AWS Engineer Puts Windows 10 on Arm on Apple Mac M1 -- and It Thrashes Surface Pro X (zdnet.com) 107
An Amazon Web Services (AWS) virtualization engineer has shown what Windows 10 on Arm could be like if Microsoft licensed its Arm-based OS to the public rather than just to Windows 10 manufacturers. From a report: With Apple's new M1 Arm-based system on chip, Mac users who need to use Windows 10 can't run Microsoft's Arm-based version of Windows using Apple's Bootcamp. The key obstacle is that Microsoft doesn't license Windows 10 on Arm to any entities other than its own Surface group and Windows 10 on Arm OEMs like HP, Asus and Lenovo. Technically, there's nothing stopping owners of the M1 MacBook Air, MacBook Pro 13-inch or Mac mini from running Windows 10 on Arm, as Apple's software engineering chief Craig Federighi recently pointed out. [...]
But Microsoft's reluctance to create a license for Windows 10 on Arm for end users hasn't stopped creative engineers from putting together a working example of what things could be like if it did. AWS principal engineer Alexander Graf did just that, using the open-source QEMU virtualization software for Windows on Arm. QEMU emulates access to hardware such as the CPU and GPU. [...] "Who said Windows wouldn't run well on #AppleSilicon? It's pretty snappy here," Graf wrote in a tweet. Graf previously worked on the Kernel Virtual Machine (KVM) for Linux distribution SUSE for over a decade. Now he's a KVM developer at AWS, which this week announced new Mac instances for AWS Elastic Compute Cloud (EC2) based on Nitro System, an AWS hypervisor for EC2 instances. [...] A developer using the handle @imbushuo on Twitter has posted Geekbench versions 4 and 5 scores that compare Windows 10 on Arm on an M1 computer with the Microsoft-made Surface Pro X. Windows on an M1 got a single-core score of 1,288 and multi-core score of 5,685 whereas the Surface Pro X's scores were roughly 800 and 3,000 in those respective benchmarks.
But Microsoft's reluctance to create a license for Windows 10 on Arm for end users hasn't stopped creative engineers from putting together a working example of what things could be like if it did. AWS principal engineer Alexander Graf did just that, using the open-source QEMU virtualization software for Windows on Arm. QEMU emulates access to hardware such as the CPU and GPU. [...] "Who said Windows wouldn't run well on #AppleSilicon? It's pretty snappy here," Graf wrote in a tweet. Graf previously worked on the Kernel Virtual Machine (KVM) for Linux distribution SUSE for over a decade. Now he's a KVM developer at AWS, which this week announced new Mac instances for AWS Elastic Compute Cloud (EC2) based on Nitro System, an AWS hypervisor for EC2 instances. [...] A developer using the handle @imbushuo on Twitter has posted Geekbench versions 4 and 5 scores that compare Windows 10 on Arm on an M1 computer with the Microsoft-made Surface Pro X. Windows on an M1 got a single-core score of 1,288 and multi-core score of 5,685 whereas the Surface Pro X's scores were roughly 800 and 3,000 in those respective benchmarks.
No, he didn't. (Score:5, Insightful)
He put it on QEMU on the Mac. If you say someone put an operating system on a machine, that implies the bare metal. Put QEMU in the headline, you tossers.
Re: (Score:2, Interesting)
Actually; this was Type2 (Score:5, Insightful)
Re:Actually; this was Type2 (Score:4, Informative)
He was running directly on the CPU, but not any of the other hardware (literally).
Re: (Score:3)
He was running directly on the CPU, but not any of the other hardware (literally).
And posting CPU benchmarks, not benchmarks from any of the other hardware.
While I do agree with the basis of your objection, I'm having a hard time crying foul too loudly.
Re: (Score:2)
AWS Engineer Puts Windows 10 on Arm on Apple Mac M1
That at least matches the headline, if you read "Apple Mac" as adjective modifiers for M1.
Re: (Score:1)
He was running directly on the CPU, but not any of the other hardware (literally).
The CPU, RAM and MMU were all running at basically bare-metal speed. The only interesting bit of hardware not included is the GPU. It's not possible to map the integrated video card over to a hypervisor instance but you can map other PCIe/thunderbolt devices to a hypervisor instance. Getting drivers for an eGPU nVidia/AMD for Win10-Arm would likely be far easier as well.
Re: (Score:2)
QEMU is capable of running as a type 2, but under Linux and BSD, it is a hybred due to kernel drivers that move it closer to the metal. Apparently, Apple is doing that too.
Meanwhile... (Score:1)
...Linus moans and weeps because it is "too hard" to put Linux on M1.
Re: (Score:2)
Linus wants native Linux support, not a VM. Running Linux on a VM is easy on any machine that supports a VM. In fact, you can even run Linux on a web browser.
The hard part is hardware support and working drivers.
Re: (Score:2)
Apple wants everything at least under their Hypervisor anyway. I doubt that will change - it's exactly what Microsoft wanted with secure boot. They've actually gone a long way toward stopping secondary sales of stolen hardware by preventing iCloud-locked devices from being wiped and reused. It's just that it's a very user-hostile, anti-repair thing to do.
Re: (Score:2)
Porting Linux to a new CPU is pretty straightforward if basic functionality is all you are after, the problem is with all of the proprietary IP blocks for which no specifications are made available, severely limiting the kind of performance and utility you might otherwise be able to get out of it. It shouldn't be difficult to make Linux run on the M1 as a standard ARM64 core, but without support for the parts people really care about (the GPUs, AI accelerators, etc.) there's not much point. Someone would re
Its not about being hard, its about politics (Score:2)
...Linus moans and weeps because it is "too hard" to put Linux on M1.
Its not about being hard, its about politics. He wants drivers to be open. If he were willing to accept closed drivers then by working with Apple we might get something like Apple's Boot Camp where Apple helps install the other OS and adds the necessary Apple written closed drivers. Its what Apple has done for Windows on Intel for many years.
Re: (Score:2)
Re: (Score:2)
Also Linus has to make sure it works with existing protocols and programs long term. This is very different than making it run long enough to run some benchmarks.
1. Apple engineers are more than capable of doing so.
2. The main hurdle preventing Linus from working with Apple on that is probably his willingness to sign an NDA for the closed source drivers. As for Linux itself Apple supports and works with open source every day.
Re: (Score:2)
1. That is irrelevant for Linus to support the M1 in Linux. No one said anything about capability or technical ability here.
2. Also somewhat irrelevant whether Linus has to sign an NDA or not. The driver has to be open source for him to use it in Linux. I’m sure if he really wanted to use an M1 for his personal device he would sign the NDA. That is a different goal than Linux supporting the M1 for everyone
3. Yes Apple supports open source all the time. That does not mean the M1 is currently too close
Re: (Score:2)
2. Also somewhat irrelevant whether Linus has to sign an NDA or not. The driver has to be open source for him to use it in Linux.
Untrue, closed proprietary drivers exist for Linux.
3. Yes Apple supports open source all the time. That does not mean the M1 is currently too closed for Linux to support it.
ARM Linux already runs on M1 Macs and greatly benefits from the ARM ISA implementation improvements, unified memory, etc.
What is missing are drivers for things like the GPU, Neural Engine, Image Signal Processor, etc; things needed to get performance on par with macOS. Things that could be supported using closed proprietary drivers.
Re: (Score:2)
Re:No, he didn't. (Score:4, Informative)
Re: (Score:2)
Regardless of any such distinction—which isn't to say that you're wrong, though other replies are questioning your claim—why does it matter if what we're trying to do is assess capabilities of the new machine? Compatibility and performance: those are the factors that matter. If a user with machine X can (a) run their desired software (b) faster than they can on other machines, then machine X is objectively and—by many measures subjectively—better.
If a machine can use an inferior tech
Re: (Score:2)
How do you think QEMU works?
Seems to happen fairly often... (Score:3, Informative)
Duplicate of this post from several days ago:
https://apple.slashdot.org/story/20/11/27/2343223/developer-successfully-virtualizes-windows-for-arm-on-m1-mac
Re: (Score:2)
Re: (Score:2)
Netcraft confirms it, msmash is a goldfish.
Windows (Score:2)
Microsoft could license Windows (Score:5, Insightful)
For ARM Macs. Just like Apple could license Mac OS for PCs.
Re: (Score:2)
I don't get why the problem is a Microsoft licensing issue. It's more that Apple won't let anybody else use their M1 chips - though you'd think some other company can't come up with an ARM implementation that bests the M1 and sell that chip to licensed Windows OEM's, and... problem solved. And Linux could port to those machines too. Are the Apple chip designers so much better than the competition? Hard to imagine.
Re: (Score:2)
If Microsoft really wanted ARM devices, they'd design their own. They just can't let go of x86 backward-compatibility, because it's their only major value. At least if they did it, they wouldn't kill the CPU emulation a couple years later (like original Rosetta).
Re: (Score:2)
Re: (Score:2)
I'll admit to apparently being over a year out of date on Surface news. But also, it's probably non-news? It must not be doing well without x86 software because I haven't heard a thing about it.
Re: (Score:2)
Re: (Score:3)
Are the Apple chip designers so much better than the competition?
Well they just managed to get dibs! on the first batches of TSMC 5nm.
though you'd think some other company can't come up with an ARM implementation that bests the M1
Another company couldn't literary have come (yet) with a better ARM implementation, simply because the 5nm process is new and only starts being available now. With only Apple and HiSilicon producing chips, and the latter only making silicon for smartphones, M1 is literary the first processor out of TSMC's fab targetting laptops. More will come later, eventually.
(Though, in my opinion, it's valid to argue whether M1 is truly a completely se
Re: (Score:2)
Apple's M1-based machines aren't different from other such custom ARM laptop
The ARM cores aren't drastically different but how does the GPU, ML processor, DSP, Image Processor, etc in the M1 machines' SoC compare to other ARM laptops?
Re: (Score:2)
For ARM Macs. Just like Apple could license Mac OS for PCs.
They could, but then they'd actually need to get it running and that would require Apple to open up their M1 chip. Windows on ARM faces the same challenges as Linux on ARM does working on the M1. You can't just change the arch type from armv7 to m1 and then press the button labelled "click here for profit".
Re: (Score:2)
I was talking about the x86 version of Mac OS
Re: (Score:2)
That is my point. There's nothing to license for Apple to run Mac OS on a PC. There is however plenty to license for Microsoft to get Windows on ARM to run on the M1.
Re: (Score:2)
They could sell licenses of Mac OS to install on any x86 PC.
I guess you could attach... (Score:5, Funny)
... a plow to a Mercedes Benz 500sel and call it a neat tractor, but why would anyone want to do that?
Re: (Score:2)
5nm (Score:2)
That is assuming the M1 has comparable power consumption to the processor in the Surface Pro X.
M1 is on TSMC 5nm process. Of course it's going to beat the crap out of Surface Pro X, while consuming less power.
But that's about the only advantage of this one-trick pony.
As soon as other chip maker start to sign 5nm manufacturing deals with TSMC, and start licensing, e.g.: ARM's Cortex A78C designs, the advantage of M1 will evaporate.
Re: (Score:2, Insightful)
don't cross the memes! (Score:2)
Will Apple license the M1 to OEMs eventually? (Score:2)
If they did they could make a lot of money and MS would have no excuse not to port Windows unless things got stuck in a Catch 22 scenario. However I suppose the main question would be would the sale of those CPUs make more money than Apple would lose through lost sales of Macs?
Re: (Score:2)
Re: (Score:2)
Re: (Score:1)
Its political, not engineering time (Score:2)
Apple's scarce resource is engineering time. They don't have the spare cycles to support other companies using their secret sauce.
-jcr
Funny, for many years they did so for Windows on Intel. Supplying the Boot Camp utility to install Windows and closed Apple drivers for Mac specific hardware. The biggest obstacle to doing this for Linux are senior Linux devs not accepting closed drivers. Its a political problem, not an engineering time problem.
ARM vs x86 (Score:2)
Funny, for many years they did so for Windows on Intel.
Intel/AMD64 is a much more standardized architecture.
Intel Mac hardware basically behaves exactly as would a Dell with an extra "aluminium" paint job.
You get the same type of firmware (UEFI), the same way to probe and autodetect hardware, the same buses.
Even down to the same actual hardware: You'd find similar GPUs in both.
You could whip out a Linux USB stick and even if it wasn't optimized for the hardware it should still mostly work (won't maybe support the custom touchpad controller or apple-specific USB
Proprietary drivers do not harm Linux (Score:2)
Funny, for many years they did so for Windows on Intel.
Intel/AMD64 is a much more standardized architecture.
Irrelevant. M1 Mac hardware is hardly a mystery to Apple. A Boot Camp for ARM Linux would not be challenging for Apple. As tech demos are showing the M1 Macs run the existing Linux ARM code. So the situation is closer to your "almost a Dell" argument than you suggest. You get the basic functionality but you aren't leveraging the extra hardware in the M1.
Linus is tired having to fight company that actively prevent their hardware being used by others.
Again, its just politics. The main roadblock is insistence on open drivers. Apple is not against their hardware being used, they are against releasing what
Custom vs. standard. (Score:2)
Irrelevant. M1 Mac hardware is hardly a mystery to Apple. A Boot Camp for ARM Linux would not be challenging for Apple.
The "to Apple"/ "for Apple" are the relevant parts.
This poses a major problem in an open development environment like Linux.
As tech demos are showing the M1 Macs run the existing Linux ARM code. So the situation is closer to your "almost a Dell" argument than you suggest.
Yes, we know it's doable. BUT:
Some geek on the internet wrote a specific hack to get Linux running on a specific ARM machine.
!=
x86-based PC are an absolutely standardized platform. Here are the ACPI and UEFI specs. As long as you follow them you could plug a USB stick into any random PC (be it a node inside a giant HPC cluster or an ultra-thin Lenovo laptop) and it should work.
Unreasonable to expect others adopt your religion (Score:2)
Some geek on the internet wrote a specific hack to get Linux running on a specific ARM machine.
In a short amount of time with little to no documentation. Which suggests, contrary to your claim, that M1 is pretty similar to existing devices, be they PCs or ARM Chromebook or whatever. Similar at least to what is necessary to run Linux. Where M1 is wildly different is with respect to optional stuff, like support for the Neural Engine.
Open drivers are a necessity in order to be able to develop a Linux kernel.
Not for the optional stuff, like a GPU driver, a Neural Engine driver. That merely effects performance, prevent fair comparisons for macOS and Linux. The absence of drivers
Re: (Score:2)
Funny, for many years they did so for Windows on Intel. Supplying the Boot Camp utility to install Windows and closed Apple drivers for Mac specific hardware.
They weren't very good though, the trackpad driver is complete crap and there isn't one for the fingerprint sensor or for customizing the touch bar.
Re: (Score:2)
there isn't one for the fingerprint sensor or for customizing the touch bar.
And if nothing in the Windows API uses them why would a driver be needed? Apple added drivers to get Windows working, not to expand the capabilities of Windows. The later is up to Microsoft to initiate, as it is with Linux.
Re: (Score:2)
And if nothing in the Windows API uses them why would a driver be needed?
Because the hardware doesn't work otherwise. You can't unlock your device with the fingerprint id, the trackpad experience is crap because there's no precision touch driver, it's up to the hardware manufacturers to provide drivers to make their hardware work with operating systems.
Re: (Score:2)
And if nothing in the Windows API uses them why would a driver be needed?
Because the hardware doesn't work otherwise. You can't unlock your device with the fingerprint id, the trackpad experience is crap because there's no precision touch driver, it's up to the hardware manufacturers to provide drivers to make their hardware work with operating systems.
You are a little confused over who is asking who for a favor. Apple provides an operating system. It is not their responsibility to make other operating systems work. However if those other operating system vendors are willing to work with Apple then apparently cooperation is possible.
The current problem is that the Linux community is presenting Apple with a list of favors and a list of conditions Apple must meet while delivering those favors. Conditions unacceptable to Apple. The party that wants a favo
Re: (Score:2)
You are a little confused over who is asking who for a favor. Apple provides an operating system. It is not their responsibility to make other operating systems work. However if those other operating system vendors are willing to work with Apple then apparently cooperation is possible.
Apple provides both the hardware and the software. They certainly do make other operating systems work, that's why they provide Windows drivers for things like the Apple trackpad and why they provide BootCamp.
The current problem is that the Linux community is presenting Apple with a list of favors and a list of conditions Apple must meet while delivering those favors.
Where is this list you speak of? Or do you mean asking them for the specifications for their hardware so the Linux developers can write the drivers for them because there's not stable kernel ABI to provide drivers for? If so then I totally agree, and in fact I think Apple's position makes perfect sense
Re: (Score:2)
Yeah, let's replace an open kernel and drivers with closed source crap.
A closed driver does close one line of the kernel's code, nor any other open code. People have been successfully using closed drivers on Linux for many years.
Why even bother going to Linux in the first place if you're going to just bring it down to the level of, "Just trust us."
Many of us just want a *nix environment, the politics aren't the draw. If a closed driver gets wireless, GPU, etc working, no problem. Nothing has been brought down except political ideology, the hardware is more functional and not one line of open source has been lost. The closed source merely plugged a hole open source couldn't.
Re: (Score:2)
Re: (Score:2)
Apple have a pretty dominant #1 or #2 position in the smartphone market around the world so they don't need to. That doesn't apply in the desktop market where they're very much a niche player.
ignore the man behind the curtains (Score:5, Funny)
Re:ignore the man behind the curtains (Score:4, Informative)
All valid points, but only as commentary. Microsoft doesn't sell Windows for ARM because ARM-based platforms are non-standardized. They can sell Windows for PCs because PCs have enough in common that you can reasonably make an OS to run on them. In ARM-land, OSes aren't for "ARM", but for specific ARM platforms, or subsets thereof. You might have a Linux distribution that runs on a handful of different implementations, and then you'll have some third parties who produce their own tarball because the official installer doesn't support it. For example, I can run Debian on my ancient Pogoplugs, but only by going to the forums at doozan.com and downloading the stuff to make it happen from there (kernel, u-boot, and tarball.)
Re: (Score:2)
On the other hand, many people have gotten Linux to run on many ARM platforms. It is available to anyone who wants to write the necessary drivers or who wants to design their hardware to work with existing drivers.
Here's an ARM platform. Now, go get the Windows for ARM developer kit and put the distro together. OH WAIT, you can't because MS doesn't offer it.
It's a lot of work on Linux, but the compilers (including cross compilers) kernel, and userspace are all out there. Given that, plus documented hardware
Re: ignore the man behind the curtains (Score:2)
Sure, that works for Linux. But Microsoft isn't going to adopt that model.
Re: (Score:2)
And THAT is the real reason. They don't wanna.
Their choice and their right to make it, but let's not pretend it's anything else.
Re: (Score:2)
Microsoft is in the business of selling software. They have to support it. They can't reasonably support the boot process if they're not responsible for it, but they can't reasonably support windows without supporting the boot process.
In short, it's just not a viable model for commercial software — not just for Microsoft.
I would rather argue that operating systems should not be commercial to begin with.
Re: (Score:2)
Wind River manages OK with VxWorks. MS used to do it with WinCE
Re: (Score:2)
Wind River manages OK with VxWorks. MS used to do it with WinCE
Not with end users, in either case. Only with corporations. I used to do tech support where everyone who called in was a sysadmin. That was great. Supporting the general public, on the other hand, sucks ass.
Re: (Score:2)
Sure, they only supported Corporations, but in MS's case, not even that anymore.
There are good public policy reasons why only supporting corporations is a problem in itself, but that's a different discussion.
Re: (Score:2)
Microsoft doesn't sell Windows for ARM, because they don't want people to realize that they could check their email and do zoom calls just as easily on a $40 arm box as on a $1000 box that sounds like a jet engine. This might also cause consumers to question why the OS for the device costs 2-3x as much as the hardware.
They don't sell it because they don't want to support it.
Re: (Score:1)
But they _do_ sell it, that is why this could be tested. It is just that nobody is buying.
Re: (Score:2)
This might also cause consumers to question why the OS for the device costs 2-3x as much as the hardware.
As the Raspberry Pi has proven, it doesn't take much for a couple of people to throw together some hardware and have it work.
As HURD has proven it takes a lot to do the same thing with an OS.
While some people are waiting for the year of Linux on Desktop, others are waiting patently for the year of HURD to not be Beta.
Year of Linux on Desktop did arrive ... (Score:2)
While some people are waiting for the year of Linux on Desktop ...
The Year of Linux on Desktop did arrive. It's just not quite what people expected. Turns out it was the Microsoft Windows Desktop. ;-)
Here's Linux on the Microsoft Windows App Store: https://www.microsoft.com/en-u... [microsoft.com]
Re: (Score:2)
Oh My Gosh!
What happened to Plan 9???
Re: (Score:2)
It's beloved by computer scientists, who consider it a perfect work of pure theoretical operating system design. It is not loved by people who want to do anything useful. There haven't been any significant updates for Plan 9 or Inferno in five years.
Re: (Score:1)
How much QEMU and how much M1? (Score:2)
Does anybody have any performance stats for QEMU vs Win10 running on the same hardware?
I'm wondering how much of the performance gain is due to QEMU rather than the ARM processor.
Re: (Score:2)
QEMU typically causes a smallish performance LOSS vs running on the bare metal.
Windows NT for the Powerpc (Score:5, Informative)
Microsoft has had a bad experience moving of the x86 Intel/AMD CPUs.
They had tried with NT for the PowerPC, which failed, as PowerPC users were either using MacOS, or some sort of Unix (perhaps a Linux) for server stuff.
Windows CE, which became Windows (version) Mobile. Was designed to run on many other CPUs as well. However they had a rocky attempt to gain any traction.
The big thing about Windows popularity is its backwards compatibility, and its large collection of software that relies on this.
many of your old DOS apps built in the 1980's can still run in windows, without DOSBox or some sort of emulation. You can still run most Apps built for windows 3.1 in windows 10 (assuming that the Window 3.1 app was 32bit, or you are using a possible 32 bit edition of Windows 10, to run old 16bit programs)
Spanning Generation around keeping compatibility of the x86 platform as bonded Intel and Microsoft into a game of pushing old architecture and designs, not because these companies cannot do something better, but because their market will not allow them to change. Because no one wants to buy new software again.
Apple Users kinda need to do that every decade or so. Motorola to PowerPC (MacOS to OS X) to Intel to M1. But Apple doesn't have a big presence in the corporate market. Where companies have invested millions of dollars in legacy software, and want to use them as long as they possibly can.
The only sign of change on this trap to Microsoft, is the fact of the rise of Mobile devices by Apple and Alphabet, the rise of web services by Amazon, where Microsoft pushing Azure as well. Where a lot of software being built today is no longer so tied to the machine language of the Computer that you are working on, and is distributed across a set of higher end computers. Making the transition off Intel possible, but I don't see it happening soon though.
Perhaps when Mobile Devices begin to Outperform Desktop PC's because of new architecture advancing faster than the old one.
Window on a 2nd arch is necessary, internally (Score:3)
Microsoft has had a bad experience moving of the x86 Intel/AMD CPUs.
Not really. From day one WinNT was designed to be cross platform so they need to keep it working on some contemporary non-x86 CPU. Its software, getting it running on PowerPC is nothing like building PowerPC based hardware. Plus this helps keep the code clean and helps find bugs. Bugs that are hard to manifest on one architecture sometimes manifest quite easily on another. When porting x86 software to PowerPC I often found lots of bugs the x86 devs didn't even know were there. Eventually when stuck on a har
Re: (Score:2)
I ran Windows NT 4 on an Alpha UDB box back in the day and while it was 32 bit on a 64 bit platform, it performed pretty well. The x86 emulation was there and it worked but as MS stopped supporting the Alpha over the years I found relying on x86 emulation became a significant impediment.
Yesterday I got Windows 10 ARM64 running on my M1 Mac mini following the instructions on the web and it is lightning fast compared with my work i7 Windows 10 laptop. More shocking was that the x86-64 emulation actually wasn'
will apple let you direct boot non macos on ARM ma (Score:2)
will apple let you direct boot non macos on ARM macs?
Or will they lock the boot loader to mac os only?
Re: (Score:2)
It's locked to booting signed code by default, much like UEFI is, but can be disabled to boot anything you want, theoretically. But there is still the issue of hardware driver support, firmware blobs, etc. None of which Apple will go out of their way to make available.
Licensing.. make it compulsory (Score:1)
It will help mitigate the patent/copyright problem
On ARM? (Score:5, Funny)
Why would I want Windows 10 on ARM? I don't even want Windows 10 on x86.
Re: (Score:2)
What about Windows 10 on a potato?
Re:On ARM? (Score:4, Informative)
Re: (Score:2)
Re: (Score:2)
Potatoes have eyes. Eyes to gather telemetry on you.
Re: (Score:2)
Re: (Score:2)
Sokath! His eyes uncovered!
Re: (Score:2)
Yes, you're clearly the target market being thought of here.
Pot, meet Kettle (Score:2)
Yeah, just like Apple doesn't license OSX to any entities other than...itself.
Licensing to individuals not required ... (Score:2)
Microsoft doesn't license Windows 10 on Arm to any entities other than its own Surface group and Windows 10 on Arm OEMs like HP, Asus and Lenovo ... But Microsoft's reluctance to create a license for Windows 10 on Arm for end users
To get ARM Windows on M1 we don't need it licensed to end users. We just need Apple to be added to the list of hardware vendors that Microsoft will work with. Its far from ideal but a Boot Camp like configuration could be a factory build-to-order option from Apple.
"Thrashes" (Score:1)
Can you imagine a beowulf cluster of these? (Score:1)
M$ has egg on it's face (Score:2)
Just...wow. Not sure what to say here other than this has got to be embarassing for Micro$oft.
Separation of OS and Apps (Score:1)
You know...If the original sentence in Microsofts' monopoly trial stood. There would have been a Microsoft OS division and a Microsoft Apps division.
Microsoft OS division would have been all over the ARM platform for an OS. Over the years, there would have probably been others as well. Apple HW, alternate x86 platforms, who knows.
And MS Office and Outlook would be on everything as well. Years ago.
Still think it would have been a Plus + Plus for users and all.
LSW (Score:2)