Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.

maflynn

macrumors Haswell
May 3, 2009
73,682
43,740
Cost -- since Apple is already investing in the processor for the iPhone, it seems they would save money by not buying Intel chips. (And of course Apple wouldn't pocket the profit, they would pass it to us consumers.;))

Switching platforms is a risky proposition that What would the benefits of such a swtich be? Would they sell more MBPs? Because at the end of the day, selling more must be the goal, I can't see how losing Windows compatibility, virtualization technology, and having developers need to recode all their apps (which means some will just abandon the platform) as a means to increase sales.
 

poorcody

macrumors 65816
Jul 23, 2013
1,340
1,585
Someone sent me this article which does a really in-depth look at the topic. It was written in January 2015 so some things have changed with A10, but it is still mostly relevant:

http://appleinsider.com/articles/15...ck-from-moving-intel-macs-to-custom-arm-chips

A few highlights:
  • Apple pays Intel between $180-$300 per Intel processor, and the A6 (time of article) costs Apple about $25 each
  • Quote: "Apple has significant experience in porting its own OS, frameworks, apps and development tools to new hardware architectures. It moved Mac OS from 68k to PowerPC, then ported NeXT software from Intel to PowerPC to deliver OS X, when it then ported back to Intel. iOS is essentially OS X ported to ARM, with significant optimizations for mobility."
  • In 2010, Apple transitioned the Apple TV from an Intel x86 processor to an A4, modifying the underlying OS X port and making significant hardware architecture changes. They then dropped the price from $299 to $99.
The article also points out some of the risks and pitfalls. Interesting read all around...
 

Dwalls90

macrumors 603
Feb 5, 2009
5,509
4,624
I could see A chips being used in a hybrid iPad and Macbook product, that could switch in between iOS and Mac OS. But it would be scary to think about power users losing x86 for the higher end Macs.
 

argentum47

macrumors regular
Jun 18, 2014
210
498
Someone sent me this article which does a really in-depth look at the topic. It was written in January 2015 so some things have changed with A10, but it is still mostly relevant:

http://appleinsider.com/articles/15...ck-from-moving-intel-macs-to-custom-arm-chips

A few highlights:
  • Apple pays Intel between $180-$300 per Intel processor, and the A6 (time of article) costs Apple about $25 each
  • Quote: "Apple has significant experience in porting its own OS, frameworks, apps and development tools to new hardware architectures. It moved Mac OS from 68k to PowerPC, then ported NeXT software from Intel to PowerPC to deliver OS X, when it then ported back to Intel. iOS is essentially OS X ported to ARM, with significant optimizations for mobility."
  • In 2010, Apple transitioned the Apple TV from an Intel x86 processor to an A4, modifying the underlying OS X port and making significant hardware architecture changes. They then dropped the price from $299 to $99.
The article also points out some of the risks and pitfalls. Interesting read all around...

Interesting read. But like the article says, even if they can do it, doesn't mean they should do it. Probably not just yet...

Apple is probably capable of (may even have prototypes) porting macOS to ARM. But it will likely be unprofitable unless

- A# processors become stronger than x64 processors of similar power consumption, running absolutely identical tests (please no Geekbench)
- There is a capability to provide seamless emulation/virtualization of x64 codes
- 3rd party apps can be recompiled seamlessly without any (or with very minor) effort

Because right now, personal computer pretty much means x64. If it's not going to be compatible with it, it might as well just be categorized as a new platform (like gaming consoles or iPad Pro). That might change maybe in like 5~10 years, but not now.
 
  • Like
Reactions: fatalogic

EnesM

macrumors 6502
May 7, 2015
447
246
Knowing Apple, switching to ARM would be done with at least a couple of major developers, like Adobe. So once released, you'd have Adobe apps compatible and ready to go.

As for virtualization, like Parallels...I don't know if it's even possible, or if it is what hit it would have on CPU to emulate x86?
 

Closingracer

macrumors 601
Jul 13, 2010
4,317
1,849
But think about the cost advantages... Intel charges a premium as the number of cores increase in the chip... For Apple, the cost differential to produce a quad-core chip versus a dual-core chip would be minimal...

So Apple could put, as an example, quad-core processors in across their whole line -- something that would have been cost-prohibitive using Intel chips in most models. Heck, they could do six cores and dedicate two cores to Intel emulation.



And you'll lose customers... I for one WILL NOT buy a Macbook based off ARM cpu. Just go look at the windows 8 experiment with it and I can tell you Apple will fare even worse.
 

Closingracer

macrumors 601
Jul 13, 2010
4,317
1,849
Someone sent me this article which does a really in-depth look at the topic. It was written in January 2015 so some things have changed with A10, but it is still mostly relevant:

http://appleinsider.com/articles/15...ck-from-moving-intel-macs-to-custom-arm-chips

A few highlights:
  • Apple pays Intel between $180-$300 per Intel processor, and the A6 (time of article) costs Apple about $25 each
  • Quote: "Apple has significant experience in porting its own OS, frameworks, apps and development tools to new hardware architectures. It moved Mac OS from 68k to PowerPC, then ported NeXT software from Intel to PowerPC to deliver OS X, when it then ported back to Intel. iOS is essentially OS X ported to ARM, with significant optimizations for mobility."
  • In 2010, Apple transitioned the Apple TV from an Intel x86 processor to an A4, modifying the underlying OS X port and making significant hardware architecture changes. They then dropped the price from $299 to $99.
The article also points out some of the risks and pitfalls. Interesting read all around...


It's more than just that.... All ill point out is windows 8 RT. You'll have a better chance of Apple making an Android device than Apple using an ARM chip in a Macbook. Perhaps a hybrid device but no way will a traditional Macbook will use arm chips.... Same reason why you can't compare an intel to an AMD chip made in the same year or even different generation chips made by the same company with the same clock speed you can't compare an Intel cpu to an ARM chip clocked at the same speed.
 
  • Like
Reactions: SteveJobzniak

mathemabeat

macrumors member
Sep 1, 2009
92
52
Cincinnati, OH
Just go look at the windows 8 experiment with it and I can tell you Apple will fare even worse.

It's more than just that.... All ill point out is windows 8 RT.

But the difference is that Microsoft had no intentions of moving their entire OS/apps platform over to ARM. RT was just designed as a super low cost Windows OS for ARM cpus. And like pretty much everything other than (x86) Windows and Office, MS didn't seem interested in actually pursuing these side projects for any length of time. They, much like google, have a long list of abandoned projects that, quite frankly, were never destined to succeed.



I think there is lots of reasons Apple might look to slowly migrate their desktop OS to ARM:

Cost: Since they control CPU design and production, they can cut out having to buy from Intel.
Performance: Their Axx series chips are beating the pants off the ARM competition and year over year the performance is increasing.
Low Power: Lets not forget, this was one of the reason's Apple switched from PowerPC to Intel to begin with.


Looking back to the 80's for some insight, Commodore was able to beat their 8 bit competition in the market because they built everything in house with their own chip designs. They controlled the whole process and fabrication and wasn't reliant on what their suppliers were doing. They bought what ever designs they needed and manufactured it themselves.

Intel has been dragging ass for years now making small 5 to maybe 10% increases between cpu generations since AMD hasn't been a threat to their desktop performance crown. I don't see Apple buying AMD (especially with the tricky x86 licensing thing in place for AMD), nor even really switching to their CPUs since they lack the mobile element that Apple is so reliant on. AMD is ok on the desktop front, but atrocious on the mobile side. No one buys AMD laptops bar the odd $199 door buster type deal.


I can totally see Apple looking into what it would take to move their desktop and laptops over to their Axx architecture. Whether consumers would like that, I have no idea. But if there was any one company that could (and has done this kind of migration before, several times), it's Apple.


I think we shouldn't be so narrow minded about this.
 

Closingracer

macrumors 601
Jul 13, 2010
4,317
1,849
But the difference is that Microsoft had no intentions of moving their entire OS/apps platform over to ARM. RT was just designed as a super low cost Windows OS for ARM cpus. And like pretty much everything other than (x86) Windows and Office, MS didn't seem interested in actually pursuing these side projects for any length of time. They, much like google, have a long list of abandoned projects that, quite frankly, were never destined to succeed.



I think there is lots of reasons Apple might look to slowly migrate their desktop OS to ARM:

Cost: Since they control CPU design and production, they can cut out having to buy from Intel.
Performance: Their Axx series chips are beating the pants off the ARM competition and year over year the performance is increasing.
Low Power: Lets not forget, this was one of the reason's Apple switched from PowerPC to Intel to begin with.


Looking back to the 80's for some insight, Commodore was able to beat their 8 bit competition in the market because they built everything in house with their own chip designs. They controlled the whole process and fabrication and wasn't reliant on what their suppliers were doing. They bought what ever designs they needed and manufactured it themselves.

Intel has been dragging ass for years now making small 5 to maybe 10% increases between cpu generations since AMD hasn't been a threat to their desktop performance crown. I don't see Apple buying AMD (especially with the tricky x86 licensing thing in place for AMD), nor even really switching to their CPUs since they lack the mobile element that Apple is so reliant on. AMD is ok on the desktop front, but atrocious on the mobile side. No one buys AMD laptops bar the odd $199 door buster type deal.


I can totally see Apple looking into what it would take to move their desktop and laptops over to their Axx architecture. Whether consumers would like that, I have no idea. But if there was any one company that could (and has done this kind of migration before, several times), it's Apple.


I think we shouldn't be so narrow minded about this.


They won't.... Compatiblty is an issue and your comparing Apple to oranges.... A 2.5ghz dual core Intel chip and an Apple theoretical cpu wouldn't be the same.... Plus you have issue with ports. Apple doesn't have the market share to dictate what their consumers to use.
 

maflynn

macrumors Haswell
May 3, 2009
73,682
43,740
As for virtualization, like Parallels...I don't know if it's even possible, or if it is what hit it would have on CPU to emulate x86?
Virtualization and emulation are two different things, Intel CPUs have virtualization technology built right into the CPU, so virtualized environment will run near the speed of the guest host OS. Emulation is emulating in software the processor, so the emulator has to change ARM CPU instructions to Intel. Its a very computationally expensive process. Back in the PPC days, we had emulators and they ran like crap. They served a purpose, but at this stage, I'd rather use a PC then buy an ARM based Mac
[doublepost=1474382086][/doublepost]
All ill point out is windows 8 RT
One of the biggest complaints with RT, is that it couldn't run all your apps. It looked like windows, it behaved like windows, but it didn't have the apps you want.

Nearly the same thing will happen if they port OS X over to ARM and start making ARM Macs. People will see it looks like a Mac, it behaves like one, but won't have the apps available.
 
Last edited:

poorcody

macrumors 65816
Jul 23, 2013
1,340
1,585
Virtualization and emulation are two different things, Intel CPUs have virtualization technology built right into the CPU, so virtualized environment will run near the speed of the guest OS. Emulation is emulating in software the processor, so the emulator has to change ARM CPU instructions to Intel. Its a very computationally expensive process. Back in the PPC days, we had emulators and they ran like crap. They served a purpose, but at this stage, I'd rather use a PC then buy an ARM based Mac
This I think may be the primary issue holding it back. I think they will need to emulate an Intel processor to support existing binaries and virtualization, and how well that goes is of course hard to say. While emulation years ago was terrible, tomorrow's chips may make the overhead become negligible. Also whereas putting processors with more cores costs a lot when you are buying from Intel, adding cores to your own chips is very cheap, so Apple would have an advantage there with trying to catch-up with performance.

One of the biggest complaints with RT, is that it couldn't run all your apps. It looked like windows, it behaved like windows, but it didn't have the apps you want.

Nearly the same thing will happen if they port OS X over to ARM and start making ARM Macs. People will see it looks like a Mac, it behaves like one, but won't have the apps available.
This to me is actually where Apple has a huge advantage, because OS X is very different than Windows. I was actually developing a commercial software package under NeXTSTEP (precursor of OS X) when they switched from Motorola to Intel chips. The process wasn't just seamless to the user, it was to the developer as well. You literally just loaded your app into the development environment and recompiled: it generated a "fat binary" with code for both processors, and everything "just worked." And that was 20 years ago. OS X was built for portability from the ground-up... the BSD UNIX core of it was also designed for portability... it's mostly a matter of recompiling and updating low-level drivers and such. That's why I think it's very doable, and you will not have the same issues Windows RT had where a lot of code had to be rewritten (and thus never was).
 

whg

macrumors regular
Aug 2, 2012
236
153
Switzerland
Apple doesn't need to create new hardware to run OS X on an ARM based machine. They already have that: Imagine an iPad Pro running OS X. We already have the keyboard and mouse support, too. The mouse just only works in selected RDP apps (like Jump Desktop). With this, the iPad Pro could indeed replace a real notebook PC. If they keep compatibility with current iPad apps we would have a really great and practical SP4 Pro replacement without the Windows handicap.
 

maflynn

macrumors Haswell
May 3, 2009
73,682
43,740
it's mostly a matter of recompiling and updating low-level drivers and such.
Its never as easy as just re-compiling. There may be a lot of work for some, and there may be little work for others, but the developers will still need to go through their code and make changes. This was a painfully slow process when Apple moved OS X from PPC to Intel. I can't see how much as changed since then

I think they will need to emulate an Intel processor to support existing binaries and virtualization,
So you're saying Apple will create an emulation later that will then support virtualization? I don't think that's possible or feasible. Again back in the PPC days the emulation was software and software is ALWAYS slower then hardware. You cannot expect good performance from an application that is emulating a CPU that will then run Windows 10

Given what you, and others posted. What will the benefit of Apple dropping Intel and going with ARM?

You mentioned Apple has to do X, Y, Z to make ir feasible, why go down that path in the first place? Why spend all that money on trying to transition over to a different platform that will assuredly anger customers and developers. Consider the negative press and how the competitors will further jump on Apple's decision.

They'll be seeing a drop in sales not an increase imo
 
Last edited:
  • Like
Reactions: SteveJobzniak

poorcody

macrumors 65816
Jul 23, 2013
1,340
1,585
Given what you, and others posted. What will the benefit of Apple dropping Intel and going with ARM?
If the numbers in the article I linked are accurate, look at it this way: take a $2000 MacBook Pro. Apple sells it to a retailer for probably around $1000 (assuming standard margins). If they are paying $250 for the Intel processor out of the $1000, imagine if they replaced it with a $25 processor? That is a huge margin gain -- CEO's dream of that kind of cost reduction. Or a $1000 laptop, wholesaling at $500, with a $180 Intel processor... an even bigger gain.

Given what you, and others posted. What will the benefit of Apple dropping Intel and going with ARM?Its never as easy as just re-compiling. There may be a lot of work for some, and there may be little work for others, but the developers will still need to go through their code and make changes. This was a painfully slow process when Apple moved OS X from PPC to Intel. I can't see how much as changed since then
I think maybe you are thinking of the progression from OS 9 to OS X -- that was horribly painful because they basically replaced the operating system on whole. Like I said as a developer I went through a transition in processors with the OS and it was remarkably easy. OS X was written very well from the very beginning with a layered approach to make it very portable. Don't forget OS X was already ported to ARM for iOS, so a lot of the work is already done. OS X/NeXTSTEP has gone from Motorola 68000 series, to Intel, to PowerPC, and back to Intel already.

So you're saying Apple will create an emulation later that will then support virtualization? I don't think that's possible or feasible. Again back in the PPC days the emulation was software and software is ALWAYS slower then hardware. You cannot expect good performance from an application that is emulating a CPU that will then run Windows 10
I sort of remember Rosetta (Apple's emulation of PPC when they switched to Intel) was actually decent (though as someone said, at the time Intel was much faster than PPC.) But yes, performance is the key issue, and there will always be a loss with emulation. The question is can the performance loss reach a point where it is negligible. History would seem to suggest with other technologies this would be likely, eventually. Probably not today though.
 

SteveJobzniak

macrumors 6502
Dec 24, 2015
489
780
Wow you're exhausting. Are you aware that you're talking to a vacuum and that Apple executives aren't actually reading this thread? If you really want to convince the people you NEED to convince, then perhaps start by sending an email to Tim Cook?

Macs are super tiny fraction of Apple's income. They don't give a damn about margins about a market they don't need. They've let MacBooks and Mac Pros stagnate for YEARS, because they really, really don't care about the macOS market. It's just kept around as a nice "halo effect" to keep Apple semi-relevant in the desktop market, but other than that they need the Mac market as little as a human needs 10 legs. All their energy goes into boosting iPhone sales.

Furthermore, your margins are wrong. Mac computers have notoriously slim retailer margins. Apple pockets something like 90% of the final sales price.

http://images.apple.com/newsroom/pdfs/q3fy16datasum.pdf

Apple's Income in Q2 2016 (before iPhones temporarily lost some popularity):

65% from iPhone.
11.9% from services like iCloud subscriptions, App Store sales commissions, etc.

10.1% from Mac computers. <---
8.7% from iPads.
4.3% from other products and services (like Apple Watch, Apple TV boxes).

But as I and others have pointed out, an ARM based (non-x86/x86_64) chip would kill the Mac as a relevant desktop. Let's look at the most relevant pros and cons.

+ Cost of the A10 is lower than an Intel chip.
+ Wattage of the A10 is lower than an Intel chip.
+ Floating point and fixed point math performance is comparable to Intel chips.
- ARM lacks the hundreds of hardware-based math functions that apps rely on for massive speed boosts; Intel's SSE, MMX, hardware-accelerated crypto, etc.
- Right now, Apple's ARM CPU cores can't talk to each other directly. So a multi-core machine would be slow.
- Apple's ARM chips have ZERO ability to communicate with Intel's desktop bus standards, which means NO PCI Express, NO Thunderbolt, NO SATA, NO DDR RAM. And forget about Intel granting Apple licenses to implement their intellectual property! Intel granted a manufacturing license to AMD decades ago, and they regret that now.
- All hardware peripheral manufacturers would need to design brand-new "ARM Mac"-specific graphics cards, soundcards, storage controllers, etc, that work with Apple's proprietary data bus.
- An ARM Mac would forever completely lack BootCamp support. No ability to use Windows would LOWER Mac sales. BootCamp is a bigger selling point than you seem to realize.
- There will be no more native CPU-accelerated OS virtualization (such as Intel's "VT-x" technology), which means that products like VMware and Parallels and VirtualBox would not work anymore.
- Trying to emulate an "Intel x86" CPU to run old Intel Mac applications would be insanely inefficient and would cause them to run at something like 5-10% of the speed you would get on an Intel CPU. Just look at how extremely slow all modern CPU emulators (like QEMU) are; and those are emulating x86 on x86! This won't ever change. Emulating a CPU in software is always going to translate 1 cycle on the native hardware to something like 20-50 cycles when emulated instead.
- If they want to emulate an advanced, modern x86 CPU, they would have to spend years writing an emulator to accurately emulate the hundreds of hardware-accelerated functions, registers and undocumented legacy bugs and hardware quirks of Intel CPUs.
- Apple has already made the mistake of going proprietary once; it was called "PowerPC" and it meant that they were stuck with a dead-end CPU that nearly killed Apple since they couldn't provide faster machines, and they were forced to make a painful switch to Intel, which killed off all Macs in existence. They did it for a great reason: Great performance gains and future safety. But going to ARM would mean less performance and less safety. So why do it!?
- They would throw away all existing applications, a lot of which will never be updated by their developers again. All developers would have to re-compile and TEST their applications (and watch out for compiler bugs) on the new ARM based Macs, which means that every developer in existence would need to buy at least one ARM Mac. How much do you want to bet they'd rather just say "**** off, Apple, you money pinching whores, throwing out perfect Intel x86 hardware for your own proprietary, locked-down ****? **** off, I am leaving!".
- They would fragment their existing market into Intel and ARM, and piss off all existing customers. Many of which would abandon Apple in disgust.
- Apple would isolate themselves from the rest of the planet. Meaning they would lose access to the wide array of Windows/Linux/BSD applications that have been written for Intel machines and are therefore super-trivial to port to current Macs. We can thank Apple's transition to Intel for the fact that apps on Macs are thriving nowadays, and are developed alongside Windows versions, despite Macs being a smaller market. With the switch to Intel, Apple gave familiar hardware to a legion of pre-trained developers. Some of which simply installed Hackintosh versions of OS X on their regular PCs to be able to join the Mac platform and develop for it. The great Intel app availability took Macs out of obscurity and into the mainstream. That would all be gone if Apple tries to stubbornly isolate themselves from the rest of the world.
- The huge barrier-to-entry for software developers would strangle the development of new/updated apps, and that would quickly lead to the death of the Mac. And in case you don't realize it: Profit margins are irrelevant on dead products that nobody wants.
- They would lose all the software and hardware compatibility, performance and clockwork Intel CPU upgrades year after year that made Intel Macs the best selling machines in Apple's history.


Well... Good luck with your board meeting with Apple. Looks like you're onto a winner! :) Keep on fighting the good fight against all the unbelievers! :)
 
Last edited:

maflynn

macrumors Haswell
May 3, 2009
73,682
43,740
I think maybe you are thinking of the progression from OS 9 to OS X
No, I'm thinking specifically of the transition to intel and how Apple mentioned it was a simple process, but in actuality, it took some time for apps to be native to X86 instead of PPC Apple asked a lot of their developers but the goal was simple, sell more Macs to customers.

That is a huge margin gain
I'm not disputing the savings, but what's not worked into the numbers is the R&D needed to make that possible, and you're missing my whole point, yes Apple will increase the margins, that's great but will people buy them?

What is Apple's end game in doing that, increasing margins or increasing sales? I don't think you can have both [with Apple changing to ARM]

You seem hung up on having Apple save even more money but the fact remains at this stage Apple is seeing a huge reduction in Mac sales. There were stories about a 40% drop in laptop shipments a few months ago. Apple should be less concerned about sales margins and more concerned about designing a computer that people want to buy.
 
  • Like
Reactions: SteveJobzniak

oneMadRssn

macrumors 603
Sep 8, 2011
6,089
14,195
Wow you're exhausting. Are you aware that you're talking to a vacuum and that Apple executives aren't actually reading this thread? If you really want to convince the people you NEED to convince, then perhaps start by sending an email to Tim Cook?

Macs are super tiny fraction of Apple's income. They don't give a damn about margins about a market they don't need. They've let MacBooks and Mac Pros stagnate for YEARS, because they really, really don't care about the macOS market. It's just kept around as a nice "halo effect" to keep Apple semi-relevant in the desktop market, but other than that they need the Mac market as little as a human needs 10 legs. All their energy goes into boosting iPhone sales.

Furthermore, your margins are wrong. Mac computers have notoriously slim retailer margins. Apple pockets something like 90% of the final sales price.

http://images.apple.com/newsroom/pdfs/q3fy16datasum.pdf

Apple's Income in Q2 2016 (before iPhones temporarily lost some popularity):

65% from iPhone.
11.9% from services like iCloud subscriptions, App Store sales commissions, etc.

10.1% from Mac computers. <---
8.7% from iPads.
4.3% from other products and services (like Apple Watch, Apple TV boxes).

But as I and others have pointed out, an ARM based (non-x86/x86_64) chip would kill the Mac as a relevant desktop.

+ Cost of the A10 is lower than an Intel chip.
+ Wattage of the A10 is lower than an Intel chip.
+ Floating point and fixed point math performance is comparable to Intel chips.
- ARM lacks the hundreds of hardware-based math functions that apps rely on for massive speed boosts; SSE, MMX, hardware-accelerated crypto, etc.
- Right now, Apple's ARM CPU cores can't talk to each other directly. So a multi-core machine would be slow.
- Apple's ARM chips have ZERO ability to communicate with Intel's desktop bus standards, which means NO PCI Express, NO Thunderbolt, NO SATA, NO DDR RAM. And forget about Intel granting Apple licenses to implement their intellectual property!
- All hardware peripheral manufacturers would need to design brand-new "ARM Mac"-specific graphics cards, soundcards, storage controllers, etc, that work with Apple's proprietary data bus.
- An ARM Mac would forever completely lack BootCamp support. No ability to use Windows would LOWER Mac sales. BootCamp is a bigger selling point than you seem to realize.
- There will be no more native CPU-accelerated OS virtualization (such as Intel's "VT-x" technology), which means that products like VMware and Parallels and VirtualBox would not work anymore.
- Trying to emulate an "Intel x86" CPU to run old Intel Mac applications would be insanely inefficient and would cause them to run at something like 5-10% of the speed you would get on an Intel CPU. Just look at how extremely slow all existing CPU emulators are; and those are emulating x86 on x86! This won't ever change. Emulating a CPU in software is always going to translate 1 cycle on the native hardware to something like 20-50 cycles when emulated instead.
- If they want to emulate an advanced, modern x86 CPU, they would have to spend years writing an emulator to accurately emulate the hundreds of hardware-accelerated functions, registers and undocumented legacy bugs and hardware quirks of Intel CPUs.
- Apple has already made the mistake of going proprietary once; it was called "PowerPC" and it meant that they were stuck with a dead-end CPU and had to make a painful switch to Intel, which killed off all Macs in existence. They did it for a great reason: Great performance gains. But going to ARM would mean less performance. So why do it!?
- They would throw away all existing applications, a lot of which will never be updated by their developers again. All developers would have to re-compile and TEST their applications (and watch out for compiler bugs) on the new ARM based Macs, which means that every developer in existence would need to buy at least one ARM Mac. How much do you want to bet they'd rather just say "**** off, Apple, you money pinching whores, throwing out perfect Intel x86 hardware for your own proprietary ****? **** off, I am leaving!".


Well... Good luck with your board meeting with Apple. Looks like you're onto a winner! :) Keep on fighting the good fight against all the unbelievers! :)

Great points.

I remember having PPC Macs, and while I loved them, it was super frustrating to find apps that worked. It was no surprise that most apps were made by Apple. However, I always kept a Windows machine as well for those little single-purpose utilities that could be easily found in opensource or somewhere.

When Apple switched to Intel, the immediate benefit I saw was that all of a sudden, all the open source Linux apps were rebuilt for OSX and it's UI. This seemed to happen overnight, as the changes were trivial to those developers now that it was all on x86. This was huge. I no longer needed to keep a Windows machine just for useful little utilities. Within a few years, tons of developers were making great Mac apps alongside their Windows counterparts.

A change to ARM would completely undo that. Unlike in the smartphone market, there is very little incentive for developers to support anything other than x86.

As someone above said, the biggest implication of Mac switching to ARM is that I will no longer buy a Mac.
 

poorcody

macrumors 65816
Jul 23, 2013
1,340
1,585
You know maybe I should point out I am making the assumption that using A# chips would be seamless to the user: new apps would be "fat binaries" that transparently ran on "old and new" processors, and existing binaries would run via emulation. I realize many doubt it could be done seamlessly; I think I find it plausible because when NeXTSTEP (precursor of OS X) did it, it really was painless to the user, and even most developers. But they didn't have the emulation half which I think is the bigger roadblock.

99% of developers don't deal with anything related to the underlying processors. We mostly use high-level languages that talk to APIs. The APIs are not tied to the processor like they once were. Only low-level OS routines and compilers deal with processor directly. That's why portability is more plausible than it once was.

I totally agree that if the user was disrupted in even a small way it wouldn't be worth it, since most of the advantages would seem to go to Apple. I actually started this thread with the intention of trying to fathom any significant advantages to the user...
[doublepost=1474397640][/doublepost]
Wow you're exhausting. Are you aware that you're talking to a vacuum and that Apple executives aren't actually reading this thread? If you really want to convince the people you NEED to convince, then perhaps start by sending an email to Tim Cook?
I'm not trying to convince anyone of anything... I just find it an interesting idea to contemplate from the technical and business perspective. It's all academic, just for fun...

Furthermore, your margins are wrong. Mac computers have notoriously slim retailer margins. Apple pockets something like 90% of the final sales price.

http://images.apple.com/newsroom/pdfs/q3fy16datasum.pdf
There is nothing in that link about margins.
 
Last edited:

SteveJobzniak

macrumors 6502
Dec 24, 2015
489
780
Yes, new apps would be very fat binaries (PowerPC+Intel32+Intel64+ARM), but that implies that there would be any new apps, after all developers and users abandon Apple. Intel (not Apple) is the one to thank for 90% of all apps available for Mac today. Because the switch to Intel opened up the platform to developers who weren't already Apple fanatics. Countless apps and major software development libraries are ports from Windows, Linux and BSD thanks to everything running on Intel CPUs and being trivial to port to OS X. And countless new developers came to us from the Wintel world.

Apple would lose all of that if they betray their users and go proprietary.

The only way Apple could even *begin* to pull off a switch to ARM, would be to make the next iPad Pro run a dual-boot of iOS and macOS as a "test pilot" to get the ball rolling, but then that would suffer from a total lack of apps, so it would be more like a toy/curiosity than a serious OS. Not to mention that it isn't a touch-optimized OS and would be too terrible for Apple to release.

Even if they manage to make an iPad Pro run macOS for fun, there's still no way they'll be able to retain developers and users if they switch all computers to ARM, though. They'll become a fringe product again by losing all Hackintosh developers, all cross-platform libraries, all users who want to run old apps, all users who want to run Windows, and all hardware support (say goodbye to PCI Express and Thunderbolt), etc. The barrier to software development would also be so high that they would strangle all future software development growth, which would quickly kill off the Mac platform.

Oh and it would be slower than Intel CPUs.

Apple stands to gain nothing from this. Higher profit margins on lower overall sales = less money. Lower overall sales = the death of Mac desktops.
[doublepost=1474398740][/doublepost]
There is nothing in that link about margins.

Uhh yeah? I never implied that the link was about margins. That link was about Apple's revenue by product category.

As for margins, you said "take a $2000 MacBook Pro. Apple sells it to a retailer for probably around $1000 (assuming standard margins)", and I simply told you not to assume standard margins since Apple's retailer margins are notoriously slim. That's why so few non-official stores sell Apple's products. Furthermore, Apple forces retailers to set certain prices, to not undervalue their products in the eyes of consumers.

Here's info about the profit margins for stores that resell Apple products: https://www.quora.com/What-is-the-average-gross-margin-of-retailers-of-Apple-products

Apple's 3rd party retailers only get an average margin of 8%. Some product margins are as low as 2-5%. So for a $2000 MacBook Pro, Apple sells it to the retailer for about $1850. Due to those slim margins together with the general overhead costs of running a store, most 3rd party retailers actually break even or lose money when selling Macs, which forces them to instead make their money by selling addons (printers, software) and services (insurance, repairs, etc). That can still be lucrative, since Mac users are notoriously rich, big spenders.

Anyway, where was I... oh yeah, I was about to install macOS Sierra! :)
 
Last edited:

poorcody

macrumors 65816
Jul 23, 2013
1,340
1,585
You're right about the margins... that is interesting... this is why I like discussions, you learn things. :)

As for compatibility, I will just say while binary compatibility is useful in the UNIX world, developers are primarily dealing with OS/API issues when porting, not the underlying processor. Most developers would not care if the processor changed as long as the language and API calls were the same. Then it's just a recompile.

Anyway, where was I... oh yeah, I was about to install macOS Sierra! :)
You are a brave soul... I keep OS updates under a virtual machine until version ".1"! ;)
 

leman

macrumors Core
Oct 14, 2008
19,530
19,709
A very likely implication would be a noticeable performance hit. The A10 is designed for low-power scenarios, I would be very surprised if it would scale up to laptop/desktop environment.

P.S. And please don't quite me Geekbench scores. I want to see some real-world application stuff. Geekbench is not comparable across architectures.
 
  • Like
Reactions: Samuelsan2001

Essenar

macrumors 6502a
Oct 24, 2008
553
186
GeekBench is a really stupid metric to go by, so please do some more research because the A10 is not faster than most x86 processors out today. It's quick... But not, "replace a Skylake i5 in a notebook" quick and it can't handle x86 architecture which is like... A huge aspect to the processor.
In general, the more architecture a chip has, the more heat it generates.
This is why AMD GPU's are in Mac Pro's (other than the Nvidia lawsuit) and why AMD gaming GPU's are much slower/hotter than Nvidia gaming GPU's. AMD gaming GPU's include FPP and Open CL computation instructions on the die, which causes the chip to run hotter because it just has more stuff in it.
An A10 that runs x86 would be larger, run hotter, and thus slower.
 
  • Like
Reactions: Closingracer

SteveJobzniak

macrumors 6502
Dec 24, 2015
489
780
You are a brave soul... I keep OS updates under a virtual machine until version ".1"! ;)

I'm writing from Sierra now. ;) Went perfectly apart from two third party apps suddenly requiring the discrete GPU. But they seem fixable.

As for compatibility, I will just say while binary compatibility is useful in the UNIX world, developers are primarily dealing with OS/API issues when porting, not the underlying processor. Most developers would not care if the processor changed as long as the language and API calls were the same. Then it's just a recompile.

The problem isn't cross-platform compilation. The problem is everything I listed in the two large posts above (one and two), such as driving away users by losing features like Windows and virtualization and all old existing Intel software and all hardware peripheral compatibility, and driving away/preventing new developers from even joining in the first place since ARM is a fringe platform. And perhaps above all; the risk of being stuck in a dead end again a la PowerPC.

In general, the more architecture a chip has, the more heat it generates. An A10 that runs x86 would be larger, run hotter, and thus slower.

Correct.
 

leechers

macrumors member
Oct 19, 2012
40
31
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.