Slashdot Mirror
iPhone 5 GeekBench Results
EGSonikku writes "The iPhone 5 has been benchmarked using the GeekBench tool. According to the results, Apple's claim of 2x higher performance over the iPhone 4S seems accurate. The results show the iPhone 5's A6 CPU is dual core and clocked at 1.2GHz, and is paired with 1GB of RAM. Despite the fact that the Samsung Galaxy S3 has a quad core CPU at 1.4GHz, and twice as much RAM, it seems the iPhone 5 is faster than the S3, or any other Android handset." Meanwhile, Samsung has launched a marketing campaign that compares some of the hardware specs and features between the new iPhone 5 and the GS3.
I'd rather it were the same thickness as the old model if the battery would last longer. Who exactly is it that thinks so they're so horribly thick?
Please note the summary is obviously about the "International" version of the Galaxy SIII.
The USA version of the Galaxy SIII, and the Evo LTE, and the One X all use the faster Qualcomm S4 chip, not the Tegra 3 they are trying to compare against. And "twice the RAM" should generally have nothing to do with performance.
What does this all mean? Generally, that the high-end [USA] Android phones perform easily as well as the new iphone 5.
The android logo on an iPhone story? Really?!?!
Since it is faster than all the other phones I can get all my phone calls done faster. That's the way it works.
Plus, all the video encoding gets done that much faster while I text and drive.
I've decided that my next phone (soon, I hope) is going to be the S3. I'd been holding out with my iPhone 4 for a while, waiting (like many others, I suspect) to see what Apple would wow us with for the iPhone 5. Needless to say, I wasn't that impressed, though to be honest, part of me really didn't expect to be, given that there are only so many innovations they could have come up with. What could they have done? An even bigger screen? NFC? A phone you could roll up? The first two would hardly have been groundbreaking and the latter is tech that doesn't really exist yet.
Still, at the end of the day, I'm sure I could be happy with the 5, but I'm ready to play with a new toy. I've never had an Android device before, but got a chance to play with a tablet and some phones over my vacation, and I liked what I saw.
Captcha: revenues
That ad makes me giggle. Samsung is so deathly afraid of Apple that they are flaunting all of their silly useless(to me... I guess... maybe someone can use them) gadgets in hopes that people will think the I5 is inferior. The numbers will speak for themselves, and Samsung is wasting their advertising dollars... they should save up to pay their patent debts.
Seriously though, I never liked the Mac Vs PC ads, I feel like if you can't sell your product on its own merit, you shouldn't release ads trashing the other guys. When you have an awesome product, people will buy it... when you stoop to trash talk, you're showing your weakness. Apple showed their weakness with the MacVPC ads. Samsung is showing theirs with this.
Also, if Android didn't almost require 2GB of memory to run I'd feel like that is a lot. My 1GB android devices slug up so fast it is silly. If Android had the memory management of iOS, 2GB would scream.
Silly large companies...
There are a ludicrous number of errors here. The summary says that the CPU is clocked at 1.2 GHz, which the screenshot clearly shows is not the case - it's 1 GHz. The quad-core Galaxy S III only has 1GB of RAM, and the LTE variant with 2GB of RAM doesn't have a quad-core CPU. And both the HSPA+ and LTE Galaxy S III's score well above 1600 on Geekbench when actually running on all cores - the test results that are below 1600 and are no-doubt included in this "average" are custom tests run on fewer cores, which is clearly shown if you actually browse the results.
At least in the US, the carriers seem determined to ensure that you upgrade every two years anyway, so it's not like you're going to be stuck with a phone which is all that old. It seems more like "fast enough" is simply a responsive GUI and a generally imperceptible execution time for the kinds of activities you do on a phone. I'm not running CFD models, transcoding movies, or running a popular web service on the thing - I'm tweaking photos, or asking it to make simple calculations my HP48 might do, streaming media or rendering a web page (without flash; thanks Steve).
Now that a couple of generations have past for Android and iOS, the options for switching are getting far more expensive and time consuming. Switch all my media to a new program for syncing - major PITA. Re-buy all my apps (not an insignificant endeavor) for the other platform - $$$. Learn where the fuck the Android/iOS developers decide to put some obscure setting I want to change? Heck, even just setting up my icons and replicating a useful look & feel means dropping at least a couple, if not several, hours.
Megapixels, streaming video chat, resolution, memory amount, memory speed - the numbers mean almost nothing. They mean even less when you can't even run the opposing OS on the hardware. But I suppose everybody has to have a ruler handy at some point.
Is it just my observation, or are there way too many stupid people in the world?
Just wanted to fess up to a typo in the story. I accidentally typed that the iPhone 5 runs at 1.2GHz, meant to type 1.02GHz.
- "Scientia non habet inimicum nisp ignorantem"
Everyone I've seen with an iPhone has a ridicilously huge rubber case protecting the fragile thing.
Well the cases are not all that large that I have seen, but let's proceed as if they were.
Why is thin such a big deal when everyone has a case that makes it NOT thin?
Because the combination of a thinner device + a case is still thinner than the thicker device + a case. If the case, as you claim, is a constant - then thinner really does mean thinner to the user.
However one thing of note with the iPhone 5 is that it has a metal back again. I'm going to drop using a case with the iPhone5 since it should hold up better to drops (I never used a case with the original iPhone and never had an issue). Other people may also choose to stop using cases.
One other factor you forgot about is weight, the new phone is lighter - that does matter to people, I jog for instance and the iPhone 4 really produces a lot of pull in the pocket.
"There is more worth loving than we have strength to love." - Brian Jay Stanley
Needless to say, I wasn't that impressed
Why? It is in fact very impressive hardware; it's simply the case that most of the details about it were leaked beforehand.
I do not know what aspect of the phone would fail to impress compared to current top-end Android phones unless you were into huge screens. The main thing I wanted was a great camera upgrade from the iPhone4; the iPhone 5 has an excellent camera. It runs iOS apps quite quickly, and has a somewhat larger screen without being physically huge.
I just don't understand the pure spec-based comparison that takes place without consideration of what software you might want to run...
"There is more worth loving than we have strength to love." - Brian Jay Stanley
I know I have a lot of money tied up in software for my phone. Whether it be remote control software, or specialty apps which are only available for a premium, or just games I paid for - there's a $100-150+ in software I would have to re-buy. I don't want to have to think about switching my media management over. Not that iTunes isn't a steaming pile of shit on Windows, but I've finally gotten it to work acceptably (most of the time) with my 80+GB of music, 400+GB of movies, audio and ebooks, podcasts, etc. I'm sure there are better managers, but the number of hours required to switch that stuff into another management app just makes my insides curl. I'm doubly tied as I have an iOS tablet.
At this point, the "competitor" from Android would have to be pretty fucking amazingly better to make it worth while to switch, and while the S3 is very nice and there are things about it I like better, it's hard to find a reason for the extra expense and time to switch.
Is it just my observation, or are there way too many stupid people in the world?
One should note that the score given for the SGS3 is an average score from thousands of benchmarks which they range everywhere form 1271 to 2211.
The Iphone 5 however only has a single result, and that's on a phone that is probably not burdened by a bunch of crap which seemingly tends to give really varying results..
I won't trust this before they have at least 250 benchmarks done after the release.
I don't have a case for my 4S, I didn't have one for my 3GS, and I didn't have one for my original iPhone, which I got from my brother when he upgraded to a 3G. My brother also is on his third phone and doesn't use a case. In all that time only one's ever fallen on the ground. My friend asked to hold it, and immediately dropped it onto a concrete floor when I handed it to him. It was the original iPhone. It put a small dent in the corner of the case, but it didn't really damage it. I'd hardly call the device fragile.
The population of iPhone owners seems pretty evenly split between people with cases and people without. I certainly appreciate a device that looks good and feels good in my hand. I'm not really concerned with breaking it since I look after my things. A lot of other iPhone users are the same.
It's really wrong to compare specs between Android and iOS devices directly without considering how the underlying systems are actually used.
For instance, an Android phone needs more memory than an iOS device as it tends to have more background processes. iOS has a tighter control over memory so it simply does not need as much to accomplish most things (unless you start getting into talking about image processing applications).
Also, what about the performance difference between Android apps and iOS apps? Android apps have to rely on a garbage collector to reclaim memory, iOS uses ARC which means memory is reclaimed without that overhead. Not to mention the VM in Android.
Also how many Android apps are written in such a way as to take advantage of all those cores? With so many Android devices still being on 2.x, lots of developers target that spec. iOS developers at worst are targeting about two versions back, currently switching from iOS4 to iOS5 as the lowest level supported - that means use of a LOT of libraries that actually make use of multiple cores for many tasks.
I can see comparing specs from on Android device to another or one iOS device to another, but comparing specs between an iOS device and an Android device seems kind of pointless unless you are giving very specific parameters for a task either might accomplish. Running GeekBench is not really a task a user would do every day...
"There is more worth loving than we have strength to love." - Brian Jay Stanley
There seem to be plenty of S3's in their database that still beat the iP5. There are dual and quad core variants of the S3. Though it doesn't matter because it seems the S3 is actually still faster according to the current, real data on their site.
Yes let's just ignore the fact that all we've heard for years from Android users is how fast the CPU is in their phones and how important it is to them, like it actually matters. Now the shoe is on the other foot it suddenly isn't a reason to buy or upgrade a phone.
The summary is bollocks. The iPhone 5 is faster than the dual-core Galaxy S III. The quad-core Galaxy SIII is faster than the iPhone 5.
Actually, you weren't listening. We buy Android phones because we want to buy Android phones. Got it?
It's like with PCs. Smartphones are so fast nowadays that whatever you buy is good enough to do 90% of the things people want a smartphone to do. So even a 100% speed increase is no compelling reason to upgrade for most people.
-- Cheers!
Smartphones are so fast nowadays that whatever you buy is good enough to do 90% of the things people want a smartphone to do.
That's because smart phones are basically 5 year old PC's with small screens.
But for some people the new network (LTE) will be radically different, especially if the 3G in your area has serious congestion issues.
This might be because samsung is marketing a dual core and quad core phone under the same brand, despite the obvious difference in capability. That is, without a doubt, my biggest gripe with Samsung in the industry. A Galaxy S III should be the same everywhere, or failing that a Galaxy S III DC, or QC should be clearly the same everywhere. Having different versions of the same product is unnecessarily confusing.
Odd considering the dual core snapdragon S4 is faster than the quad core one in almost every single benchmark. Only the really parallel ones (Which face it, never happens on a smartphone) pull ahead, and even then, just.
It's easier to fight for one's principles than to live up to them.
As you say, while the S3 has a consistent edge elsewhere, the iPhone destroys the S3 in the memory bandwidth tests. But those tests are strangely inconsistent, for both devices.
The S3 is a lot slower for sequential read bandwidth (578MB/s vs 1.73GB/s), but actually faster for sequential writes (1.53GB/s vs 1.35/GB/s). It's interesting that write speed is so much faster than reading; usually read speeds are faster than writes (as with the iPhone). This appears common to many Android devices though.
OTOH, the iPhone 5 is ridiculously fast in the stdlib write test - over 6GB/s. Maybe I'm misunderstanding the tests, but I don't see how this result can be three times higher than sequential writes; I'd expect a little slower. Perhaps the iPhone has a large enough cache that the test fits within it?
Why would anyone engrave "Elbereth"?
The camera is almost exactly the same as in 4S?
No it is not.
Up to two stops better performance is a good upgrade. And also there happen to be sample pics on DPReview from an iPhone 4s that match one of the shots the iPhone 5 was demoed with - the iPhone 5 captures detail better. Also I cannot find details on how the 4s camera was constructed but I believe the iPhone 5 is a step up in terms of the lens used.
I have a DSLR and profesional compact cameras too. What I want out of a cell phone camera is an image that does not make me wish I also had a compact camera, and the iPhone 5 meets that goal (really the 4s did as well, but the 5 has a nice boost beyond even that).
"There is more worth loving than we have strength to love." - Brian Jay Stanley
As far as the end user is concerned, whether it is native code or not is irrelevant. The available apps should be compared. If they run fast, that's all the user cares about. Some theoretical e-peen contest about "oh my smartphone has a quad core CPU that is way faster" doesn't matter if the software available consumes far more resources and doesn't run as fast.
I run: Windows, OS X, Linux, FreeBSD. Just because you have a hammer, doesn't mean everything is a nail.
Meanwhile, no matter the hardware specs, iOS will keep being more responsive and iOS phones will keep getting software updates for years after launch. Clock speed and number of cores has stopped being relevant even in phones (it's not really relevant on the desktop any more as well) already.
Note: i've owned two Android phones before switching to iOS.
I apologize for the lack of a signature.
Reference counting (ARC) is EXACTLY a form of garbage collection, not particularly better or worse than any other.
It's not the same as garbage collection, it's exactly what the name says - AUTOMATED reference counting. The moment your code no longer needs an object code is inserted to release it for you. It has no cost over the code you would have written manually.
It is superior to traditional GC because there is no processor time taken in deciding what to collect, no examination of the object tree to find what is still in scope. That means no overhead, and no "pauses" in application flow as a GC fires up to collect things.
You DO realise that ARC imposes a runtime cost which some other garbage collectors do not?
Compile time feature, moron. Even the weak reference zeroing is just code inserted around properties.
You DO realise that ARC is sensitive to some forms of data structure that it cannot collect? (circular references)
It's not "sensitive" to anything, that is simply an artifact of reference counting. By the way, in almost a year of developing multiple applications using ARC you know how many circular references I have seen in real life? Zero. Over-retention is still possible, but cycles are quite rare.
And no, iOS cannot just run multiple apps at the same time to use multiple cores, as iOS only supports specifically
written background tasks
Which then run in the background doing whatever they were designed to do in the background. For instance what do you think Pandora does, genius? What happens when I have Pandora running AND have backgrounded a navigation application? Why in fact they ARE both running.
Of course the system tasks all do run in the background so you really come off as quite ignorant claiming iOS cannot do this arbitrarily when it's a limitation specifically imposed on a subset of applications on the system. A jailbroken iPhone can run any user application in the background simply by a tweak to Launchpad, not the OS or app.
it cannot just continue normal execution of a non-foreground task.
Actually it can for about ten seconds for any app even without jailbreaking. You just have to let the OS know.
You also, I bet, dont know what a process scheduler is,
I've written several thanks. That was a while ago as I moved on from such trivial things.
I do also know what an apostrophe is. Zing!
that addresses your idiocy about primary apps being slower.
Might want to watch the word idiocy when you are so prone to misunderstanding what is being said - I am talking about an foreground application that is not taking full advantage of the system resources. Pretty obviously an application that runs on one core when it could make use of two would be slower than it could be. Duh.
I kind of feel sorry for the corporate IR development teams you worked with
Imagine the concern I feel for whatever company must put up with your constant misunderstandings of technology! I sure hope you are not in charge of any iOS work for sometime to come.
Really, your UID is low enough that you should know better..
My UID is low enough you should have known to do more research rather than spout off on technologies you have not used.
I will allow you the last response, you may either choose the path of wisdom and grovel for forgiveness at your iOS 101 level of understanding, or you may continue down the path of proving beyond all doubt you enjoy staying ignorant. Your choice, but I'll respond no more as I have already spent too much time on your education.
If I were you though I would go watch all of the Stanford introductory iOS course and read some of the iOS documentation to understand how the system works. Oh and find a good white paper on what ARC does, because Damn.
"There is more worth loving than we have strength to love." - Brian Jay Stanley
Right, but the south koreans get quad core, some countries get different amounts of RAM etc.
If all you heard was the launch announcement of 'quad core*' and ignored the asterix of different countries getting different products you'd be confused by the whole thing.
As a developer by the way, this is a fucking nightmare. I work at a university, so we have, every year and every christmas people with phones from all over the world trying to use our mobile app. We need to test on the indian version, the korean version the chinese versions, the hong kong version, the taiwanese version, etc. etc. etc. And we need someone to keep track of what all the different versions are. I know the guys at big blue bubble in town who make mobile games have a big lab but I think they only care about europe and north america rather than everywhere else too.
Try 10 year old PCs. I have a Core 2 Quad here that's pretty much 5 years old and is still 5x faster than any phone in the geekbench data. Linpack is 1000x slower on any ARM than it is on a current x86 too.
Jack Dongarra published a paper how he got about 800 MFlops out of an iPad 2, using only one core, and estimates that about 1.5 GFlops should be possible. The iPhone 5 chip should run a lot faster. And no current x86 does 1500 GFlops.
I've seen that claim. There is a youtube video where linpack on iOS gets 800 MFLOPS. Linpack for android gets 70 Mflops for a comparably clocked ARM. My desktop gets 40 GFLOPS (hence the 1000x - or three orders of magnitude at least). My video cards are rated at over 1 TFLOP dpfp each. My laptop can do about 25 GFLOPS. Perhaps linpack for iOS is more representative of what the hardware can do, but it still doesn't hold a candle to a desktop/laptop.