IMO the Hackintosh is 100% dead for some time following the ARM transition. Even if someone gets a future ARM-only version of macOS running on non-ARM hardware, it will suffer all the problems of being non-native. Not worth it IMO.
Got a tip for us?
Let us know
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
Is ARM the death knell for the hackintosh?
- Thread starter glenthompson
- Start date
- Sort by reaction score
The Apple ecosystem managed by T2 (therefore will become inaccessible to non-Apple hardware with an exception of iTunes and iCloud storage) opens up much more opportunities. Apple has a banking mechanism, is totally independent from network carriers (eSIM), has unique and robust authentication mechanisms... I can imagine a hardware-as-a-service offering quite soon, even with unlocking additional cores on the CPU or allowing access to cloud processing power for additional fee.
Yeah, about that time we either had a Power Computing or Umax clone at work.
There will be a few years of new MacOS Intel releases. The last MacOS Intel releases will get security updates for years after that. And they will continue working for years after that. Hackintosh is dead when new PC hardware is not supported by the last MacOS version anymore. To be precise, when you cannot buy any new PC that would be supported by that last MacOS version. That will be a very long time, probably 10-15 years. And of course if people start building ARM PCs, which is not unlikely within 10-15 years, then Hackintosh could suddenly come back.
[automerge]1593177479[/automerge]
[automerge]1593177479[/automerge]
I can remember someone making Atari's OS run on PowerPC Macs. If I remember correctly you could get an Atari computer with a 68060 processor, but it turned out a 150 MHz PowerPC Mac was the fastest Atari computer ever
I agree with some of the above comments - the hackintosh scene has been largely dead for some time now. It really hasn't been worth doing (beyond tinkerers and "because I can"ers) for a good few years.
Especially since if you're building a hackintosh to save costs, you're probably buying AMD gear considering the bang-for-buck vs intel chips at the moment, and hackintosh on AMD is always a pain. Most people have worked out that by the time you buy the full system, using compatible parts, add in a decent display, you may as well just buy an iMac, upgrade the RAM and add external storage. The couple of hundred you may save just isn't worth the crap you have to go through at every OS update.
Especially since if you're building a hackintosh to save costs, you're probably buying AMD gear considering the bang-for-buck vs intel chips at the moment, and hackintosh on AMD is always a pain. Most people have worked out that by the time you buy the full system, using compatible parts, add in a decent display, you may as well just buy an iMac, upgrade the RAM and add external storage. The couple of hundred you may save just isn't worth the crap you have to go through at every OS update.
I thought opencore was creating a new interest in hackitoshes?
I'm pretty sure MacOS will not work on anything before ARMv8.3-A.
Outside Apple that's Marvell ThunderX3 and Fujitsu A64FX, maybe the next Neoverse Core.
Outside Apple that's Marvell ThunderX3 and Fujitsu A64FX, maybe the next Neoverse Core.
There's no "outside Apple" in this case. Apple employs custom chips based on Arm architecture and by tying them to the T2 lock out anything non-Apple from running the OS.
There's a few people with the right answer in this thread.
ARM is IP, Apple buy that IP from ARM which is the processor core which they can then use inside their own Silicon.
The Raspberry Pi Broadcom SOC uses ARM IP, it has a whole load of other IP embedded in the silicon around it, some off the shelf, some custom all directly connected to the processor.
Apples SOC will include their own combination of IP along with the processor (which they can further customise if they want).
ARM doesn't mean anything in particular other than it implements (a version of) the ARM instruction set.
As an example, you can buy lots of different ARM Cortex M4 microcontrollers, different members of the same family will have different peripherals, different vendors chips will have different peripherals and although they run the same basic instruction set, you can't take the binary from one vendors Cortex micro and run it on another, it just won't work (save for the example where memory is mapped to the same place and no peripheral access is done).
Making the code run on a different Cortex processor means porting the code so that it works with the new processors peripherals and recompiling.
Both ARM, both incompatible with each other.
And the situation scales up, the same problem with the Apple Silicon, MacOS will be tied to peripherals inside the Apple silicon. Now while there may be instances where drivers could theoretically be written to replace drivers for some peripherals, it's hard to imagine that there will be not be core parts of the OS that are inherently tied to peripherals that the OS expects to always be there, regardless of generation, and that is linked into parts like the kernel.
Just because two people speak the same language doesn't mean they can understand each other.
ARM is IP, Apple buy that IP from ARM which is the processor core which they can then use inside their own Silicon.
The Raspberry Pi Broadcom SOC uses ARM IP, it has a whole load of other IP embedded in the silicon around it, some off the shelf, some custom all directly connected to the processor.
Apples SOC will include their own combination of IP along with the processor (which they can further customise if they want).
ARM doesn't mean anything in particular other than it implements (a version of) the ARM instruction set.
As an example, you can buy lots of different ARM Cortex M4 microcontrollers, different members of the same family will have different peripherals, different vendors chips will have different peripherals and although they run the same basic instruction set, you can't take the binary from one vendors Cortex micro and run it on another, it just won't work (save for the example where memory is mapped to the same place and no peripheral access is done).
Making the code run on a different Cortex processor means porting the code so that it works with the new processors peripherals and recompiling.
Both ARM, both incompatible with each other.
And the situation scales up, the same problem with the Apple Silicon, MacOS will be tied to peripherals inside the Apple silicon. Now while there may be instances where drivers could theoretically be written to replace drivers for some peripherals, it's hard to imagine that there will be not be core parts of the OS that are inherently tied to peripherals that the OS expects to always be there, regardless of generation, and that is linked into parts like the kernel.
Just because two people speak the same language doesn't mean they can understand each other.
Since you can’t put iOS on any hardware except Apple, I think it’s safe to assume that macOS for ARM will not be installable on anything but Apple hardware. The limitations to iOS are coming to ARM MacOS, its only a matter of time.
That’s not remotely possible. Apple cpu will have its own unique instruction set with custom co-processors like the gpu, neural engine, audio engine, and Secure Enclave, etc., and without those present you won’t even be able to boot macOS for arm. Once apple stops making intel systems you’ll get a few more years of intel hackintosh as apple continues to support those models. Probably in 6-7 years there will be no intel builds of macOS.
I think it's going to be over soon.
As soon as Apple stop offering support for old intels, hack users won't be able to upgrade to a newer os.
In addition, a bigger blow is going to be obsolete drivers support and limit of drivers offering only upto last intel parts will pretty much make using old Mac OS with future parts impossible.
As soon as Apple stop offering support for old intels, hack users won't be able to upgrade to a newer os.
In addition, a bigger blow is going to be obsolete drivers support and limit of drivers offering only upto last intel parts will pretty much make using old Mac OS with future parts impossible.
The question is: How long many versions of MacOS will they support on Intel? According to Tim: "several". It will be a few years before they're able to phase out the Intel Macs I think. As long as they're producing Intel versions of MacOS 11.x a hackintosh is possible.
Even if possible at all it will be extremely hard to get macOS ARM running on non apple hardware for sure but I do not think anything of the above are valid reasons for that.
There will not be an T2 requirement anytime soon (current iMacs don't have them), if at all as ARM Macs won't have a T2 chip but rather implement similar functionality into the SOC itself (a T2 chip is basically an A10). However even if there was it can hardly be made a uncircumventable requirement - think about SMC. While it is used (among other things) to implement a secure boot chain this security cannot be enforced the other way around: if any hierarchically superior process is compromised, so are those being loaded afterwards. Examples to this would be Thunderspy, where the Macs EFI is compromised, and thus macOS is too, regardless of any non firmware updates, or last years iPhone boot rom exploit, which leave any iOS device up to the iPhone 10 (A11) open for jailbreaks regardless of the iOS version.
This means if you emulate the T2 in software (like VirtualSMC.efi) the system cannot do anything about it.
What remains to be seen is whether there will be any apple silicon specific instruction sets (like e.g. SSE 4.1, AVX, etc. on Intel) which cannot be emulated without having too much overhead. The same is true for the GPU side of things. However even those Apple GPUs builds upon known instruction sets from Imagination Inc., while there could very well be apple specific instructions as well, just like with the CPU.
However to make any hackintoshing feasible you would ideale need a SOC which is at least close to being on par with apple offerings, and ideally superior to deal with the unknowns mentioned above. And currently there isn't any in the consumer space, and possibly neither anywhere else (do those super computer ARM server SOC have integrated GPUs?).
All that said at least apples XNU kernel can already boot on ARM since several years. However it is an old version and there is no GUI. In this regards it might be more feasible to run macOS software on other OS' with Darling or similar solutions.
Last edited:
Oh man, I remember Macworld Magazine in those days with several full page ads for Mac clones.
If we are going by the traditional definition that a Hackintosh is macOS running on non-Apple hardware then it is finished.
My guess is that Apple will find a secure hardware solution to prevent macOS Big Sur to be installed on non-Apple Arm devices.
Using the T2 Security Chip as a reference is no longer valid. That chip exists because it was paired with Intel CPUs. With Apple in full control of the Apple Silicon SoCs, they can put something in their custom silicon that will make it effectively impossible to install macOS Big Sur (and later) on unqualified Arm devices. T2 was an interim stopgap measure.
It is worth pointing out that Apple already does this with iOS. You can't install iOS on a non-Apple device. Heck, you can't even install an unsigned version of iOS on a compatible Apple device once Apple stops signing the IPSW blob.
My guess is that Apple will find a secure hardware solution to prevent macOS Big Sur to be installed on non-Apple Arm devices.
Using the T2 Security Chip as a reference is no longer valid. That chip exists because it was paired with Intel CPUs. With Apple in full control of the Apple Silicon SoCs, they can put something in their custom silicon that will make it effectively impossible to install macOS Big Sur (and later) on unqualified Arm devices. T2 was an interim stopgap measure.
It is worth pointing out that Apple already does this with iOS. You can't install iOS on a non-Apple device. Heck, you can't even install an unsigned version of iOS on a compatible Apple device once Apple stops signing the IPSW blob.
Thing is, hackintoshes are about two things: More performance, less price.
If Apple makes their arguably best ARM chip designs in the world, and all you can do is to hackintosh macOS on weaker hardware: What's the point?
If Apple makes their arguably best ARM chip designs in the world, and all you can do is to hackintosh macOS on weaker hardware: What's the point?
i think that this "arm being the death knell for the hackintosh" is missing the point, because arm desktops from apple will rise non-x86 desktops and workstations again (since that time we missed powerpc and sparc there) - as a mostly gnu/linux user (and osx eventually) and risc cpu fanboy, would be great seeing non-x86 workstations and desktop computers on sale in the computer shops abundant in each corner, specially knowing how amazingly better in performance and consumption are risc cpu when compared with those painfully obsolete x86-based - it might be more the start of the end of x86 architecture than the end of hackintoshes
We can always hope that x86 hegemony will be broken, but that would require someone to make a fast enough consumer non-x86 CPU. So far Apple is the only company that does it, and they will not license their architecture to anyone.
Given the 10th Gen Intel based Macs that exist now (particularly, the 27" iMac), you'll likely be able to keep running Intel versions of macOS on Hackintoshes that are at newest, 10th Gen, just fine. Where it will become ugly and hard to do will be when there's Intel hardware that Apple doesn't have in their Macs. It won't be impossible to, say, make an 11th Gen Intel based Hackintosh when Apple doesn't have 11th Gen chips in any Mac (and won't). But it will start to require more third party drivers to be concocted rather than being able to use what Apple has baked in natively. But, yes, the hard limit will be when Apple puts out a macOS release that doesn't support the Intel x86-64 processor architecture.
As for a timeline for that, let's suppose the following:
- Intel Hackintosh Hardware is available that is similar to current Apple Macs - Another 6 months to 1 year
- More modifications are required for hardware newer than Apple's final set of Intel-based Mac hardware - Between 1 year from now until Y date ("Y date" being defined as the release date of the first macOS release to drop support for x86-64 versions of macOS); with increasing difficulty
- Intel Hackintoshes are effectively dead
And, no, there will not be ARM64 based Hackintoshes. Apple may use ARM64 instruction sets on their processors, but it's not like you're going to be able to run ARM releases of Big Sur and onwards on a Qualcomm Snapdragon or a Samsung SoC. Apple's designs are custom enough that it'd be like trying to run a PowerPC Mac OS X release on an Xbox 360 or some other 3rd-party PowerPC box.
Suffice it to say that, now is a fantastic time to buy 10th Gen Intel hardware and build yourself a Hackintosh as after this it's going to get increasingly more difficult before Apple stops producing x86-64 releases of macOS altogether.
Nope. It's not just that Apple's Silicon runs ARM64 instruction sets. That's about all it has in common with other ARM64 SoCs. It's to the point where you'd need substantially more than drivers to get Apple Silicon Big Sur and beyond to run on a system that wasn't using the Apple SoC.
You're talking about the ability to run other operating systems on Apple Silicon hardware. And toward that end, I agree. But the discussion here is more on running Apple Silicon based releases of macOS on non-Apple ARM64-based hardware.
Not necessarily. It works effortlessly because the CPUs that people use for their Hackintoshes are of the same families that Apple is using on actual Intel Macs. Once Apple stops producing Intel Macs, then the work to get those processors going will be a bit more difficult (probably not unlike how it is to get an AMD Hackintosh going) and they'll get more and more difficult the longer time passes after those final Intel Macs are produced. I'd say that anyone wanting to build a 10th Gen Hackintosh should do so now.
It won't be possible to get Apple Silicon macOS releases to run on non-Apple Silicon ARM64 hardware and it will be impossible to get it to run on a non-ARM64 system without an emulator. It'll be much easier to get an x86-64 release of macOS to run on an ARM64 system via emulation, but that really won't be worth it.
The T2 has nothing to do with whether you can or can't Hackintosh. It's possible that once Apple limits x86-64 releases of macOS to Macs with a T2 chip (which I can totally see happening along the road to eventual Intel Mac obsolescence from the macOS system requirements side of things), it may be impossible to get Hackintoshes going, but even then, I find that highly unlikely.
No, you're wrong. Hackintoshing is not currently dead. This transition will see it dying, but it is not dead yet. Those wanting an affordable and more powerful alternative to the 2020 27" iMac can build an Intel Hackintosh that's more powerful. But it's not dead and it hasn't been dead just because you don't see a point to it.
Also, the notion that Hackintosh users are going for AMD to save costs is ridiculous; AMD Hackintoshing is a completely different realm of tinkering than is required for Intel Hackintoshes and those wanting to Hackintosh aren't necessarily game for that. They want a Mac more powerful than an Intel iMac and they want to pay less for it than the nonsense that Apple charges. That doesn't mean that they're going to opt for AMD to save money and then quit altogether because AMD Hackintoshing is a pain in the a$$.