note that iOS already supports external displays -- i use it every time i plug my iPhone in my Pioneer AppRadio head unit in the car. external display + tap events.
You're absolutely right, I'm just obsessed with the new smart connector and trying to imagine what they're planning for it. I'm thinking more and multiple monitors through the connector acting as part of some dock (imagine dropping your iPhone into a little dock that is connected to a keyboard, display and touchpad or the iPhone acts as the touchpad). Just shows, though, that the OS is already primed in some ways to support a proper desktop setup, which will annoy the trolls and those who believe an OS is only "real" when it's got a pointing device, keyboard and external display.
I mean even it's very foundations were .... lets design it to slow down the user to as slow as possible (QWERTY keyboard design reasoning).
That's not true at all. QWERTY was designed to place the commonly used keys far apart, so as to prevent typewriter jamming, but that's a big difference from intentionally making the typist as slow as possible.
I'm willing to bet that Apple's ARM processors are even more powerful than the benchmarks show because they are in mobile units with no fans and probably under-clocked. Put a couple in an iMac and we would probably see Core i7 performance.
if you truly believe one of the best apple analysts is an idiot and keyboards are unnatural and unneeded, how about you go a month without a keyboard, or self-ban?
A keyboard is very unnatural, and the only reason we are so reliant on it is because the software research in machine learning or AI has not advanced enough to make it disposable.... but we already have a computer that works great with a keyboard.... it is the Macs. If you are data entry heavy then choosing the iPad to begin with is like choosing a square peg into a round hole. If it were natural I would not have had to take typing class way back in high school. First slow, then faster.... then eventually becoming an expert. Where the iPad will excel is in re-engineered processes, not sitting down and hacking out an n-page document.... Or inputting a new customer into the Customer information system (which hopefully only has to be done once). I personally am a software developer which right now means lots of typing - so no... I would not use an iPad for work it would not be the right use case. The hottest area right now is in the area of AI and Machine learning.... and that is not just for self-driving cars. It is for being able to do things like allowing the computer to take an image of the store shelf, decoding it, figuring out what items are what then counting them and updating the inventory. It is for being able to communicate verbally with your computer without having to dumb down what you are saying to the level of communicating with a complete moron (i.e. learning context - similar to the way you communicate with someone you know well). A lot of what we do as data entry is redundant and repetitive....
I can see Apple making an ARM powered MacBook...especially if it can emulate x86 Intel environment for legacy apps...maybe via a "transparent" VM. I'd guess we're 3-4 yrs away from that.
Don't see MacbookPro laptops changing anytime soon.
I can see Apple making an ARM powered MacBook...especially if it can emulate x86 Intel environment for legacy apps...maybe via a "transparent" VM. I'd guess we're 3-4 yrs away from that.
Don't see MacbookPro laptops changing anytime soon.
We are probably 2 years away... and it won't have to use a VM..... LLVM (which was misnamed from the beginning) will allow the application to be written once and run on different platforms natively..... The compiler technology is already there.... there is some work on the application store that has to be done and potentially a way to package applications in portable bitcode that are compiled down during installation (assuming you don't submit your application to the app store). The only areas that it will not work for are anything running in VMWare and Parrallels (Windows x86 Linux). A cross CPU architecture VM would be disasterously slow.
Tim Cook has been galavanting around the cosmos with just his iPhone and iPad Pro, telling anyone within earshot that this new iPad Pro can and will replace PCs.
More than any other product, the iPad Pro is going to pull iOS and OS X together at an even greater pace. iOS is going to be borrowing desktop features from OS X to make this transition possible.
For me, iCloud made the use of my iPad in grad school possible. I agree with Gruber that a trackpad might be pretty useful in using the Pro as a desktop (after all, the pencil is nothing if not a pointer... the trackpad would alleviate lifting my huge guns every time I wanted to select something).
I'm not sure about the whole file system needing to be exposed. Working with iCloud is a weird hybrid of that paradigm.
I'm still at the point of 'making it work.' It's not elegant and doesn't offer the ease of use that OS X offers when I'm at home on my 27" iMac.
Right now, having not tried out the iPad Pro, the best scenario for my work is an iMac at home, iPhone in my pocket, Laptop at school. If the iPad Pro's keyboard is easy to type on and the size makes it easier to use multiple apps, I could be swayed.
Seems like it's just a matter of time. The iPad Pro, with its keyboard and pencil and size and multitasking...... looking more and more like a laptop every day.
Why is there no emulator for running iOS apps on a Mac? There are so many useful little Apps that I'd love to use side-by-side with OS X apps on my desktop Mac.
NO, the foundations are already being laid so that applications / app store is seamless whether it is ARM based on Intel based.
There are a lot of major third party apps that do not use the Mac app store or Xcode exclusively for development such as Adobe CC. Apple can't leave all those people out in the cold. I know the majority of the people on this forum do not use that suite but it is the most popular graphics application suite for Mac. There are many others such as Firefox and Chrome, 1password, littleSnitch, Skype, Transmission, VLC, Google Earth, Mathmatica, R, Dropbox, Spotify, HandBrake, TextMate, and Plex. I'm sure there are others.
There are a lot of major third party apps that do not use the Mac app store or Xcode exclusively for development such as Adobe CC. Apple can't leave all those people out in the cold. I know the majority of the people on this forum do not use that suite but it is the most popular graphics application suite for Mac. There are many others such as Firefox and Chrome, 1password, littleSnitch, Skype, Transmission, VLC, Google Earth, Mathmatica, R, Dropbox, Spotify, HandBrake, TextMate, and Plex. I'm sure there are others.
They have to use something that uses the same foundations - whether there is an additional layer between them or not. C, C++, Objective-C and Swift all compile through the C-Lang compiler which is built on top of the LLVM. Areas that obviously would have problems is if someone used embedded x86 assembly language. The UI components calls would be common to both platforms.
Even if the ARM processors were used they would be in part of the lineup (the lower end) which is why this type of architecture would have to be in place for.
Google would probably be very happy with it since they have tried to push other browser companies to incorporate pNaCL support across their platforms - which is basically portable LLVM bitcode.
"Apple's iPad Pro & powerful A9X CPU pose threat to Intel". Only when A9X CPU becomes x86 for obvious reasons. There is something much more important than speed: compatibility withe the rest of the world. That is why is so great, besides its interface.
The A9x is a great mobile processor for tablets, but there is a reason why Apple didn't put it in a MacBook. These benchmarks are nothing new to Apple, as they more then likely tested both the A9x and Intel Core M in the new MacBook. Ultimately the Intel chip won out and ran OS X more effectively. Believe me, if Apple could save money by putting an A series chip in a laptop or desktop they would do it.
The ultimate would be an Apple App store. When you open up iTunes, you have to toggle between iPhone and iPad stores. Then there is the App store for OS X.
This is inelegant and I'm sure Steve is spinning in his grave. In an ideal future, Apple would be making all the critical components it needs to create awesome hardware, including the CPSs and memory chips and monitors. Once the hardware is homogenized, software wouldn't be too far behind.
I don't see any problems in having an OS that is specific to various hardware needs (wearables vs desktop vs touch vs whatever). I don't think Apple needs one App store as it's better to shop specific apps for your Apple TV and Apple Watch. But what I'm saying is it'd be nice to be able to run those apps across all these future apple devices.
I think that's where Apple will be within 10 years.
The A9x is a great mobile processor for tablets, but there is a reason why Apple didn't put it in a MacBook. These benchmarks are nothing new to Apple, as they more then likely tested both the A9x and Intel Core M in the new MacBook. Ultimately the Intel chip won out and ran OS X more effectively. Believe me, if Apple could save money by putting an A series chip in a laptop or desktop they would do it.
The intel chip won out for the Macbook because the consumer confusion of using two different chip architectures in the same line would be more damaging than the positive of being able to build the same Macbook for $200 cheaper.
Does anyone think the next MacBooks may switch to Apple processors? And if they did how much of a cost savings could Apple transfer to potential buyers over the current MacBook 1 model? And if there is an in-house OS X ARM variant (very likely given past history) how much battery life could an ARM MacBook 1 get? Anyone?
I don't think the next generation MacBooks will switch over to Apple processors. Though they give some performance boost, it isn't huge (unlike the 68000 to PPC jump) and x86 doesn't have a deadend roadmap (like the PPC to x86 jump).
Cost:
This is really hard to estimate. But here are some super rough numbers:
A 13-inch Macbook w/ Retina is listed at $1300. If Apple makes a 25-40% margin, their cost is about $780-975. The an i5 processor's suggested price is about $200, let's guess that Apple is paying $150-175. Intel's has about a 40-60% margin, so their production cost is maybe $80-120. Let's assume apple can design a core with performance as good as Intel in about 80-100% of the area (maybe ARM cores can be a little leaner), so roughly $65-100. Apple's chip foundry gets a 30-50% margin(?), so they sell to Apple at $90-200. So Apple can swap a $150-175 part for a $90-200 part. So the production cost for apple is $720–1000, assuming the same margins, the price would be about $1200. So maybe $100 cheaper? Lots of guesses in there.
Battery life:
I am not sure how much more efficient a beefed up ARM core is compared to an x86. But, let's assume the ARM core uses 20% less power for the same performance (I think 20% is aggressive). I think a laptop CPU uses 10-20% of the toal power (the screen is the biggest consumer). So a laptop which gets 9 hours of battery life today could get about 9:46–10:42. Significant, but not huge.
NO, the foundations are already being laid so that applications / app store is seamless whether it is ARM based on Intel based. There will be no need for Rosetta. The difference is the whole development environment was moved to work on top of LLVM and the compilers generate a portable "binary code / assembly code" which then is made runnable on an ARM or Intel. What Apple does not want to do is confuse users using Macs like Microsoft did with Windows RT. Once all the pieces are in place the user will not have to be aware or care if they are running intel or arm. Developers will submit their applications to the app store in bitcode (LLVM transportable) and then the user buys the application and installs it and it is "compiled/assembled down" to Intel or ARM native executables.... all nice and slim.
Instead of building an ARM-based MAC, would this allow Apple to merge the Mac App Store / iOS App Store for truly Universal Apps across Mac & iOS devices? That way, those who do use Macs can have the option run iOS Apps, obviously not unadultered, but optimized for KB / Mouse / trackpad?
Does anyone think the next MacBooks may switch to Apple processors?
Well I wish they would.. Wishes don't often come to life though. That wish is contingent upon a couple of things though, more RAM would be key and it needs to run a full version of Mac OS. Oh and I would need HomeBrew to bring ARM support online.
And if they did how much of a cost savings could Apple transfer to potential buyers over the current MacBook 1 model?
Knowing Apple they would transfer cost saving into more profits. Given that, let's say it costs Apple $100 for each A series chip, that value is probably high but I'm not thinking by much. The reason for the high cost is that they are coop'ing the bleeding edge processes to build these chips and that does cost extra. In any event that still would save Apple considerable money, anywhere from $160 to $250 dollars a machine depending upon what they actually pay Intel for a chip. Nothing to sneeze at really.
The savings can go beyond that due to the higher integration and Apple specific support built into the chips. For example direct connection to flash though a built in controller.
And if there is an in-house OS X ARM variant (very likely given past history) how much battery life could an ARM MacBook 1 get? Anyone?
The discussions about battery life don't mean much when it goes beyond 12 hours in my mind. However iPad Pro already tells us what it might be like, the likely hood of a similarly sized battery is a reality so that sets one point on the potential run time. The big thing with Mac OS, and users like me for example, is that we almost never run the platform with just one app doing work. Multitasking means more power usage if you actually multitask.
The really interesting thing about this new A9 is that it apparently hardly ever throttles in testing. If that holds up in real life then the chip would be a huge win in a Mac Book right now. It means full performance all day long no matter what you do. Also when considering a laptop ere is the possibility of better thermals and thus higher clock rates so A9X would demolish Intels chip even more.
As an aside any ARM based laptop from Apple needs to support Linux well. Here I mean running Linux in a VM or doing a native install as the primary OS.
They have to use something that uses the same foundations - whether there is an additional layer between them or not. C, C++, Objective-C and Swift all compile through the C-Lang compiler which is built on top of the LLVM. Areas that obviously would have problems is if someone used embedded x86 assembly language. The UI components calls would be common to both platforms.
Even if the ARM processors were used they would be in part of the lineup (the lower end) which is why this type of architecture would have to be in place for.
Google would probably be very happy with it since they have tried to push other browser companies to incorporate pNaCL support across their platforms - which is basically portable LLVM bitcode.
Linux platforms have been moving to LLVM.
Now even Microsoft has made a commitment to it.
I wouldn't count Rosetta++ out. Keep in mind that with a laptop you can drop a fan-cooled heat sink on the package...and double them up...
People don't want a converged OS. Did the utter failure of Windows 8 teach people nothing?
Well not yet they don't. Windows is a mess because it is Windows, that isn't a base to really build things upon.
That being said I don't want to loose the power that Mac OS gives me. Likewise I can't dismiss that iOS has a lot of short comings that Apple needs to address if they ever expect to get pass the consumption device mentality currently directed at iOS. File system and file access are two big issues with iOS that half to be addressed soon if they really want professionals outside of certain niches to adopt iOS.
And the A10 will be a dual core, same with A10X. Just simpler that way.
Being simple isn't what it is all about. What it is all about is engineering die space so that you get maximum benefit for the aspects that you believe are important. I haven't even seen die shots of this. Chip but I can reasonably assume that there is a lot of die space on A9X supporting the GPU. The next die shrink could easily allocate more space to more CPUs. Apple will allocate space as they see fit to deliver the performance they think they need.
Gruber pointed out what I think is the iPad Pro's biggest fault - and it's the fault of iOS, not the iPad Pro itself. The lack of keyboard functionality and bugs surrounding keyboard input in general will undermine iPad Pro adoption for productivity users until thoroughly addressed.
The good thing here is that this is mostly an issue of software. I do wish that Apple would abolish MiFi and free things up so that more hardware can be supported by iOS devices.
Comments
note that iOS already supports external displays -- i use it every time i plug my iPhone in my Pioneer AppRadio head unit in the car. external display + tap events.
You're absolutely right, I'm just obsessed with the new smart connector and trying to imagine what they're planning for it. I'm thinking more and multiple monitors through the connector acting as part of some dock (imagine dropping your iPhone into a little dock that is connected to a keyboard, display and touchpad or the iPhone acts as the touchpad). Just shows, though, that the OS is already primed in some ways to support a proper desktop setup, which will annoy the trolls and those who believe an OS is only "real" when it's got a pointing device, keyboard and external display.
I mean even it's very foundations were .... lets design it to slow down the user to as slow as possible (QWERTY keyboard design reasoning).
That's not true at all. QWERTY was designed to place the commonly used keys far apart, so as to prevent typewriter jamming, but that's a big difference from intentionally making the typist as slow as possible.
if you truly believe one of the best apple analysts is an idiot and keyboards are unnatural and unneeded, how about you go a month without a keyboard, or self-ban?
A keyboard is very unnatural, and the only reason we are so reliant on it is because the software research in machine learning or AI has not advanced enough to make it disposable.... but we already have a computer that works great with a keyboard.... it is the Macs. If you are data entry heavy then choosing the iPad to begin with is like choosing a square peg into a round hole. If it were natural I would not have had to take typing class way back in high school. First slow, then faster.... then eventually becoming an expert. Where the iPad will excel is in re-engineered processes, not sitting down and hacking out an n-page document.... Or inputting a new customer into the Customer information system (which hopefully only has to be done once). I personally am a software developer which right now means lots of typing - so no... I would not use an iPad for work it would not be the right use case. The hottest area right now is in the area of AI and Machine learning.... and that is not just for self-driving cars. It is for being able to do things like allowing the computer to take an image of the store shelf, decoding it, figuring out what items are what then counting them and updating the inventory. It is for being able to communicate verbally with your computer without having to dumb down what you are saying to the level of communicating with a complete moron (i.e. learning context - similar to the way you communicate with someone you know well). A lot of what we do as data entry is redundant and repetitive....
Don't see MacbookPro laptops changing anytime soon.
I can see Apple making an ARM powered MacBook...especially if it can emulate x86 Intel environment for legacy apps...maybe via a "transparent" VM. I'd guess we're 3-4 yrs away from that.
Don't see MacbookPro laptops changing anytime soon.
We are probably 2 years away... and it won't have to use a VM..... LLVM (which was misnamed from the beginning) will allow the application to be written once and run on different platforms natively..... The compiler technology is already there.... there is some work on the application store that has to be done and potentially a way to package applications in portable bitcode that are compiled down during installation (assuming you don't submit your application to the app store). The only areas that it will not work for are anything running in VMWare and Parrallels (Windows x86 Linux). A cross CPU architecture VM would be disasterously slow.
Tim Cook has been galavanting around the cosmos with just his iPhone and iPad Pro, telling anyone within earshot that this new iPad Pro can and will replace PCs.
More than any other product, the iPad Pro is going to pull iOS and OS X together at an even greater pace. iOS is going to be borrowing desktop features from OS X to make this transition possible.
For me, iCloud made the use of my iPad in grad school possible. I agree with Gruber that a trackpad might be pretty useful in using the Pro as a desktop (after all, the pencil is nothing if not a pointer... the trackpad would alleviate lifting my huge guns every time I wanted to select something).
I'm not sure about the whole file system needing to be exposed. Working with iCloud is a weird hybrid of that paradigm.
I'm still at the point of 'making it work.' It's not elegant and doesn't offer the ease of use that OS X offers when I'm at home on my 27" iMac.
Right now, having not tried out the iPad Pro, the best scenario for my work is an iMac at home, iPhone in my pocket, Laptop at school. If the iPad Pro's keyboard is easy to type on and the size makes it easier to use multiple apps, I could be swayed.
Seems like it's just a matter of time. The iPad Pro, with its keyboard and pencil and size and multitasking...... looking more and more like a laptop every day.
Why is there no emulator for running iOS apps on a Mac? There are so many useful little Apps that I'd love to use side-by-side with OS X apps on my desktop Mac.
NO, the foundations are already being laid so that applications / app store is seamless whether it is ARM based on Intel based.
There are a lot of major third party apps that do not use the Mac app store or Xcode exclusively for development such as Adobe CC. Apple can't leave all those people out in the cold. I know the majority of the people on this forum do not use that suite but it is the most popular graphics application suite for Mac. There are many others such as Firefox and Chrome, 1password, littleSnitch, Skype, Transmission, VLC, Google Earth, Mathmatica, R, Dropbox, Spotify, HandBrake, TextMate, and Plex. I'm sure there are others.
There are a lot of major third party apps that do not use the Mac app store or Xcode exclusively for development such as Adobe CC. Apple can't leave all those people out in the cold. I know the majority of the people on this forum do not use that suite but it is the most popular graphics application suite for Mac. There are many others such as Firefox and Chrome, 1password, littleSnitch, Skype, Transmission, VLC, Google Earth, Mathmatica, R, Dropbox, Spotify, HandBrake, TextMate, and Plex. I'm sure there are others.
They have to use something that uses the same foundations - whether there is an additional layer between them or not. C, C++, Objective-C and Swift all compile through the C-Lang compiler which is built on top of the LLVM. Areas that obviously would have problems is if someone used embedded x86 assembly language. The UI components calls would be common to both platforms.
Even if the ARM processors were used they would be in part of the lineup (the lower end) which is why this type of architecture would have to be in place for.
Google would probably be very happy with it since they have tried to push other browser companies to incorporate pNaCL support across their platforms - which is basically portable LLVM bitcode.
Linux platforms have been moving to LLVM.
Now even Microsoft has made a commitment to it.
Got my iPad pro today and I'm amazed by how fabulous it is. Looking forward to getting the keyboard and pencil to see if I can do real work with it.
Only when A9X CPU becomes x86 for obvious reasons. There is something much more important than speed: compatibility withe the rest of the world. That is why is so great, besides its interface.
The ultimate would be an Apple App store. When you open up iTunes, you have to toggle between iPhone and iPad stores. Then there is the App store for OS X.
This is inelegant and I'm sure Steve is spinning in his grave. In an ideal future, Apple would be making all the critical components it needs to create awesome hardware, including the CPSs and memory chips and monitors. Once the hardware is homogenized, software wouldn't be too far behind.
I don't see any problems in having an OS that is specific to various hardware needs (wearables vs desktop vs touch vs whatever). I don't think Apple needs one App store as it's better to shop specific apps for your Apple TV and Apple Watch. But what I'm saying is it'd be nice to be able to run those apps across all these future apple devices.
I think that's where Apple will be within 10 years.
The A9x is a great mobile processor for tablets, but there is a reason why Apple didn't put it in a MacBook. These benchmarks are nothing new to Apple, as they more then likely tested both the A9x and Intel Core M in the new MacBook. Ultimately the Intel chip won out and ran OS X more effectively. Believe me, if Apple could save money by putting an A series chip in a laptop or desktop they would do it.
The intel chip won out for the Macbook because the consumer confusion of using two different chip architectures in the same line would be more damaging than the positive of being able to build the same Macbook for $200 cheaper.
Does anyone think the next MacBooks may switch to Apple processors? And if they did how much of a cost savings could Apple transfer to potential buyers over the current MacBook 1 model? And if there is an in-house OS X ARM variant (very likely given past history) how much battery life could an ARM MacBook 1 get? Anyone?
I don't think the next generation MacBooks will switch over to Apple processors. Though they give some performance boost, it isn't huge (unlike the 68000 to PPC jump) and x86 doesn't have a deadend roadmap (like the PPC to x86 jump).
Cost:
This is really hard to estimate. But here are some super rough numbers:
A 13-inch Macbook w/ Retina is listed at $1300. If Apple makes a 25-40% margin, their cost is about $780-975. The an i5 processor's suggested price is about $200, let's guess that Apple is paying $150-175. Intel's has about a 40-60% margin, so their production cost is maybe $80-120. Let's assume apple can design a core with performance as good as Intel in about 80-100% of the area (maybe ARM cores can be a little leaner), so roughly $65-100. Apple's chip foundry gets a 30-50% margin(?), so they sell to Apple at $90-200. So Apple can swap a $150-175 part for a $90-200 part. So the production cost for apple is $720–1000, assuming the same margins, the price would be about $1200. So maybe $100 cheaper? Lots of guesses in there.
Battery life:
I am not sure how much more efficient a beefed up ARM core is compared to an x86. But, let's assume the ARM core uses 20% less power for the same performance (I think 20% is aggressive). I think a laptop CPU uses 10-20% of the toal power (the screen is the biggest consumer). So a laptop which gets 9 hours of battery life today could get about 9:46–10:42. Significant, but not huge.
NO, the foundations are already being laid so that applications / app store is seamless whether it is ARM based on Intel based. There will be no need for Rosetta. The difference is the whole development environment was moved to work on top of LLVM and the compilers generate a portable "binary code / assembly code" which then is made runnable on an ARM or Intel. What Apple does not want to do is confuse users using Macs like Microsoft did with Windows RT. Once all the pieces are in place the user will not have to be aware or care if they are running intel or arm. Developers will submit their applications to the app store in bitcode (LLVM transportable) and then the user buys the application and installs it and it is "compiled/assembled down" to Intel or ARM native executables.... all nice and slim.
Instead of building an ARM-based MAC, would this allow Apple to merge the Mac App Store / iOS App Store for truly Universal Apps across Mac & iOS devices? That way, those who do use Macs can have the option run iOS Apps, obviously not unadultered, but optimized for KB / Mouse / trackpad?
The savings can go beyond that due to the higher integration and Apple specific support built into the chips. For example direct connection to flash though a built in controller.
The discussions about battery life don't mean much when it goes beyond 12 hours in my mind. However iPad Pro already tells us what it might be like, the likely hood of a similarly sized battery is a reality so that sets one point on the potential run time. The big thing with Mac OS, and users like me for example, is that we almost never run the platform with just one app doing work. Multitasking means more power usage if you actually multitask.
The really interesting thing about this new A9 is that it apparently hardly ever throttles in testing. If that holds up in real life then the chip would be a huge win in a Mac Book right now. It means full performance all day long no matter what you do. Also when considering a laptop ere is the possibility of better thermals and thus higher clock rates so A9X would demolish Intels chip even more.
As an aside any ARM based laptop from Apple needs to support Linux well. Here I mean running Linux in a VM or doing a native install as the primary OS.
They have to use something that uses the same foundations - whether there is an additional layer between them or not. C, C++, Objective-C and Swift all compile through the C-Lang compiler which is built on top of the LLVM. Areas that obviously would have problems is if someone used embedded x86 assembly language. The UI components calls would be common to both platforms.
Even if the ARM processors were used they would be in part of the lineup (the lower end) which is why this type of architecture would have to be in place for.
Google would probably be very happy with it since they have tried to push other browser companies to incorporate pNaCL support across their platforms - which is basically portable LLVM bitcode.
Linux platforms have been moving to LLVM.
Now even Microsoft has made a commitment to it.
I wouldn't count Rosetta++ out. Keep in mind that with a laptop you can drop a fan-cooled heat sink on the package...and double them up...
That being said I don't want to loose the power that Mac OS gives me. Likewise I can't dismiss that iOS has a lot of short comings that Apple needs to address if they ever expect to get pass the consumption device mentality currently directed at iOS. File system and file access are two big issues with iOS that half to be addressed soon if they really want professionals outside of certain niches to adopt iOS.
Being simple isn't what it is all about. What it is all about is engineering die space so that you get maximum benefit for the aspects that you believe are important. I haven't even seen die shots of this. Chip but I can reasonably assume that there is a lot of die space on A9X supporting the GPU. The next die shrink could easily allocate more space to more CPUs. Apple will allocate space as they see fit to deliver the performance they think they need.
The good thing here is that this is mostly an issue of software. I do wish that Apple would abolish MiFi and free things up so that more hardware can be supported by iOS devices.