Got a tip for us?
Let us know
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
Apple Interested in Developing ARM-Based Mac Processors and iPhone Modems in House
- Thread starter MacRumors
- Start date
- Sort by reaction score
Yeah, the iCloud and Store, where Apple makes the majority of its ongoing income is hosted on an array of A10 chips.
Give me a break.
This had been rehashed at least 100 times.
ARM is not in the same class as the current generation i5, let alone an i7 or Xeon.
ARM is great as a tablet and mobile processor. It is not an x86 competitor.
Research and you will see how many companies went out of business trying to compete with Intel making processors.
Oh, remember Apple was one of those failed attempts with AIM and PowerPC.
Lack of virtualization and threads along with a host of other things make it ARM unsuitable for a desktop.
ARM is not in the same class as the current generation i5, let alone an i7 or Xeon.
ARM is great as a tablet and mobile processor. It is not an x86 competitor.
Research and you will see how many companies went out of business trying to compete with Intel making processors.
Oh, remember Apple was one of those failed attempts with AIM and PowerPC.
Lack of virtualization and threads along with a host of other things make it ARM unsuitable for a desktop.
Don’t underestimate x86. It’s a big ole Ford F-350 turned into an RV motor home with full kitchens baths and however many rooms you want....and it can also be a dump truck, cargo hauler, bus or whatever else you need. If you dig deep in the engine there’s still some Model T parts in there! What most people NEED is a good SUV sized vehicle though.
ARM is like a Honda.. it’s a fancy motorbike that’s been upgraded into an SUV sized vehicle. Under the hood it’s still motorbike parts..
there’s some big things that need to happen for ARM to replace x86, especially for developer and media creation tasks. There’s no ARM processors out there that can do the sheer brute force bit-hauling of x86... if Apple burns that bridge with their own hardware they’ll get screwed totally by Intel... Intel holds grudges a LONG time.
I don't think emulation will be difficult technically speaking, but legally Intel has already told Microsoft to watch out. I think Intel would be smart to license their emulation just like ARM licenses its designs, it seems inevitable to move away from x86 at a hardware level eventually.
You’re overlooking a small detail in “control over the supply chain” for chips. Who is going to manufacture them?
Apple relies on TSMC to manufacture chips it designs. Apple relies on Intel to manufacture chips designed by Intel. If Apple starts designing the chips now designed by Intel, who is going to fab them?
Probably not Intel, though it has dipped its toe into the foundry world. And the core of Intel’s business, which has extraordinarily high barriers to entry and which Intel does better than anyone, is fabricating chips. Maybe Apple could persuade Intel to operate as a foundry for Apple, but it’s not likely.
TSMC? That’s its business and it’s already familiar with Apple’s designs and needs. But, IIRC, Apple already represents a scary-high 19% of TSMC’s business. That’s verging on the point of Apple “owning” TSMC in the event things go sideways with TSMC’s business, which doesn’t make the finance and accounting people comfortable. Tack on the business represented by what Intel currently does, and TSMC would be all but a subsidiary of Apple.
Maybe GloFo, but there’s a big ramp up there and many of the same concerns as with TSMC.
Not Apple. Not unless it’s crazy. Apple doesn’t (or certainly shouldn’t) want to be in the business of developing process technology and building and operating fabs. There is a reason the IBM got out of this business. There is a reason AMD was forced out of this business. There is a reason that Intel is the last remaining company in the world that both designs and fabricates high performance logic.
Today, just building a leading edge fab costs ~$8B. And then you have to do t again every few years. That’s to say nothing of the thousands of chemical, materials science, and other engineers, PhDs, factory workers, and support staff needed to develop the technology and keep the place running. None of which Apple currently has. Cash hoard or not, I can’t for the life of me think of why Apple would want to dump $50+ billion cash just to get into that business. Which is also why I don’t see Apple placing itself in a position of being forced, by dint of being an overwhelming part of its foundry’s portfolio, into buying or dumping billions of dollars year after year to prop up a foundry.
Apple may or may not need Intel’s designs. Though, for reasons discussed previously, I believe Apple’s current ARM designs probably don’t scale or perform as well as Intel’s in the high performance desktop and laptop environment as simplistic GeekBench scores might lead some to believe. But Apple does need, or in any case would be financially foolish to abandon, Intel’s manufacturing services.
Actually Apple, Qualcomm and Samsung have the most expensive ARM license.
It gives the access to the architecture and allows them to modify it at long as they pass compatibility tests.
There is plenty stopping them from extending. They must pass *ALL* compatibility suites.
They also can't extend in a direction that would make other incompatible.
No extensions in the instruction set.
Finally how many of you understand processors and processor technology?
Virtualization, shadow registers, instruction set design, cache coherency (MOSEI, MOSEIF, MESI, etc.)?
I have been designing chips for more than 25 years from 3.5 micron down to 14nm FinFET.
Apple switching from Intel to ARM isn't going to happen in their general computing machines.
ARM doesn't have the performance, features, etc.
The comparison is a dual Core i5 to a 6 core ARM in a synthetic benchmark.
Wait until yo ned to move real datasets around that require more than the 2-3GB of memory in a phone.
Wait until you need to connect real peripherals and real disks.
As far as OS porting?
iOS is nothing but a stripped MacOS with a different interface layer.
They are both BSD.
You’re overlooking a small detail in “control over the supply chain” for chips. Who is going to manufacture them?
Apple relies on TSMC to manufacture chips it designs. Apple relies on Intel to manufacture chips designed by Intel. If Apple starts designing the chips now designed by Intel, who is going to fab them?
Probably not Intel, though it has dipped its toe into the foundry world. And the core of Intel’s business, which has extraordinarily high barriers to entry and which Intel does better than anyone, is fabricating chips. Maybe Apple could persuade Intel to operate as a foundry for Apple, but it’s not likely.
TSMC? That’s its business and it’s already familiar with Apple’s designs and needs. But, IIRC, Apple already represents a scary-high 19% of TSMC’s business. That’s verging on the point of Apple “owning” TSMC in the event things go sideways with TSMC’s business, which doesn’t make the finance and accounting people comfortable. Tack on the business represented by what Intel currently does, and TSMC would be all but a subsidiary of Apple.
Maybe GloFo, but there’s a big ramp up there and many of the same concerns as with TSMC.
Not Apple. Not unless it’s crazy. Apple doesn’t (or certainly shouldn’t) want to be in the business of developing process technology and building and operating fabs. There is a reason the IBM got out of this business. There is a reason AMD was forced out of this business. There is a reason that Intel is the last remaining company in the world that both designs and fabricates high performance logic.
Today, just building a leading edge fab costs ~$8B. And then you have to do t again every few years. That’s to say nothing of the thousands of chemical, materials science, and other engineers, PhDs, factory workers, and support staff needed to develop the technology and keep the place running. None of which Apple currently has. Cash hoard or not, I can’t for the life of me think of why Apple would want to dump $50+ billion cash just to get into that business. Which is also why I don’t see Apple placing itself in a position of being forced, by dint of being an overwhelming part of its foundry’s portfolio, into buying or dumping billions of dollars year after year to prop up a foundry.
Apple may or may not need Intel’s designs. Though, for reasons discussed previously, I believe Apple’s current ARM designs probably don’t scale or perform as well as Intel’s in the high performance desktop and laptop environment as simplistic GeekBench scores might lead some to believe. But Apple does need, or in any case would be financially foolish to abandon, Intel’s manufacturing services.
Because they were having performance issues that IBM could not solve.
IBM was interested in performance and not low power.
The Power architecture was never good for laptops and portables.
So when IBM could not deliver power efficient processors, Apple switched to Intel.
Except all of these still have had their issues.
Rosetta couldn't full emulate all PPC software (mainly high performance or G5 based) and MSs Windows 10 ARM only supports 32bit software and ARM drivers (don't know about performance).
Apple is so slow at everything it does due to it's size so don't expect anything radical anytime soon. What you are cheering for here is iPod in notebook casing. I'm sure it will happen down the road but it will be expensive and it will not do the job that MacOS can.
App gap would be ridiculous to overcome and while your "Architectural changes" theory makes sense you have to remember that app game is in x86. It's completely different going from anything irrelevant to x86 and then from x86 to irrelevant. Apple notebooks surged thanks to Intel so app developers can cover all desktop OSes at once. Good luck making Adobe, Microsoft, Autocad, Avid, BMD and many others switch to Apple's new architecture whatever that may be. Some cannot bother to code for Mac today while the discrepancies between MacOS and Windows are minimal compared to the past.
If you are going to emulate x86 then what's the purpose of switching to new architecture in the first place? Windows is supporting ARM to try to conquer tablet market and mobile market not to try to do it on desktop like Apple is implying here.
App gap would be ridiculous to overcome and while your "Architectural changes" theory makes sense you have to remember that app game is in x86. It's completely different going from anything irrelevant to x86 and then from x86 to irrelevant. Apple notebooks surged thanks to Intel so app developers can cover all desktop OSes at once. Good luck making Adobe, Microsoft, Autocad, Avid, BMD and many others switch to Apple's new architecture whatever that may be. Some cannot bother to code for Mac today while the discrepancies between MacOS and Windows are minimal compared to the past.
If you are going to emulate x86 then what's the purpose of switching to new architecture in the first place? Windows is supporting ARM to try to conquer tablet market and mobile market not to try to do it on desktop like Apple is implying here.
The software side is not the problem. First of all, most developers already make their software cross platform (ready), so it is usually just a matter of a recompile for the majority of apps.
All of macOS framework and api's will remain the same. Many of the big developers have been acutely aware of frequent architecture changes. Remember that even x86 architecture has evolved during the years, and you have to be a super specialist to be on the edge on SSE, AVX etc. It is much easier to use third party or system frameworks, and let the compiler be the expert on the instruction sets. The few lines of assembly code left in legacy code, are not the hardest work to change anymore.
Microsoft has for a long time almost begged developers to avoid machine specific assembly code. The reason have been the failed transitions to Itanium and how it went with Windows RT. Now they have focused a lot on their Windows Store and the .Net framework. Both which do not cater to hardware dependent code.
In Linux/BSD/other unices there has always been a wide variety of architectures, and so software already are agnostic to it. Some do have hardware specific non portabel code, but it is rare and usually someone else have added additional code for other architectures if the software is popular enough.
For whatever that is left, x86 emulator will take care if it until the main applications are ported.
My prediction is simple, if Apple see that Intel are not able to give them a CPU that are able to be comparable to their own developed ARM chip, they will start the transition program. We are not there yet as far as I know, but lots of Intel and Apple roadmaps are hidden deep down in NDA land, and very few people know in detail what will be in 2-4-10 years from now. The Apple A processor are very capable, but the Intel chips have the SIMD instructions and lots of speciality which are missing.
All of macOS framework and api's will remain the same. Many of the big developers have been acutely aware of frequent architecture changes. Remember that even x86 architecture has evolved during the years, and you have to be a super specialist to be on the edge on SSE, AVX etc. It is much easier to use third party or system frameworks, and let the compiler be the expert on the instruction sets. The few lines of assembly code left in legacy code, are not the hardest work to change anymore.
Microsoft has for a long time almost begged developers to avoid machine specific assembly code. The reason have been the failed transitions to Itanium and how it went with Windows RT. Now they have focused a lot on their Windows Store and the .Net framework. Both which do not cater to hardware dependent code.
In Linux/BSD/other unices there has always been a wide variety of architectures, and so software already are agnostic to it. Some do have hardware specific non portabel code, but it is rare and usually someone else have added additional code for other architectures if the software is popular enough.
For whatever that is left, x86 emulator will take care if it until the main applications are ported.
My prediction is simple, if Apple see that Intel are not able to give them a CPU that are able to be comparable to their own developed ARM chip, they will start the transition program. We are not there yet as far as I know, but lots of Intel and Apple roadmaps are hidden deep down in NDA land, and very few people know in detail what will be in 2-4-10 years from now. The Apple A processor are very capable, but the Intel chips have the SIMD instructions and lots of speciality which are missing.
In regards to Apple using TSMC, or Samsung or any other foundry for chip production, it's not the same as waiting for Intel to design, produce and release their chips. Apple, in designing their own silicone, has the ability to work with multiple sources for production, and can dictate when they need deliverable product. They don't have that ability with Intel. Or they can certainly say they want product on X date, but when Intel fails to hit that date, Apple doesn't really have any way to recover from the failure.
And the thing about designing and producing chips for laptop or desktop use, the numbers needed are going to be such small volume, compared to iPhone / iPad / Watch / ATV needs that it's not likely to present an issue for any manufacturer. Apple's recent acquisition of Toshiba's chip unit is likely a step in the direction of their owning their own fab, or at least one they can use more exclusively to avoid roadblocks in bringing product to market.
Apple could buy / build as many chip foundries as they need and still have billions in cash stockpiled, if that is what they ultimately want or need to do.
Yes it can, and of course it has a performance penalty. Their processors would have to reach a certain level before the performance penalty is negligible, but even then there are differences. Microsoft have already created an ARM based Windows with X86 emulation, it’s not like it’s out of the question especially when those products come to market.
Big downgrades are par for the course with a Mac mini update aren't they?
Yes Power PC went so well they switched to Intel.
The Mac App Store is a mess because it is a flawed business model not because its difficult to develop for.
Unsurprisingly most software developers don't want to keep coughing up a chunk of their revenue to the richest company in the world and given any choice about it they won't.
No, no they wouldn't. Best case scenario is they maintain where they are now.
What’s the obsession with an ARM based Mac mini about? Is it just desperation for a new Mac mini or is there a reason? The Mac mini processors are moving to quad-core this year, there’ll be a significant performance boost there once it’s refreshed.
If current treatment of the pro market is any indicator of the future, I suspect this won't matter all that much to Apple. Pro market is dwarfed by the consumer and prosumer markets that are satisfied with less featured and lower cost software substitutes.
I don't get any sense that Apple is eyeing pro market in any given field (that also demands x86) as critical to future success. At best I think this market can expect to get niche and irregular updates using x86 processors, such as the serviceable mac pro alluded to earlier this year. The consumer and prosumer markets for PCs are so clearly held back by Intel at this point.
In what way? What do consumers and prosumers need in an x86 machine that isn't actually available from Intel?
Apple doesn't care about professional users much, that much is fairly clear, they also don't seem to have any understanding of what professional users want or need, that isn't Intel fault.
Blasphemous! The thought of throwing any Apple device against the wall would never occur to me.
[doublepost=1506712092][/doublepost]
Exactly!
