With iPhone 8, Apple's Silicon Gap widens as the new A11 Bionic obliterates top chips from...

135

Comments

  • Reply 41 of 83
    tjwolf said:
    cincymac said:
    birko said:
    Salivating at the possibility of A12 in macbook
    Maybe Apple has been working on a MacBookAir with A12 ... :)
    I really doubt an ARM based Mac will happen in any near future.  The problem isn't the power - the problem is software.  The Mac is used to *create* software for many platforms - not just iOS and Mac.  I use it to create Java-based applications that run on all sorts of platforms.  Others use Macs to write Python- C, C++ based applications.  None of these languages/tools will be available (at least initially) on an ARM-based Mac.  And then there are of course the commercial third-party applications - x86 based Macs already occupy a pretty small market  compared to Windows-based machines.  An ARM-based Mac would occupy an even smaller niche - how many 3rd-party vendors would bother writing/selling applications for it?

    I guess if an ARM-based Mac could run iOS apps and help the open source community port development tools (e.g. JVM, Apache stuff, etc.), there's a chance - but it'll take a lot of time.
    I agree. I used to think one was coming but no more. It's iOS / ARM for mainstream computing and macOS / Intel for more heavy-lifting computing.
    williamlondonwatto_cobra
  • Reply 42 of 83
    tmay said:
    Aren't the A11's scores better because the iPhones are running at a lower resolution, or am I missing something?
    Sucks for Samsung to have 4K resolution in a handheld device that no human eye can resolve; i.e., it's a bad design decision to drive that many pixels, but it was done for marketing to people who purchase based on a list of specs.
    That ultra-high resolution seems to be a great selling point. The Youtubers are always claiming how wonderful the Samsung displays are and they point out to some insanely high ppi as a result. I thought the human eye was limited to a certain resolution but I must be wrong because some of those Android smartphones have displays that are far beyond Retina displays. With my poor eyesight they'd be of no use to me but it's likely most young people have very good eyesight. Then again, it also depends on what a person is using their smartphone for. You don't need high resolution to do Snapchat, Twitter or browsing the internet. One thing they do say is rather important for outdoor use and that's the amount of nits in a display. It's being shown the Galaxy Note 8 has twice the nits of the iPhone X which is quite amazing. Twice as bright is a huge leap. I don't know if the brighter display eats up a battery a lot faster but it's still something to think about.
  • Reply 43 of 83
    radarthekatradarthekat Posts: 3,842moderator
    "The same way that iOS has positioned Apple as "the Microsoft" of premium mobile devices, its Ax series chips have made it "the Intel" of mobile. In fact, the volume and value of iOS hardware actually make Apple far larger than the WinTel PC empire ever was."

    This was exactly my point back in late 2015 as Apple stock was languishing.  In fact, my writing on the subject was picked up by Phillip Elmer Dewitt and published without edits in his Fortune article about the law of large numbers.  Here's that again...

    Law of Large Numbers Debunked

    Among the arguments why Apple shares cannot outperform the market or its peers has been the oft repeated law of large numbers; the claim that Apple is too big to meaningfully grow and that its market cap, at over $600 billion, is so big that there aren't enough investment dollars to move the needle.

    But October 2015 earnings unwittingly provided an irrefutable counter argument by taking the combined market caps of GOOGL and AMZN to $800 billion.  The market seems to have no trouble adding $80 billion to these two companies, whose combined profits are a fraction of Apple's, but won't allow the same for a single company.  It was Microsoft's year 2000 valuation, north of $600 billion at the peak of the dotcom bubble and stagnant for the decade thereafter, that has since been used as the poster child for what happens to the company with the world's highest market cap.  The street is convinced that will be Apple's fate. 

    What the market doesn't seem to understand is that vertically integrated Apple, in terms of the profits it generates and markets it addresses, is equivalent to the entire PC industry of the 1990s, including MSFT, Sony, Toshiba, IBM's PC division, Compaq, HP, and all the other PC makers.  Adjusted for inflation, MSFT's year 2000 valuation alone would today be $850 billion, against Apple's current $650 billion.  How much higher when you add in all the PC makers from 2000?

    Apple's market cap, given the scope of its business, and adjusted for inflation, is very conservative.  As usual, the market is wrong.

    -----

    And here we are, with Apple's market cap within $25b of $850b.  Damn, sometimes I split the arrow!
    tmay2old4funwatto_cobra
  • Reply 44 of 83
    brucemcbrucemc Posts: 1,541member
    cropr said:
    It's nice that the A11 outperforms the CPUs of other vendors, but the main question is: does it really matter?  I am using my iPhone 6s for calling, web browsing, emailing, checking my calendar, taking notes, paying, chatting, taking photos, checking bank account, playing music, checking public transport timetable.  And for none of these actions my iPhone 6s feels sluggish.   How impressive the A11 may be, I am not at all convinced that I will be more productive with a faster CPU.  
    One might say that the A11 is needed for face recognition, but I fail to see the advantage of face recognition versus TouchID.  I am pretty sure ApplyPay with face recognition will be more cumbersome ans slower to use than the current ApplePay.  The fact that I have to turn the phone to my face is inherently slower than putting my finger on the home button.
    If smart device management, Siri, machine learning mean something to you, yes it matters. If 4K video, augmented reality, Metal 2 and 3D gaming means something, yes it matters again. Apple didn't create the A11 for execs and managers send animated poop to their subordinates, A11 means a forward jump for the whole iOS platform, just like the A7 in iPhone 5s was a forward jump to 64-bit.
    Fully agree.  Anyone who claims that current smartphones have more than enough (CPU/GPU) power are only looking into the past, not the future.  Cropr's list there is probably true for me as well right now (I don't do much mobile gaming) - but already we have AR that requires reasonable power (hence why it is A9 and above, which is already more powerful than the vast majority of Android handsets in existence).  

    Looking forward, it doesn't take much imagination on what new functionality could come to the smartphone that requires more advanced silicon than is in the A11.  How about Siri utilizing neural chip so that voice recognition and on-board functions can be done locally without going to the cloud.  Machine learning on device for instantaneous context awareness, and less reliance on cloud for some functions (with inherent increase in privacy).

    Due to the mobile nature of smartphones, the sensors that they already have, and that the functions they do increases - smartphones will continue to require & utilize more silicon performance when compared with the plateau of "most" PC use cases.
    tmaymizhou2old4funwatto_cobra
  • Reply 45 of 83
    cincymac said:
    birko said:
    Salivating at the possibility of A12 in macbook
    Maybe Apple has been working on a MacBookAir with A12 ... :)
    there might be some truth there.  Its interesting to me that Apple putting out a minor tweak of Sierra this year. Maybe this is because the bulk of the macOS team is working on a port. 
    tmaywatto_cobra
  • Reply 46 of 83
    Tim obviously will not be up to the challenge.   But, will TIm's Apple?
    I much prefer a Company that is not reliant on a single person/founder for its future success.

    Look what has happened to firms whose founders brought them incredible success.  Their Companies were designed to fulfill the initial vision, not to expand on it.

    Jobs' greatest accomplishment wasn't the Mac, or iPod, or iPad or even the iPhone, it was building a culture that made those products possible.
    tmaySolimizhouRayz2016radarthekatchiawatto_cobra
  • Reply 47 of 83

    birko said:
    Salivating at the possibility of A12 in macbook
    Apple won't use the Ax in Macs until they outperform Intel's best chips.  Putting Ax processors in MacBooks will only diminish Apple's buying power with Intel.

    Besides, a new class of Ax processors would be required to operate MacOS applications.  You don't want to force developers to recompile for a new processor design when all of their experience/expertise is firmly established in existing technology.  The exception to that statement is a processor that can operate two different OSs ala MacOS 9/MacOS X.
    morkywilliamlondonwatto_cobra
  • Reply 48 of 83

    birko said:
    Salivating at the possibility of A12 in macbook
    Apple won't use the Ax in Macs until they outperform Intel's best chips.  Putting Ax processors in MacBooks will only diminish Apple's buying power with Intel.

    Besides, a new class of Ax processors would be required to operate MacOS applications.  You don't want to force developers to recompile for a new processor design when all of their experience/expertise is firmly established in existing technology.  The exception to that statement is a processor that can operate two different OSs ala MacOS 9/MacOS X.

    You need to factor in the possibility that Apple's buying power won't diminish...but rather evolve. "We wont buy any more x86 chips, but will purchase all the A15 chips you can make".
    edited September 2017 jeff_cook2old4funwatto_cobra
  • Reply 49 of 83
    irnchriz said:
     cropr said:
    It's nice that the A11 outperforms the CPUs of other vendors, but the main question is: does it really matter?  I am using my iPhone 6s for calling, web browsing, emailing, checking my calendar, taking notes, paying, chatting, taking photos, checking bank account, playing music, checking public transport timetable.  And for none of these actions my iPhone 6s feels sluggish.   How impressive the A11 may be, I am not at all convinced that I will be more productive with a faster CPU.  
    One might say that the A11 is needed for face recognition, but I fail to see the advantage of face recognition versus TouchID.  I am pretty sure ApplyPay with face recognition will be more cumbersome ans slower to use than the current ApplePay.  The fact that I have to turn the phone to my face is inherently slower than putting my finger on the home button.
    Not sure about how you use Apple Pay but I find the following happens:

    When I double click to launch it I often end up on my home screen and have to press sleep then try again (I find it works better on my Apple Watch although thats a bit awkward to use with fixed readers).  Anyhoo, I have to look at the screen to make sure its ready to Pay then touch my finger and present the phone to the reader, so having to look at the phone to pay isn't going to be any different, maybe a bit more successful to invoke as Im not going to end up on the home screen.
    You're not supposed to click the home button when using Apple Pay with your phone. You just hold it up to the terminal and place your finger on the home button when ready. 
    Rayz2016chiawatto_cobra
  • Reply 50 of 83

    birko said:
    Salivating at the possibility of A12 in macbook
    Apple won't use the Ax in Macs until they outperform Intel's best chips.  Putting Ax processors in MacBooks will only diminish Apple's buying power with Intel.

    Besides, a new class of Ax processors would be required to operate MacOS applications.  You don't want to force developers to recompile for a new processor design when all of their experience/expertise is firmly established in existing technology.  The exception to that statement is a processor that can operate two different OSs ala MacOS 9/MacOS X.
    There is no reason to put it in the smaller MacBook.   They will put it iPad Pro and continue to build out iOS for iPad to have the same desktop features as MacOS.  Eventually it will to replace the need for MacBook as iPad Pro gains more MacOS ported features. MacBook and MacBook Pro can continue along in parallel as this happens until it makes no sense to buy them anymore.    Give me a trackpad and adjustable hinge on a smart keyboard that attaches to the 12.9 inch and a few more Apple native MacOS apps ported over and I'm good. Terminal, etc.  They already are taking the steps to go in that direction.   Dock, Multitasking, Files (aka Finder lite). 

    MacOS is already feeling like it is falling behind in features in some way.  HomeKit, lack of App availability.  I think if Apple had an opportunity to get people off MacOS and only one common iOS platform (TV, iPhone, iPad, iPad Pro (Desktop replacement) ), then they would take it. If they can get people to move from MacOS on MacBook to iOS on iPad Pro by providing enough Desktop features in iOS, then I think they would be in a much stronger market position on the Desktop.   More Apps, more secure OS, less SW development efforts using a unified SW platform.

    I dont have a magic 8 ball, however indications do seem like they are heading in this direction.  I have no pressing reason to upgrade to High Sierra. It just not that much of a feature release for me and I suspect others.  The MacOS team has obviously focused on something else over the past year. 
    edited September 2017 tmaymorkyStrangeDays
  • Reply 51 of 83

    birko said:
    Salivating at the possibility of A12 in macbook
    Apple won't use the Ax in Macs until they outperform Intel's best chips.  Putting Ax processors in MacBooks will only diminish Apple's buying power with Intel.

    Besides, a new class of Ax processors would be required to operate MacOS applications.  You don't want to force developers to recompile for a new processor design when all of their experience/expertise is firmly established in existing technology.  The exception to that statement is a processor that can operate two different OSs ala MacOS 9/MacOS X.

    You need to factor in the possibility that Apple's buying power won't diminish...but rather evolve. "We wont buy any more x86 chips, but will purchase all the A15 chips you can make".
    very good point.  The color of Apple's money is the same, especially when Intel is chomping at the bit to build chips for iOS based devices.  I'm pretty sure Apple orders a lot more Ax chips vs Intel native chips.  I think if Apple put Intel in a position where they could build
    Ax chips for Apple's desktop platform (whatever OS that is) vs Intel native chips for Apple's desktop platform, they would gladly take their money. In the end, I don't think Intel will care what Apple is putting in their Desktop platform (be it iOS or Mac) as long as Apple is buying silicon from Intel when they compare the revenue opportunity.   For example,If Intel tries to raise the prices on x86 chips for Apple because Apple is not buying enough and Intel is making Ax chips for Apple, then Apple  could just threaten to move their Ax business away from them for leverage until they come to their senses.    I don't think Intel wants to push that.   If Apple begins to move away from Intel on the desktop, then Intel will need to get more competitive and make it more attractive to stay on x86.  Trying to raise prices to Apple, isn't going to work.  That will just drive the opportunity to do business and the money away even further.  When your customer is moving away from your product the solution is not to raise prices.   What are tying doing to raise the prices 10x??  What about AMD are they going to raise prices 10x when Apple comes calling for a x86 replacement because Intel jacks up their prices out of protest? ...oh and btw, an opportunity Global Foundries to pick up Ax building business too?  Not going to happen 
    edited September 2017 watto_cobra
  • Reply 52 of 83
    morkymorky Posts: 200member
    jeff_cook said:

    birko said:
    Salivating at the possibility of A12 in macbook
    Apple won't use the Ax in Macs until they outperform Intel's best chips.  Putting Ax processors in MacBooks will only diminish Apple's buying power with Intel.

    Besides, a new class of Ax processors would be required to operate MacOS applications.  You don't want to force developers to recompile for a new processor design when all of their experience/expertise is firmly established in existing technology.  The exception to that statement is a processor that can operate two different OSs ala MacOS 9/MacOS X.
    There is no reason to put it in the smaller MacBook.   They will put it iPad Pro and continue to build out iOS for iPad to have the same desktop features as MacOS.  Eventually it will to replace the need for MacBook as iPad Pro gains more MacOS ported features. MacBook and MacBook Pro can continue along in parallel as this happens until it makes no sense to buy them anymore.    Give me a trackpad and adjustable hinge on a smart keyboard that attaches to the 12.9 inch and a few more Apple native MacOS apps ported over and I'm good. Terminal, etc.  They already are taking the steps to go in that direction.   Dock, Multitasking, Files (aka Finder lite). 

    MacOS is already feeling like it is falling behind in features in some way.  HomeKit, lack of App availability.  I think if Apple had an opportunity to get people off MacOS and only one common iOS platform (TV, iPhone, iPad, iPad Pro (Desktop replacement) ), then they would take it. If they can get people to move from MacOS on MacBook to iOS on iPad Pro by providing enough Desktop features in iOS, then I think they would be in a much stronger market position on the Desktop.   More Apps, more secure OS, less SW development efforts using a unified SW platform.

    I dont have a magic 8 ball, however indications do seem like they are heading in this direction.  I have no pressing reason to upgrade to High Sierra. It just not that much of a feature release for me and I suspect others.  The MacOS team has obviously focused on something else over the past year. 
    I agree with gregg thurman that they will have to outperform Intel's best, but I don't think they are far off from that, especially if you consider they wouldn't just use an iPhone/iPad chip in a top-end Mac - there would be custom silcon for it. In general I believe you are correct on the indications: I think there is a long term one-architecture strategy at Apple. If anyone remembers the 2006 transition from PowerPC to Intel, it was incredibly smooth via the use of universal binaries and Rosetta. Apple's control of the development stack has only become more complete since then, and the legacy classic MacOS cruft made possible by Carbon support is now long gone. We could expect a transition to ARM from Intel would go even smoother.

    To your point, I just don't know if they are going to evolve iOS into a version that can take on the Mac's desktop features and capable of running Mac apps, or to simply migrate MacOS to ARM. The former is enticing because it creates the possibility of using your phone as your PC to run all desktop Mac software, with only the more power-needy users buying dedicated laptops and desktops. The only loss here would be for Boot Camp/VMware/Parallels users, but I doubt that would be a show-stopper for Apple.
    watto_cobra
  • Reply 53 of 83
    -repost with fixed links-

    Geekbench 4:

    iPad Pro 12.9 (A9X) - Source
    • Single-Core: 3206
    • Multi-Core:  5411

    MacBook (Core m3-6Y30) -  Source
    • Single-Core: 2768
    • Multi-Core: 5343
    ----------------------------------------------

    Anandtech iPad Pro review using Spec2006 - Source
    Ultimately I think it’s reasonable to say that Intel’s Core M processors hold a CPU performance edge over iPad Pro and the A9X SoC. Against Intel’s slowest chips A9X is competitive, but as it stands A9X can’t keep up with the faster chips. However by the same metric there’s no question that Apple is closing the gap; A9X can compete with both Broadwell and Skylake Core M processors, and that’s something Apple couldn’t claim even a generation ago. That it’s only against the likes of Core m3 means that Apple still has a way to go, particularly as A9X still loses by more than it wins, but it’s significant progress in a short period of time.
    So despite having a higher Geekbench score, the 6Y30 was determined to have greater performance. So I'm hesitant when someone uses Geekbench claims for an x86/ARM comparison, but it still shows that Apple has been making considerable amounts of progress in trying to catch up to Intel's Y/U processors.
  • Reply 54 of 83
    morky said:
    The only loss here would be for Boot Camp/VMware/Parallels users, but I doubt that would be a show-stopper for Apple.
    I wonder what Microsoft is thinking about that when it is Microsoft itself who is putting in great effort to support iOS  and pushing hard to get  Windows on ARM.  I wonder what they are predicting is going to happen long term on the desktop.   Again, no magic 8 ball here.  However, it looks like hedge bet to me and are no longer going all in on Wintel for desktop.   Why do they continue to push Windows for ARM and why get aggressive about supporting iOS?  

    Maybe its because iOS has been eating away at personal computing for a while.  For many people the phone or the tablet replaced their desktop.

    Is pushing iPad Pro further toward the Desktop by making it more MacOS like in CPU power, storage, RAM and UI is going to attract even more people away from MacOS?  I think it can and will.   Its clear where all the development effort was and was not over the past year, when you compare what happened on iPad vs MacOS feature updates.  Its going to keep heading in that direction I bet.  

    I think Intel needs to retreat and focus the Server market long terms Apple keeps eating away at the Desktop. Intel has strong competition of late from AMD. Both performance and price.  They need to stay on top of that.  

    In my mind, they have already demonstrated that  they won't be able to compete on the low power personal space when they waived the white flag on Atom based phones and tablets and shuttered that business. Laying of ten's of thousands in the process. Microsoft too waived their white flag when they gave up on Windows Mobile, Nokia (laying off ten's of thousands as well)  and have aggressive adopted iOS for their productivity apps.  Factor in the Windows on ARM push even in an angry Intel in the background  and I think I may know what they are thinking in going to happen to the Desktop over time too.
    morky
  • Reply 55 of 83
    chiachia Posts: 713member
    It will be interesting to know how Apple decides what direction to take in development of its processors.
    Do they evaluate what works smoothly and not so smoothly, then look at how they can improve the performance in that area.
    Alternatively do they think "what will be a good feature to have in an Apple Device", prototype it, then work towards designing a processor that delivers that feature at the right size, power consumption and price?

    Perhaps an answer lie in studying the evolution of Apple processors from their first A4 to the current A11.

    Apple explicitly expressed the latter thinking when they designed the custom controller chip for  the 5K Retina iMac.
    watto_cobra
  • Reply 56 of 83
    chia said:
    It will be interesting to know how Apple decides what direction to take in development of its processors.
    Do they evaluate what works smoothly and not so smoothly, then look at how they can improve the performance in that area.
    Alternatively do they think "what will be a good feature to have in an Apple Device", prototype it, then work towards designing a processor that delivers that feature at the right size, power consumption and price?

    Perhaps an answer lie in studying the evolution of Apple processors from their first A4 to the current A11.

    Apple explicitly expressed the latter thinking when they designed the custom controller chip for  the 5K Retina iMac.
    to me, clearly apple is saying it can get a huge leg up on their competition if we have control over the silicon and start over from a clean sheet. I'm looking forward to see what they do on iPad Pro.  They already have more HP then they know what to do with already.  Bring over the Mac Apps and features. Looking forward to iOS12. ;-)
    watto_cobra
  • Reply 57 of 83
    Intel has to be afraid by now: it would not be outlandish to think Apple is going to use its own silicon for its own server farms (iCloud, Siri), which would call into question Intels future. Small error in the table: the iPhone SE has the same 2GB RAM the iPhone 6S has.
    How would they manufacture enough for that? They can’t even keep up with iPhone demand.
    Meeting iPhone demand isn't processor bound.
    watto_cobra
  • Reply 58 of 83
    nhtnht Posts: 4,522member
    EngDev said:
    -repost with fixed links-

    Geekbench 4:

    iPad Pro 12.9 (A9X) - Source
    • Single-Core: 3206
    • Multi-Core:  5411

    MacBook (Core m3-6Y30) -  Source
    • Single-Core: 2768
    • Multi-Core: 5343
    ----------------------------------------------

    Anandtech iPad Pro review using Spec2006 - Source
    Ultimately I think it’s reasonable to say that Intel’s Core M processors hold a CPU performance edge over iPad Pro and the A9X SoC. Against Intel’s slowest chips A9X is competitive, but as it stands A9X can’t keep up with the faster chips. However by the same metric there’s no question that Apple is closing the gap; A9X can compete with both Broadwell and Skylake Core M processors, and that’s something Apple couldn’t claim even a generation ago. That it’s only against the likes of Core m3 means that Apple still has a way to go, particularly as A9X still loses by more than it wins, but it’s significant progress in a short period of time.
    So despite having a higher Geekbench score, the 6Y30 was determined to have greater performance. So I'm hesitant when someone uses Geekbench claims for an x86/ARM comparison, but it still shows that Apple has been making considerable amounts of progress in trying to catch up to Intel's Y/U processors.
    Conversely Intel has been making considerable amounts of progress in improving it's performance per watt to the point where apple to apple comparisons are possible.
    tmay
  • Reply 59 of 83
    Aren't the A11's scores better because the iPhones are running at a lower resolution, or am I missing something?
    The CPU score has nothing to do with resolution
    And its easy to test GPU independently of resolution. Get it to write into  buffer or something (off screen). Still beats all comer.
    Of course, it beats them even more if the rez is lower, that goes without saying.
    But, often that's not even what happens, the GPU writes to a scale that's bigger that the actual screen and then the hardware downscales to the actual resolution (that makes it easier on the devs).
    The X has a higher resolution I believe so what will Androites find next.
    edited September 2017 watto_cobra
  • Reply 60 of 83
    Aren't the A11's scores better because the iPhones are running at a lower resolution, or am I missing something?

    Yes, you are missing a lot of things. There are benchmarks which are NOT dependent on screen resolution. Even in ALL of those benchmarks, A11 is a champion. Guess which SoC is second in most of those benchmarks? No prices, if you guessed A10 from Apple.
    Then why does the iPhone 8 have a better overall score than the 8s, which in turn scores better than the X? Obviously resolution is playing a factor here, so the test is not a good Apples to Apples comparison. The test should pick a baseline resolution and measure all devices based on it. For example, you could say a laptop with an Nvidia 1060 "obliterates" the laptop with Nvidi 1080 at 4K because the "performance" is so much better, but that wound be entirely inaccurate because the 1080 is much more powerful.
    Those are on off scores, lets see averages before concluding anything. The silicon lottery means there can be a substantial variance in performance even outside natural variance from tests to tests (10%).
    In theory, the chip the best performance should be in the 8+, then the X and the 8.
    But, I got the feeling the rest of the SOC (like the neural core and higher resolution screen) on the X has to work harder (because the phone does more thing). This impacts thermals in the phone and thus slightly depresses the SOC score.

    watto_cobra
Sign In or Register to comment.