iOS 10, OS X 10.12, Apple A10 chip, and where this is all headed

[throAU]

I remember the transition from PPC to Intel, and it seemed like overnight people started clamoring for Macs because it would now run Windows and many builds of Linux. The Macintosh marketshare would not be as healthy as it is today, (regardless of the iPhone halo affect), if they stayed on the PPC or moved to the Ax processor.

I was one of them. Additionally, if i don't like OS X, i still have nice hardware to run something else on. THAT, is a massive selling point to potential PC switchers who are on the edge. They have the comfort of being able to run Windows on it if they don't like it.

Switch to ARM, and the options go away.

 

[maflynn]

i still have nice hardware to run something else

Agreed, and I find myself running Windows on my iMac quite often, and at times for work, I need windows, so I'd need to look at a different solution.

 

[grahamperrin]

… Apple was going to switch to Intel back during the G4 days …

I recall vague speculation (I was an avid reader of Mac-related magazines and so on) but Apple's announcement of the switch was a complete surprise to me. It seemed to me as if Apple had kept the secret very well. Maybe I read the read the wrong magazines

… The Macintosh marketshare would not be as healthy as it is today, (regardless of the iPhone halo affect), if they stayed on the PPC or moved to the Ax processor.

+1

What was Ax?

 

[/V\acpower]

Apple have been steadily taking Mac OS X in the same direction that Microsoft did with Windows 8, trying to bring features of a mobile platform to the desktop. If they continue this pattern we can expect to see more and more iOS features (Siri is a good example of this, though I believe that Siri would actually be a useful feature). Other aspects will probably include more dumbed down software, making the desktop environment progressively less useful. Apple don't seemed to have learned from Microsofts monumental screwup, that customers do NOT want a mobile OS on their desktop, but they are trying to bring it in by stealth one piece at a time. Features like their full screen start screen (Launch Pad) that looks like something off an iPad, and is very similar to the well hated full screen start menu on Windows 8, are a good example of what has been happening to OS X for years.

Ah please, Launch Pad is only a very good way to find and launch app, way way better than the old "find your application folder" ever was.

Windows 8 home screen became the tablet interface. In OS X the good old reliable desktop is still the default interface. Launch Pad is simply a functionnality of that desktop.

Launch Pad "look" very superficially like an iPad home screen, but it doesn't make the mac work like an iPad at all.

Plus, i'm very curious about what change in particular to OS X UI is actually in the process of "making the desktop environment progressively less useful." ?

 

[Zarniwoop]

Apple pays Intel a hefty $250 USD per core-m3 processor that has A9X level performance. It costs Apple ~$40 to make an A9X processor. R&D included. You see, Apple invests certain amount of money to design an Axx chip. Then, the more they can produce those, the less is the cost per chip. Putting an ARM to Mac is a no brainer, any Excel sheet that Tim Cook runs will show it.

A person who buys Macbook (or even Air) doesn't care too much about performance. If their Office, Facebook and Twitter works... they really wont care less if it runs on Intel or ARM.

The more problematic area are the iMac and Pro users. But you know what, they can have their Intel's. If there's a Mac software made for ARM only, it can be emulated.. a desktop chip is powerful enough to run normal ARM software. And those who need specialized Intel software for power use, wont's buy a Macbook anyway. If Apple takes the ARM route, it's not that difficult to have Office and some other key apps to be compiled for both, Intel and ARM, natively.

But, I think, macOS 10.12 is not mature yet for this transition. Perhaps next, macOS 10.13 with new filesystem and third revision of Metal will open the doors for ARM & Intel side by side life.

PS. AMD has a technology to put ARM & x86 chip on the same motherboard... even on the same SoC. Who knows what AMD & Apple are "Cook"ing together... custom Zen with Apple ARM?

 

[leman]

It should be possible to 'compile' the x86 binary code to ARM binary. AFAIK, there are open-source tools which an translate from x86 to LLVM IR. Using a toolchain like this would allow one to seamlessly run any x86 code on a CPU of a different architecture, without the performance penalty commonly associated with emulators.

Again, I'd like to point out that I am not a big fan of a possible ARM transition (not at this point anyway),for obvious reasons. But it is worth noting that such a transition — if Apple considers it — might be much less technically challenging than many here assume. The software infrastructure and the tech required to do so seamlessly already exists.

 

[maflynn]

recall vague speculation (I was an avid reader of Mac-related magazines and so on) but Apple's announcement of the switch was a complete surprise to me. It seemed to me as if Apple had kept the secret very well. Maybe I read the read the wrong magazines

There were rampant rumors of Apple maintaining an x86 version of OS X up to date and the sites I had frequented at the time (I forget now) had various discussions on the what if. I will say the switch was a surprise for many people, but I won't say it was a complete surprise as there were enough rumors and speculations occurring.

 

[rnbwd]

Historically, Apple has chosen quality over price - fairly consistently. I'm sure intel is doing whatever they can to keep Apple as a customer, and (to best of my limited knowledge) intel processors - at least subjectively - appear superior to most arm processors - although the gap is growing smaller. Macbook's are also 'subjectively' better than PC's - the numbers don't translate into experience. I remember 5-6 7-8 years ago when I first got a macbook how people thought it was silly to buy such an overpriced machine - and yet - I still own that laptop and it works great - while my friends with other laptops had to replace them every few years. Just this year my boss purchased a windows laptop for the printer driver - but we both had macs running virtual machine at the same time - with a quarter of the allocated RAM /etc. that the laptop claimed to have - it ran twice as fast on my macbook. I don't know the technical details - but I'm absolutely certain that the metrics discussed in popular forums / reviews / advertisements are not a reflection of the reality of using the product.

Anyways - i think apple will choose what works best - arm / intel - probably depends on the device. I can see the smaller macbooks running arm while the macbook pro's will probably keep intel for the foreseeable future. I'm excited about APFS - but most of that will be implemented in the SSD drives / RAM / Cache - I'm not sure what difference running ARM / Intel would be for the new FS.. but if there is a difference then it's a more compelling argument.

 

[lysingur]

Anyone who seriously thinks Apple would consider this switch to ARM in the near future doesn't understand the reasons for the switch the last time.

The switch is already happening with iPad Pro, a tablet with a keyboard and a stylus. For one thing, computing has changed. Consumers care less about speed now than cost because computers have reached a speed that can satisfy most needs of your regular consumers. So your whole argument here is anachronic, not to mention plain wrong (see below).

By the time that Apple released the first Intel iMac, they were claiming a 2-3x advantage in raw processing power over the G5 version. With the first Macbook pro, they were claiming 4-5x over the fastest G4 Powerbook.

http://barefeats.com/pentium4.html (so, where did you get your 2-3x number from? Care to share?)

Rubbish. Absolute rubbish. The Windows to iOS+OS X ratio is 1.5:1 and even then it is totally disingenuous to include iPhones in this comparison.

1.5:1 of what? Unit shipment? Web clients? It's not disingenuous to include iOS at all as ARM chips used in iPhones and iPads help spur research.

What do you base this on? OS X and iOS have always had huge amounts of code in common.

Based on the fact that both software and hardware are converging on the two platforms.

HAHAHA. I love this. The "only" thing missing is the thing that is by far the most important, maybe not even possible and certainly not within your 3 year prediction.

Based on the speed increase of the ARM processors, it's certainly possible, if not probable.
[doublepost=1467153703][/doublepost]

I was one of them. Additionally, if i don't like OS X, i still have nice hardware to run something else on. THAT, is a massive selling point to potential PC switchers who are on the edge. They have the comfort of being able to run Windows on it if they don't like it.

Switch to ARM, and the options go away.

With so many softwares that people regard as essential on Windows now available on the cloud, this has already become less and less of a problem.
[doublepost=1467154140][/doublepost]

I remember the transition from PPC to Intel, and it seemed like overnight people started clamoring for Macs because it would now run Windows and many builds of Linux. The Macintosh marketshare would not be as healthy as it is today, (regardless of the iPhone halo affect), if they stayed on the PPC or moved to the Ax processor.

With cloud computing, cross-platform compatibility and being able to run multiple platforms on one machine have become less and less of an issue. It'll certainly be even less of an issue in three years when cloud services will have been even more mature and the availability of even faster wireless technologies more widespread.

 

[grahamperrin]

… iPad Pro, a tablet with a keyboard and a stylus …

http://web.archive.org/web/20151225234047/http://www.apple.com:80/ipad/compare/? no mention of the phrase key …

 

[HatMine]

Since APFS will be used for both iDevices and Macs (+ Apple TV and Apple Watch, I believe), and since Swift code can be compiled to run on all Apples devices, they are really starting to tie all of their machines together. A Retina MacBook with an ARM-processor would make sense, although it's hard to tell when they would like to release it. At best they may have one ready by Q3 2017; by then APFS will be out, and there'll be even more apps coded in Swift that will run on the ARM MacBook.

But that's just a wild speculation. It will be very interesting to see what Apple will do with the 2016 MacBook line-up. Personally I'd like to see a rMBP 13" with a Quad-core i5, DDR4 memory and a P3 display.

 

[lysingur]

http://web.archive.org/web/20151225234047/http://www.apple.com:80/ipad/compare/? no mention of the phrase key …

It's just one way Apple makes consumers pay more for full functionality. Both the Smart Keyboard and the Apple Pencil are marketed as "for iPad Pro" and I don't see any other iPad Pro out there except the aforementioned iPad Pro.

 

[grahamperrin]

@lysingur thanks, now I see http://web.archive.org/web/*/https://www.apple.com/smart-keyboard/ (nine months … I had no idea that Apple made a tablet keyboard)!

 

[mildocjr]

Intel is quietly updating their naming convention for this release, it will include the Intel i10-1210 and i10-1210X chips which will allow the new file system to be bootable. It will also come in mobile processors that will be included in the iPhone 7 providing 3.5 Ghz of quad-core compute power.

 

[SlCKB0Y]

http://barefeats.com/pentium4.html (so, where did you get your 2-3x number from? Care to share?)

Firstly, Apple didn't switch because of Pentium 4, they switched because of the future roadmap of Intel vs. that of PPC along with the massively more energy efficient processors.

The speed claims I mentioned were those cited by Apple when they released their Core Duo iMac and Macbook Pro at Macworld 2006.

1.5:1 of what? Unit shipment? Web clients? It's not disingenuous to include iOS at all as ARM chips used in iPhones and iPads help spur research.

Estimates put total number of Windows computers and servers at about 1.5 billion and iOS + OS X at just over 1 billion.

Based on the fact that both software and hardware are converging on the two platforms.

I'm not sure what you mean by saying the software is converging since they are both based on the same underlying OS. I would argue that the two platforms are undergoing integration not convergence, and this allows them to work much more harmoniously together.

Based on the speed increase of the ARM processors, it's certainly possible, if not probable.

It's not even remotely possible. The highest end of ARM currently performs in line with the lowest end of Intel. It won't come close to matching the high end of Xeons within your 3 year period, nevermind the multiples it will need in speed in order to emulate Intel (unless you think Apple would switch without compatibility).

Even after you consider all this, and even if ARM can massively scale up in terms of speed, there is no guarantee they can do this and keep all the power efficiences.

 

[Zarniwoop]

Just for fun: For the price of one Four core Xeon, Apple could put ten A9Xs in to a desktop PC. That would be twenty core beast with ~50W TDP. All the GPUs are GPGPU ready. This fugure is actually MacBook Pro level of pricing and thermal conditions.

 

[wrldwzrd89]

Let's be honest here: The one thing holding Apple AX processors back from true desktop use is a "Turbo" implementation, like Intel's Turbo Boost / AMD's Turbo Core. While the lack of this feature isn't holding iOS back at all, and in fact a strong case can be made that iOS simply doesn't need it, macOS definitely does.

 

[Zarniwoop]

Sure the previous ten SoC's system is not very practicsl solution, but lets say that there is going to be a four, six and eight core versions of next A10 chip. That will definetelly rival Intel.

Going to in-house ARM chips is one of the few options Apple has to keep Macs competitive and (Apple level) proftiable.

 

[HatMine]

Sure the previous ten SoC system is not very practicsl solution, but lets say that there is going to be a four, six and eight core versions of next A10 chip. That will definetelly rival Intel.

If they make, say, an A10X with 4 cores and a high clockspeed, then yes, they would rival Intel. But probably not the "higher end" Intel i-processors (like the ones in the MacBook Pros).

I don't find it unlikely that Apple will make a MacBook with an AX-processor instead of an Intel m-processor. Knowing the power of the ARM processor in the Ipad Pro, it wouldn't be strange if it would outperform the MacBook Airs and the MacBook Retina.

 

[cynics]

We need to keep in mind the features current Wintel ahem Intel CPU's offer. Skylake for example supports QuickSync FF-Mode, HEVC, a bunch of new features for RAW image processing plus certain hardware support like DDR4, PCIe 4, Sata Express, etc etc...

Plus of ton of patents on certain tech that Apple would need to reinvent. For example, Intels QuickSync...

Nothing is impossible but a lot of post here sound like we could just drop in an ARM CPU and call it a day. Even if Apple made an ARM CPU that was literally faster it would be a massive downgrade without A LOT of work, which I'm sure Apple is more than capable of but I don't think the time is right....for quite a while for that matter...

 

[HatMine]

We need to keep in mind the features current Wintel ahem Intel CPU's offer. Skylake for example supports QuickSync FF-Mode, HEVC, a bunch of new features for RAW image processing plus certain hardware support like DDR4, PCIe 4, Sata Express, etc etc...

Plus of ton of patents on certain tech that Apple would need to reinvent. For example, Intels QuickSync...

Nothing is impossible but a lot of post here sound like we could just drop in an ARM CPU and call it a day. Even if Apple made an ARM CPU that was literally faster it would be a massive downgrade without A LOT of work, which I'm sure Apple is more than capable of but I don't think the time is right....for quite a while for that matter...

They could literally just take an Ipad Pro, attach a keyboard to it, and call it a Mac. The one thing holding them back is the fact that they would need to make macOS run on ARM, and all which that means.

But thanks to swift, which can compile apps to run on ARM AND x86, the app compatibility isn't as big of a problem as it could've been. Like I've said before, I wouldn't be surprised if we saw an ARM MacBook in 2017 or 2018.

However, what's more likely to happen is that they start to make iOS even more like macOS (feature wise), and continue to place their hopes on the Ipad Pro.

 

[cynics]

They could literally just take an Ipad Pro, attach a keyboard to it, and call it a Mac. The one thing holding them back is the fact that they would need to make macOS run on ARM, and all which that means.

But thanks to swift, which can compile apps to run on ARM AND x86, the app compatibility isn't as big of a problem as it could've been. Like I've said before, I wouldn't be surprised if we saw an ARM MacBook in 2017 or 2018.

However, what's more likely to happen is that they start to make iOS even more like macOS (feature wise), and continue to place their hopes on the Ipad Pro.

Valid point however it doesn't address any of the proprietary features Intel has built in.

I'd still be weary of future capability with newer tech though. Thunderbolt for example is Intel tech, and as far as I can tell requires an x86 processor (citation needed). Any sort of dedicated graphics. The very fast moving m.2 standards (Samsung recently releasing their 3200mbs r/w cards or whatever they are at very competitive prices I might add). I just personally feel if Apple wants to remain even remotely hardware competitive they are going to need to stick with x86 for a while.

PLUS the software, and assuming that can be addressed easily as mentioned above its still another bump in the road.

I'm also leery of real world performance, not on a tablet but in a desktop doing desktop task. Look at AMD's venture into ARMv8, there might be a place for it in the market just not in my Mac. Currently Apples ARM is compared to Intels Core M, which is impressive however...well....yeah. Even if it was 100% better thats still abysmal compared to what we currently have available.

However I do agree with you, if/when we do see it it will be in a MB, maybe even a new line or nMB if nothing else just to test the waters.

Regardless my argument will eventually become moot. Apple will eventually do something, my argument however is that we are a long way off to seeing anything more then a glorified ultrabook and even that I feel is a ways out. Again just my opinion...

 

[Zarniwoop]

Apple ARM A9/X SoC already has a bunch of co-processsors; DSP (digital signal processor for audio effects and sound processing), ISP (multipurpose Image & video processor and encoder), GPU (500-768 GFLOPs GPGPU), NVMe controller and DDR4. There's also M9, but for laptop it's not so important, besides SIRI.

 

[Janichsan]

But thanks to swift, which can compile apps to run on ARM AND x86, …

…which does absolutely nothing for thousands of legacy applications, of which the vast majority is not written in Swift.

 

[wrldwzrd89]

…which does absolutely nothing for thousands of legacy applications, of which the vast majority is not written in Swift.

Furthermore, there's no incentive to port these applications to Swift until the API stabilizes, which hasn't happened yet.