1) Where are the people that laughed at Apple for buying PA Semi?
2) I love how I hear everyday from people that Apple isn't innovating because they define it as releasing an entirely new product category. Now I don't expect the average person to understand how this chip is innovative and well ahead of the market but I do expect them to at least understand that [I]unseen[/I] innovations are still innovations.
3) I don't think 2x A8's is the right way to go. Better to use more cores, or simply up the clock as Anand states these handheld devices are clocked too low to take full advantage of this "desktop class" chip.
4) I do think a small, 12", low-power notebook could happen. Between the Apple's ARM chips now being 64-bit, it's performance-per-Watt, the high cost of Intel's CULV CPUs in the MBA, the universality of web-based apps, and the Mac App Store I think it's not out of the realm of possibility that Apple could update Xcode to allow Mac apps to be compiled for x86_64 and AArch_64 for MAS. I wouldn't expect Apple to cut out the old-school download and install method but allow developers who have a viable product that will work on ARM to adjust and recompile like with the transitions from PPC. I also wouldn't expect a Rosetta-like option since the performance envelope will be going the other direction in both the chip and by not starting near the top of the performance line. This move could get Apple to make a MBA-like device that comes in several hundred less expensive than today with more than double the battery life for the current size/weight, that fits the needs of the user who isn't playing [I]Resident Evil 5[/I] on a 12" machine or using the Adobe Suit 6. If Chromebooks can gain some traction without MS Office and Adobe apps then so can an ARM-based Mac-like notebook.
5) QFT: "Looking at [Apple's A7 SoC] makes one thing very clear: the rest of the players in the ultra mobile CPU space didn't aim high enough. I wonder what happens next round." ~Anand Shimpi
Would be cool if a future iteration of the iPhone could be wirelessly connected to a keyboard, monitor, external storage and mouse to play the part of an instant Mac mini.
An A8 chip would cost about $12, allowing Apple to sell an A8 laptop for $699 (and at a higher profit margin). It would also have about 120 hours of battery life. Oh, and if they put 8 cores into the thing, it would run circles around Haswell.
And since the ARM is basically a PPC, the SDK's are all there, and much of the code is already written. It would just need an updating.
This is what I have been saying and some people said that it was not possible. I believe Apples next move is put their own processors in to their laptop products. They do not need Intel at this point. I would not be surprise if they did not have a OSX port for their processors already.
Get me an iPad with keyboard that snaps apps to the side like the surface and I'll replace my Notebook with an iMac and take the iPad on the go! Especially considering the next iPad will have an A8 thats likely going to be comparable with a low/middle-class notebook. I'd love a lighter mobile solution but as long as I need a computer on the go I won't have two computers and therefore have to use the Retina MacBook Pro to be able to do all my work. But on the go the power of an iPad would be more than enough!
So at this point would iOS be the inhibiting factor in unleashing the A7 power?
Not at all, for the most part iOS is Mac OS with a slightly different GUI and a more restricted execution environment. That being said I'd love to see a Mac OS device that uses A7/A8. Even better if the machine supports easily booting other operating systems.
I agree. Fragmentation is bad. Apple will not just convert MBA to A7 chip and keep Intel on MBP, I believe.
Apple seems to like lateral moves. If release an A-chip-based notebook it might not even be called a Mac or use Mac OS X, at least not at first. This could allow them to build a parallel project and let developers get apps converted, work out any issues, and further advance their A-chips to a point that it could eventually allow them to make the switch much further down the road.
The question is how many hypothetical A8X could you stick in a Mac (MBA, MBP, iMac, MacPro) for the same thermal specs as the equivalent Intel chips.
I don't think Apple would work it that way. I'd think they make it at least equivalent performance (if they were going to replace an existing product) but then use all the gains to balance a reduction in materials cost, weight and size, as well make the battery life much, much longer (if we're talking about a notebook).
How about maybe! It really depends upon what Apple delivers.
Battery life and weight on the Air is already terrific, and performance has been improving. Why harm performance by going to ARM, around the same timeframe as MB Airs NEED more performance due to going retina (whenever that happens)?
Did you not read the article, the performance potential is there. You assume that in a laptop that at they would use the same exact hardware. A faster RAM interface would have a dramatic impact.
And your Intel software would no longer run. And workarounds would involve major work for developers, and fat binaries what waste expensive SSD space.
Intel compatibility is only of concern to a few. As to fat binaries that is looking to the past.
So it would mean MASSIVE fragmentation for developers, and massive headache for users.
Baloney.
The kind of thing you only do if the benefit is HUGE, or if you HAVE to (like the PPC->Intel transition). And you do it for ALL the Macs in the lineup. ARM-based MacBook Pros, iMacs and Mac Pros? Makes no sense any time soon.
The benefits would be huge, laptops that are 2-300 dollars cheaper being one benefit. Piloting your own ship being another.
Apple already has an ARM-based portable that IS a good idea. They don't need to add one that isn't.
And your Intel software would no longer run. And workarounds would involve major work for developers, and fat binaries what waste expensive SSD space.
Two points:
According to comments in the original Anandtech article there are ARM emulators for x86 that yield 40%-60% degradation over native x86. These would be adequate for legacy apps where no source code exists. ARM and/or Emulator performance improvement could make x86 on ARM more practical for current apps.
As to fat-binaries, Apple has done a lot of work in Xcode as to application packaging -- for both OSX and iOS. I suspect that at some point, Apple will deliver Fat Packages to the App Stores which will be able to skip unneeded code for the specific device during an intelligent download/install/update/sync process.
The increased thermal headroom of a MacBook Air compared to an iPhone or iPad would allow a design to run at faster speeds for more of the time (most of the processing time in an iPhone or iPad is actually spent in "race to idle" mode - long battery lives are only achieved by being idle most of the time). I don't think they would use two A8s, they would probably fab an A8X or clock the A8 faster due to the increased thermal headroom.
But I don't think Apple is in the right place yet to release a ARM device running Mac OS X - all the software is compiled for x86-64 currently, not ARM. There would have to be a period of time where ARM was enabled for all software builds in XCode for Mac OS X. I believe Apple uses the threat of ARMing themselves to make Intel provide its CPUs at a reasonable price.
Would there be technical and/or bargaining advantages by include an A8 in addition to an x86 chip on a Mac device?
An A8 chip would cost about $12, allowing Apple to sell an A8 laptop for $699 (and at a higher profit margin). It would also have about 120 hours of battery life. Oh, and if they put 8 cores into the thing, it would run circles around Haswell.
And since the ARM is basically a PPC, the SDK's are all there, and much of the code is already written. It would just need an updating.
120 hours is too high. The CPU is only part of a laptop's power draw. The DRAM, GPU, screen, power supply, wireless, etc... all take up a lot of power. Even if the CPU took zero power you might only see a 10-20% battery life improvement.
I think it is also a little far to say 'ARM is basically a PPC' – they are both RISC cores, but the ISAs are very different, and, more importantly, their memory models are different. Porting can be done, but it is very non-trivial.
And while I am sure Apple could manage the transition (as they did with 68K->PPC and then PPC->x86) it would still be tricky. PPC gave a huge performance boost over 68K and x86 gave a future which PPC simply didn't have. The ARM cores do give good performance per watt, but don't give a huge raw performance boost over x86.
Apple would also have to improve the 'uncore' part of the A* line to be competitive with x86 - things like more scalable cache coherency, more memory channels, etc... All of this can be done, but is non-trivial.
On the other hand, the A* line could be less expensive, and it could allow some interesting integration paths that x86 doesn't support – e.g. integration with stacked memory in the package.
You are assuming that there would be fragmentation. There are a number of ways Apple could pull this off. One way is to make it an iOS laptop. The second would be to make it a Mac laptop that runs iOS apps in a window. That is just two approaches.
Apple will not just convert MBA to A7 chip and keep Intel on MBP, I believe. Can the "Ax" chip become powerful enough for MBP in the next iteration? It would be interesting if it can. How about MacPro? Somehow, it feels hopeless for Apple to stop using Intel chips.
You can't look a machines designed for the pro market in the same way as for the general consumer. The fact is, as demonstrated by the iOS world, is that most consumers have zero need for i86 compatibility. Even more so with Windows compatibility. The world is a changing and Apple is in the drivers seat.
Confuscious says, the comment section is not what you think it is... It's not meant to talk about the article, but rather to respond to the first comment.
Gaming. In AppleTV. High Performance Gaming. With gesture control. Gotta be what Apple is thinking. It's ARM, therefore iOS compatible, where there are more and more immersive games. That's gotta be where they're taking this.
I've stated all along that Apple's A7 is light years ahead of what Samsung and Qualcomm are doing.
I also stated previously that the A7 is so good that they don't really need to improve it. An A8 could simply be Apple's first quad core SoC using 4 A7 cores. Or they could ramp up the A7 clock to 2.0GHz and slightly tweak the design. Both would again give them a decent performance increase without a lot of effort.
I think the next big re-design will be reserved for the A9. Then again, Apple surprised us with the A6 & A7 and maybe they will again with the A8.
Comments
2) I love how I hear everyday from people that Apple isn't innovating because they define it as releasing an entirely new product category. Now I don't expect the average person to understand how this chip is innovative and well ahead of the market but I do expect them to at least understand that [I]unseen[/I] innovations are still innovations.
3) I don't think 2x A8's is the right way to go. Better to use more cores, or simply up the clock as Anand states these handheld devices are clocked too low to take full advantage of this "desktop class" chip.
4) I do think a small, 12", low-power notebook could happen. Between the Apple's ARM chips now being 64-bit, it's performance-per-Watt, the high cost of Intel's CULV CPUs in the MBA, the universality of web-based apps, and the Mac App Store I think it's not out of the realm of possibility that Apple could update Xcode to allow Mac apps to be compiled for x86_64 and AArch_64 for MAS. I wouldn't expect Apple to cut out the old-school download and install method but allow developers who have a viable product that will work on ARM to adjust and recompile like with the transitions from PPC. I also wouldn't expect a Rosetta-like option since the performance envelope will be going the other direction in both the chip and by not starting near the top of the performance line. This move could get Apple to make a MBA-like device that comes in several hundred less expensive than today with more than double the battery life for the current size/weight, that fits the needs of the user who isn't playing [I]Resident Evil 5[/I] on a 12" machine or using the Adobe Suit 6. If Chromebooks can gain some traction without MS Office and Adobe apps then so can an ARM-based Mac-like notebook.
5) QFT: "Looking at [Apple's A7 SoC] makes one thing very clear: the rest of the players in the ultra mobile CPU space didn't aim high enough. I wonder what happens next round." ~Anand Shimpi
Fragmentation is a serious issue to be sure.
But . . .
An A8 chip would cost about $12, allowing Apple to sell an A8 laptop for $699 (and at a higher profit margin). It would also have about 120 hours of battery life. Oh, and if they put 8 cores into the thing, it would run circles around Haswell.
And since the ARM is basically a PPC, the SDK's are all there, and much of the code is already written. It would just need an updating.
Could Apple stick 2X A8s in a slim MacBook Air?
What I'd give to see the fully working ARM port of OS-X that for sure Apple has in a lab somewhere deep in the bowels of One Infinity Loop.
One would be good enough. The biggest hurdle to SoC performance these days is bandwidth to RAM. If Apple solves that an A8 would be fine.
But they do. It debuted on June 24, 2003 on the Mac Pro G5. ARM and the PPC are both RISC chips made from the same cloth.
Ever wonder how Apple was able to release a fully 64-bit OS with the A7? Wonder no more. The 64-bit version of iOS is 11 years old.
Not at all, for the most part iOS is Mac OS with a slightly different GUI and a more restricted execution environment. That being said I'd love to see a Mac OS device that uses A7/A8. Even better if the machine supports easily booting other operating systems.
Apple seems to like lateral moves. If release an A-chip-based notebook it might not even be called a Mac or use Mac OS X, at least not at first. This could allow them to build a parallel project and let developers get apps converted, work out any issues, and further advance their A-chips to a point that it could eventually allow them to make the switch much further down the road.
I don't think Apple would work it that way. I'd think they make it at least equivalent performance (if they were going to replace an existing product) but then use all the gains to balance a reduction in materials cost, weight and size, as well make the battery life much, much longer (if we're talking about a notebook).
That is a matter of opinion.
Two points:
Would there be technical and/or bargaining advantages by include an A8 in addition to an x86 chip on a Mac device?
An A8 chip would cost about $12, allowing Apple to sell an A8 laptop for $699 (and at a higher profit margin). It would also have about 120 hours of battery life. Oh, and if they put 8 cores into the thing, it would run circles around Haswell.
And since the ARM is basically a PPC, the SDK's are all there, and much of the code is already written. It would just need an updating.
120 hours is too high. The CPU is only part of a laptop's power draw. The DRAM, GPU, screen, power supply, wireless, etc... all take up a lot of power. Even if the CPU took zero power you might only see a 10-20% battery life improvement.
I think it is also a little far to say 'ARM is basically a PPC' – they are both RISC cores, but the ISAs are very different, and, more importantly, their memory models are different. Porting can be done, but it is very non-trivial.
And while I am sure Apple could manage the transition (as they did with 68K->PPC and then PPC->x86) it would still be tricky. PPC gave a huge performance boost over 68K and x86 gave a future which PPC simply didn't have. The ARM cores do give good performance per watt, but don't give a huge raw performance boost over x86.
Apple would also have to improve the 'uncore' part of the A* line to be competitive with x86 - things like more scalable cache coherency, more memory channels, etc... All of this can be done, but is non-trivial.
On the other hand, the A* line could be less expensive, and it could allow some interesting integration paths that x86 doesn't support – e.g. integration with stacked memory in the package.
You can't look a machines designed for the pro market in the same way as for the general consumer. The fact is, as demonstrated by the iOS world, is that most consumers have zero need for i86 compatibility. Even more so with Windows compatibility. The world is a changing and Apple is in the drivers seat.
Well that made me sit up.
Possibilities galore.
Gaming. In AppleTV. High Performance Gaming. With gesture control. Gotta be what Apple is thinking. It's ARM, therefore iOS compatible, where there are more and more immersive games. That's gotta be where they're taking this.
I've stated all along that Apple's A7 is light years ahead of what Samsung and Qualcomm are doing.
I also stated previously that the A7 is so good that they don't really need to improve it. An A8 could simply be Apple's first quad core SoC using 4 A7 cores. Or they could ramp up the A7 clock to 2.0GHz and slightly tweak the design. Both would again give them a decent performance increase without a lot of effort.
I think the next big re-design will be reserved for the A9. Then again, Apple surprised us with the A6 & A7 and maybe they will again with the A8.