Purported iPhone 5 benchmark score doubles fastest iDevices, outperforms Android's best

123468

Comments

  • Reply 101 of 145


    Just ran geekbench on my evo lte  running ICS 4.01 (Qualcomm S4 with dual core Krait). I score right around 1560 everytime. This is a 6 month old phone BTW. Anything running 4.1 jellybean with the s4 will smoke the 1601 score.


     


    THe score from the iphone 5 is not that impressive. What is impressive though is that it does it at only 1ghz. I assume the pipeline is shorter and wider than the S4 in order to compete at such a low clock rate. Whats going to be interesting to see is if apple can scale the processor so that future A6's will be able to achieve higher clock rates with the same power consumption.


     


    Most of the time in the CPU world when you have a short and wide data pipeline you have problems scaling it even with shrinking the die size. That will be apples real rabbit from the hat, getting this thing up to 1.5ghz or above.

     0Likes 0Dislikes 0Informatives
  • Reply 102 of 145
    The most interesting point isn't which hardware provides superior performance. The critical point is that (apparently) a 1.02 GHz dual-core processor rivals 1.4 GHz quad-core processors sharing the same 32nm high-k + metal gate LP process. Furthermore, Apple has approximately doubled the performance without compromising power efficiency.

    The technology Apple has (apparently) incorporated into the Apple A6 SOC is almost unprecedented. The technology in the Apple A6 SOC will allow Apple to scale their processors to a degree that is virtually unfathomable. Apple can now far exceed the current roadmap for existing processors while competitors scramble for an advantage such as using ARM Cortex A-15 based processors and incrementally improving process technology.

    If Apple thought such was necessary they could almost certainly easily double performance by creating a quad-core version of the Apple A6 SOC. If Apple though such was necessary they could almost certainly improve performance by increasing the clock speed of the Apple A6 SOC. Apple has considerable room to consider their next steps.

    ***************************************************************************************************************************************************************************************************************************

    If the Geekbench benchmarks are correct the Apple A6 processor is the processor story of the decade. Any knowledgeable technology pundits should be gushing over the Apple A6 which likely demonstrates the paucity of hardware engineering knowledge amongst the technorati.
     0Likes 0Dislikes 0Informatives
  • Reply 103 of 145

    Quote:

    Originally Posted by mercury99 View Post


    IPhone 5  1601


    http://browser.primatelabs.com/geekbench2/1030202   


     


    2-core US Galaxy S3  (not overclocked):  1918


    http://browser.primatelabs.com/geekbench2/1034550


     


     


    4-core International Galaxy S3 (not overclocked):  2087


    http://browser.primatelabs.com/geekbench2/1038481


     


     


    4-core Galaxy Note 2 (not overclocked):  2987


    http://browser.primatelabs.com/geekbench/103874



     


    Browsing around, I find some EXTREMELY inconsistent and questionable results.  Just hitting some high and low points.  Anywhere from 1 to 4 cores and speeds of 1400 and 1800.  How does a single core 1400 double the results of a 4 core 1400?  Either something is wrong w/Geekbench or people are spoofing numbers.  browse around after following one of these links. 


     


    1857 (1 core, 1400) - 950 (4 core, 1400) - 2433 (4 core, 1800) - 872 (2 core, 1400)

     0Likes 0Dislikes 0Informatives
  • Reply 104 of 145


    The best analysis would be Intel vs AMD.  Even with the latest gen 8 core AMD CPUs running at 3GHZ, it's still no matched for Intel Core i5/i7 running 2/4 cores and under 3GHZ in real world applications performance.

     0Likes 0Dislikes 0Informatives
  • Reply 105 of 145
    anonymouse wrote: »
    It is a lot like Microsoft's actions. And a lot, as mentioned above, like the Skyhook incident.

    Basically, and contrary to derekmorr's attempt to portray it as an innocuous and benevolent act on Google's part, what's going on here is that Google is telling companies, if you want to be part of the "OHA", if you want access to information we give our OEM's, if you want to be able to use Android branding, not only do your "Android" phones have to comply with our arbitrary demands (such as not using Skyhook for location data), but thou shalt not make any other phones based on open source code, even if you don't want to brand those as Android, or we'll kick your teeth in, kneecap you, and pull your Android certification so fast you won't have a chance to blink.

    Well, interestingly, since they aren't going after forks of Android, but members of the OHA, they should just rename it the "Closed Android Alliance" and I wouldn't have any issues with that. In the end, you're either a Google partner or you're "other", but you can't be both. Here's what I find ironic: Microsoft doesn't nor could they get away with preventing Windows Phone handset makers from not making handsets for other operating systems. in other words, we live in a world where Microsoft seems to be the more permissive licensor of mobile operating systems. I wouldn't have pictured that 15 years ago.
     0Likes 0Dislikes 0Informatives
  • Reply 106 of 145

    Quote:

    Originally Posted by SSquirrel View Post


     


    I see a ton of people thieving software from the Android market and reselling it as their own on Alibaba's market.  This is a completely separate issue yes, but it is also wrong.  Google is reminding them of agreements they made and are expected to honor to retain membership.  I don't see what the problem with that is.  The OHSA is what is allowing OEMs to get the 3 month early access to Android updates so they can shorten the lead time on new updates for their phones right?  It behooves Google as well as their OEM members to abide by the policies so the consumer actually gets more timely updates. 


     


     


    More specifically, they don't want those companies making other phones based on offshoots of THEIR software also being memebrs of the OHSA.  Yes Android is ostensibly open source, but there are parts that are still protected and the software is not shared during the creation stages.  They could choose to leave the OHSA and still go thru w/their plans relating to Alibaba, but it would have other consequences for their Android based phones, like access to the Google products IIRC.  No one is forced to sign these agreements, but if they do, they are expected to abide by them. 


     


    Not the same as MS, MS was forcing the price of a Windows install for every computer OEMs sold, even if Windows was not on the system.  Android is free.  That kills your entire point right there.



     


    Is Android open or is it not? Is Google playing hardball, in extra-legal ways to protect it's IP or is it not. The answers are obvious.


     


    So, anyone with a shred of honesty can stow their carping on litigation and stop BS'ing us all about openness. I for one am tired of the lies, and it's reached a point where only the delusional can still argue that Android is open or that Apple is somehow a bad guy for protecting its IP, unlike the Google which is all love and sharing with everyone. Bull.

     0Likes 0Dislikes 0Informatives
  • Reply 107 of 145
    cnocbuicnocbui Posts: 3,613member


    Wired is stating that the IP5 benchmarks fall short of the SGS III, which they give as scoring 1723.


     


     


    Quote:


    Although the iPhone 5 crushed its iOS predecessors and most leading Android devices, it fell short of the the quad-core Samsung Galaxy S III, which scored a 1723.



     




    http://www.wired.com/gadgetlab/2012/09/first-iphone-5-benchmarks/

     0Likes 0Dislikes 0Informatives
  • Reply 108 of 145
    I must admit, I'm in awe of (some) AI's commenter's abilities to spin anything to fit an anti-Google agenda.
    anonymouse wrote: »
    "Voluntarily", you mean after threats from Google?

    No, I mean back in 2009, when Acer voluntarily joined the Open Handset Alliance: http://www.acer-group.com/public/News/2009/20090601.htm
    In other words, in Google Newspeak, the "Open Handset Alliance" is actually closed to Aliyun or whatever Android forks exist.

    No, the OHA is not closed to Aliyun. In fact, Andy Rubin has asked Alibaba to join the OHA and said that they'd help Alibaba fix their Android compatibility issues:

    "So if you want to benefit from the Android ecosystem, then make the choice to be compatible. Its easy, free, and we'll even help you out. But if you don't want to be compatible, then don't expect help from OHA members that are all working to support and build a unified Android ecosystem."

    Source: https://plus.google.com/112599748506977857728/posts/hRcCi5xgayg (and since I know Google sites scare some people on here, he's a non-Google site: http://www.ibtimes.com/articles/385042/20120917/google-against-aliyun-os-alibaba-andy-rubin.htm)
    I don't see the problem of Acer wanting to support Android and "other", but Google does because it supposedly claims it doesn't want Android "fragmentation" to occur (a specious argument given that Android handsets are not always upgradable to newer Android releases, but that's a different topic). I understand what the OHA original intent was: Google didn't want their manufacturing partners to fork and create completely different flavors of Android out of a self interest-driven need for product differentiation, despite Android's supposed open source nature

    The problem is that Acer agreed not to do this. Back in 2009. Acer is free to leave the OHA and ship Aliyum devices. Or they can work with Alibaba to fix Aliyum's compatibility problems.

    Android was intended to be flexible, but the community has an interest in maintaining compatibility. Non-OHA members can use AOSP however they wish, but they won't benefit from working with OHA members. Google uses the benefits of OHA membership to hold together most of the Android ecosystem. No one is forced to join, but there are benefits to doing so, and it's a lot of work to replicate the ecosystem if you're not a member (Amazon is the only company that's tried and pulled it off). I'd encourage you to take a look at the AOSP FAQs http://source.android.com/faqs.html.

    And what do you mean by "Android's supposed open source nature" ? It's an open-source project published under OSI-approved licenses. The mailing lists, source repo, code review system, restore images, bug tracker, etc are all open. That's an open project.
     0Likes 0Dislikes 0Informatives
  • Reply 109 of 145
    waybacmac wrote: »
    I may be late to the game in this discussion but I feel that many people have missed a key point in all this chip versus performance debate. Android phones are using basically off-the-shelf processors, though I believe some are somewhat modified. Apple is using a highly customized, virtually unique chip optimized to do what the phone needs to do. Yes, this means there can be a massive performance difference. A dual-core chip can be faster, more efficient, than a quad-core.

    I don't think it's that big a stretch to use high performance car engines to illustrate the point. A 2012 Ferrari California 30 F1 is a mighty street machine. It is a production car off the shelf at Maranello. It has a naturally aspirated V8, 4297cc, with a rated power of 489.7 PS @ 775rpm. But as impressive as these specs are, they pale in comparison to a truly custom built Ferrari F1 Tipo 056 racing engine used in Formula 1. A naturally aspirated V8 of 2400cc (about half as big as the California) putting out about 735 PS (about a third more) @ 18,000rpm. If such a wide performance gap can be accomplished with a mechanical engine, it certainly can be done with a digital one. It's all a matter of optimizing the design to maximize the performance for a particular set of desired functions.

    Your analogy reminded me of a Ferrari California vs a Cadillac Sportswagon:
     0Likes 0Dislikes 0Informatives
  • Reply 110 of 145

    Quote:

    Originally Posted by derekmorr View Post



    I must admit, I'm in awe of (some) AI's commenter's abilities to spin anything to fit an anti-Google agenda.

    No, I mean back in 2009, when Acer voluntarily joined the Open Handset Alliance: http://www.acer-group.com/public/News/2009/20090601.htm

    No, the OHA is not closed to Aliyun. In fact, Andy Rubin has asked Alibaba to join the OHA and said that they'd help Alibaba fix their Android compatibility issues:

    "So if you want to benefit from the Android ecosystem, then make the choice to be compatible. Its easy, free, and we'll even help you out. But if you don't want to be compatible, then don't expect help from OHA members that are all working to support and build a unified Android ecosystem."

    Source: https://plus.google.com/112599748506977857728/posts/hRcCi5xgayg (and since I know Google sites scare some people on here, he's a non-Google site: http://www.ibtimes.com/articles/385042/20120917/google-against-aliyun-os-alibaba-andy-rubin.htm)

    The problem is that Acer agreed not to do this. Back in 2009. Acer is free to leave the OHA and ship Aliyum devices. Or they can work with Alibaba to fix Aliyum's compatibility problems.

    Android was intended to be flexible, but the community has an interest in maintaining compatibility. Non-OHA members can use AOSP however they wish, but they won't benefit from working with OHA members. Google uses the benefits of OHA membership to hold together most of the Android ecosystem. No one is forced to join, but there are benefits to doing so, and it's a lot of work to replicate the ecosystem if you're not a member (Amazon is the only company that's tried and pulled it off). I'd encourage you to take a look at the AOSP FAQs http://source.android.com/faqs.html.

    And what do you mean by "Android's supposed open source nature" ? It's an open-source project published under OSI-approved licenses. The mailing lists, source repo, code review system, restore images, bug tracker, etc are all open. That's an open project.


     


    Spin it however you like. It's not open by any standard that is meaningful. Google has turned the concept of open source on it's head to mean, "you'll do what we tell you to do with it, or we'll crush you like a fly."

     0Likes 0Dislikes 0Informatives
  • Reply 111 of 145


    I see a LOT of qrguing back and forth about "whether benchmarks matter".  To me, the answer is YES, bot only when comparing simmilar systems. 


     


    Just like Horsepower matters, but I dont drive an engine around, I drive a car,  you cannot say which car is faster by just looking at their horsepower,  a 350 HP Civic is going to be faster than a 400HP Suburban. 


     


    What this means to the smartphone scene is:


     


    From most/many performance comparisons between the Iphone 4s and the Galaxy S3,  it's pretty obvious that the two phones perform approximatly the same (and, depending on the specific comparison, the Iphone 4s may actually be faster).


     


    Now think about the numbers.  one of the current FASTEST android phones available now is approximatly the same performance(real world) as the 1 year old Iphone 4s.


     


    The benchmarks for the S3 are more than DOUBLE the Iphone 4s,  YET the Iphone 4s is it's equal (if not faster than the S3). Now think about this,  Apple just improved the performance of the Iphone by ~2.5 times. 


     


     


     


     


     


     


     


    http://www.youtube.com/watch?v=o8Qlls346Hk


    http://www.youtube.com/watch?v=_ItI9qxwioE


    http://www.youtube.com/watch?v=msPHKs8cmbs


    http://www.youtube.com/watch?v=4VzAupqQ6Bo&feature=related


     


     


    These videos show the "back-and-forth" between the Iphone 4s and the Galaxy S3


    and that they are approximatly equal in performance

     0Likes 0Dislikes 0Informatives
  • Reply 112 of 145
    anonymouse wrote: »
    Spin it however you like. It's not open by any standard that is meaningful. Google has turned the concept of open source on it's head to mean, "you'll do what we tell you to do with it, or we'll crush you like a fly."

    Amazon's Kindle Fire and Baidu's Yi product lines solidly refutes your argument. Android is open-source. But if a company joins an alliance and agrees not to ship incompatible variants, then they can't cry foul later.
     0Likes 0Dislikes 0Informatives
  • Reply 113 of 145
    anonymouse wrote: »
    Spin it however you like. It's not open by any standard that is meaningful. Google has turned the concept of open source on it's head to mean, "you'll do what we tell you to do with it, or we'll crush you like a fly."

    "Any meaningful way"? Being able to look at the source code and edit it in any way possible is not meaningful enough?
     0Likes 0Dislikes 0Informatives
  • Reply 114 of 145
    mac_128mac_128 Posts: 3,454member
    Why do benchmarks matter at all for smartphones at this point?

    I'm a diehard IPhone user, will never use anything else as long as Apple doesn't screw up the iCloud sync services with my Mac, and iDevices.

    But seriously, we're talking about the Internet here, schedules, and a phone. How fast do any of these things have to be? LTE is going to make a difference in browsing the Internet for customers who have reliable access to it, but the biggest bottleneck is the wireless speed, and signal. And for everything else, the iPhone 4 was, and is about as fast as I need one of these things to be. I honestly don't see much difference in the 4S, though I'm sure the 5 will feel peppier. The original iPhone was too slow. It felt slow. The 3G did nothing to improve upon that. The 3GS was a nice speed bump, but still sluggish. The 4 hit a sweet spot, and unless Apple bogs it down with dibilitating code under iOS 6, it should keep running at its speedy usual. But honestly, as long as its fast enough to do what it's advertised to do ... who cares?

    Its not like I'm 3D modeling, or crunching massive photoshop files on my phone. Benchmarks are for computers, where certain tasks require massive amounts of speed to make them tolerable Sure it's nice to know what a person is working with on a phone, but in the real world, most consumers aren't going to care! or even notice the difference. Just like the Mac v. PC wars, marketers are going to make an issue of it, since its something tangible to promote. But in the end it means nothing via-a-via the end user experience. And, as has been pointed out earlier, ultimately it's the software that will determine how quick the hardware can execute anyway, no matter what it's capable of.
     0Likes 0Dislikes 0Informatives
  • Reply 115 of 145

    Quote:

    Originally Posted by derekmorr View Post





    Amazon's Kindle Fire and Baidu's Yi product lines solidly refutes your argument. Android is open-source. But if a company joins an alliance and agrees not to ship incompatible variants, then they can't cry foul later.


     


    Really, are they allowed to call it Android? No, they aren't. Oh, right, because they didn't kowtow to Google's unilaterally imposed terms and don't include the "proprietary" parts. Sorry, but an operating system that is defined by it's "proprietary" parts, and conditions that forbid handset makers to use third party services like Skyhook isn't open, unless we go through the looking glass or down the rabbit hole.


     


    The "Open Handset Alliance" isn't open, nor is it an alliance. It's a sham the very name of which is meant to deceive people into thinking that it is both of those things. It would be more accurately named the Google Handset Hegemony. "Open" and "Alliance", in this case, are just a bit of hand-waving on Google's part, Sort of like "German Democratic Republic" or "People's Republic of China".

     0Likes 0Dislikes 0Informatives
  • Reply 116 of 145
    anonymouse wrote: »
    Really, are they allowed to call it Android? No, they aren't. Oh, right, because they didn't kowtow to Google's unilaterally imposed terms and don't include the "proprietary" parts. Sorry, but an operating system that is defined by it's "proprietary" parts, and conditions that forbid handset makers to use third party services like Skyhook isn't open, unless we go through the looking glass or down the rabbit hole.

    The "Open Handset Alliance" isn't open, nor is it an alliance. It's a sham the very name of which is meant to deceive people into thinking that it is both of those things. It would be more accurately named the Google Handset Hegemony. "Open" and "Alliance", in this case, are just a bit of hand-waving on Google's part, Sort of like "German Democratic Republic" or "People's Republic of China".

    That's really going off the deep-end.

    You're right. They can't call it Android if it doesn't pass the CTS (http://source.android.com/faqs.html#is-compatibility-mandatory) -- because it isn't Android if it isn't compatible. There's nothing in the Open Source Initiative's definition that requires open-source projects to let anyone use their trademarks, or companies to help their competitors.

    I find it rather curious that you care so much about Android since you apparently don't use it. In any event, it's clear that I can't have a rational conversation with you since you twist and distort anything I say. I'm done wasting my time. Enjoy your evening.
     0Likes 0Dislikes 0Informatives
  • Reply 117 of 145
    jragostajragosta Posts: 10,473member
    mcrs wrote: »
    One of the most recent RDF was when SJ explained the "death grip" issue of Iphone. He was faulting the users for holding the phone the wrong way and the fact that signals did fluctuate from time to time- only he neglected to mention that the lost signals are happening so frequently and so predictably for certain batches of Iphone, and the fact that Apple engineers were ever so quietly fixing the antenna trouble by modifying its design. SJ should just admitted that it was just a design issue instead of having to warp Iphone's flaw with his usual RDF. 

    This infamous RDF moment was also captured in one publication when an iPhone user Aram from Arizona State had written a short email to SJ:
    Hi Mr. Jobs,


    I love my new iPhone 4 (nice work) but when I put my hand on the steel bands I lose all reception.

    It appears to be a common issue. Any plans to fix this?



    SJ responded to this email in an even briefer fashion, he wrote:
    Just avoid holding it that way. 

    2010 wants their post back.

    That wasn't a real issue even 2 years ago. Why are you bringing it up now?

    ssquirrel wrote: »
    Browsing around, I find some EXTREMELY inconsistent and questionable results.  Just hitting some high and low points.  Anywhere from 1 to 4 cores and speeds of 1400 and 1800.  How does a single core 1400 double the results of a 4 core 1400?  Either something is wrong w/Geekbench or people are spoofing numbers.  browse around after following one of these links. 

    1857 (1 core, 1400) - 950 (4 core, 1400) - 2433 (4 core, 1800) - 872 (2 core, 1400)

    And that's the problem with comparing ANY benchmarks from multiple sources (especially anonymous ones). For a benchmark to have any chance of being useful, it has to be run under controlled conditions. Simply pulling benchmarks off the 'net from multiple different sources is largely useless.

    anonymouse wrote: »
    Is Android open or is it not? Is Google playing hardball, in extra-legal ways to protect it's IP or is it not. The answers are obvious.

    So, anyone with a shred of honesty can stow their carping on litigation and stop BS'ing us all about openness. I for one am tired of the lies, and it's reached a point where only the delusional can still argue that Android is open or that Apple is somehow a bad guy for protecting its IP, unlike the Google which is all love and sharing with everyone. Bull.

    Google is fully open in a marketing-speak sense.

    Other than that, it's tightly controlled. Heck, Google didn't even release the source code of some versions.
     0Likes 0Dislikes 0Informatives
  • Reply 118 of 145
    jragostajragosta Posts: 10,473member
    derekmorr wrote: »
    That's really going off the deep-end.
    You're right. They can't call it Android if it doesn't pass the CTS (http://source.android.com/faqs.html#is-compatibility-mandatory) -- because it isn't Android if it isn't compatible.

    Yes, that's the theory.

    Unfortunately, in the real world, it's not true. Fragmentation is a massive problem for Android and there are many, many examples of things that don't work on some phones that are labelled Android.
     0Likes 0Dislikes 0Informatives
  • Reply 119 of 145

    Quote:

    Originally Posted by derekmorr View Post





    That's really going off the deep-end.

    You're right. They can't call it Android if it doesn't pass the CTS (http://source.android.com/faqs.html#is-compatibility-mandatory) -- because it isn't Android if it isn't compatible. There's nothing in the Open Source Initiative's definition that requires open-source projects to let anyone use their trademarks, or companies to help their competitors.

    I find it rather curious that you care so much about Android since you apparently don't use it. In any event, it's clear that I can't have a rational conversation with you since you twist and distort anything I say. I'm done wasting my time. Enjoy your evening.


     


    I care about large corporations lying. And I care about people playing semantic games on their behalf to perpetuate those lies, whether they be paid minions or true believers.


     


    Oh, it's not Android, but it's android and if we sow a little semantic confusion, ignore the very real differences between Android and android, and pretend that the "OHA" actually follows "standards" instead of Google's fiats, wave around some nice, feel-good, open-source verbiage, no one will notice what hypocrites we are. I'll give Google credit that they have mastered the art of bullshit, and getting people to swallow it. If pointing that out is "twisting and distorting", and I'm sure it is on your side of the looking glass, so be it.


     


     


    Quote:


    “When I use a word,’ Humpty Dumpty said in rather a scornful tone, ‘it means just what I choose it to mean — neither more nor less.’



    ’The question is,’ said Alice, ‘whether you can make words mean so many different things.’



    ’The question is,’ said Humpty Dumpty, ‘which is to be master — that’s all.”




    ? Lewis Carroll, Through the Looking Glass


     0Likes 0Dislikes 0Informatives
  • Reply 120 of 145
    mcrsmcrs Posts: 172member


    I think you're right about the Iphone undocumented feature, the death grip. Even the former Russian President Medvedev [now a Russian Premier] thought it was a cool feature. 


     


     



     


     


    Quote:

    Originally Posted by Tallest Skil View Post


    ????


     0Likes 0Dislikes 0Informatives
Sign In or Register to comment.