Jump to content

End of Hackintosh nearing (slowly)?


frankiee
 Share

40 posts in this topic

Recommended Posts

Look here: https://www.bloomberg.com/news/articles/2017-02-01/apple-developing-new-mac-chip-in-test-of-intel-independence

 

Sure, for a while Apple has to support older macs without that chips as well. But within a few years, when macOS only supports - and needs! - those custom chip its over I guess.

 

What do you think? Would it be possible to emulate this functionality with a kext, just like FakeSMC (kinda) emulates the SMC chip? Guess even if possible, a LOT harder to do so.

Link to comment
Share on other sites

I personally don't believe in a soon to be presented Apple computer powered only with an ARM chip, that article it's about a chip for specific features similar to the one present on the 2016 touchbar models, probably it will as described, a chip for battery saver powernap and maybe something else, today many users are buying Macs, even cause they can run Windows too, and Apple will not want to lose such users...

 

But probably, not in the near future Apple could build x86 chips too, who knows!

 

BTW long life to the hackintosh :)

Link to comment
Share on other sites

I personally don't believe in a soon to be presented Apple computer powered only with an ARM chip, that article it's about a chip for specific features similar to the one present on the 2016 touchbar models, probably it will as described, a chip for battery saver powernap and maybe something else, today many users are buying Macs, even cause they can run Windows too, and Apple will not want to lose such users...

 

Well, they could still build machines that can run bootcamp, in that case windows simply ignore the extra chips.

 

But things might get hairy if macOS demands that such an extra chip is installed. Building hackintoshes will be virtually impossible then, unless someone goes the extreme way of emulating such hardware. Good thing is that this transition - IF it happens - will take another several years for sure.

 

And I am also not so sure if we don't see a Mac with ARM only in the near future. Not highend machine of course (well, does Apple still build anything highend? lol). I mean, I bet they have an ARM version of macOS running for quite a while, just like they did with the transitition to Intel.

 

 

BTW long life to the hackintosh :)

 

Word!

Link to comment
Share on other sites

If we use 10.12 as any indication of their future, I think we have more to worry about with the OS itself being usable instead of what they use for hardware.

 

Ahhh yeah. Exactly! This one gives me worries as well. Especially Apple QA is not what it used to be, and since  apparently there isn't even a dedicated macOS team anymore, I also think this might not get much better. Actually I  did something I never did before now, and downgraded my OS back to El Capitan. Wasn't much fun, but works much better for me.

 

Not only that, also the neglection of the "pro" users in general is not good, both regarding hardware (trashcan, anyone?) and software (useless new "features" like even more emoticons, dumbing down things all overall the place). I mean actually the lack of a proper workstation aka "Mac Pro" (at least what _I_ regard as a proper workstation ...) is the one main point that drove me to the hackintosh scene. I admit I am not so much the "hacker / tinkerer" type of person, I actually have to work with my machine and earn some money to pay my rent ... But hey, at least I learned a lot in the process (but I am still kinda stupid I guess) ;)

  • Like 2
Link to comment
Share on other sites

Ahhh yeah. Exactly! This one gives me worries as well. Especially Apple QA is not what it used to be, and since  apparently there isn't even a dedicated macOS team anymore, I also think this might not get much better. Actually I  did something I never did before now, and downgraded my OS back to El Capitan. Wasn't much fun, but works much better for me.

 

Just curious, what do you not like about Sierra? I've found it to basically be a slicker El Capitan. What features did they remove? I still hate the post-Mavericks UI, but not a lot I can do about that.

 

These extra processors don't sound like a big deal IMO. Worst case scenario, sleep won't work. Oh no! Whatever will I do when my desktop computer that's designed to always be connected to a power source can't go to sleep!

Link to comment
Share on other sites

I think lots of you remember the WWDC 2005 keynote and the "Performance per wat" explanation by Steve Jobs about switching the PowerPC chips with Intel. I wonder where the current ARM chips stand in such sheet, not to mention the Intel value in that sheet has also changed over the years (it's a matter of 12 years after all).

At the current point, there's no ARM chip that can compete with any Intel Core i3 chip, mobile or desktop (which Apple don't even use), except maybe in term of power consumption (with corresponding productivity of course) , what about i5 or i7...

Poor Steve Jobs is probably already started turning in his grave...

 

Edit: IMO, it's more likely to see Intel chips with AMD graphics core inside them than ARM chips in all Apple computers in the near future.

  • Like 5
Link to comment
Share on other sites

Just curious, what do you not like about Sierra?

 

First of all, there seems to be absolutely nothing new that is actually useful (for me at least - maybe the new tabs feature, but I dont miss that actually). And at least for me it does not seem as stable / reliable as 10.11.6. Well, based on my experience, at least since Yosemite no macOS is really stable until a 1x.x.5+ revision, so maybe I also was too early to migrate. Since I run a production machine, I do not update to any x.x.0 Version anymore, waaaay to risky.

 

For example, there are even more graphical glitches with NVIDIA web driver, but also 3rd party Issues, like my parallels macOS VMs not functioning properly anymore. Yeah, you could blame parallels for that (but not sure whose fault it is), but I don't feel like buying a new version each and every year. And with 10.11 it "just works". And those graphical glitches I mentioned are there even when using the native NVIDIA driver on a real mac! For example, do some fast user switching on a machine with NVIDIA graphics and you will see what I mean. So also on my "real" MacBook Pro, fast user switching basically means I have to restart my machine after that.

 

 

What features did they remove?

 

One point is the console app. At least _my_ most common two usecases were 1) inspecting boot log, and 2) opening console.app after some glitch / issue, to see what just happened. Both not possible anymore, at least not without using some rather crude workarounds (without using the console at all). Console has become also useless, because it basically spams me to death with irrelevant stuff, and I had to make some extensive filtering to tame the output a bit. I mean it could have been much better. Feels a bit like what they did with the DU app in El Capitan, i.e. looks more pretty but is also much more useless. And yeah, the old DU still works in El Cap with a patch, but apparently not anymore in Sierra.

 

I had also problems with some 3rd Party PDF app, and another thing is that - at least with Web Driver - some Apple apps totally ceased to work, namely Instruments, iBooks and iBooks Author. And since I have an actual job that involves producing such iBooks, that was the final dealbreaker for me.

 

But ultimately, my problem with macOS and Apple in general is not really so much a specific version, i.e. Sierra. It's more like they neglect macs (especially for "pro" folks who need to earn money with the platform) in general in favor to all that iOS stuff, and that Apple QA has gotten way worse in recent years. Anyone remember Snow Leopard? Well, that was quality software in my eyes. Absolutely rock solid and reliable! Not any more ... If only the alternatives I have would not be even more worse (I hate Windows) or downright unusable (all Linux).

At the current point, there's no ARM chip that can compete with any Intel Core i3 chip, mobile or desktop (which Apple don't even use), except maybe in term of power consumption (with corresponding productivity of course) , what about i5 or i7...

Poor Steve Jobs is probably already started turning in his grave...

 

Yeah, at the current point! But this gap is narrowing more and more, so I am not so sure that we don't see this special configuration in other Macs as well, or even ARM only Macs (but this may take longer of course). Guess we will have to wait and see.

Link to comment
Share on other sites

First of all, there seems to be absolutely nothing new that is actually useful (for me at least - maybe the new tabs feature, but I dont miss that actually). And at least for me it does not seem as stable / reliable as 10.11.6. Well, based on my experience, at least since Yosemite no macOS is really stable until a 1x.x.5+ revision, so maybe I also was too early to migrate. Since I run a production machine, I do not update to any x.x.0 Version anymore, waaaay to risky.

 

For example, there are even more graphical glitches with NVIDIA web driver, but also 3rd party Issues, like my parallels macOS VMs not functioning properly anymore. Yeah, you could blame parallels for that (but not sure whose fault it is), but I don't feel like buying a new version each and every year. And with 10.11 it "just works". And those graphical glitches I mentioned are there even when using the native NVIDIA driver on a real mac! For example, do some fast user switching on a machine with NVIDIA graphics and you will see what I mean. So also on my "real" MacBook Pro, fast user switching basically means I have to restart my machine after that.

 

 

 

One point is the console app. At least _my_ most common two usecases were 1) inspecting boot log, and 2) opening console.app after some glitch / issue, to see what just happened. Both not possible anymore, at least not without using some rather crude workarounds (without using the console at all). Console has become also useless, because it basically spams me to death with irrelevant stuff, and I had to make some extensive filtering to tame the output a bit. I mean it could have been much better. Feels a bit like what they did with the DU app in El Capitan, i.e. looks more pretty but is also much more useless. And yeah, the old DU still works in El Cap with a patch, but apparently not anymore in Sierra.

 

I had also problems with some 3rd Party PDF app, and another thing is that - at least with Web Driver - some Apple apps totally ceased to work, namely Instruments, iBooks and iBooks Author. And since I have an actual job that involves producing such iBooks, that was the final dealbreaker for me.

 

But ultimately, my problem with macOS and Apple in general is not really so much a specific version, i.e. Sierra. It's more like they neglect macs (especially for "pro" folks who need to earn money with the platform) in general in favor to all that iOS stuff, and that Apple QA has gotten way worse in recent years. Anyone remember Snow Leopard? Well, that was quality software in my eyes. Absolutely rock solid and reliable! Not any more ... If only the alternatives I have would not be even more worse (I hate Windows) or downright unusable (all Linux).

 

 

Yeah, at the current point! But this gap is narrowing more and more, so I am not so sure that we don't see this special configuration in other Macs as well, or even ARM only Macs (but this may take longer of course). Guess we will have to wait and see.

how about use both

dmesg and console,

html and pdf and ibook,

amd and nvidia,

arm and x86,

debian and mavericks.

Link to comment
Share on other sites

how about use both

dmesg and console,

html and pdf and ibook,

amd and nvidia,

arm and x86,

debian and mavericks.

 

Yeah and why not use quintuple boot? One system for each task? Or needing to use 5 computers instead of one? Do some graphic stuff in Illustrator for example, like creating a svg Icon and reboot into Linux for testing it on my local web server? Reboot 5 minutes later to adjust that graphic? Reboot 2 minutes later to adjust some CSS? Rinse and repeat?

 

Maybe I dont get your reply, but I think you are suggesting that I should go to straight to the asylum :surprised:

 

And, I mean if my clients wants that I produce an iBook, I cannot say: "see, that does not work with my super duper macOS Sierra, so I give you PDF". And no, using multiple GPU brands to maybe work around that issue also does not sound like a proper solution.

 

So, why not use one system that does all I want? And that system actually exists - at least until now - and is called "OS X El Capitan".

 

Besides my graphic jobs, I am doing also a lot of (frontend) web development, and OS X used to be perfect for both of these. (Graphic Design is the main reason - due to lack of software - why all Linux are not usable for me).

Link to comment
Share on other sites

I worked up to mid 2016 still with Leopard, so even if they will switch, it is not too much of a problem. If a new version of OS X requires a new special component, we'll see.

 

However, with the Macbook Pro 2016 Apple screwed away a lot of users. I found nearly nobody who had a positive review, and even in the business network XING you can talk completely open about building a hackintosh as an alternative. This wouldn't be possible a while ago. Apple has gotten so much critics for the Mac line, if they will announce to switch away from Intel this would be a desaster. The last persons who bought Intel Macs will be really annoyed, and the ones who know the history of Apple they know that old architectures like 68K or PPC were very fast discontinued. If this will happen again, perhaps Mac users will switch to Windows because they know they will stay at x86 and AMD64, and on Windows you'll get the most programs like for the Mac. In this article was stated the Apple stock got +6.1%, but the customers choice will be more important, so this value says nothing.

 

ARM was never a platform for desktop computers, some may remember the british Acorn Archimedes RISC computers. They were wery fast, but they didn't took off. But Microsoft bought the task bar of the RISC OS. ARM is used for smartphones, tablets, Android minicomputers for TV and experimental things like the Raspberry Pi and such things. ARM is also used in embedded systems. But an ARM desktop system which is matching the speed of an Core i7 with 8 cores and 4 GHz? I can't imagine this.

 

@ frankiee:

 

I have a septaboot system:

 

Windows XP

Windows 7

OS X Sierra

OS X Mavericks

OS X Snow Leopard

Zorin OS 9 (Ubuntu 14 derivate)

Android-x86

 

All one one 1 TB harddisk, and there are still about 300 GB free for data. If I need something else, I boot into it, but my default is Mavericks. I also have no alternative to OS X, I'm no Windows fan and there's no software for Linux. I tested 30 audio players, and no one is comparable with iTunes (version 10.7 ;-).

  • Like 2
Link to comment
Share on other sites

 

@ frankiee:

 

I have a septaboot system:

 

Windows XP

Windows 7

OS X Sierra

OS X Mavericks

OS X Snow Leopard

Zorin OS 9 (Ubuntu 14 derivate)

Android-x86

 

All one one 1 TB harddisk, and there are still about 300 GB free for data. If I need something else, I boot into it, but my default is Mavericks. I also have no alternative to OS X, I'm no Windows fan and there's no software for Linux. I tested 30 audio players, and no one is comparable with iTunes (version 10.7 ;-).

 

Septaboot? OMG, you are a brave person :)

 

Actually, I even have over 25 systems in total, but most are mainly here for testing my webpages, and I strongly prefer to use VMs for this if it makes sense (but have also a lot of tablets and smart phones of all varieties, since here it is better to have "the real thing"). So why don't you use just Parallels or VMware? Imho a lot more managable and convenient - if you don't have a good reason to boot natively of course.

 

But for work I prefer to have only *one* system and that is of course OS X. Also have windows dualboot (for occasional gaming only).

Link to comment
Share on other sites

It's not only installing multiple operating systems on one drive, I've installed Windows XP and Windows in one partition ;-) That's possible because it's a German version, and XP has real localized names. SO 'Documents and Settings' is 'Dokumente und Einstellungen oder 'Program Files' is 'Programme'. So there are no name conflicts, and it works. It's even possible to share the pagefile.sys. ZorinOS and Android-x86 are also installed in one partition using EXT3 filesystem. The Android installer creates an own directory, so there are also no name conflicts.

 

The tricky part is the bootloader configuration. I'm using Clover, EasyBCD, and GRUB 2. It's possible to implement all operating systems in GRUB 2, but the most important ones could be directly loaded by Clover.

 

I also don't need all operating systems, but if I do, I have them all on one computer. And it's fun ;-)

  • Like 1
Link to comment
Share on other sites

Yeah, but still VMs may be more convenient, I mean you not only have all systems on your computer, you even can run them all at once  :)  (If you have enough RAM of course). Very convenient for side by side testing.

 

Plus, you have things like snapshots and suspend, makes life a lot easier, esp. if you do something by accident that ruins your installation. Or you install something, play with it, and ditch it by rewinding to a former snapshot, so it leaves absolutely no unwanted leftovers. Plus its more safe since each VM runs in its own sandbox.

 

And I cannot mix Windows Versions, since it is hard to impossible to install multiple IEs on one Windows system, and it can have weird side effects even if you somehow manage to do it. Same problem with Safari. At least for testing its best to have an unmodified vanilla installation, but of course YMMV.

Link to comment
Share on other sites

If you are using an english version of Windows XP you can mix it with a newer one, because the names are the same. Windows XP and before has full localized filenames, modern versions are using english names and a localization over it. In Windows it's called 'MUI'. The first Windows version which uses MUI is Windows XP Professional x64 Edition.

 

OS X also uses english names in the filesystem and a localization over it. So if I enter in the Terminal cd /Programme/Dienstprogramme, I'm getting an error No such file or directory. I have to enter cd /Applications/Utilities, then it works. For beginners in the Terminal this may be annoying. Or the keyboard. In Clover or in single user mode, the US keyboard is selected. Finding the keys is also a bit difficult.

 

But I like fiddling around, that's why I have so many OS'es on the drive. Not long ago, I got two Toshiba Libretto 50ct, one of the smallest laptops ever. It's about the size of a VHS tape. It features a Pentium I with 75 MHz, 16 MB RAM, a 6,1" truecolor display and an integrated Soundblaster Pro, which makes it ideal for DOS gaming. The docking station has VGA, Serial and Parallel. Floppy or bootable CD will be connected using PCMCIA. It also has a hardware-based hibernation mode which works even in DOS. Bad is, it uses 8,4 mm high harddisks. So I used an 8 Gig Compactflash card. It's possible to partition and format it, also installing DOS 6.22 (I have the original disks ;-) works. But it won't boot because the BIOS allows drives up to two or four Gig. After some contacts I got EZ-Drive, it allows to boot from harddrives bigger that size. Now I have MS-DOS working and had to install everything from scratch, mouse driver, soundblaster variables, everything. And also the fun with the great memory configuration including 640K, EMS and XMS. Some games won't work with EMM386 or so, and I integrated some really tricky features to change between memory configurations - luckily I'm very good in DOS script programming. It's also weird: with the simple command echo G=FFFF:0000| DEBUG >NUL you can reboot a  computer running MS-DOS ;-) Somehow, MS-DOS is funnier than modern operating systems.

 

If anyone is interested in building a DOS system, I can provide my system. It's not too big to be uploaded, but the MS-DOS version is German. Using a compactflash card instead of a harddrive is perfect because they're absolutely quiet, not expensive and data exchange with modern systems is very easy.

  • Like 1
Link to comment
Share on other sites

Apple will not go into ARM chips with macOS unless it's still planning iOS convergence: thy have forked a lot of ARM code when they spun off iOS and rewriting that code for more functionality is inefficient compared to just glossing over those features with simpler application "updates".

Apple also won't go into ARM chips because it would take them a sizeable chunk of their bank to catch up with Intel on performance. Worse still is the fact that all the stuff Intel did recently is definitely still under patent so every single modern method in the chip must be reinvented. (This is why it took AMD so long to catch up with Ryzen; if you take too long to get to the patent office, you have to think of a new way to do exactly the same thing, often a less efficient way just to avoid the patent)

Link to comment
Share on other sites

I can't imagine that Apple will not release ARM-iMacs. Here are some reasons:

 

 

- Macs are rather uninteresting for Apple, the money comes from iPhones, iPads and iTunes. It would be easier and cheaper to reject Macs.

 

- architecture switching would result in a lot of work. The complete system, all drivers, programming tools, Apple software have to be reprogrammed. Companies must be convinced to switch to the new architecture. Imagine: Adobe should reprogram all its software to a platform which is only used by a few users, and their software works better on Windows or OSx86-systems because it's optimized for nVidia? Hobby programmers will have much less interest to learn a new processor architecture.

 

- architecture switching would need a complete hardware redesign. In comparison, the last Mac Pro was a failure and wasn't updated since 2011. The last Macbook Pro also was a failure, with some features which were not accepted by the buyers.

 

- Changing to a new architecture will be an immediately sign that the last expensive Macs bought will be outdated soon. Buyers won't like that.

 

- On ARM-Macs Windows won't work natively. Another negative issue.

 

- ARM-Macs will never have the same performance like Intel-Macs.

 

 

And... there are so many ARM-based experimental platforms on the market. Operating systems like Windows, Linux and Android are working on the Raspberry Pi. What if some guy developes a board which is compatible enough to run ARM-OS X? This would be the biggest laughter ever. It's funny enough that Apple tried on the Intel-Macs so hard to make the hardware as different to PCs (EFI, GPT, SMC, TPM) and already version 10.4.4 was even working on a low-end PC with a Pentium 4. Now are not even reprogrammed kernels are neccessary, just a bootloader.

 

I'm sure, even if Apple would switch to ARM, there will be enough people who want to get it working on other platforms, like the beginners of the OSx86 scene more than 10 years ago.

I don't think iOS is working on ARM-platforms, but probably because it's too uninteresting. Android is Open-Source, so it's much easier to modify. And you are not forced to have contact to Apple.

  • Like 1
Link to comment
Share on other sites

I can't imagine that Apple will not release ARM-iMacs. Here are some reasons:

I have to disagree here

 

- Macs are rather uninteresting for Apple, the money comes from iPhones, iPads and iTunes. It would be easier and cheaper to reject Macs.

Unfortunately this is true. But they cannot ditch them at least not now. So it might be cheaper to unify platforms as a first step. They are already doing it. Apparently there is no dedicated macOS team anymore, and virtually every new macOS features are basically iOS ports.

 

- architecture switching would result in a lot of work. The complete system, all drivers, programming tools, Apple software have to be reprogrammed. Companies must be convinced to switch to the new architecture. Imagine: Adobe should reprogram all its software to a platform which is only used by a few users, and their software works better on Windows or OSx86-systems because it's optimized for nVidia? Hobby programmers will have much less interest to learn a new processor architecture.

You are abolutely wrong here. It is not as much work as you think. Modern software - except the low the level stuff - isn't dependent on any specific architecture anymore and that is true especially for Apple. Basically it is just only a recompile. And they don't need to "repogram" anything, because I bet they already have an ARM version of macOS running deep in some secret basement. They already did it with Intel based OS X which existed long before anyone saw the official release. Apple already did switch platform two times before with great success (68k -> PPC -> X86) and they are experts in doing that transition. Adobe software already runs better on windows since quite some time, and NO they won't have to repogram anything as long as the underlying APIs does not change. And no again, at least for basic app development you don't have to know anything about the underlying CPU architecture. It only matters if you do assembler and maybe drivers. And assembler is only used sparingly nowadays on desktop OSes, maybe for some tiny performance critical parts.

 

- architecture switching would need a complete hardware redesign. In comparison, the last Mac Pro was a failure and wasn't updated since 2011.

No comparison, the nMP failed because Apple forgot that form follows function. Same goes for the MBP (which isn't a failure at all, like it or not)

 

- Changing to a new architecture will be an immediately sign that the last expensive Macs bought will be outdated soon. Buyers won't like that.

No, beause if that happens, it will happen bottom up, not top down. I.e. with the least powerful hardware first, like MB (not MBP).

And again, this already happened in the 90s and 2006. Nothing new.

 

- On ARM-Macs Windows won't work natively. Another negative issue.

True, but is it really that important? Do not overestimate the significance of Bootcamp (at last not what Apple thinks about it)

 

- ARM-Macs will never have the same performance like Intel-Macs.

iPad pro is already coming close to the perf. of low end macs. Apple has made huge jumps regarding the performance of their CPUs.

 

It's funny enough that Apple tried on the Intel-Macs so hard to make the hardware as different to PCs (EFI, GPT, SMC, TPM) and already version 10.4.4 was even working on a low-end PC with a Pentium 4. Now are not even reprogrammed kernels are neccessary, just a bootloader.

And you are wrong again! Apple didn't try hard at all to make the hardware as different as possible, because for the most part, Intel Macs are just PCs. EFI, GPT, and TPM (the latter one AFAIK even does not exist on Macs) are no Apple exclusive technologies at all. The only "special" hardware is indeed SMC, and this is also the reason that FakeSMC is absolutely mandatory to boot. The only one! (besides the bootloader of course)

 

I'm sure, even if Apple would switch to ARM, there will be enough people who want to get it working on other platforms, like the beginners of the OSx86 scene more than 10 years ago.

Yeah, surely some want, maybe they can.

Link to comment
Share on other sites

EFI, GPT, and TPM (the latter one AFAIK even does not exist on Macs) are no Apple exclusive technologies at all. The only "special" hardware is indeed SMC, and this is also the reason that FakeSMC is absolutely mandatory to boot.

1) Their EFI was extended quite heavily, which is the only reason one needs a 'bootloader' (not actually) on spec-conforming machines.

2) TPM was present on early Intel Macs, but removed quite early.

3) Not even SMC is special, Chromebooks have the chip called 'EC' (Intel name). Though it has its own protocol.

  • Like 2
Link to comment
Share on other sites

Apple also won't go into ARM chips because it would take them a sizeable chunk of their bank to catch up with Intel on performance. Worse still is the fact that all the stuff Intel did recently is definitely still under patent so every single modern method in the chip must be reinvented. (This is why it took AMD so long to catch up with Ryzen; if you take too long to get to the patent office, you have to think of a new way to do exactly the same thing, often a less efficient way just to avoid the patent)

 

There's a lot of misunderstanding and misconceptions in this thread - so, pardon me if this gets somewhat wordy or lengthy.

 

Apple has bought a chip design company not so long ago, back in 2008.

 

https://www.wired.com/2008/04/four-reasons-ap/

 

PA Semicondutor are the folks behind the A-series of CPU chips that are powering iOS devices. They have been creating not only the A-Series CPUs, but also the M-series motion processors, and the W-series wireless chips. 

 

Apple also recently bough a chip fab/foundry (back in 2015), that used to belong to Samsung. Likewise, Foxconn just voiced interest in purchasing the Toshiba RAM memory business, mainly their foundries. (Foxconn is basically a surrogate of Apple).

 

http://appleinsider.com/articles/15/12/14/apple-buys-former-maxim-chip-fab-in-north-san-jose-neighboring-samsung-semiconductor-

 

http://www.reuters.com/article/us-foxconn-china-idUSKBN168401

 

This just to make an introductory point that Apple is very serious abou owning and controlling every aspect of the production of their devices - so, Apple not just designing but potentially manufacturing future CPUs to be used in Macs is a real possibility -- and those will most assuredly be ARM based chips, as the performance/Watt ratio is way ahead of intel x86 chips.

 

What does that mean, really?

 

Aside from the commodity value, availability and fabrication that Intel can offer, the huge advantage of their chips is the x86 compatibility which makes the Mac easily able to run Windows, Linux, and other x86 OS'.

 

Switching to an ARM platform would mean that Apple would lose that advantage.

 

Or does it?

 

Not if AMD has anything to say about that (and this is where - surprise - the AMD rumors are starting to make sense).

 

https://www.extremetech.com/computing/182790-amds-next-big-gamble-arm-and-x86-cores-working-side-by-side-on-the-same-chip

 

In 2014, AMD announced an initiative that would culminate in project SkyBridge - an ambidextrous ARM architecture, with ARM and x86 chips operating side by side.

 

The first of such proof-of-concept implementations were limited to 32-bit x86, but that was 3 years ago, and the ultimate goal was 64-bit compatibility, and compatibility with modern x86_64 architecture. In other words, I believe that if Apple were to go with an ARM CPU for future Macs, this is the path they would follow - and an alliance with AMD would make perfect sense in that regard. 

 

That's all nice and good, but wouldn't this ultimately mean that Apple will go that route (good for them), and that the folks in the Hackintosh camp will be basically screwed, as we'd have no more compatible hardware?

 

Well, maybe not.

 

Last year, at the WinHEC hardware conference in Shenzhen, a big deal was being made (sort of) of the 'Return of Windows to ARM', with Qualcom (ARM) powered Windows 10 PCs supposedly to be shipping in 2017 - with other manufacturers following suit.

 

Obviously, especially if Apple makes an impact with pushing their hardware to ARM, the rest of the PC industry following won't be far behind - particularly if these new ARM chips are pin-compatible with regular ARM CPUs and provide ambidextrous cores on the same chip. Most likely, the usual suspects (hp, Dell, Asus, etc...) will provide desktops and laptops based on the new architecture, with the usual motherboard makers not too far behind.

 

Technically, this would open the door to Hackintosh/ARM -- albeit clearly some sort of advanced fakeSMC or the likes will need to be appropriately updated.

 

Could Apple screw this up?

 

Sure. Starting with a custom SOC with custom support and peripheral processors for all the relevant ports, thus cutting us off from using commodity parts for USB, SATA, NVMe, etc..  which is kinda what they have already done on the new MacBook Pros, which are using an Apple peripheral architecture.

 

In the near future (4-5 years after such an introduction) MacOS will still support the older architecture (even just the regular x86_64 that we are all currently on), so this is not an immediate issue. Furthermore, considering how slow-moving Apple is, consistently, it would take at least another 2-3 years before such an ARM CPU implementation would actually be realistic, particularly since there are still performance and compatibility hurdles to be overcome.

 

Overall, I think we are safe for at least another 5 years, probably longer.

 

Interesting bit of Trivia -- Apple originally formed and created ARM holdings, when they used the very first ARM chips in the Newton PDA. Their ownership was then sold after Jobs killed the Newton.

 

 

And imho, this is exactly the plan. The writing is on the wall all overall the place.

 

 

No, it's not - consolidation of the OS/UI platforms into a single platform isn't the plan. Possibly consolidating the hardware architecture, but not the OS architecture.

  • Like 1
Link to comment
Share on other sites

 Share

×
×
  • Create New...