Apple mulling transition away from Intel chips for Macs



  • Reply 81 of 119
    v5vv5v Posts: 1,357member


    Originally Posted by bsimpsen View Post

    Remember that Apple's elimination of Intel also eliminates Intel's hefty margins. That immediately places Apple at a cost advantage against other PC manufacturers.


    This is good news for shareholders, irrelevant to Mac buyers. Apple eats up cost savings in margins, not reduced product prices. As evidence please see the iPad Mini and any iOS device with more than the base amount of storage. Mac prices don't change, except for build-to-order options which become more comically egregious with each passing year.

  • Reply 82 of 119
    jragostajragosta Posts: 10,473member

    Really, it sort of is. Macs don't sell because of Boot Camp.

    Actually, they do. I know a number of people who bought Macs because of Boot Camp. I know even more who bought them because they allow virtual machine performance which is close to native performance. ARM-based Macs would lose both of those advantages.
    Honestly, I think this is more of a Worst-Case-Scenerio than where they are headed. Much like what they did with PowerPC. They kept with the PowerPC Version until it made no sense for them to continue supporting it. They made the transition to Intel, and life carried on (even though my wife said it was the end of the Mac). I see them keeping a version of OS X running on their Ax Series of chips, and probably have an idea what it would take to make it work well with them. I still think it is just a matter of if Intel drops the ball, and then Apple has to respond.

    The difference is that when they switched from PPC to Intel, they were switching to an architecture that was comparable in performance. In fact, since IBM and Motorola had really slowed down PPC performance, the new Intel Macs were as fast (or faster) as the PPC Macs they replaced. Equally important is the fact that they could now run WIndows apps at native speed rather than at the speed of a turtle chasing a snail through quicksand.

    I have no idea about any of this stuff (hardware), would somebody fill me in? I mean, I know the basics, don't treat me like a child. Give me 10 (or as much as you can) reasons why switching to an ARM architecture would be bad on a Mac...


    1. Performance would stink
    2. Most existing apps would not run. Given how long it took major players like Adobe and Microsoft to create native Intel versions, this would be a major problem.
    3. ARM is slow enough that they could not have a Rosetta-type solution to run Intel apps on an ARM system - so they'd have to wait for the apps to be ported.
    4. Windows apps currently run at native speeds via virtual machines. Under ARM, they'd have to be emulated - which would make problem #1 even worse.

    Repeat those 4 problems ad nauseam.

    lkrupp wrote: »
    Who says it has to do Intel emulation? Apple surprised the tech world when it revealed it had a working verso of OS X running on Intel processors. Who says they don't have OS X running on custom, Apple designed ARM in the labs? And I don't agree that leaving Windows behind would be a big problem.

    Lots of people use Macs to run Windows. Parallels claims to have 3 M Mac customers. When you add in other virtual machine software, you're probably looking at over 10 M customers. That's a lot.

    And while putting Mac OS X on ARM would not be a terrible problem (other than the fact that the fastest ARM chip is far slower than the slowest Intel chip currently used by Apple), but the apps would be a problem. How long did it take Adobe to release a native version of Creative Suite (and even that did not use Apple's technologies)? How long did it take Microsoft to release a native version of Office? How long did it take AutoCAD to reappear on Macs? And given that ARM is so much slower than Intel, users would suffer - even if Apple could make a decent emulator.
  • Reply 83 of 119
    dunksdunks Posts: 1,254member


    Originally Posted by Jeeves Staub View Post

    If I was negotiating with Intel to get the best processor prices possible, I'd say the same thing.


    I agree. Got to keep one eye on the metagame!

  • Reply 84 of 119
    dunksdunks Posts: 1,254member


    Originally Posted by Bregalad View Post

    I think you underestimate the number of Mac users who occasionally need to run a Windows app. Boot Camp/Parallels/VM Ware/VirtualBox has allowed people who are tied to one or more Windows applications to switch and allowed Mac users to integrate into Windows dominated environments. I've been a Mac owner since the early 1990's, but I have a virtualized Windows installation for software that doesn't have a Mac equivalent.


    I used to have a cheap PC desktop, but got sick and tired of having two computers.


    Today Windows has a huge library of software, much of it unavailable for Mac. iOS also has a huge library of software, much of it also unavailable for the Mac.


    When most computing goes mobile the old formats will likely lose their relevance and new ones will take their place. A seamless product lineup from iPhone to desktop running the same software might look very appealing, but there are still some big bumps in the road.


    Almost all iOS and Android users also have a notebook/desktop computer of some sort and 95% of those run Windows.


    iOS is openly hostile to collaboration and iCloud is a syncing service not a sharing service. Apple's sandbox that locks documents away in a private silo only accessible to one app and one Apple ID is a troubling sign.


    Most of the world cannot afford to buy Apple hardware. Apple has always been happy to sell only to those who can afford to pay premium prices, but that means they have never been able to define the file formats or tools that the rest of the world uses.



    Premium price or not I never got this much quality software or capability out of a windows machine.

  • Reply 85 of 119


    Originally Posted by Tallest Skil View Post


    They sure didn't do it thanks to Boot Camp! No one is going to buy a Mac just to run Windows. They'll drop $400 on a piece of crap Windows desktop.


    Unless they also want a Mac for running Mac OS X. In which case, they don't need to buy a separate Windows machine at all. I wouldn't underestimate the appeal of Boot Camp to these people. I remember the zeitgeist when Apple first moved to Intel and introduced Boot Camp: people from the Windows world began justifying the purchase of their first Mac because Apple offered Boot Camp and Windows as a "safety blanket."

  • Reply 86 of 119
    tbelltbell Posts: 3,146member


    Originally Posted by leesmith View Post

    Don't do it!



    I think I agree. However, interesting enough Microsoft is developing apps for the ARM processor. If enough of the important Windows apps run on ARM, it is possible that the ARM version of Windows would also be supported on an ARM based Mac. 


    As an attorney though, I rely on some Windows software. If I can't run it, I have to go Windows (at least currently). 


    Apple going Intel made a lot of inroads for Apple. It no longer had to play the specs game explaining how the lower spec'd Power PC chip was faster than higher spec'd Intel chips. Further, the costs of Macs went down when Apple was able to use more over the counter parts. Many people who needed Windows program could than run then much more easily. Further, more applications came to the Mac. 


    Apple, however, could leverage its iOS app store to allow such Apps to more easily run on a Mac. 

  • Reply 87 of 119


    Originally Posted by CU10 View Post

    Apple's decline may start when they get off intel's roadmap and go on their own.


    Intel made Apple relevant in a way.


    You mean by turning Apple into another commodity Wintel PC assembler?


    Nice theory. Except for the fact that they've never been "on intel's roadmap" for any of their iOS devices, and yet these devices drive Apple's earnings. If the post-PC vision ever becomes a reality, how important will being "on intel's roadmap" be for Apple? How important will intel be in the post-PC world?

  • Reply 88 of 119
    ksecksec Posts: 1,569member
    The potential savings from switching to ARM are billions per year. Calculating at 10 Million Macs and a saving of $100 per devices. The actual figures could varies since we dont know what sort of discount Intel are giving Apple.

    Someone mentioned the Ars Chromebook review, but Jeff Atwood also mention somehow the Nexus 7 despite having inferior hardware performs better. Which may be a sign that ChromeOS is not as optimize. It is important to point out x86 has the YEARS of engineering and tuning on performance and ARM still have a little to catch up.

    The Rumors pops out basically because, 1. AMD decided to make a 64Bit ARM chip and 2, ARM announced their A57 to be around 30% faster then current A15 per clock. And A57 is designed to run up to 3Ghz. It is not hard to imagine A57 will be on the same level of performance with Core2Duo per clock. Think of it like a 8 Core Core2Duo running at 3Ghz on your desktop, but in terms of performance i am pretty sure by 2014 to 2015 they should have more or less worked it out.

    And if x86 is REALLY that important, i am sure AMD will give a helping hand there.
  • Reply 89 of 119

    Remember the day, when there was nearly no way, to view files from a pc, or view mac files on a pc? The old PowerPC days, and the 68030 days. I'm not looking forward to that happening again.

  • Reply 90 of 119
    MarvinMarvin Posts: 14,952moderator
    According to one of the sources, the recently-returned Bob Mansfield has long been interested in making a more consistent experience across iOS and OS X, a goal that could take shape as former mobile software chief Scott Forstall was recently ousted.

    Is Forstall going to get blamed for everything from now on? Maybe Apple will build a car and a TV now that Forstall isn't holding them all back.

    When it comes to ARM, I don't see why every big company wouldn't move to it. As soon as you buy chips from a 3rd party, you are putting yields, price, performance and features of the core part of your product under someone else's control.

    Apple can't force Intel to make their IGP decent - that's why they have had to keep shifting back and forth between Intel's junk and GPUs from AMD/NVidia.

    When Apple bought up the chip design firm and started doing this themselves, it was a clear sign they are tired of having other companies dictate the standard of their core components. This 'Swift' architecture is early days but a pretty good effort:

    Of course you can't take the chip from an iPhone, put it in an iMac and call it a day but consider performance per watt. What dual-core 1.3GHz x86 chip runs in a TDP of 2-5W? The iMac CPU has up to a 95W TDP.

    Now consider Moore's Law. Right now we look at what we need now and we could not run productive apps comfortably on a phone. But, even if processing power doubles every 2 years, the processors we have will be 30x faster in 10 years. Apple is at 32nm with the A6 and Intel is at 22nm - it shrinks 70% each time so next up are 15nm, 10nm, 7nm, 5nm:

    Further might be possible:

    But just assume 30x for the CPU alone. Now assume that we have fully heterogenous computing where code runs on the GPU just as it does on the CPU but faster so now you are at least double again to 60x.

    Even if you take a 50% hit on emulation speed, there is little reason to complain about x86 performance. Apple is not going to be alone in this. Just look at the population growth. Can a single company (Intel) delivery the kind of yields to satisfy shipments in the billions? What if they are late, should the world's computer technology sector just wait? That's what we have to do now because nobody can fairly compete with Intel.

    More and more companies are clearly becoming frustrated by Intel. Modern games consoles don't use Intel and they are fairly powerful machines. The Wii-U uses an ARM CPU to run the OS and they all use IBM PPC CPUs.

    If Intel has a roadmap to 60x then there might be no reason to switch but if Apple has a faster roadmap, there is because Apple could make an ARM chip 2^5 (32x) faster in the time Intel makes an x86 chip 2^3 (8x) faster. Even with a 50% hit with emulation, Apple would still be faster than Intel native. There are undeniable downsides but it's all just legacy. Companies will have to change their software to accommodate new input methods and operating systems in due time anyway.
  • Reply 91 of 119
    Its been reported already that Apple has ported OS X to ARM for testing and was favorably impressed with it's performance on a 32 bit processor. Keep in mind that the A6X has quad core GPUs for driving a high resolution 2048 x 1536 pixel display at a high performance rate. That's on a SOC.. It's easily possible as Apple has an instruction set level license for ARMv8. With the recent A50/A57 64 bit chips appearing, a transition could be closer than you think.
  • Reply 92 of 119
    wozwozwozwoz Posts: 261member
    It sounds dumb. Last week, Apple's Tim Cook was criticising Windows for adopting the same paradigm across desktop Windows and tablet Windows (surface): anyone can make a car that flies, or some such thing ... but would it really do either well?

    This piffle article now suggests that Apple do precisely that. Would you really want the same chip powering your iPad mini as that powering your Mac Pro? Nah. Would it be able to run Windows too? Of course not. Would it have to emulate everything .. probably very slowly - yes. What would be the advantage ??? Maybe that a Macbook Air could run cooly for a bit longer. Who cares ... by then, Intel will have something that runs cooler and longer anyway.

    Go find something useful to do ... like take on Google on search. :)
  • Reply 93 of 119


    Originally Posted by Bregalad View Post

    Apple is never going to choose a chip that's twice as big as the competition. Smaller, thinner, lighter applies to the inside of the Mac just as much as the outside.

    They already did field a chip that was much larger than the competition. Analysts were surprised by the large die size of the A5, and reasoned (correctly) that Apple could afford the extra silicon because they'd dispensed with the margins of a Qualcomm or NVIDIA.

  • Reply 94 of 119


    Originally Posted by v5v View Post


    This is good news for shareholders, irrelevant to Mac buyers. Apple eats up cost savings in margins, not reduced product prices. As evidence please see the iPad Mini and any iOS device with more than the base amount of storage. Mac prices don't change, except for build-to-order options which become more comically egregious with each passing year.

    I think it would be good news for both. If Apple is able to improve performance at the same price and/or reduce price at the same performance while maintaining margin, everybody benefits, except the middleman who's been cut out.

  • Reply 95 of 119


    Originally Posted by renner.david.b View Post

    Its been reported already that Apple has ported OS X to ARM for testing and was favorably impressed with it's performance on a 32 bit processor. Keep in mind that the A6X has quad core GPUs for driving a high resolution 2048 x 1536 pixel display at a high performance rate. That's on a SOC.. It's easily possible as Apple has an instruction set level license for ARMv8. With the recent A50/A57 64 bit chips appearing, a transition could be closer than you think.

    I'd read that Apple has an architectural license, but wasn't sure about an instruction set license. The latter would be ideal, as the license cost is lower and Apple seems to have the talent to design from scratch. This would decouple Apple from ARM's development timeline.

  • Reply 96 of 119
    asciiascii Posts: 5,936member

    I'm not sure you really want to go up against Chipzilla. They will figure out *some* way to make x64 as power efficient as ARM, just you watch, they'll pull something out of the hat.

  • Reply 97 of 119

    I am sure that Apple are always exploring alternatives to the architecture, keeping close to AMD while trying to steer Intel to a roadmap that suits Apple. It would be remiss of a company that big to do anything other.


    If and it is a very big if, ARM were to catch up and overtake Intel then it may not be completely silly to consider the options. I know that means running x86 in emulation and there is a Russian company that claims it is running emulation of x86 on ARM at 40% efficiency and expects to get to over 80% within a couple of years.


    On the other hand Intel have been pushed by Apple over the years to produce a roadmap that suits Apple aspirations and have had some success (assuming Haswell and Broadwell deliver) especially in the graphics department.


    I would be concerned if Apple were not exploring options, they have a pretty good track record of adopting technologies that meet their needs while accommodating the majority of their current and anticipated user base. 

  • Reply 98 of 119
    realisticrealistic Posts: 1,154member


    Originally Posted by bsimpsen View Post


    ARM runs OS X natively already, so emulation is pointless. I believe you meant that ARM would have difficulty emulating the X86 instruction set. As Apple is designing the processor, emulation of the X86 instruction set is probably more a legal problem than a technological one. The PA Semi folks have already implemented two different instruction sets, PowerPC and ARM. X86 would not be beyond their grasp. However, they can't get a license to run X86 by purchasing AMD (there's a caveat in AMD's license that prevents inheritance of the license). There may be another way for Apple to emulate X86 efficiently and legally, or it may be that by the time they envision making the switch, they also envision X86 compatibility being unnecessary.


    I think it's important to consider the leverage Apple gains by removing Intel's 60% gross margin from the bill of materials.

    ARM doesn't run iOS otherwise there would already be Macs running on ARM chips. Show us any links saying, showing or even claiming that ARM chips can run OS X.

  • Reply 99 of 119


    Originally Posted by melgross View Post

    I don't see your argument as being correct. Some of your facts are simply wrong as well. For one, Apple isn't "experimenting" with selling iPads without cell connectivity. They've always sold them that way, and usually WiFi only models have gone on sale first. I don't know where you get the idea you have.

    You're making assumptions of what would be a winner, and what won't. How do you know this? You don't. You just think it's true. Maybe it is, and maybe it isn't. How much of your own real money would you bet on it?

    I doubt that Apple isn't investigating every possible option out there. They have to. But most of what they do never is used. And we don't know what they are doing. A few enthusiastic engineers are making statements that are their own desire, perhaps, nothing more.

    Apple sold 18 million "computers" last year, about $25 billion worth. They aren't going to abandon that business so soon.

    Convincing potential customers to buy an iPad is great, I've got a bunch. But it will be some time before the computer becomes that truck SJ was talking about. And when it does, we won't be using OS X on ARM, we will be using iOS 12 on ARM.

    I got the idea I have because Apple is currently shipping Mini iPads and iPad 4s WIFI ONLY. No cell capability versions. They never had the time gap between the two kinds like this. 


    Hundreds of Millions of users are finding that they can get along just fine and dandy without MS Office on their iOS Devices. That includes a lot of Fortune 1000 commercial users. The longer this goes on, the more IT managers and users are realizing that X86/Windows is not as essential as once believed as gospel truth.


    Where's the tipping point where enough users believe they can exist without X86/Windows? Is it here now with nearly half a billion iDevices in users hands, or will it come next year? This can happen without iOS devices being the "truck" computer. There may be a place for Intel/Win in the future, just as there are mini-computers and IBM big iron still running and needed. We may just as well see Intel/Macs for as long as well. I think portables will lose the Intel chip due to (1) battery life differences, (2) the lesser need for a "truck" computer for most users.


    Finally, iOS apps haven't experienced the "bloatware" situation that is common to Intel computers. I remember doing some pretty fast work with Photoshop on a Mac IIx, with 8 Mg of RAM and a 16 MHz clock. The needs for terra-flops of speed and giga-bytes of memory today are crazy.  :-)

  • Reply 100 of 119


    Originally Posted by melgross View Post

    I love it when people say "just recompile". Sure. So you have gone and found Apple technologies and listed them, wonderful. Have you ever programmed on a serious level? If you have, then you would know that "just recompile" is a very naive accessment of the problem for large amounts of code. You are talking about major work. Work that many developers may not want to do again.

    And please don't make statements that you can't support. Rumors don't mean that something is actually happening. Even if they are, it means little.


    in most cases it would just be a recompile.


    Certainly there are important exceptions, things like photoshop and office.  But for the most part apps already are built to support MacOS X on Intel and iOS on ARM.  And developers who know what they're doing don't write code that would care which CPU it ran on.


    And Xcode already knows how to build binaries that support multiple CPU architectures.  Announce at WWDC and ship in the fall and 90% of shipping apps would already be updated by the time machines shipped.  I'm sure people like Adobe and Microsoft would take longer, but that wouldn't matter too much because if Apple did this, the first machines would most likely be more affordable MacBook Airs. which already don't target photoshop users.


    as for the cpu power that keeps being mentioned.  Certainly Apple wouldn't do something like this without a more powerful CPU than the iPad 4 has.  But a 4 core 64 but ARM CPU should be plenty powerful for a more affordable entry level machine.  


    Building their own CPUs has proven to be quite important for Apple in the iOS world.  It wouldn't' be at all surprising if they wanted to bring that to the MacOS world. 

Sign In or Register to comment.