Contribute
Register

Hackintosh in the future

Also fact is, the community could port macOS 15 when it's released (well if it's Apple Silicon only) to x86, like Ubuntu to the M1.
 
There is serious competition for Apple, Intel and AMD processors. Qualcomm is testing Snapdragon X Elite processors based on the Oryon core (ARM).

Zrzut ekranu 2024-04-2 o 15.38.59.png


Zrzut ekranu 2024-04-2 o 15.39.18.png



Do you think it will be possible to make a Hackintosh out of them?
 
Last edited:
Do you think it will be possible to make a Hackintosh out of them?
It is all going to come down to if someone creates drivers for MacOS for the parts. The reason that Hackintosh is possible is there are drivers for the hardware.
 
To be honest, I don't think it's over YET. Instead of x86 it will be ARM. Microsoft already released their Windows 11 ARM Developer Preview : https://www.microsoft.com/en-us/software-download/windowsinsiderpreviewARM64 so it means that we could still TRY getting the ARM MacOS image on usb, and use it in a ARM based PC. But x86 Hackintosh will soon die, which should be obvious.
Windows is made to work with generic ARM while MacOS is made to work on Apple devices. While windows will work in most ARM devices, MacOS will not unless someone make a hard hack to make it possible, and it will be way worse than opencore. Also worth noting Apple Silicon have some other stuff that could make this even worse.

Also fact is, the community could port macOS 15 when it's released (well if it's Apple Silicon only) to x86, like Ubuntu to the M1.
Ubuntu have source-code available, so it's portable. Without the source code is close to impossible to just "port".
 
It is all going to come down to if someone creates drivers for MacOS for the parts. The reason that Hackintosh is possible is there are drivers for the hardware.
Another BIG factor is that x86 macOS runs on generic Intel hardware, not on custom chips with special accelerators not found on regular PCs. We "only" need VirtualSMC to make for Apple-specific components.
ARM macOS runs on AppleSilicon, not on generic ARM. We'd need a "VirtualNeuralEngine", "VirtualMediaEncoder", "VirtualSecureEnclave" and any further "VirtualWhatever" to make up for all the proprietary accelerators which macOS expects. Good luck with that!
 
Another BIG factor is that x86 macOS runs on generic Intel hardware, not on custom chips with special accelerators not found on regular PCs. We "only" need VirtualSMC to make for Apple-specific components.
ARM macOS runs on AppleSilicon, not on generic ARM. We'd need a "VirtualNeuralEngine", "VirtualMediaEncoder", "VirtualSecureEnclave" and any further "VirtualWhatever" to make up for all the proprietary accelerators which macOS expects. Good luck with that!

Some of this was in the T and T2 which was the beginning of AppleSi in Mac.

Then in M series, the units of T2 were integrated into a set of chips, which is where the SSD controller moved out of the "drive" the way we think about NVMe and SATA and the drive flash became an adjunct to the chipset with secure enclave intermediating to protect storage as part of a total system chipset. The smart thing to do was solder it all to improve product reliability.

Basically, what M did was rearrange the edge of modularity to be at the Thunderbolt port, rather than PCI / DMI. Security and reliability are vastly improved and in the Mini form factor, an entire system is at the price-point of a top end GPU.

Meanwhile, Intel in desktop is a gamer hodgepodge, while on Xeon, Intel was pulling an IBM 360 / 370 trick by sub-licensing access accelerators to milk buildouts of server farms. This would have been in the roadmaps Apple saw as a top tier design win for Intel back in 2006+.

The flourish of 2019 Mac Pro was a huge middle finger to Intel and Microsoft where Apple could claim they were making the most beautiful PC while end-running Intel towards a vastly more performant architecture which harmonized with Apple mobile. And if the shift went wrong they could stay in the fold while working it out.

Apple built a hack that far exceeds any PC hackers dream: they hacked the entire architecture and eliminated the need to futz with all the Intel BS to get far and away class-leading performance at an insanely great price point.

HACK AS IT HAS BEEN KNOWN SINCE 2006 IS DEAD AND WILL NOT RETURN BECAUSE THE PC WAY OF THINKING ABOUT MODULARITY IS NOW POINTLESS. It will become pointless for gaming too.

Thunderbolt, which dates back to 2000ish serial bus design and failed IO initiatives at Intel via Compaq and MSFT, was seized upon by Apple as sort of a poison pill. It let the full DMI plus Southbridge be moved out of the box, setting the stage for the Studio and the death of Mac Pro, which was half server farm and half workstation.

It's like the elimination of the CD drive which absolutely nobody misses.

Genius!

The next step in gaming is obviously AR / VR but it will take 10 more years.
 
Some of this was in the T and T2 which was the beginning of AppleSi in Mac.

Then in M series, the units of T2 were integrated into a set of chips, which is where the SSD controller moved out of the "drive" the way we think about NVMe and SATA and the drive flash became an adjunct to the chipset with secure enclave intermediating to protect storage as part of a total system chipset. The smart thing to do was solder it all to improve product reliability.

Basically, what M did was rearrange the edge of modularity to be at the Thunderbolt port, rather than PCI / DMI. Security and reliability are vastly improved and in the Mini form factor, an entire system is at the price-point of a top end GPU.

Meanwhile, Intel in desktop is a gamer hodgepodge, while on Xeon, Intel was pulling an IBM 360 / 370 trick by sub-licensing access accelerators to milk buildouts of server farms. This would have been in the roadmaps Apple saw as a top tier design win for Intel back in 2006+.

The flourish of 2019 Mac Pro was a huge middle finger to Intel and Microsoft where Apple could claim they were making the most beautiful PC while end-running Intel towards a vastly more performant architecture which harmonized with Apple mobile. And if the shift went wrong they could stay in the fold while working it out.

Apple built a hack that far exceeds any PC hackers dream: they hacked the entire architecture and eliminated the need to futz with all the Intel BS to get far and away class-leading performance at an insanely great price point.

HACK AS IT HAS BEEN KNOWN SINCE 2006 IS DEAD AND WILL NOT RETURN BECAUSE THE PC WAY OF THINKING ABOUT MODULARITY IS NOW POINTLESS. It will become pointless for gaming too.

Thunderbolt, which dates back to 2000ish serial bus design and failed IO initiatives at Intel via Compaq and MSFT, was seized upon by Apple as sort of a poison pill. It let the full DMI plus Southbridge be moved out of the box, setting the stage for the Studio and the death of Mac Pro, which was half server farm and half workstation.

It's like the elimination of the CD drive which absolutely nobody misses.

Genius!

The next step in gaming is obviously AR / VR but it will take 10 more years.
This was only a good solution for Apple. For users, it limits the possibility of expansion and the freedom to use any components.
 
This was only a good solution for Apple. For users, it limits the possibility of expansion and the freedom to use any components.
Please re-phrase that: "The freedom to use any [PC] components." As freely as a PC let you use them.

But how free is that? It's quite a mixed bag.

Mac was never PC and no one with clear thinking is confused about this. It's an obvious a point of history. And it's Apple's main selling point!

The edge of compatibility of Mac SW with PC hardware has always been very limited.

Mostly PC "freedom" means add-on storage and USB.

Consider where the Intel Mac began:

2006 15in Macbook Pro

"1.67 dual core, 512M RAM / 80G HD"

This was in the era of what we think of as modern PCs: DDR, PCIe, SATA, USB.

There are no benchmarks listed because Geekbench had not yet occurred. You had to boot Windows to bench. But if you could Geekbench 6 it, the score would be something like 175 SC / 250 MC, where an M1 is 2500 / 10000. Its USB limits storage to 30 MB/s compared to 1000 MB/s for USB 3 gen 2 common today.

—And the benchmark don't consider the accelerators that add amazing performance to M series.

In time of M-seriees, built-in RAM, storage and graphics are hugely plentiful. And USB is has kept up.

Since Mac was built from Intel, there have been 13 Core series gens, with 4 gens of RAM and 4 gens of PCI.

(Interestingly, two of the PCIe gens occurred in the last 3 years.)

With backwards compatibility there's always a cut-off and where a cutoff hasn't yet occurred there are many tradeoffs. Everyone knows this, but some people hang onto to an idea that PC is infinitely backwards compatible because Windows looks basically the same as it did in 2000.

So an argument about "expansion freedom" pre-M series is not strong.

I'll repeat myself:

//Basically, what M series did was rearrange the edge of modularity to be at the Thunderbolt port//

The module you upgrade is now the integrated unit of a cube. If you want a generational advance, you replace the cube which has a price-point overlapping a high-end mobile phone or GPU.

(Interestingly—part 2: the phone's total system performance is close to the cube, it's optimized to run on a battery and has one simple port. The GPU has become a phat accelerator for 3D and neural nets. Apple is keeping up with these trends.)

The cube's serial IO port lets you expand plenty of stuff, including storage that's connected by inexpensive, fast 1GB/s USB, which because of chip evolution is running at 20x the speed and 20x the capacity that hard drives were when Intel Macs were introduced. Everything in silicon is 20x the OG MBP.

If you have lots of spinners, connect them via USB using cheap outboard case. A 5Gbps USB PORT is fast enough to run 2 of the latest 20T spinners at full speed at the same time, and 1 Thunderbolt port has enough bandwidth to run 8 5Gbps USB ports in tandem.

As cube evolves, if you don't need all the latest hottest chips, choose an older device readily available in the aftermarket. Users of old PCs are happy with massively down-rev kit. Users of Macs can do the same. The Thunderbolt 3 mini has been around for 7 years.

And as to old software. Old Windows is as hamstrung by incompatibility as old macOS.

The more you think about it, the less meaning there is to the idea of "expansion freedom" of Intel. The idea is drained of any substance.

I don't understand how anyone doesn't see the M series as an excellent "solution" for everyone who wants Mac, across the board. And it's the natural evolution of PC, it's just not a PC.

If you want freedom to have some true meaning, the dream of "hackintosh" lives on, in the form of Linux, and there is Linux on everything. Inc ARM.
 
This was only a good solution for Apple. For users, it limits the possibility of expansion and the freedom to use any components.

Freedom and expansion are an illusion.

In the next 20 years we will be back where we were in the 1970's using terminal workstation.

Buy a Mac with Mx chip resistance is futile!
 
Last edited:
Back
Top