“KGI Securities analyst Ming-Chi Kuo has issued a new report predicting that Apple would begin to put its custom ARM-based systems on chip (SOCs) in new Mac OS X systems starting this year,” Mark Hibben writes for Seeking Alpha. “While the exact time line for the introduction of ARM SOCs into Macs may be in doubt, their introduction appears increasingly inevitable. The most compelling reason for this is the lower cost of ARM processors relative to Intel processors of equivalent capability.”
“Increasingly, news reports have confirmed that Samsung and TSMC will have their own 14-16 nm processes in production by mid-year. Kuo’s analysis assumes this,” Hibben writes. “While the technical details of the process implementation for the foundries remain unknown, I consider it probable that these processes will achieve performance parity with Intel’s most advanced 14 nm process, currently only used for the Broadwell-based Core processors.”
“Thus it appears that Apple is at the threshold of a critical gate for deciding whether to pursue an ARM-based Mac. Without a process advantage, Intel’s chip offerings become much less compelling,” Hibben writes. “There is, of course, the issue of backward compatibility for Mac apps created for Intel-based Macs, but this is much less of an issue than some Intel bulls would have investors believe.”
“This may stem from ignorance of the app creation process for Mac OS X and iOS, so that it’s assumed that transitioning to a different platform is monumentally difficult for developers. For most developers with apps on the Mac App Store, it won’t be. If the developer has followed Apple guidelines and has used only tools provided by Apple in the Xcode developer environment, transitioning an app that’s already on the Mac App Store from Intel to ARM will be as simple as pushing a button,” Hibben writes. “Of course, not all Mac OS X software is on the App Store, and many developers use software tools not provided by Apple. Here, the situation is not either/or. Apple can continue, and will continue, to offer Intel-based Macs along side ARM-based Macs.”
Read more in the full article here.
MacDailyNews Take: As we wrote yesterday:
There is no reason why Apple could not offer both A-series-powered Macs and Intel-based Macs. The two are not mutually exclusive…
Why Apple dumping Intel processors would be disastrous – January 14, 2015
KGI: Apple is designing its own processors for Mac – January 14, 2015
Apple A9-powered MacBook Air? – December 16, 2014
Why Apple will switch to ARM-based Apple A-series-powered Macs – August 27, 2014
Intel-powered Macs: The end is nigh – August 4, 2014
Intel’s Broadwell chips further delayed; not shipping for most Macs until early-mid 2015 – July 9, 2014
Apple will inevitably drop Intel for their own A-series processors in the Mac – June 26, 2014
How long before Apple dumps Intel from MacBook Air? – June 26, 2013
64bit quad-core A8, A9 or A10 – must be well between the Atom and i3 processors. Perhaps close to i3 even. Yet a iMac low end – soldered ram, slower chip can not be much use fro a Desktop can it?
For those who are able to disern that 4 is bigger than 3 but lack the knowledge of processor computing capability, your ‘analysis’ is right on but computing is so much more than a comparison of product names and numbering schemes. It might be worthwhile to learn about processor capabilities and computing requirements before you spout off.
I am no programmer but wouldn’t the Mac OS X code have to be totally rewritten to work on an ARM-based processor?
I suspect it would merely need to be recompiled, with a new compiler Apple already has nearly ready to go if they are actually working on this. Thus the “simple as pushing a button” line in the article.
Sorry, I misread your post as “Mac apps in OS X” rather than the OS itself. Still, probably parts of it will be similarly simply recompiled, while much of it is likely to be done then from the ground up. But that’s also where iOS integration may come in. I don’t expect the desktop will be replaced by an iOS version, but the basics of handling icons and names (and to some degree folders) are at least already there. So Yes. And No. And Maybe in places, but not in others.
Ha! Whomever voted you down and me up also misread your question. 😀
No. iOS and OS X share a lot of the same code. So, most of the work is done. In fact… I’m sure they’ve already got versions of Mac OS X running on ARM.
You can safely bet on that no matter what the immediate plans are. Clearly Apple would much prefer to have Macs with their own chips if at all possible do they will be forever testing the boundaries of so doing even, for pure experimentation I’m sure they would acquire enormous amounts of knowledge for future developments and interoperability.
iOS, which is a simplified variant of OS X, is obviously already compiled for ARM. So, it would not be a stretch to host OS X on ARM.
Apple had OS X running on Intel years before Intel-based Macs were released. I strongly suspect that Apple R&D has been testing OS X on ARM for quite a while.
I don’t care about the lower cost of ARM processors. I think it’s unlikely that any substantial cost savings would be passed on to the end user.
Apple probably could make Xcode compile existing code to ARM binary apps. Maybe it would be easy and flawless, and maybe not. Small apps and games might be easy to convert while the productivity software developers might have a more difficult time. It’s hard to say since I’m not a programmer and this is all speculation anyway.
Even if the code conversion is easy, the entire process would be disruptive to the end users. Sorry, your existing software may not work and you may have to upgrade or buy new licenses.
So Apple would get a cheaper chip. End users get confusion and extra cost. Unless the ARM chips provide some radical advantage I can’t see this as a great idea.
Why did you say so much about things you said you are not knowledgeable about? 😉
Because I am knowledgable, just not a hard core programmer. ;0p
Perhaps somebody who is a developer could fill us in on the possible advantages of an ARM-based mac and the difficulties involved in re-compiling software. I’d be interested to hear it.
Because that is what most of us do most of the time on MDN?
If switching to ARM CPUs reduces Apple’s production costs, then that is a strong reason to do it as long as the compromises are reasonable. Apple maintains gross margins, so that cost reduction would eventually make it to the consumer.
If you want advantages over Intel, then think Metal and similar deeply integrated OS-CPU functions for speed and efficiency. Think minimal or no fans. Think extended battery life on laptops. Apple went from Motorola 68000 series CPUs to PPC CPUs to Intel CPUs over a period of a little over ten years. Apple is not afraid to make major changes in its hardware architectures.
Some people are worrying about Windows compatibility on ARM. Fusion and Parallels Windows emulators will provide the necessary functionality (as they do for many people today). Boot Camp, of course, will not be available on ARM-based Macs. That will affect a small percentage of users and I sympathize with those folks.
Apple has been on the ARM RISC path for years. Intel CPUs were just a stopgap to deal with the failure of the AIM alliance to maintain PPC parity with Pentium processors in the early 2000s. When Apple decides that ARM CPUs are ready for prime time, you will see them debut in the Mac lineup. My guess would be the MacBook Air.
We have been there before. I have witnessed several of these: 68k to PPC; System 9 to OS X; PPC to Intel; 32-bit to 64-bit.
There is nothing massively disruptive about these transitions. I have made each and every one of them, without major hassle (although there was some minor hassle there).
Most professional people have 4 – 5-year replacement cycle for their Macs (although a small percentage runs them longer). This has to do with support for old software and old OS, more than the old machine slowing down too much, or the new machines being so much faster. Every major transition Apple has made before included some legacy support for at least 3-4 years. By the time this legacy support is discontinued (System 9, Rosetta, etc), you are ready for an upgrade to a new machine, new OS, new software. Most major software players (Adobe, Microsoft, Avid, etc) had grudgingly worked with Apple, migrating their flagship tools to the new platforms. Also, they all offer some sort of upgrade pricing (if not an annual subscription with the “Up-To-Date” plans). This means that the migration usually doesn’t present cost any higher than the regular, planned hardware/software upgrade.
Let us not forget, OSX is now running on ALL Apple hardware (except the iPod nano / shuffle). It is just the UI that is different between multi-touch and keyboard-mouse-display devices. I’m pretty sure there there is another Project Marklar somewhere at Apple, running Mac OS X on A-family processors.
The only other major transition I remember was System 7, which was a major departure from all that came before it. I don’t even remember why, but everything back then was being “certified” for System 7.
Most have been pretty smooth. The only “major” disruptions I recall was for printer (and scanner!) drivers, but it didn’t personally affect me. I doubt that would be an issue at all today.
Even without the replacement cycles you mention, I doubt we would see much difference in availability and upgrade paths for whatever we want to do. You might lose some “legacy” software, but something will be ready to step in its place before Apple dumps the legacy support, like you say.
I’m ready. I’ve been threatening to get a new Mac for some time. The new iMac 5k is really tempting. This makes me want to wait and see just what is next – again. 🙁
You may be right. The disruptions caused by the previous transitions weren’t always “massive” but they did happen. It’s the 3rd party developers that will be the problem.
For example, printer and scanner drivers for older, but still perfectly serviceable, devices might not get an update. And we’re not necessarily talking about inexpensive devices either.
HP and Xerox won’t be in a hurry to re-write their drivers for older machines that are still in service, but the businesses that rely on them may still be under contract or won’t be in a position to replace functioning hardware just because a new kind of Mac was invented.
EFI’s PrintSmith software was only recently updated so the system can work on Macs running 10.7 and later. That’s not Apple’s fault exactly, but the end users that relied on the software to run their business had no other option but to maintain 10.6 machines.
So if this comes to pass, I hope you are correct and Apple implements some kind of Rosetta-like translation. I still see little advantage in making the change.
68 to pc was more than a little hairy for many. A company I worked for was advised to rebuild their Macs with pc processors. That was a total disaster that led to the machines to barely manage 30 minutes without a crash. Those ‘experts’ however claimed it would all be see less and were no where to be seen as we designers were pilloried for not being able to avoid crashes and long work overruns. Pardon me for feeling nervous expect i ally with some of Apples recent software problems.
For people who don’t think this would be disruptive, you might want to take a look at the results of the current MDN poll: Has Apple’s software quality slipped in the last few years?
Apple has a lot going on right now along with some major transitions and core components that have been re-written from scratch that they’re still trying to work out the bugs with.
And while some people are calling for a Sno-Semite release, there’s little likelihood of that given some major things Apple is working on right now (like a completely new file system).
So now people are expecting Apple to not only develop OS X on two architectures, but also expect 3rd party developers to do the same? I don’t think so.
Also, it’s not like as if we’re seeing A-series blowing away what’s coming from Intel. Intel is still way ahead despite what could be considered as a major setback from not correctly reading the trend of low-power/heat demand. However, Intel is now all-in on this and their roadmap is incredibly impressive.
If anything, I’d expect Intel to be more involved with Apple’s iOS devices than less involved with their OS X products.
First and foremost, Samsung currently makes most of Apples A-series chips for iOS devices. Pray tell, why would anyone want to drive significantly more business in that direction????
Moreover, the TOTAL performance of ARM chips isn’t competitive with Intel’s offerings, and it’s certainly not adequate for power users who buy the most profitable Macs today. Don’t we have enough evidence from Microsoft’s failed “Windows RT” project to show that forking OSX would be ludicrous????
For those lightweight customers who prefer minimal power computing and want to accept iOS limitations, then they are welcome to do so today. There is nothing stopping you.
Apple strategy needs to be simple: iOS for RISC, Mac for CISC. iOS only for mobile and consumer devices; Macs for professionals and power users. Long may it remain so.
MDN: find another theme to push, this constant proposal to follow the MS Windows 8 path makes no sense. I advocate that Apple split up the iOS and Mac OS software teams so they can concentrate on building BETTER software for their respective customers rather than dumbing down both to the lowest common denominator. And get Ive as far away from software as you can throw him!!!
Using the Windows RT debacle as your rationale is pitiful. Besides, the Windows RT experiment failed due to a combination of poor performance and no apps.
Apple will ensure better performance due to a more efficient OS, custom ARM CPUs, and several additional years of ARM development since Microsoft’s RT failure.