Of course Windows is supporting ARM now, so there's no reason that by the time the switch happens that Windows wouldn't run on it. It will not be the entire lineup, and probably be something in the Air category first.
OK This is going to another one of those subjects where nobody on here thinks it will happen, then, what-do-you-know...it does!
I think what some of you fail to see, is that as soon ARM processors are fast enough to become 'adequate' (by that I mean good enough for ordinary people to watch films and browse etc), the line between tablet and laptop will blur and Intel will just fade away as a niche processor for high end users.
My guess is that the first 'laptops' with ARM will be IOS based machines targeting low end users.
Well that's my prediction, feel free to re visit this post in 3 years.
Apple's been down this road before... it would really surprise me if they went 100% into a non-intel design. How fast we forget history.
Yes, how fast. 11 years ago, die hard Apple users would have said "It would really amaze me if they went 100% into a non-PPC design.'
This OS has been ported and sold to run on at least 4 platforms 68K SPARC, HP-RISC, PPC, X86 and ARM.
Moving the Mac fork to ARM is purely a capability curve decision. we've had this discussion before... ARM is improving faster within it's own space, but in the overall performance/power/price curve Intel has the edge. If there is a change, Apple will adjust. And given that Apple enjoys tweaking the chip to match it's performance requirements, moving away from stock Intel (say a custom chip built by, say, AMD) would be 100% within their design parameters.
The real game changer is the move away from 'WinTel compatibility' as a must have for a set of customers. Not too long ago, Boot Camp was a key marketing bullet. You don't see it now. If that continues, I think the key issue is Apple 'controlling' its key components that tightly integrate with their OS and app compiler platforms... that leads to 'inhousing' CPU design, which at that point if the ARM price performance power curve is right, they'll move. If not, it wouldn't surprise me for Apple to buy a controlling share of AMD, or strike a deal with Intel for a x86 design fork unique only to Apple.
Lots of people will argue that ARM is inappropriate for the Mac, and that Intel is a better choice. These arguments are irrelevant. Apple is targeting the consumer market, not the high end professional. ARM chips are more then sufficient for this market.
I don't agree with this. Even consumers are doing video editing / processing, photo editing, audio conversion, running multi-apps, number processing, etc.
If Apple transitions to ARM, I don't believe it will be 100%. I believe their low-end machines will, but the rest will be Intel. ARM isn't going to catch up to Intel. Intel has so much speed in their back pocket right now, they could literally jump another 1-2ghz overnight if they wanted to. They are simply waiting for AMD to catch up.
Assuming Apple transitions to ARM 100%, their computers are going to continue to be inferior to PC/Intel computers. That also assumes that Apple doesn't care about the developers that got them to where they are at today. We (developers) pine for performance. No one likes to build 50 times a day on slow hardware. Also, consumers aren't dumb, they still take numbers into account when purchasing items. A huge portion of Apple's consumer base will drop off if this happens, as their software isn't so much better than the competition that consumers will completely ignore performance on the other side.
I'm not so sure. I'm not a processor expert but I believe there is something fundamental about ARM chips which makes them pretty hard to compete with when taking into consideration power consumption/speed.
As laptops are already running red hot, power consumption becomes more and more important, favouring ARM designs going forward.
Intel may one day come out with a quantum processor, but that technology is still a long way away. Till then, my money is on ARM (which it is literally as I own ARM stock).
This is fairly spot on if you ask me.
I am also far short of being a chip expert, but even the tiniest amount of searching will show that the intel x86 designs are really not that good. They get good performance expressly because they don't care about things like heat dissipation and efficiency. It's all about raw power. They are neither efficient nor particularly well designed because they operate under almost none of the constraints that ARM chips have.
Anything can happen, but intel is on top right now because of the special "Wintel" relationship and because of history.
They certainly aren't on top because they are the best designed chips around.
I'm not so sure. I'm not a processor expert but I believe there is something fundamental about ARM chips which makes them pretty hard to compete with when taking into consideration power consumption/speed.
As laptops are already running red hot, power consumption becomes more and more important, favouring ARM designs going forward.
Intel may one day come out with a quantum processor, but that technology is still a long way away. Till then, my money is on ARM (which it is literally as I own ARM stock).
It's my understanding that ARM is great at the low end but can't be scaled to 3 and 4GHz like Intel chips to get the same performance per watt.
ARM based Macs will come but Intel based Macs will remain.
While hiring AMD veteran Jim Mergard adds more CPU design expertise at Apple, it does not necessary mean they will drop Intel. Hec, they may license the basic x86 design from Intel or AMD and make their own desktop CPUs.
Time will tell.
I think you're mostly on the right track. I do not, however, see Macs running ARM any time soon. Instead, I see an iPad Pro (perhaps even with a keyboard and MBA configuration) which would be one step above an iPad, but still far below a Mac. And they could make something that looks like an iMac (albeit probably with a much smaller screen due to GPU limitations) that could use faster processors since battery life wouldn't be an issue.
But I just don't see any chance of Macs switching. Why is it that all of these ARM proponents keep talking about ARM performance improving but assume that x86 performance will not?
Apple discontinued Rosetta with Lion. This was a big step in the move to ARM.
That doesn't make any sense.
If they do switch to ARM, they're going to need some kind of compatibility layer for apps that have not yet been ported. So they'd need to resurrect Rosetta. So how does canceling Rosetta help them down the path toward a future which would require Rosetta?
I guess I'll be switching back to pcs then. Cause I want a processor that's powerful and compatible with windows.
Stupid move IMO
I like ARM in phones Etc
Hopefully it's just be like a version of the MacBook Air and not all MacBooks
The change will not be any time soon, as stated.
And by the time it *does* happen, whatever Apple chooses WILL be powerful enough. Apple isn't stupid, LOL. They understand the market better than anyone.
It's my understanding that ARM is great at the low end but can't be scaled to 3 and 4GHz like Intel chips to get the same performance per watt.
This is correct. In fact at those speeds (currently) Intel trounces ARM. Inherently x86 and ARM were designed to do very different things. As technology progresses the line blurs, but at it's core many things remain the same with each respective design, which is why such a move seems odd. Intel probably could catch up to ARM in PPW before ARM could catch Intel on performance. Dark Horse: Intel still has a license for ARM, though they maintain they have no intention of using it.
I am also far short of being a chip expert, but even the tiniest amount of searching will show that the intel x86 designs are really not that good. They get good performance expressly because they don't care about things like heat dissipation and efficiency. It's all about raw power. They are neither efficient nor particularly well designed because they operate under almost none of the constraints that ARM chips have.
Anything can happen, but intel is on top right now because of the special "Wintel" relationship and because of history.
They certainly aren't on top because they are the best designed chips around.
This is what the danger of "a little knowledge" looks like. There is a small theoretical advantage to ARM simpler instruction set.
But this is swamped by execution. The real reason ARM was historically so much better at mobile, is because that is what they have been building for, for years, while Intel OTOH concentrated on high performance.
In case you haven't noticed, Intel is waking up. Check out the Motorola Razr i, with an Intel Atom inside. It was the only phone to regularly beat the new A6 iPhone on benchmarks, and it is still gets good battery life.
Next year, Intel is releasing a brand new Atom Architecture (first in 5 years) it will be arriving on Intels mature 22nm process and coming in varieties ranging from 1 core to 4 core.
After that Intel is planning to start doing Tic-Toc iterations in mobile, like they do on desktops (where the obliterated AMD).
In three years most new Android phones could be running Intel Atoms and making iPhones looks slow in comparison and all the talk will be about how Apple is going to switch iOS to Intel and ditch non competetive ARM.
The truth is we don't know, but anyone writing off Intel in Mobile or saying an Apple switch of OSX to ARM is inevitable, is clueless.
Nothing a few years out in the CE or computer world is inevitable. For all we know in a few years time Intel may have chips for the iPhone that blow ARM away, use less power and Apple uses those in their mobile devices instead, and keeps using improved Intel chips in Macs. Nothing is guaranteed here.
I agree. I think the article's point is that Apple is preparing to move the Arm into the Macs, which they will do if and only if it makes sense. And as you say, they might surprise us the other way and move Intel into the iOS world, if and only if that makes sense. But I think the take-home point is there are may different ways the technology may go in the next few years, and Apple does a pretty good job staying light on its feet unlike some of its competitors.
Comments
Of course Windows is supporting ARM now, so there's no reason that by the time the switch happens that Windows wouldn't run on it. It will not be the entire lineup, and probably be something in the Air category first.
OK This is going to another one of those subjects where nobody on here thinks it will happen, then, what-do-you-know...it does!
I think what some of you fail to see, is that as soon ARM processors are fast enough to become 'adequate' (by that I mean good enough for ordinary people to watch films and browse etc), the line between tablet and laptop will blur and Intel will just fade away as a niche processor for high end users.
My guess is that the first 'laptops' with ARM will be IOS based machines targeting low end users.
Well that's my prediction, feel free to re visit this post in 3 years.
No way. A full x86 compatibility is a must. Besides power. Intel power. No toy-like ARM power.
Originally Posted by mfryd
Apple discontinued Rosetta with Lion. This was a big step in the move to ARM.
Moving entirely to x86 was a… big step in moving to ARM?
Quote:
Originally Posted by zunx
No way. A full x86 compatibility is a must. Besides power. Intel power. No toy-like ARM power.
It's always difficult to foresee the future, especially because it's not the same as what you can see now.
Lots of engineers in the know didn't see the transistor as a replacement of the vacuum tube back in the 50's.
How could such an unreliable low power device replace the far superior vacuum tube, that's impossible.
I must admit, the future of ARM is a lot easier to predict.
J.
Quote:
Originally Posted by cpenzone
Apple's been down this road before... it would really surprise me if they went 100% into a non-intel design. How fast we forget history.
Yes, how fast. 11 years ago, die hard Apple users would have said "It would really amaze me if they went 100% into a non-PPC design.'
This OS has been ported and sold to run on at least 4 platforms 68K SPARC, HP-RISC, PPC, X86 and ARM.
Moving the Mac fork to ARM is purely a capability curve decision. we've had this discussion before... ARM is improving faster within it's own space, but in the overall performance/power/price curve Intel has the edge. If there is a change, Apple will adjust. And given that Apple enjoys tweaking the chip to match it's performance requirements, moving away from stock Intel (say a custom chip built by, say, AMD) would be 100% within their design parameters.
The real game changer is the move away from 'WinTel compatibility' as a must have for a set of customers. Not too long ago, Boot Camp was a key marketing bullet. You don't see it now. If that continues, I think the key issue is Apple 'controlling' its key components that tightly integrate with their OS and app compiler platforms... that leads to 'inhousing' CPU design, which at that point if the ARM price performance power curve is right, they'll move. If not, it wouldn't surprise me for Apple to buy a controlling share of AMD, or strike a deal with Intel for a x86 design fork unique only to Apple.
Quote:
Originally Posted by zunx
No way. A full x86 compatibility is a must. Besides power. Intel power. No toy-like ARM power.
for you maybe, but not for the typical (read: most) Apple Macbook (Air) users.
Intel never had the best pure performance, just the best pricing model given the WinTel duopoly, and the ability to deliver in quantity.
Quote:
Originally Posted by mfryd
Lots of people will argue that ARM is inappropriate for the Mac, and that Intel is a better choice. These arguments are irrelevant. Apple is targeting the consumer market, not the high end professional. ARM chips are more then sufficient for this market.
I don't agree with this. Even consumers are doing video editing / processing, photo editing, audio conversion, running multi-apps, number processing, etc.
If Apple transitions to ARM, I don't believe it will be 100%. I believe their low-end machines will, but the rest will be Intel. ARM isn't going to catch up to Intel. Intel has so much speed in their back pocket right now, they could literally jump another 1-2ghz overnight if they wanted to. They are simply waiting for AMD to catch up.
Assuming Apple transitions to ARM 100%, their computers are going to continue to be inferior to PC/Intel computers. That also assumes that Apple doesn't care about the developers that got them to where they are at today. We (developers) pine for performance. No one likes to build 50 times a day on slow hardware. Also, consumers aren't dumb, they still take numbers into account when purchasing items. A huge portion of Apple's consumer base will drop off if this happens, as their software isn't so much better than the competition that consumers will completely ignore performance on the other side.
Quote:
Originally Posted by monstrosity
I'm not so sure. I'm not a processor expert but I believe there is something fundamental about ARM chips which makes them pretty hard to compete with when taking into consideration power consumption/speed.
As laptops are already running red hot, power consumption becomes more and more important, favouring ARM designs going forward.
Intel may one day come out with a quantum processor, but that technology is still a long way away. Till then, my money is on ARM (which it is literally as I own ARM stock).
This is fairly spot on if you ask me.
I am also far short of being a chip expert, but even the tiniest amount of searching will show that the intel x86 designs are really not that good. They get good performance expressly because they don't care about things like heat dissipation and efficiency. It's all about raw power. They are neither efficient nor particularly well designed because they operate under almost none of the constraints that ARM chips have.
Anything can happen, but intel is on top right now because of the special "Wintel" relationship and because of history.
They certainly aren't on top because they are the best designed chips around.
It's my understanding that ARM is great at the low end but can't be scaled to 3 and 4GHz like Intel chips to get the same performance per watt.
Originally Posted by SolipsismX
It's my understanding that ARM is great at the low end but can't be scaled to 3 and 4GHz like Intel chips to get the same performance per watt.
That would partially explain the heat and terrible battery life of even the 1.5GHz Snapdragon processors.
I think you're mostly on the right track. I do not, however, see Macs running ARM any time soon. Instead, I see an iPad Pro (perhaps even with a keyboard and MBA configuration) which would be one step above an iPad, but still far below a Mac. And they could make something that looks like an iMac (albeit probably with a much smaller screen due to GPU limitations) that could use faster processors since battery life wouldn't be an issue.
But I just don't see any chance of Macs switching. Why is it that all of these ARM proponents keep talking about ARM performance improving but assume that x86 performance will not?
That doesn't make any sense.
If they do switch to ARM, they're going to need some kind of compatibility layer for apps that have not yet been ported. So they'd need to resurrect Rosetta. So how does canceling Rosetta help them down the path toward a future which would require Rosetta?
Quote:
Originally Posted by Seankill
I guess I'll be switching back to pcs then. Cause I want a processor that's powerful and compatible with windows.
Stupid move IMO
I like ARM in phones Etc
Hopefully it's just be like a version of the MacBook Air and not all MacBooks
The change will not be any time soon, as stated.
And by the time it *does* happen, whatever Apple chooses WILL be powerful enough. Apple isn't stupid, LOL. They understand the market better than anyone.
As for running Windows. Who cares?
Well, change is inevitable.
*shrugs*
Quote:
Originally Posted by SolipsismX
It's my understanding that ARM is great at the low end but can't be scaled to 3 and 4GHz like Intel chips to get the same performance per watt.
This is correct. In fact at those speeds (currently) Intel trounces ARM. Inherently x86 and ARM were designed to do very different things. As technology progresses the line blurs, but at it's core many things remain the same with each respective design, which is why such a move seems odd. Intel probably could catch up to ARM in PPW before ARM could catch Intel on performance. Dark Horse: Intel still has a license for ARM, though they maintain they have no intention of using it.
Quote:
Originally Posted by Gazoobee
This is fairly spot on if you ask me.
I am also far short of being a chip expert, but even the tiniest amount of searching will show that the intel x86 designs are really not that good. They get good performance expressly because they don't care about things like heat dissipation and efficiency. It's all about raw power. They are neither efficient nor particularly well designed because they operate under almost none of the constraints that ARM chips have.
Anything can happen, but intel is on top right now because of the special "Wintel" relationship and because of history.
They certainly aren't on top because they are the best designed chips around.
This is what the danger of "a little knowledge" looks like. There is a small theoretical advantage to ARM simpler instruction set.
But this is swamped by execution. The real reason ARM was historically so much better at mobile, is because that is what they have been building for, for years, while Intel OTOH concentrated on high performance.
In case you haven't noticed, Intel is waking up. Check out the Motorola Razr i, with an Intel Atom inside. It was the only phone to regularly beat the new A6 iPhone on benchmarks, and it is still gets good battery life.
Next year, Intel is releasing a brand new Atom Architecture (first in 5 years) it will be arriving on Intels mature 22nm process and coming in varieties ranging from 1 core to 4 core.
After that Intel is planning to start doing Tic-Toc iterations in mobile, like they do on desktops (where the obliterated AMD).
In three years most new Android phones could be running Intel Atoms and making iPhones looks slow in comparison and all the talk will be about how Apple is going to switch iOS to Intel and ditch non competetive ARM.
The truth is we don't know, but anyone writing off Intel in Mobile or saying an Apple switch of OSX to ARM is inevitable, is clueless.
Quote:
Originally Posted by Ireland
Nothing a few years out in the CE or computer world is inevitable. For all we know in a few years time Intel may have chips for the iPhone that blow ARM away, use less power and Apple uses those in their mobile devices instead, and keeps using improved Intel chips in Macs. Nothing is guaranteed here.
I agree. I think the article's point is that Apple is preparing to move the Arm into the Macs, which they will do if and only if it makes sense. And as you say, they might surprise us the other way and move Intel into the iOS world, if and only if that makes sense. But I think the take-home point is there are may different ways the technology may go in the next few years, and Apple does a pretty good job staying light on its feet unlike some of its competitors.
Using the app store this would not be an issue.
1) Compile creates a FAT version of the binary.
2) Downloading to ARM machines would strip the intel code and vice versa.
Quote:
Originally Posted by SolipsismX
It's my understanding that ARM is great at the low end but can't be scaled to 3 and 4GHz like Intel chips to get the same performance per watt.
That seems to be in direct contradiction with the ARM 64bit Cortex A57 roadmap (up to 3GHz).
J.