 - link
This. So much this.LG G2X good bye. I tried to like the Android echosystem, but fragmentation and support is just pathetic. It, like the Tbold, was a *freaking Flagship device* 1 year ago, and have been essentially 'off support' for their eintire product life.
The HTC "ONE" and Galaxy SIII don't even use the same processors/gpu/memory from carrier to carrier from what I can tell.
How the hell can you even tell what you're buying? How much research should you have to do to buy a flagship product and expect it to work for the duration of your contract?
I bought the G2X on the basis of a ton of research for a consumer purchase, including very favorable reviews on this site - and I believe Anandtech is hands-down the best technical review site on the planet.
But there are too many variable that go into a phone purchase that go beyond the specs and hardware. I have learned the hard way that manufacturer support, carrier support, and the carrier-manufacturer partnership are AS IMPORTANT AS THE HARDWARE in picking a phone that will have value for the term of the contract.
And if it's an Android, very close behing the manufacturer of the phone itself, you have to evaluate the relationship of the component manufacturers to the open source community. NVIDIA - fuck you.
Because if you think 'community support' somehow makes the carrier-manufacturer relationship irrevelant, you beter think again. Open source developers can only hack without drivers, and when it gets hard, they lose interest, sell their phones on craigslist, and develop for other platforms.
Couple that with the frequent releace cycle of Andrioid, and it's really screwed. Because the community gets ICS support 60% complete for the G2X, well here comes Jelly Bean. And the 3 guys that still had interest? They say 'screw it' and go get a Nexus.
We need review sites to do a better job grading manufactures, carriers, and component suppliers with respect to smart phones... because those things matter as much as the hardware.
And from what I see, Apple benchmarks so far ahead of the rest of the industry on those measures, it isn't even funny.
That Optimus G hardware looks compelling? Run screaming from the building. Because you will never, ever be supported on that phone.
TL;DR: Andriod is a fragmented mess. The review community must look beyond hardware and software. Fuck NVIDIA.
robinthakur - Monday, September 24, 2012 - link
Wow, it's actually rare that I find someone who experienced the same problems with Android that I did with the Galaxy 3. I think they would seem really good if you've only ever used Android, but compared to the very mature iOS with its App library and the more polished iTunes, I would take an iPhone any day.LG are known to be pretty bad for support generally, but then you only usually find this out after you buy it, and by then a new version of Android has come out that can't easily upgraded to, your phone is no longer the top of the benchmarks. Android phones seem to depreciate like crazy compared to iPhones too, which I found rather alarming. I think the support which you are effectively paying for through the higher cost of the iPhone plus the fully QA'd software like iTunes and iOS is worth paying for, especially if you are tied to a 2 year contract. Hope you are happy with your choice.
tuxRoller - Tuesday, September 25, 2012 - link
The simple answer is buy whatever the current Nexus is OFF CONTRACT (the Google model, not the verizon model).The more thorough answer is to research popular models and buy those off contract (in case you aren't getting the message, buy your phones off contract--- it costs a bit more upfront, but less in the long run and you get much more flexibility but you really need to stick with GSM). Then you can either rely on kies (which I"ve never used but have heard is pretty good), or run the latest AOKP/AOSP ROM of your choice.
Frankly, the problem is the expectations of smartphones. Although Apple's model insulates people from this (on all sides, not just the mobile area), essentially, the problem is that which Window's faces. Smartphones ARE COMPUTERS. Just as Windows runs on radically different hardware so does Android. This means that the carriers need to get out of the way of the upgrade process (why are they involved so much anways? I thought the telephony subsystems on these SoC ran their own RTOS?). I think this will happen in time b/c you can currently upgrade your phone outside the carrier approval process.
Android, being linux based, should be able to run on any of these SoC as long as the drivers are provdied (which the OEM's can handle).
KoolAidMan1 - Saturday, September 22, 2012 - link
Nailed it. Of course, anyone with your opinion is *obviously* brainwashed...richpmd - Wednesday, September 26, 2012 - link
Maps works fine for me too. Also see the Anandtech review on iOS 6 which found the Maps to be quite serviceable.Fluffles - Tuesday, October 2, 2012 - link
There's a quick fix for that!:http://www.tomtom.com/en_us/products/car-navigatio...
Or even quicker!:
http://itunes.apple.com/us/app/motionx-gps-drive/i...
Yea but I mean I'm buying a 200$ smartPHONE for only the maps, maybe ill get two and use the other one as a phone, music player, video player, web browser, video caller, wireless hotspot, address book, document editor, gaming device, social networking device, quick camera fix, video recorder, voice recorder, personal reminder, alarm clock, voice assistant, send SMS, send email etc... But the maps suck so all that other stuff doesn't matter. Logic at its best.
aNYthing24 - Friday, September 21, 2012 - link
Nice custom design by Apple.tipoo - Saturday, September 22, 2012 - link
I'm starting to wonder how custom it is, since we now know its max clock is 1.2GHz vs 800MHz in the 4S, that's 1.5x the speed right there, fix the memory controllers (which were a sore point in A9) and you might have 2x without that much customizing.michael2k - Saturday, September 22, 2012 - link
You still have to improve performance by 33% to see a 100% improvement:33% improvement is still a good improvement in the course of 11 months.
In other words, at 800 MHz the A6 should still be 133% of the A5 and at 1.2GHz, a 50% clock increase, you get to 200%.
Given the improvements as measured can be even higher, Apple's accomplishment isn't trivial.
lowlymarine - Saturday, September 22, 2012 - link
Of course, as we've all learned by now browser benchmark performance is largely a measure of how well you cheat (see: IE9 SunSpider scores), so I take any comparison to Android handsets with a grain of salt, especially in light of this: http://arstechnica.com/apple/2012/09/developers-cl...But those are still some darn impressive numbers compared to the outgoing 4S. I'm looking forward to TI's and Samsung's Cortex-A15 CPUs even more now, since at the end of the day the iPhone "5" still has too tiny a screen, no simultaneous voice and LTE, and runs iOS.
jameskatt - Saturday, September 22, 2012 - link
Funny.Waiting for second best is what Android users look forward to.
tuxRoller - Saturday, September 22, 2012 - link
I wouldn't expect the Cortex A15s to be radically better than krait in pure computation, though they have other advantages (the biggest, IMHO, are the virt extensions).@AT:
Thanks for including other benchmarks than just sunspider! I'd still like to see the Kraken benchmarks and Firefox browser included, but this is certainly better.
mavere - Saturday, September 22, 2012 - link
Performance dropped when there's a network connection. Does this include WiFi-on, cellular off?mavere - Saturday, September 22, 2012 - link
Ignore me. I'm stupid. lolwhodakat - Saturday, September 22, 2012 - link
"Hmmm, yeah the camera, the CPU, the GPU, the build quality, the look, the screen but... on the maps app there is cloud cover in Scotland, so yup its trash!""Specs are everything! - Well.... you can cheat these benchmarks....."
"Yeah but can you flash the boot loader and change the power on graphic? Ha iOS is a closed crappy OS!"
(For the record, maps has been working like a charm since the first beta.)
gorash - Saturday, September 22, 2012 - link
Funny the GS3 beat iPhone 5 on Geekbench.You've obviously never used the new Maps, Apple fanboy apologist.
serversurfer - Saturday, September 22, 2012 - link
No, it didn't.http://www.pcmag.com/article2/0,2817,2410034,00.as...
I'm not aware of any benchmarks where the GS3 bests the iPhone 5.
tynopik - Saturday, September 22, 2012 - link
GS3 = Samsung Galaxy S 33GS = ancient iPhone
KitsuneKnight - Saturday, September 22, 2012 - link
The article also included the GS3. Scroll down.tynopik - Saturday, September 22, 2012 - link
I was giving the benefit of the doubt because he said "I'm not aware of ANY benchmarks where the GS3 bests the iPhone 5.", and right there on that that site, the GS3 clearly wins the Geekbench Floating Point benchmark 2315 to 2099 and then the simplest of googling reveals even more tests.I think part of the confusion is the quad-core vs dual-core GS3
KitsuneKnight - Saturday, September 22, 2012 - link
> I think part of the confusion is the quad-core vs dual-core GS3Maybe Samsung should stop giving different regions drastically different phones, all branded similarly? They seem to have issues trying to support so many different variants (as I've had the personal displeasure of experiencing).
cserwin - Saturday, September 22, 2012 - link
How much research should a consumer be expected to do the buy a "Flagship" product? That HTC and Samsung GSIII products have different freaking displays, processors, memory, battery, radios, etc. from carrier-to-carrier.Seriously, "part of the confusion is the quad-core vs dual-core GS3"?
How can that even be possible?
If I'm watching a Galaxy S3 commercial and then go buy the dual-core version.... that can happen.
Really? Tho, seriously, tho. Really? It's like the Andrioid device makers are trying to make it impossible for the consumer to be informed.
As a consumer shopping for a *flagship* Android phone, you can research for hours, for days! And when you make your purchse, without the most careful attention to detail, you could have no clue if the product bought has anything to do with the products you researched.
Impulses - Sunday, September 23, 2012 - link
I get what you're saying, but you're either stretching it too far or you downright suck at research. But yeah, with Android it's way easier to get burned by buying a phone that falls out of favor with the manufacturer or even users/developers...It's not too hard to pick out which are gonna be the most popular models with the most support though. The Tbolt was overhyped and delayed, the G2X was never gonna be such a thing (LG's known for flaky software builds and it came out on the smallest carrier only).
American SGS3/One X have a different processor in order to have built in support for LTE, but performance between Krait/S4 and Tegra 3 is very very comparable so I don't see what the big deal is. S4 is slightly more efficient but that's about it.
You'd certainly never be able to tell both versions apart unless you started benchmarking them. Same displays, same batteries. If you don't wanna do any research and you don't care for having different options then clearly Apple's the perfect option for you.
Nothing wrong with that either.
robinthakur - Monday, September 24, 2012 - link
There arent really any games for either which take full advantage of their graphical capabilities, so you would probably never know the difference unles you were running benchmarks lolserversurfer - Sunday, September 23, 2012 - link
Sorry, I was referring to the overall benchmark, not just a sub-test of said benchmark.doobydoo - Saturday, September 22, 2012 - link
'Funny the GS3 beat iPhone 5 on Geekbench.'Is it funny? Given that:
a) It lost on every single other benchmark (and there's loads)
b) Anandtech already pointed out that Geekbench can't be used to compare cross platform.
KoolAidMan1 - Saturday, September 22, 2012 - link
Fandroid apologists are hilariously sad. All the spin you guys do whenever a new iPhone comes out is weird.Why are you all so mad, why do you even care?
robinthakur - Monday, September 24, 2012 - link
Which version of the S3? The international Quad core with good graphics or the American Dual Core with middling graphics? What do you use the power for, Angry birds? If you think the S3 beats the iPhone 5 in anything in the real world, jog on, you are the apologist here. I have used the new maps and by the way they work fine.jjj - Saturday, September 22, 2012 - link
again too few CPU bound tests , i wouldn't call this perf preview,maybe browsing & GPU perf :(kyuu - Saturday, September 22, 2012 - link
No question, it's an impressive GPU. By all acounts, the CPU seems to be roughly equal to the S4, though.Won't deny it, would love to have it in my phone. But since that GPU is wrapped up in the iPhone 5, I'll have to pass.
manders2600 - Saturday, September 22, 2012 - link
Can we get the date of those benchmarks, and the OS version used?I ask because my VZW Galaxy Nexus running AOKP Jellybean (stock clocks and voltages) scored way better in Sunspider than the one listed here. Haven't tried the rest of the benchmarks, but I'd imagine all of these are old, taken when the phones were first released.
If the benchmarks are not going to be updated to show the actual current performance of the devices, then at least let people know that.
I'm also noticing a lack of Geekbench scores comparing the iPhone 5 to any other device, in contrast to all of the other benchmarks.
manders2600 - Saturday, September 22, 2012 - link
(eg: I see there are included benchmarks for the iPhone 4S running iOS 6.0, which is marginally worse than my own GNex, but much better than the one listed in the chart)manders2600 - Saturday, September 22, 2012 - link
Also, just clocked my GNex to 1.4 GHz and scored 1525 on Octane, basically doubling the 4S's perfs and about 25% higher than the Krait listed.Since there's no way the GNex is faster that the S4 when clocked 100 MHz lower, my best guess would be that the difference is in the OS version (JB vs ICS).
serversurfer - Saturday, September 22, 2012 - link
Apple sell the 4S with 6.0 installed on it. Do Verizon sell the Nexus with Jellybean pre-installed?These phones are tested as they come out of the box. They're testing the performance of the phones as sold, not their own ability to tweak them.
manders2600 - Saturday, September 22, 2012 - link
It's an OTA update that started rolling out today. Much the same as the update for the iPhone 4S to iOS 6.0Also, minor updates with performance enhancements have been issued over the past year the phone has been available. If these benchmarks were run in December of 2011, when the phone was first released, they would not be applicable to the phone today, just as an iOS 5.0 benchmark would not be applicable to the iPhone 4S today.
teiglin - Saturday, September 22, 2012 - link
Not to mention, said Jelly Bean update has been available to GSM Galaxy Nexus users for months, as well as to more enterprising users of toro[plus].I agree that dating these benchmarks should be a requirement, and re-running them (at least for flagship devices/major updates, like JB on the Nexus) would be hugely valuable. Especially for benchmarks like Sunspider, where software optimization is just as relevant as the SoC in question.
All that said, I'm pretty jealous of A6 users. Those of us who don't want everything that goes with buying into Apple are stuck waiting for S4 Pro (hopefully soon) or TI/Exynos A15s (god only knows when).
serversurfer - Saturday, September 22, 2012 - link
So, given the lack of a direct response to my question, am I to assume a direct response would be something along the lines of, "No, they don't sell it with Jellybean, and while I now understand why only certain benchmarks are updated, I still don't like it"?manders2600 - Saturday, September 22, 2012 - link
No, the response would be that the GSM models ARE sold with Jellybean, and the VZW model updates to Jellybean as soon as you activate it.Impulses - Sunday, September 23, 2012 - link
In a couple of weeks or a month at worst (the update came out recently) I'm sure VZ will be selling them with the update already on board. Then again, a month or two from now they might be selling the next Nexus. :p1008anan - Saturday, September 22, 2012 - link
I was surprised by many things:--how good the A6 cores are
--how good the PowerVR SGX543MP3 graphics are
--how good the Adreno 320 and 4 Krait Cores performed (in Brian's LG Optimus G)
How much of the iphone 5's outstanding performance was due to its improved operating system and middleware versus hardware? For the results to be this uniformly good, didn't at least some of the improvement have to have come from improved software?
Anand Shimpi, thank you very much for the Mobile SoC GPU Comparison chart and the easy to understand explanation regarding how the SGX543MP3 GPU works.
Some questions for you if I could:
" Imagination Technologies' PowerVR SGX543 GPU core features four USSE2 pipes. Each pipe has a 4-way vector ALU that can crank out 4 multiply-adds per clock, which works out to be 16 MADs per clock or 32 FLOPS."
Please explain to me why 16 MADs per clock (I understand where that comes from) results in 32 Flops per clock? How can 1 MAD per cycle result in 2 flops per cycle? What am I not understanding?
"Imagination lets the customer stick multiple 543 cores together, which scales compute performance linearly."
How close to linear is it? Is the performance of three 543 cores = 2.8 times the performance of one 543 core? 2.9 times? Or some other number?
"The A5 featured a two core design, running at approximately 200MHz based on our latest news. The A5X in the 3rd generation iPad featured a four core design, running at the same 200MHz clock speed. The A6 on the other hand features a three core PowerVR SGX 543MP3, running at higher clock speeds to deliver a good balance of die size while still delivering on Apple's 2x GPU performance claim. The raw specs are below:"
Is the clock speed of the 543MP3 = [25.5 Gigahertz/19.2 Gigahertz]*200 Megahertz = 266 Megahertz?
From the Mobile SoC GPU Comparison chart, the 543MP3 has 12 SIMDs (3 graphics cores multiplied by 4 SIMDs per core). Each SIMD has 4 MADs. Therefore the 543MP3 has 48 MADs = (3 graphics cores multiplied by 4 SIMDs per core multiplied by 4 MADs per SIMD). Since each MAD has 2 flops per clock, the total number of flops is 96 per clock for the 543MP3. Multiply 266 Megahertz by 96 flops per clock and we get 25.5 gigaflops. This math computes.
Are the 48 MADs all 64 bit? Which results in two flops per MAD? Or is there some other explanation?
Many thanks in advance for your help and the help of the many commentators on this blog.
kfishy - Saturday, September 22, 2012 - link
I presume that 1 MAD equals 2 FLOPS because a MAD is algorithmically decomposed into a multiply and an add in terms of raw computation. In hardware however the multiply and the addition/accumulation would be performed in a single cycle.Lucian Armasu - Saturday, September 22, 2012 - link
Glad to see you're finally including other JS tests, as Sunspider was really getting long in the tooth. It reminds me of when the A5 scored like 4-5x in graphics over other phones in other simple graphics tests before. But those are kind of irrelevant. We need more complex tests that show a more realistic difference in performance between different devices and chips.So I'm glad you've included the V8 test, which was specifically made by Google as an alternative to Sunspider and as a much more complex and realistic test of Javascript performance. Google has been saying Sunspider is irrelevant and they won't optimize for it anymore for about a year now. That might be why Chrome tests even lower than the stock browser in some Sunspider tests. Although it's still strange Chrome score lower in V8 as well compared to the stock browser. I think Google is not where they want to be with Chrome for Android yet. Hopefully the next version of Android that will be unveiled this November and the next version of Chrome will fix that.
What's your conclusion on why the A6 CPU is so fast in these tests? Does it really have much higher IPC than Krait? Or is it because memory was a bottleneck for Krait, and Apple fixed that in the new iPhone 5? Should we expect higher performance from Krait as well when paired with better memory?
I'm also looking forward for you to test Samsung's Exynos 5 Dual with Cortex A15 and Mali T604. No idea why Samsung is dragging its feet and not releasing it with a device already.
tipoo - Saturday, September 22, 2012 - link
We now know it's clocked up to 1.2GHz, vs 800MHz in the A5, so that alone accounts for 1.5x change. The IPC doesn't need to be hugely higher to account for the rest, if they fixed the memory controllers from the A9 that might just account for the rest.Becherovka05 - Saturday, September 22, 2012 - link
Very good gpu, cpu too. I become a android user for navigation and share trading. Apple would of possibly got me if they had the maps earlier. Now I will stay with android for Swype keyboard and widgets and being able to attach files to email easy and the ease of sharing files with other apps. Still can't recommend Apple to anyone on moral stand.MykeM - Saturday, September 22, 2012 - link
The problem with moral stand is that it's usually filled with personal stance and when you put them all together you'll always end up with a lot of holes.KoolAidMan1 - Saturday, September 22, 2012 - link
Exactly. That and all businesses are shady in some way.Just focus on the product. Boycotting a tech company so you can be on some moral high horse is ridiculous
Scannall - Saturday, September 22, 2012 - link
What moral stand? Seriously. It just grates your cheese that Apple makes successful products is all.Becherovka05 - Saturday, September 22, 2012 - link
I like competition I don't like a company as large as Apple trying to squash competition, by claiming silly patents. If Microsoft did this they would get done over (and did)doobydoo - Saturday, September 22, 2012 - link
If you like competition, surely you advocate companies demonstrating originality and innovation, rather than just copying what already exists?anactoraaron - Saturday, September 22, 2012 - link
Rather than copying what already exists... are you talking about copying a dock, icon based os, copied clock, notifications (al la ICS)... or are you talking about copying 'rounded cornered rectangles' kind of copying.Every company 'copies' each other (what already exists) to an extent. Sometimes 'copied' things are no-brainers... phones have never been circular, have always had buttons on them, and can all make <gasp> calls.
I would have no problem with apple's design patent if it clearly specified a specific aspect ratio or specific dimensions of a phone ie: 3.76" x 5.15" with 15 degree rounded corners etc, but that isn't what has happened here.
There's plenty of rounded rectangles in technology from computer monitors to pc speakers, card readers, digital picture frames, etc. that pre-date the iPhone.
anactoraaron - Saturday, September 22, 2012 - link
*meant to say "rounded cornered rectangles"*cserwin - Saturday, September 22, 2012 - link
It's hard to find good explainations about 'Trade Dress' and the meaning of Apple's patent. I had a hard time wrapping my mind around it until it was explained to me like this:Coke has a registered trade dress on a glass bottle with curved sides. When you see it, you know it's a 'coke bottle'. It's an iconic shape. Nobody else can sell a beverage in a similarly shaped bottle in the United States. It doesn't matter if it's diet coke with a white lable, or regular with a red, or no lable at all. When you see that shape, it's a coke bottle. There are 2 liter, 1 liter, 20 oz, 12 oz, 8 oz sizes. The ratios of hight to width are not constant. Some are glass, some are plastic. But's it's all a 'coke bottle' and you know it when you see it.
That same form of recognition is what Apple set out to establish with the Iphone: A single piece of glass, surrounded by a thin bezel, a speaker centered above the screen, a single button beneath the screen. It's an iconic piece of design, and they registered the essential elements as 'trade dress'.
I think the thing that got Samsung in court was the single button and the placement of it in relationship to the other elements of the smartphone design.
When you see a coke bottle on a shelf, you know instantly it's 'Coca-Cola'. There is no doubt, no question. Apple set out to establish that same level of iconic status. They designed that concept, and registered it with the PTO.
What Samsung did with the Galaxy S was the equivalent of Pepsi putting their beverage in a coke bottle with a red label that says 'Pepsi'. You see the product, and it first envokes the brand 'Coke', even tho the lable says 'Pepsi'. The fact that Samsung has rounded corners on a square button or that 'Samsung' is printed between the speaker and screen were trivial - like the difference between glass and plastic coke bottles.
The Galaxy S too closely resembled Apple's design, as did subsequent Samsung products with single buttons centered on glass faces with narrow bezels.
There were other glass bottles before the coke bottle, and there have been many after. "Rectangles with Rounded Corners" is too much of a generalization to describe the iPhone's trade dress. There is more too it, and Samsung hit all the essential elements.
Becherovka05 - Saturday, September 22, 2012 - link
I agree the button is iconic of Apple, but the speaker placement is set by where your ear would be lol. Round corners are not Apple. And pinch to zoom is a fundamental was to use your phone I don't think that should be patterned and if it is should be cheap and available to everyone, like notifications that Apple copied from android. When I look at a Samsung phone I don't think that looks like a iPhone at all. And my local super market Woolworths logo does not look like apples logo.doobydoo - Monday, September 24, 2012 - link
Pinch to zoom is only fundamental because Apple made it so.And remember, Apple didn't patent 'pinch to zoom' - it patents its specific implementation, which is very different.
doobydoo - Monday, September 24, 2012 - link
Dock - not patentableIcon based OS - not patentable
'Copied Clock' - do I even need to answer this one.
Notifications - remains to be seen if they infringed Google's patent in implementation their own version. Google hasn't sued them over it yet which gives you part of your answer.
Copying things is legal, when it's legal. It's illegal, when you infringe patents / trade dress.
I don't know why you state phones have never been circular (shape wasn't anything to do with why Samsung lost), nor buttons (again, not why Samsung lost) nor making calls (again, not why Samsung lost). Ironically, the FRAND patents which Samsung is suing Apple over are for both 3G (which they tried and failed to ban the 4 over) and LTE (which they claim they will try to ban the 5 over).
The main items Samsung lost here were on the Utility patents, which has nothing to do with the shame. The decisions they did win on the design patents and trade dress were nothing to do with the shape claims. Making your final paragraph irrelevant.
To quote an article from a clearly more educated individual:
'Samsung contributed greatly to this with a post-trial statement that said: ““It is unfortunate that patent law can be manipulated to give one company a monopoly over rectangles with rounded corners, or technology that is being improved every day by Samsung and other companies.” It’s more unfortunate that the claim was a gross exaggeration that was swallowed whole by many writers.
Apple claimed that Samsung infringed on four design patents. The D’677 patent covers the overall design of the iPhone while D’305 covers the layout of icons. These claims were upheld. But the jury rejected infringement claims based on patent D’889, which covers the iPad, and rejected eight of 13 claims under D’087 and which deals specifically with the rectangles-with-rounded-corners design of the iPhone (see the relevant sections of the jury verdict form below.) '
http://techpinions.com/pinch-to-zoom-and-rounded-r...
It's only very simplistic, uninformed people who claim this case was about rounded corners, etc - exactly the type of person I wouldn't expect to find on Anandtech.
Becherovka05 - Monday, September 24, 2012 - link
http://www.youtube.com/watch?v=8QzJa_OU0tIdoobydoo - Sunday, September 30, 2012 - link
All this quote refers to is taking inspiration from other products whilst adding your own value.It does not relate to illegally infringing patents / trade dress.
Becherovka05 - Monday, September 24, 2012 - link
Most companies don't sue other company's over minor infringements. This has been a core part of apples strategy And this is the reason I dont own a apple product. Its not that I dont think they are good for people that want a fun phone with good games thats easy to use. And the Ipad is really good. Apple still sucks a55 as a company. And even Woz said this last win over samsung was BS. Your link is bias and there are plenty with the opposite bios. Jobs wanted to go "Thermo Nuclear" on android, they have proven this to be the case. And if this was just about Samsung I wouldnt have a problem, but its been every company thats produced a competing phone. They just dont want to have competition. I like competition hence I dont like a company that tries to remove it.Becherovka05 - Monday, September 24, 2012 - link
Note: The original version of this post said the jury had rejected all claims regarding the rounded-corner design. The jury in fact rejected all claims only regarding willful infringement. On the simple question of infringement, the jury rejected a majority of claims, but did accept five regarding the iPhone. The corrected version appears above.doobydoo - Sunday, September 30, 2012 - link
I quoted the amended article, it needs no further qualification.That your only argument to the article is that it is 'bias', is frankly only representative of your own bias.
The 5 claims which were accepted did not related to rounded corners, as the quote I already pasted proves.
And to claim other companies aren't suing over trivial things - they've gone a step further. Samsung tried to ban the 4S over 3G - with a zero hit rate, and claims it will try to ban the 5 over LTE - both FRAND patents and Samsung is already under investigation for that.
They don't sue over 'trivial' (read: utility, design & trade dress) patents because Apple hasn't infringed any.
kyuu - Sunday, September 23, 2012 - link
Apple copied as much, if not more, from other companies' products as anyone copied from Apple. Quit pretending that Apple's legal shenanigans are anything other than anti-competitive trolling, please.doobydoo - Monday, September 24, 2012 - link
There is a difference between copying things which are in the public domain, not patented/patentable - and doing them in your own way, to outlining a 132 page document on how you plan to replicate technology which has been patented whilst offering no value or originality of your own.Impulses - Sunday, September 23, 2012 - link
If Microsoft did this they would get squashed? Heh, Microsoft would never do that, they're smarter than that. Instead they forced every Android OEM into licensing deals rather than trying to stomp them in court. Microsoft makes money off every Android phone sold except Motorola's.I agree moral stands are silly, but Apple's crusade is just as silly. Start milking the patents and stop trying to run competitors off the road. That doesn't affect my buying decisions in the least tho.
Becherovka05 - Sunday, September 23, 2012 - link
You missed my point about Microsoft, it was about monopoly not whatever your talking about.Impulses - Sunday, September 23, 2012 - link
How can Apple have a monopoly if they don't even own a majority of the market though?Becherovka05 - Monday, September 24, 2012 - link
Its about patents not just market share. They have a monopoly of patents as they were the 2nd in the market with a touch screen phone and they built an arsenal of patents ready to swash any competition around. Let me say this touch phones was about to happen anyway (LG Prada). The technology to make them viable was just happening. Multi touch was shown years earlier at TED's conference and as the guy said there "this is not new" but they still had troubles getting it in a small form factor with high power consuming components.PS When S Jobs showed multi touch he said "we invented" lmao
pmartin - Wednesday, October 3, 2012 - link
Microsoft does do this. All of the Android OEM's (except for Motorola) are paying Microsoft a fee for every Android phone they sell. Google and Samsung combined made almost $50billion in 2011. Neither is small nor in danger of being squashed.And you wouldn't think patents are silly if you were patent holder.
Belard - Saturday, September 22, 2012 - link
Does this mean I must buy an iPhone5?I admit, its easily nicer and stronger than the iPhone4. It does have a nice feel... and its size is pretty good.
But it really functions mostly the same as the iPhone4s... so I don't quite get the mad-rush to buy the iPhone5. The UI is the same (mostly) changed apps (bye bye google) and longated screen.
I still think most people just use their phones until contract is up or the unit dies... and buy the best they like for the price.
Going from Android 2.x to 4.0 has been very nice. My iPad1 (iOS 5) is so... old.
ananduser - Saturday, September 22, 2012 - link
If you ask KoolAidMan and FrenchMAC, yes you should always buy the iphone, whatever iteration, because it is the best; or so they say between the words.doobydoo - Saturday, September 22, 2012 - link
And if you ask one of the millions of Android fanboys, they will tell you they would never consider an iPhone regardless of what it can do or if it has the best performance in the thinnest package with superior battery to the SG3.If anything I'm seeing more extremism on the anti-Apple side of late.
akdj - Saturday, September 22, 2012 - link
Amen. Not sure where they are all coming from. Almost like a paid herd of a thousand or so Samsung/Android kids hanging out in tech forums JUST to remind us of Apple's inferiority. It's actually kinda bizarre. Why not just go hang out with other fans of Android. Ya know, in Android forums, chats, and group discussions? If anything, this forum is a mature group of technology enthusiasts. Regardless of OS. No reason to bag on one OR the otherKoolAidMan1 - Saturday, September 22, 2012 - link
At least we give reasons like excellent performance and the best app developers as excuses.The best you can come up with aren't even positives with Android (selection of larger screens, keyboards), they basically come down to "not Apple".
My opinions are based on numbers and objective analysis. Most of the angry Android fans base their opinions on emotion and rhetoric.
kyuu - Sunday, September 23, 2012 - link
Lol, yeah. Totally objective, no bias and emotional attachment to your preferred brand at all.Please. I have no attachment to Android (I don't even use an Android phone), but your examples of your "objective" pro-Apple arguments vs. "emotional" pro-Android arguments are just laughable.
dishayu - Saturday, September 22, 2012 - link
Good hardware. Would probably be near the top of my consideration list if it ran Jelly Bean/ICS. iOS, no thanks.Piffington - Saturday, September 22, 2012 - link
But then you would be trading off performance for widgets.KitsuneKnight - Saturday, September 22, 2012 - link
That's why he wants such good hardware.ltcommanderdata - Saturday, September 22, 2012 - link
Given the A6 CPU clock speed varies between 800Mhz and 1.2GHz, do we know if the 1.2GHz is sustainable or is it just a temporary Turbo mechanism? I'd be interesting to find out if the CPU can hit 1.2GHz when the GPU is running full tilt given they share the same die. Any idea if there is any type of single core Turbo?Are you sure that both the SGX543MP2 and SGX543MP4 are clocked at 200MHz? If I'm not mistaken, the SGX543MP2 in the A5 is clocked differently for the iPhone 4S and the iPad 2 since the iPad 2 has stronger graphics performance. Something like 200MHz for the iPhone 4S and 250Mhz for the iPad 2, one quarter for their respective CPU clock speeds, would match the performance difference. If the SGX543MP2 in the iPad 2 is clocked at 250MHz, then the SGX543MP4 in the 2012 iPad should also be clocked at 250MHz and not 200MHz since Apple said the 2012 iPad GPU is 2x faster than the iPad 2.
addictech - Saturday, September 22, 2012 - link
looks good now..lets see how it holds up after 50 apps and gamesPaulman - Saturday, September 22, 2012 - link
♪ "Anand-tech...! I think I love you."Paulman - Saturday, September 22, 2012 - link
I meant to say, this kind of in-depth and really "in-the-know" type of article is why I love you, Anandtech!Tangey - Saturday, September 22, 2012 - link
Ipad2/3 clock the gpu @ 250mhz.However can the article quote gflops for the iphone5 gpu without ever quoting its clock ? Earlier anandtech articles cited 266mhz, but this is on the low side to me. To get similar performance tp ipad3 you would need >300mhz on mp3 against 250mhz on mp4. iPhone5 offscreen is showing slightly better performance than ipad3. So 325mhz seems right to me.
Charbax - Saturday, September 22, 2012 - link
Please explain how Apple can't possibly have a team of 10 engineers who's job it is to tweak how iOS behaves so that it'll display higher numbers on the benchmarks for which they have full access to source code, so they'd know exactly how to cheat.Please explain why benchmark numbers jump for devices that get upgraded to Jelly Bean from ICS (SGS3, HTC One X, Motorola M/HD, Optimus G all around October).
Please explain why benchmark numbers jump for rooted Android devices that install certain cyanogenmod or other mod.
I think these benchmarks cannot be used to compare the hardware and software performance of Android vs iPhone.
Real performance comparisons should be done using carefully designed test suites of for example loading certain web pages, loading certain comparable app functionalities, rendering certain verifiably comparable graphics. The way to measure would be to actually verify the graphics rendered on the screen, verify the actual bandwidth usage chronology, you could measure that using high-speed cameras, for example use one of those 240fps cameras, place the devices carefully in a system that coordinates capacitive inputs to trigger and that measures timing to the fraction of the millisecond. But even then, you have to measure through the camera that rendered data/graphics is exactly same.
It sounds complicated, but I think this is how performance comparisons between different platforms must be measured if you want to do it in a lab. You can also do some more general performance tests based on the speed of usage of the user interface by actual users. There, you get users of different levels of experience to be given some tasks to do on each smartphone platform. For example "go to the new york times and look for the latest article about X" and measure time. You'd need to do this test on several different people and have them each do the same tests on each platform but in random order. Come up with as many different relevant tests as you can.
You'll find Android has been much faster than iOS for years. And that the Exynos 4412 Quad-core 1.5Ghz with Mali-400, S4 Pro Quad-core 1.5Ghz with Adreno 320, OMAP4470 1.5Ghz Dual-core with SGX544, Huawei K3V2 and Tegra3+ are all faster than Apple's A6 and A5X etc.
FrenchMac - Saturday, September 22, 2012 - link
Android has not been faster years. It s quite the opposite in fact. You cannot beat a platform where everything is built from Llvm compiler with code coming from a JIT java VM. Add to that the general lack of GPU optimizations of the UI and the lower quality of applications on Google Plays compare to those of the AppStore.Android has it s plus but it not in the raw performance area. That what I see for using both platforms at work.
Charbax - Saturday, September 22, 2012 - link
Android uses virtualization and compatible processes to reach as broad a range of hardware as possible. They can progressively optimize how the virtualization is done, hooking in more and more hardware acceleration mainly because all the different hardware platforms use more and more compatible hardware acceleration architecture. The point is Android has FOR YEARS been seconds faster than iOS at MOST actual real use cases, for example ask people of different experience levels to accomplish the top-20 most important smartphone tasks, Android has been faster than iOS at these real performance tests since Android 1.6 in 2009.akdj - Saturday, September 22, 2012 - link
"The point is Android has FOR YEARS been seconds faster than iOS at MOST actual real use cases, for example ask people of different experience levels to accomplish the top-20 most important smartphone tasks, Android has been faster than iOS at these real performance tests since Android 1.6 in 2009."Any chance you could point is to this 'study'? I'm curious as I've been developing now for Android for 4 years, iOS for 3. In my experience, as well as the dozen or so full time colleagues I regularly develop with--we've found the inverse to be true. From the top down. Whether we are talking UI fluidity, app launching, browser speed/loading, email fetching and all around application performance to memory management...iOS is clearly the victor. Whether benchmarking (artificial) or real world, perceived performance (subjective)--in an apples to apples comparison (no pun, talking about current hardware vs current hardware) Apple's devices are clearly 'faster'
I wish...I really do that Google would actually close down it's open source idea a bit. Take more control over both manufacturing AND carriers/operators. The integration between soft, middle and hardware that Apple imposes clearly benefits end user experience. Google has the ability to do the same with their Nexus devices...but when it comes to other manufacturers abiding by a simple set of rules set by Google, I think drastic changes could be made is some areas of Android. Namely fluidity, perceived speed....and software development.
FrenchMac - Saturday, September 22, 2012 - link
As a developer on both platforms, I notice the same thing. Everything in iOS feel faster. It show that android has been created before the multi touch era. Seriously, android 1.6 was soooo slow...Even the 2.x serie has't much optimization in the 3d department for the ui.
pmartin - Wednesday, October 3, 2012 - link
Please show us where you found this information?xype - Saturday, September 22, 2012 - link
You’re just sounding butthurt here. How many people will have a rooted Android device? "Real world" performance tests should be done with what most people will be using.If you had one bit of real world experience developing either software or hardware you’d understand that putting aside 10 engineers to optimize a platform for a single benchmark suite is retarded, and double retarded if you are on a tight, 12-month release cycle schedule.
And, no, Android hasn’t been "much faster" than iOS. It’s been faster in some areas, perhaps, and slower in others (UI performance being the most obvious one—you can’t "fake" smooth animations vs jerky ones). It would actually take holding more than one device in your hands to understand that, though, and implying that AnandTech is too stupid to catch any kind of "benchmark tweaking" from either Apple or any other hardware manufacturer is insulting.
All in all, you can’t handle the fact that Apple is capable of producing a solid CPU/GPU combination. Which is surprising, seeing how they have shit loads of money to throw at the problem if they wanted to.
Charbax - Saturday, September 22, 2012 - link
You're retarded if you think a $600 Billion company is going to choose not to use 10 benchmark cheating engineers if that can make some people online believe they are faster.The point is rooted Android is not necessary faster at everything, it's faster at BENCHMARKS ONLY. The point is how OS is made influences how benchmarks score. And the OS does not need to be fast everywhere, it just needs to render the benchmarks faster!!! There is no proof that all elements of benchmarks are fully rendered, they can easily code OS to skip rendering aspects, skip frames even, just show a faster scrore.
I'm not saying Apple's ARM processors are crap, I'm saying Android ARM Processors are FASTER or at the least comparable. It's just moronic to suggest that the whole world of processor designers can't match or outcompete one little Apple process design team.
darkcrayon - Saturday, September 22, 2012 - link
LOL... I mean really. Do you even read what you write?You called another poster "retarded" because, of *course* Apple hired "benchmark cheating engineers" to rig the benchmarks? :) That's really just so obvious that you'd have to be "retarded" to not see it?
I'm sure it's possible though! So, show us some real world applications where the iPhone 5 can't keep up with these FASTER (just slower in benchmarks) "Android" ARM processors.
serversurfer - Sunday, September 23, 2012 - link
"You're retarded if you think a $600 Billion company is going to choose not to use 10 benchmark cheating engineers if that can make some people online believe they are faster."Well said. I thought everyone knew the only reason Android appeared to come close to iOS levels of performance is because Google hired 10 guys to doctor the benchmark. Hell, they even went so far as to create their own benchmark that was "more realistic." I mean, how transparent is that??
I just can't believe how many sheep continue to buy Android phones, just because someone told them "it's almost as good as an iPhone." Don't these people see that Google are just a bunch of anti-competitive patent bullies? They're trying to ban Apple products because they claim they have a patent on messaging FFS. Why can't they just put out a competitive product instead? Are they afraid of innovation and competition?
Retarded indeed!!
darkcrayon - Saturday, September 22, 2012 - link
Notice that Apple hasn't posted any benchmarks, nor have they really gone on and on about the performance of their new SoC. All they really say is that it has 2X the CPU and GPU performance of the A5. If anything, they're seriously downplayed just how well this chip performs compared to the competition, even when the competition is using devices nearly twice the size to house their respective chips and huge batteries.iOS has always "felt" much faster than Android even on much more modest CPUs. Now that Apple has released a chip that has CPU performance on par or better than what's being used in current Android devices, that gap is only going to be wider.
I would like to see some real world apps that actually "do" something to compare performance. Thing is, it seems most of those types of apps exist only on iOS (advanced audio apps that are CPU limited on the #s of real time effects and simultaneous tracks for example).
Charbax - Saturday, September 22, 2012 - link
There are plenty of apps to compare actual performance. Web browsing, UI, apps, games, those run faster on Android over iOS since 2009. You measure that not by bogus faked benchmarks, you measure that by measuring newbie/expert users actual usage of the device! Perceived smoothness of UI is a completely bogus way to measure things. Andoid is real multi-tasking, Apple cannot do real multi-tasking. Android uses virtualization throughout, Apple cannot. It's just ridiculous to dismiss Android's huge advantage in being based on multi-tasking and virtualization, it's pure technological ignorance.darkcrayon - Saturday, September 22, 2012 - link
Please name specific real wdarkcrayon - Saturday, September 22, 2012 - link
Web browsing? Yeah that will really show how much faster these Android devices are... What with how laggy and choppy it is scrolling and tapping to zoom even on dual and quad 1.5ghz devices. Meanwhile a 1 ghz A8 iPad 1 has no problems scrolling and zooming like a champ (it has other problems due to its limited RAM, but that's another thing entirely)."Perceived smoothness of UI" is quite a valid way to measure things. After all, I'm assuming Google and the OEMs want it to be smooth. In fact they made and named a "project" after the idea of making things smooth.
"Apple cannot do real multi-tasking" - iOS does do real multitasking, it's a UNIX-like system, as is Android. It runs a string of background tasks like Android. The primary difference is that user facing apps must use iOS's more strict APIs for multitasking- which has its advantages (difficult to have an app spawn some kind of persistent spyware or drain the battery) as well as the obvious disadvantages. Still, most tasks that most people do are covered by it. Saying "it can't multitask" is lazy unless you really aren't familiar with iOS.
"Android uses virtualization throughout" do you mean the Java-like virtual machine? That's one of the reasons why code runs more *slowly* on Android. There's a tradeoff to that too. And no one is saying those are bad things to have or dismissing them, they're just saying iOS has overall better *performance*.
FrenchMac - Sunday, September 23, 2012 - link
Charbax was surely speaking of java virtualization because their is no hardware virtualization instructions on ARM yet like there are on intel.steven75 - Monday, September 24, 2012 - link
Wow, so much ignorance on display here.pmartin - Wednesday, October 3, 2012 - link
So you just made all the up because you know its true. My technological knowledge and first hand use of Android phones since the G1 tells me your full of s**t. A single processor phone running a bloated OS with no hardware acceleration, real multi-tasking and virtualization is going to be slow. Android OS did get close to fast or smooth until the dual core phones were released in 2010.doobydoo - Saturday, September 22, 2012 - link
Has it seriously become Android fanboys policy now to simply deny the results of objective and multiple benchmarks?'You'll find Android has been much faster than iOS for years. And that the Exynos 4412 Quad-core 1.5Ghz with Mali-400, S4 Pro Quad-core 1.5Ghz with Adreno 320, OMAP4470 1.5Ghz Dual-core with SGX544, Huawei K3V2 and Tegra3+ are all faster than Apple's A6 and A5X etc.'
What we'll find is that no phone out right now can match the performance of the iPhone 5 - why can't you just accept that reality?
joelypolly - Saturday, September 22, 2012 - link
Because they don't live in the same reality you do.. there is obviously some crossover some where on the web to a reality where Android is faster and has been for years....Or... he is a 5 year old that still has imaginary friends
Charbax - Saturday, September 22, 2012 - link
This is just pure fakeness. Go do a search for benchmark bogus, benchmark cheat etc. It's a fact for decades that software benchmarks ARE MOSTLY BOGUS since hardware/software makers can always cheat on a benchmark. Apple has access to benchmark source code, it's just pure bogus to think that Apple doesn't look at it and make sure numbers are fake when those specific benchmarks are run on the device. SGS3, LGG, HDQ, A101XS, and upcoming Tegra3+ devices are all faster and much better value for money overall compared to iPhone5.KoolAidMan1 - Saturday, September 22, 2012 - link
WOW, and they say that Apple users live in the RDF.You are pathetic.
darkcrayon - Saturday, September 22, 2012 - link
Let me guess, in your tinfoil hat conspiracy fantasy world, only Apple could possibly "cheat" on these benchmarks. Surely Samsung or nVidia or Qualcomm would never do such a thing in your world, right? :)Charbax - Sunday, September 23, 2012 - link
They all cheat. Just wait for SGS3, LG G, OMAP470 JB updated stuff, it'll all show Android way above these iPhone5 numbers, but regardless the numbers are bogus by design, a benchmark is a bogus way to measure performance. Measuring the performance of an ARM platform is NOT as easy as installing one apk or ios app and click run and note a number.eziodafirenze - Saturday, September 22, 2012 - link
we all know the s3 at&t and T-mobile versions are running a dual core chip, thats why it has lower score than the iphone5. anandtech didn't include the standard version of s3 which is using a quad core chip. this may seems unfair....thunng8 - Saturday, September 22, 2012 - link
What are you on about? The international version of the s3 with quad core results are there.doobydoo - Saturday, September 22, 2012 - link
thunng8 - stop throwing logic bombshells in the way of the Android fanboys' denial.CyberAngel - Monday, October 15, 2012 - link
LOLcserwin - Saturday, September 22, 2012 - link
Unfair is that Samsung and the carriers pull that bullshit on the consumers, but advertise as if 's3' has some universal meaning.steven75 - Monday, September 24, 2012 - link
Nailed it.pmartin - Wednesday, October 3, 2012 - link
How is it unfair for a US site to review phones for sell in the US? The versions of the S3 and One X your referring to are not or sale in the US. I'm sure there's a European website with the benchmarks your looking for.FrenchMac - Saturday, September 22, 2012 - link
In my area in France I don't encounter problems using the new Maps. I live in a small town and the satellite view is actually better than Google Maps. The precision is better. Also turns by turns direction is works well. But TomTom turns by turns works well here also (could be related).As for the performances of the A6, what I would like to see is a more real world application performance. Geekbench use C language for its tests (Android NDK) but most applications are written in Java. So there a huge performance penalty here.
What would be interesting to see is a algorithm written both in Objective C/C and in Java running on say a S3 and the iPhone 5. From my point of view a quad core processor brings not that much in real world tests (because most applications rely on one or two threads not four). And then there is the penalty of converting pseudo code to machine code with the Java Dalvik VM. How much does it cost today on modern phone. It is always a 5-10x penalty like on Java on the Desktop? Note that I really like Java as a language but still it cannot compare to native code and on a small device it could be impacting. As a developer I see huge differences in things like parsing an XML file (a common thing in mobile applications).
Using for development purpose both a Nexus 7 (not the most sluggish Android device) and iPad/iPhone I can quite tell you that iOS fly in comparison. Everything is smoother. And more that that everything seems to closely follow the fingers of the user. It's a really important point in the user experience. You don't have the feeling that is a computer but more a real objet dans is represent in the user interface. It follows the physics of finger movements.
I found the UI of Chrome better (the number of tabs is not limited) but scrolling in a page is slower than in Safari. Safari makes also a more profound usage of GPU acceleration (and iOS use GPU acceleration everywhere from the beginning).
The main thing has see as an advantage on Android is the ability to share more easily file between applications. You can also do it on iOS but is more limited (It should improves over times I hope with iCould).
xype - Saturday, September 22, 2012 - link
I doubt iCloud file sharing between applications will improve, as each application gets its own sandbox to play around in and has no access outside of that. Perhaps if Apple one day implements a "shared iCloud" sandbox, but that might take a while to get right.darkcrayon - Saturday, September 22, 2012 - link
iOS Chrome will become more smooth when Google releases an update for it. UIWebViews in iOS 6 will now render in a background thread which makes scrolling and resizing much smoother. The apps just have to be recompiled with the new SDK.FrenchMac - Sunday, September 23, 2012 - link
In fact I was talking about Chrome on Android. On iOS it use UIWebKit which is slower than Safari for security reason (no JIT).darkcrayon - Sunday, September 23, 2012 - link
Ahh ok.Yes, I know about no JIT in UIWebView, but at least in iOS 6 the page rendering will work in a background thread like it does in Safari, so scrolling isn't as "choppy". It's still a marked improvement over how Apple had you use it before (unless you used a private API call that could get your app rejected).
lilmoe - Saturday, September 22, 2012 - link
Adreno 320 on a phone is looking great. I'm also wondering about the power profile of those chips. I also can't wait for the tests to come to compare these custom chips to the coming Exynos 5 Dual SoC, I really want to know how they stack up to good ole A15 and Mali t600 series.KitsuneKnight - Saturday, September 22, 2012 - link
Yeah, I was quite surprised with how well the 320 performed. It's nice to see another phone actually competitive in the GPU race at launch. The other high end Android phones seem to be still working on beating the 4S, rather than being a threat to the 5!Penti - Saturday, September 22, 2012 - link
So it's roughly double the A9 800MHz (A5) performance with it's 1.2GHz processor. Seems to be pretty inline with 1.5GHz Krait. Nice to see that they do match 3D-performance of the A5X. To bad they don't use the A6 on the expensive iPod Touch though.tipoo - Saturday, September 22, 2012 - link
The Touch also starts at 300 vs over 600 for an off contract iPhone though, so the A5 is decent for that.Penti - Saturday, September 22, 2012 - link
Well think of how it would match the PS Vita if it had the A6, as it does have their great (LG/Sharp I guess) IPS-screen here not TN-panels like in earlier iPod Touches. You can still get a full power smart phone or gaming friendly tablets for same or less then the new iPod Touch. Galaxy SII matches the iPod Touch 5-gen here in Sweden. Sony Xperia P costs less. All the Tegra3 or Exynos-tablets with less then 8.9"displays typically costs less too. A Galaxy Tab 2 10.1 ends up at the same price. Also consider that the still selling 4-gen iPod Touch is a old device with an A4 processor. Plus the iPod Touch 4G gets beaten by even MSM8255-products like HTC One V in GLBenchmark. A5 might be decent, but still off-puting as a second device, or gaming console. It's also a steep price to pay to get an A5. Overall a high-price to pay to get into the ecosystem and use it as an entertainment device when comparable stuff with all the telephone-bits or larger screens ends up virtually at the same price.FrenchMac - Sunday, September 23, 2012 - link
But the big difference is that there is tens of milion of each version of the iPhone in the hand of users. So developer can spend more time porting games or devoloping specific games. People tends also to buy a lot of software. And so there are a lot of quality game in the AppStore.Penti - Sunday, September 23, 2012 - link
Big difference? They could have made it equal to porting games to PS Vita. I.e. not just traditional mobile games. If a phone costs as much as the Touch then the appeal of it is drastically reduced. If you can get a SGS2 for the same you get games, software / apps and even more features and a much more flexible device. It's more then (i.e. higher price) popular gaming friendly Android tablets. If you don't need a telephone it's still hard to justify the price. As a second device, or gaming device, portable computing device it gets almost impossible to justify. Despite that it should have many uses in a household. You can control stuff remotely, stream music to stereos and tvs, stream tv-shows and movies (at least to Airplay devices) and much more. When that is not defensible it becomes take it or leave it with the iPhone 5 instead while in reality people have many different devices and generation of devices in a household.tipoo - Saturday, September 22, 2012 - link
Going from 800MHz to 1200 accounts for a 1.5x change in speed, we know that the Cortex A9 was penalized from bad memory controllers, I'm wondering if getting to 2x was done just by changing those two things, and not so much the actual processor architecture. Does that seem feasible?FrenchMac - Saturday, September 22, 2012 - link
The instruction set of the A6 processor is not the same (armv7s) like an Cortex A15 processor with twice the number of floatting point registers.tipoo - Saturday, September 22, 2012 - link
Right, forgot about that. Apart from that, I wonder how extensive the customization is.zdw - Saturday, September 22, 2012 - link
One minor nitpick - you have the iPhone 4S's screen resolution as 540 width, when it's 640 across in the legend on some of the comparison graphs (GL Benchmark, etc.)Bpease - Saturday, September 22, 2012 - link
Damn, I know Sunspider is generally crap, but my old Dell D420, with a low voltage Core Duo (U2500) clocked at the same max GHz as as the A6 (1.2 GHz) scores only slightly faster 820ms vs 908ms for the A6. In terms of GPU performance the A6 creams the integrated Intel part.Given that this CPU was not cheap at launch, and sucks down 9W for the CPU alone, it is amazing to see how far Smartphones have progressed since the first iPhone / Android devices.
However, I can imagine by 2015 Smartphones running quad-core Arm V8 64-bit processors at ~ 3 GHz with an exponentially faster GPUs, combined with a wireless dock, it truly will be the post PC era.
Question is why would 95% of population need anything faster?
tipoo - Saturday, September 22, 2012 - link
That sounds not right. My parents Pentium 4 2.9GHz, which is way slower than that Core Duo, gets 750 on Chrome, and an old Core Duo laptop I have gets ~400. What are you using, IE?tipoo - Saturday, September 22, 2012 - link
Oh sorry, missed the 1.2GHz part. Maybe you're right.Zink - Saturday, September 22, 2012 - link
Locked at 800MHz in power saving mode my Sandy Bridge laptop i5 gets 710ms in Chrome.Bpease - Saturday, September 22, 2012 - link
Yep the IPC of the ULV Yonah (2006) chips is far low than the 2011 i5. However if you compare the per watt performance of the A6 / A15 class chips, they are outstanding.My main point was that my old Dell U420 equipped with a slow zif SSD, is still a sufficiently swift web / youtube browsing machine for many people.
Given another 3 years software / Hardware development and the 64-bit architecture of the Arm's V8 design to exploit 4 GB + worth of RAM, I can't see most people needing anything more than a tablet / smartphone, especially with technologies such as Intel WiDi improving rapidly. I just can't see how companies like HP will stay in business, as brand is so important in the mobile computing market.
Red Oak - Saturday, September 22, 2012 - link
You described the A6 as "very good". I think a more appropriate description would have been "Outstanding". It blew away everything on almost every benchmark. How is that not outstanding?Also, it's running on two CPU cores! There is tremendous headroom as Apple can increase the clock speed (e.g. run the baseline at 1 GHz) AND increase to 4 cores. These are the obvious steps to increasing performance another 50-100%
As someone who loves tech and innovation, I frankly thought you'd be more wowed by what Apple has done here than indicated by this review
Zink - Saturday, September 22, 2012 - link
He also called the Adreno 320 "quite competent" when it beat the A6.big dom - Sunday, September 23, 2012 - link
I can guarantee you both anand and his entire staff tha did these prelims are extremely wowed by the A6. But being an unbiased review website they need to keep those feelings bottled in otherwise people will call them one sided fanboys and ruin the validity of the site. By always politely embracing both sides of the fence anand will always have a peaceful and very informative site.Ps- the results speak for themselves. I think that in itself will wow both anand and the consumer.
darwiniandude - Saturday, September 22, 2012 - link
Maps? Who cares. Works ok for me, I've also gone to maps.google.com.au and added it as a button to my home screen for when I want google maps. Works fine also.This post is about performance, and the numbers certainly match what I've been seeing with my iPhone 5. Really quick. And while I've not done any formal testing, the battery life seems far better than the 4S. We shall see how that goes when Anand tests that later.
Quite impressive combination of performance and battery life considering the svelte form factor.
Stringtheory1 - Saturday, September 22, 2012 - link
The standard Fandroid fallback of "Yeah, so you beat a phone that's 12 weeks old, what'dya expect?" falls a bit short when last years iPhone is hanging with or BEATING THE CURRENT CROP OF PHONES. I'm amazed that the year and a half old A5 beats the One X and GS3 in many of the graphics tests.joelw - Saturday, September 22, 2012 - link
There is still a retail season to go, which includes new SOCs and new designs. The iPhone 5 is just an indicator of things to come.The fight I'm looking at is Intel verses all the ARM SOC, and power consumption in general.
I think we need a better method to quantify software power consumption for consumer devices. An industry standard set of metrics applicable to any device, with a selection of consumer realistic use cases. Making it easier for a consumer to select a device based on their real world use case(s).
Fleeb - Saturday, September 22, 2012 - link
There are discussions in SO over here: http://stackoverflow.com/questions/12506897/is-saf...doobydoo - Sunday, September 23, 2012 - link
Caching Ajax calls has nothing whatsoever to do with benchmarks.Ajax call caching cuts down on internet bandwidth usage and load speeds, but it's associated with retrieving data from an external web address, which is also limited by bandwidth.
Thus, none of the benchmarks use any ajax requests as part of their procedure.
paravorheim - Saturday, September 22, 2012 - link
Shouldn't the iPhone 4S resolution be 960 x 640, not 960 x 540?vision33r - Sunday, September 23, 2012 - link
It's amazing that so many Android folks come in here just to be defensive and attacking the new iPhone 5 A6 whipping the industry heavyweights.It's so sad to see the legions of Android phone nerds that won't admit that the Android design by nature can't come close to Apple's tightly screwed design.
We've seen plenty of fake quad-cores non A15s running as high as 1.5GHZ only to score abysmally against the A6 running lower clocks.
This is as bad as AMD throwing out 8 cores and 3GHZ CPU getting trounced by Intel running 50% slower clocks.
Battery life on Android phones are terrible right now and I can't see how innovative Android gets beyond constantly stretching screen sizes and increasing GHZ.
Samsung is good at manufacturing but in terms of design they suck hard. They make such a good silicone with Apple's design.
kyuu - Sunday, September 23, 2012 - link
"[...] I can't see how innovative Android gets beyond constantly stretching screen sizes and increasing GHZ."As opposed to what Apple did with the iPhone 5, which was... stretching the screen and "increasing the GHz"?
KoolAidMan1 - Sunday, September 23, 2012 - link
By taking his quote out of context you missed the point."Battery life on Android phones are terrible right now and I can't see how innovative Android gets beyond constantly stretching screen sizes and increasing GHZ.
Samsung is good at manufacturing but in terms of design they suck hard."
The point is that Android phones did these things at the expense of others, either making the phones massive or having bad battery life. The iPhone 5 increased size and leapfrogged everything else in performance without compromising on either size (it actually got smaller) or battery life.
The iPhone 5 is faster than devices twice its size while maintaining excellent battery life. That is good design.
lmnopz - Tuesday, September 25, 2012 - link
Too bad Iphones cant stream DLNA, connect to Samba shares, or play Grand Theft Auto 3...lmnopz - Tuesday, September 25, 2012 - link
*play GTA3 w/ a playstation controller via bluetoothlmnopz - Tuesday, September 25, 2012 - link
Also enjoy a phone that doesnt support Flash.Lol pathetic
pmartin - Wednesday, October 3, 2012 - link
Your right, those are the 3 reasons no one should ever buy a iphone, Loldarkcrayon - Wednesday, September 26, 2012 - link
GoodReader (among other apps) connects to Samba shares. I actually wish there were an all in one army knife file manager app as good as GoodReader for Android.Airplay.
Yes i sure would love to play GTA3 with a controller much larger than the iPhone itself though!
(Nevermind, I read your Flash comment, I didn't think you were joking at first :)
doobydoo - Monday, September 24, 2012 - link
Also Apple primarily improved the architecture - with its own design, rather than just 'increasing the GHz'.big dom - Sunday, September 23, 2012 - link
Am I reading some of these posts properly? some people are accusing Anand and Apple of cheating these results????? LMAO. I cant do anything but laugh to those responses. Come on guys. Anandtech remains one of the few unbiased hardware/software review sites left on the net. I highly doubt anandtech would EVER post faulty and falsified results or even if he had knowledge that the results might be skewed.My best buddy has the GS3 and I just purchased the IP5 so I have experience with both phones. Both phones are extremely smooth, extremely fluid, and internet browsing/app loading has increased dramatically. But IMO, the iphone 5 for me, coming from my iphone 4s was a night and day difference. It is dramatically faster than both my IP4 and my Ipad 3, the A6 was an astonishing achievement IMO. When I say dramatically faster I refer to web browsing, app loading, fluidity of screen swiping and browsing, 1080p movies, 1080p gaming, using the camera for stills and video, and internet download speed/response time.
I am in no way an apple fanboy, I own a 24/7 phase change cooled i7-3770k with windows 7 as my main PC. I also own a GS2 as my business phone which I chose over the Iphone 4S last year so I am what you can call a hardware fanatic, not a fanboy one way or another.
-D
OCedHrt - Sunday, September 23, 2012 - link
http://stackoverflow.com/questions/12506897/is-saf...humancyborg - Sunday, September 23, 2012 - link
As mentioned already, caching ajax POST requests has nothing to do with the Javascript benchmarks in this report.Alexosm - Sunday, September 23, 2012 - link
Thanks for great review. I wonder how the Overclocked S3, would compare to the new iPhone 5. Any chance you could compare the two? Thanks!MykeM - Sunday, September 23, 2012 - link
Why would you want or need to overclock a phone?vngo - Sunday, September 23, 2012 - link
The iphone 5 even beat out the galaxy note 2http://www.engadget.com/2012/08/31/samsung-galaxy-...
lmnopz - Tuesday, September 25, 2012 - link
Not in GL Bench, it didn't:GL Benchmark Egypt Offscreen
114 fps
thunng8 - Friday, September 28, 2012 - link
Wrong. Engadget tested glbenchmark 2.1, not 2.5. One look at the s3 result and you can see it 2.1. Iphone5 gets 135 fps in egypt 2.1http://glbenchmark.com/phonedetails.jsp?D=Apple+iP...
jconan - Sunday, September 23, 2012 - link
Even though the tests compare to the other processors ie games, but is it fair to say that the benchmark don't account for optimization eg Intel ohttp://www.anandtech.com/show/6324/the-iphone-5-pe... code for Intel processors but codes run slowly on AMD processors; Nvidia code runs better on Tegra processors but less on Tegra processors, etc... How much does java handicap android vs iOS?NickDG - Monday, September 24, 2012 - link
Just wondering if the Adreno 225 in the Galaxy S III runs at 400 Mhz because if it does, that would put the GLOPS at 25.6 which is pretty close to the new iPhone 5. I know there is a lot more involved than just GFLOPS for performance but I would think it would be a bit closer in the benchmarks. Unless of course the GPU isn't running at 400 Mhz. Or maybe GFLOPS don't scale per Mhz like I think they do.GotThumbs - Monday, September 24, 2012 - link
It appears that Apples 6th IPhone has had some significant internal development. Too bad all buyers are restricted to use Apples software and market place.I feel the DOJ is overdue in addressing Apples closed ecosystem and allow users an option to choose (like Android does) if they wish shop for content from other marketplaces. Imagine if you were only allowed to buy food, gas, clothes, etc. from only one store. Competitive pricing is non-existent and price fixing occurs (example: e-book price controls/fixing....sound familiar?).
Best wishes,
MykeM - Monday, September 24, 2012 - link
With the exception of apps, you can pretty much load the iPhone with videos (movies/TV shows), music, books etc. that you buy elsewhere provided they're the same format. All you need to do is basically import the file onto the iTunes and sync the phone.jameskatt - Tuesday, September 25, 2012 - link
Android apps don't work on Apple's iOS even when jailbroken.iOS apps won't work on Android at all.
The DOJ won't address Apple's ecosystem because it is NOT a monopoly.
Android's ecosystem likewise is not a monopoly.
The eBook prices are set by the publishers on Apple's platform. Each publisher sets its own price. Moreover, each author can set his or her own price if he self-publishes.
Amazon itself is a monopoly on eBooks. Realize that it is a publisher itself. As a publisher, it is competing with other publishers, and undercutting their prices to kill competition - just like Microsoft does.
pmartin - Wednesday, October 3, 2012 - link
I guess the DOJ should force the auto markers to uses interchangeable partsmikegonzalez2k - Tuesday, September 25, 2012 - link
One thing that needs to be noted for this test is that the Galaxy SIII does not have Android's latest OS available to it. That being said there are significant performance differences between the two. If you don't think there is than run iOS4 instead of iOS5 and compare the same iPhone5.humancyborg - Wednesday, September 26, 2012 - link
Why does it need to be noted? Jelly Bean is barely available for the GS3 at the time of me writing this (by barely, it's been released in only a few countries).Why is it when comparing with Android devices we need to compare with: latest OS that isn't really available other than hackery, custom OS, custom kernel, etc?
pmartin - Wednesday, October 3, 2012 - link
Following your logic the only Android phone that should be reviewed is the Galaxy Nexus since its the only phone using the latest version of the OS. Lol, your grasping at screws with that oneSantaAna12 - Wednesday, September 26, 2012 - link
It is amazing the way people can blindly support Apple. Makes me wonder who is getting paid to do so. The maps program sucks. Period. The phone and Apple are examples of excellent engineering, with a heavy dash of arrogance.darkcrayon - Wednesday, September 26, 2012 - link
Yup. You nailed it. It's amazing how benchmarks can blindly support Apple- haven't those benchmarks tried Maps?doobydoo - Sunday, September 30, 2012 - link
I wonder what you would be crying about if it wasn't for the maps debacle. I'm sure you'd find something.pmartin - Wednesday, October 3, 2012 - link
And your post is an example of a fanboy reduced to trolling about a maps app, lolpolzombie - Wednesday, September 26, 2012 - link
The guys from Prime Labs (Geekbench) now claim that A6 is actually working at 1.3Ghz (and that their software didn't measure the speed correctly before) ... I think we all knew that a 1Ghz A9/A15/somethingLikeThis processor wouldn't score as high as the A6 was scoring ... and even at 1.2Ghz it was beyond belief.Now, at 1.3Ghz, with a very good memory interface, I can imagine even an A9 type of architecture getting most of those scores. It all makes sense now, I can go on with my life :)
darkcrayon - Wednesday, September 26, 2012 - link
I don't understand you - at 1.2 ghz it was beyond belief, but at a less than 10% increase to 1.3 ghz it's nothing special?wheeqo - Thursday, September 27, 2012 - link
OMG! LG Optimus G for the first time I see Android device could match iPhone performance. Maybe if not because of higher resolution, Optimus G will beat iPhone 5.tulebo - Thursday, September 27, 2012 - link
I think it's about time someone at Kishonti or Anandtech comments on the fill rate numbers that the Apple devices achieve.It's obvious that the result does not reflect real fill rate, question is why?
MTexels/s implies we are measuring how fast the texture units can read texture data.
At some point it's also resonable to assume this date is written to the render buffer. I beleive we are looking at 32bpp color and uncompressed textures.
So, iPhone5 1650 MT/s x 4bytes x 1 read + 1 write = 13,2 GB/s.
That is for the GPU only. In addition the display drive and other masters on the SoC use the DRAM as well.
If we assume they do it in a diffrent way and use several reads / write and blend the pixels in the shader we get (per pas) 8 reads/1 write = 7.4GB/s or 4 reads / 1 write = 8.2GB/s.
Obviously it's clear that the device has nowhere near that memory bandwidth for the GPU, especially since the calculation assumes perfectly alligned acceses with a 100% texture cache hit.
What is surprising is that I don't think Apple needs these bogus numbers to be sucessful, the devices are awesome in so many ways that are relevant and I don't think the fill rate result is one of them.
So why not come clean and maybe reduce some of the douchebag points that have been growing.
versesuvius - Saturday, September 29, 2012 - link
Apple designing a CPU? Ho hum ho.Time for an open source CPU design software then. If the "facade company" can do it, everybody can.
Pipperox - Sunday, September 30, 2012 - link
The iPhone 5 is certainly king of the hill when it comes to the GPU, but there are tests where the SGS III International (the "true" SGS III) soundly beats it, for example:Google V8
SGS III: 1960 (stock browser)
iPhone 5: 1672
Also on Octane my SGS III gets the same numbers as the iPhone 5 BUT running one test less (this test fails due to compatibility reason), i.e. once this is fixed the score on the SGS III will be higher.
Blookie Spookwell - Tuesday, October 9, 2012 - link
I downloaded the new operating system onto my 4s and am waiting for my 5. I used the maps and compared it to my dedicated Tom Tom GPS. It performed as well as my dedicated GPS and I actually liked the look of it at least as well as my Tom Tom. My GPS is not a low end GPS and is less than a year old and I had NO problems with the IPhone maps. It was just as accurate. Maybe it is due to the fact that I live in Houston which may be mapped a little bit better than some small town in the middle of nowhere. I have been hearing about the map problems and have yet to see what everyone is talking about. Probably my location I guess.mail2mevibin - Sunday, October 14, 2012 - link
The performance shows awesome for iphone 5 with A6.....but is it affordable for everyone is a big question....
Even Intel with its beginner single core processor comes second in sunspider benchmark competing with multicore processors...