MacRumors

macrumors bot
Original poster
Apr 12, 2001
54,201
16,023


Late last week, Parallels launched a Parallels Desktop 16 for Mac Technical Preview Program, which allows M1 Mac owners to use the Parallels software. Paired with a version of Arm-based Windows that's available through the Windows Insider program, it's possible to get Windows up and running on an Arm-based Mac.


In our latest YouTube video, we installed Parallels on an M1 Mac and tested out Windows, but it didn't exactly go smoothly. When we first set up Parallels on an M1 Mac, it was functional, but about an hour into the experience we kept running into errors attempting to get Windows working. It continually froze and the performance was abysmal.

With some tinkering and some troubleshooting help from the MacRumors forums, we were able to get it working well enough to test out. Parallels on the M1 Mac is available as a Technical Preview and Windows can only be installed through Windows Insider, so neither of these are release versions of software. Given that we're working with beta software that's still in development, it's not surprising that there are issues that need to be worked out, and anyone who wants to test out Windows on an M1 Mac through Parallels should expect to deal with bugs.

When running macOS Big Sur, the M1 MacBook Pro with 8GB RAM has a single-core Geekbench score of 1719 and a multi-core score of 7384. When running Windows 10 through Parallels using the default two-core setting, the M1 Mac achieved a single-core score of 1491 and a multi-core score of 2753.

That's pretty similar to Microsoft's own Surface Pro X in terms of multi-core performance, and better when it comes to single-core performance, so when it works, speeds aren't too bad. Parallels defaults to allowing just two cores when running Windows, but you can swap over to four core operation.

With four cores activated, multi-core Geekbench scores were much better, hitting 5013. The single-core score didn't improve quite as much at 1518. You can enable eight cores too, but going from four to eight cores did little for performance (1524 single core and 5958 multi-core).

In the four-core mode, there are some performance issues, so it's best to stick to the two-core default where possible. When an app is able to open and run, performance is solid, but there are a lot of bugs to deal with and a lot of software that does not work.

Many pre-installed apps don't function well and often refuse to open, but third-party apps like Geekbench seem to operate as expected. Paint 3D, the Xbox app, and the Calendar app refused to work, for example, but the Microsoft Edge browser and Office suite of apps were functional and performed decently. Third-party apps like Spotify, Notion, and others worked well, and the Windows VM is even capable of playing older games like Civilization IV and Skyrim.

Right now, getting Arm-based Windows on an M1 Mac requires Parallels and the Windows Insider preview version of the software, and it continues to be unclear if Microsoft will make a release version of Arm Windows available for people to license.

Those who want to try out Windows (or another OS) on an M1 Mac can sign up to try the Parallels Technical Preview.

Article Link: Video: Testing Windows on an M1 Mac With Parallels 16
 
  • Like
Reactions: Lights87

SignalFi

macrumors newbie
Jun 20, 2020
2
21
Can anybody help me understand the use of Windows on ARM with a Mac?? Is there enough app support to justify adding the OS to an M1?
 
  • Like
Reactions: 0815
Comment

EatinPonies

macrumors regular
Feb 15, 2016
108
204
"...but going from four to eight cores did little for performance (1524 single core and 5958 multi-core)."

Um... the difference between 4 and 8 cores is 945 points by your own numbers. 945 points is an increase of almost twenty percent (18.85%) going from 5013 to 5958... That's not insignificant!
 
Comment

mrhick01

macrumors 6502
Sep 22, 2008
418
223
Can anybody help me understand the use of Windows on ARM with a Mac?? Is there enough app support to justify adding the OS to an M1?
Not now, but the performance of Apple's M1s and what is to come from them will influence the industry towards ARM processors and the development of operating systems to work on them.

ARM chips are proving themselves a technological iteration in power draw and overall efficiency and it's a marvel. In Dec 2020, Windows on ARM doesn't look like much. By December 2022 or 2023 however...
 
Comment

mrat93

macrumors 68000
Dec 30, 2006
1,636
564
Right now, getting Arm-based Windows on an M1 Mac requires Parallels and the Windows Insider preview version of the software, and it continues to be unclear if Microsoft will make a release version of Arm Windows available for people to license.

Wouldn’t that make Macs the best Windows machines? If M1 is already beating official ARM Microsoft-branded PCs via virtualization, imagine the native performance benchmarks if you could install ARM Windows via Boot Camp.
 
Comment

dmylrea

macrumors 68040
Sep 27, 2005
3,539
4,723
"...but going from four to eight cores did little for performance (1524 single core and 5958 multi-core)."

Um... the difference between 4 and 8 cores is 945 points by your own numbers. 945 points is an increase of almost twenty percent (18.85%) going from 5013 to 5958... That's not insignificant!
18.85% is nowhere near to the 200% you would expect with a doubling of cores. I think that was the point.
 
Comment

NewUsername

macrumors 6502
Aug 20, 2019
371
704
A solution to a problem that pretty much does not exist. There's a very limited set of apps that run on Windows ARM and I suspect that they're already available on a Mac. I doubt many would need this.
ARM Windows also runs Intel apps (though the x64 emulator is still in beta). This would thus allow us to run a ton of Windows apps, including games, in an environment potentially much more reliable than CrossOver.

I honestly don't know why Microsoft doesn't work with Apple to make Boot Camp possible with ARM Windows. Perhaps in 1-2 years? Anyway, the first Intel Macs didn't have Boot Camp from day one either, so we might still see something in the future.
 
Comment

Seoras

macrumors 6502a
Oct 25, 2007
647
1,525
Scotsman in New Zealand
The more you make of something the cheaper you can make it.
That usually holds true so I wonder if the new Apple TV will use an M1 chip instead of an A series.
A series enjoy the mass market of phones and tablets.
M1, being used in laptops where Apple isn't as dominant, will be lower in production numbers.
Adding it to Apple TV would increase production requirements and help lower the price somewhat.
 
Comment

MacLC

macrumors 6502
Oct 18, 2013
360
188
Can anybody help me understand the use of Windows on ARM with a Mac?? Is there enough app support to justify adding the OS to an M1?
If Windows can run 32-bit apps in compatibility mode, that's a plus.

If you need to test document compatibility (i.e. what Windows using customers see when they open your file), that's a plus.

If you want to run 200 games you don't need but were on sale at GOG/Steam for 75% off, that's a plus.
 
Comment

hajime

macrumors 603
Jul 23, 2007
6,381
900
Wouldn’t that make Macs the best Windows machines? If M1 is already beating official ARM Microsoft-branded PCs via virtualization, imagine the native performance benchmarks if you could install ARM Windows via Boot Camp.

Isn't BootCamp not allowed on the M1?
 
  • Like
Reactions: joelhinch
Comment

CJ Dorschel

Cancelled
Dec 14, 2019
407
806
Berlin
Not now, but the performance of Apple's M1s and what is to come from them will influence the industry towards ARM processors and the development of operating systems to work on them.

ARM chips are proving themselves a technological iteration in power draw and overall efficiency and it's a marvel. In Dec 2020, Windows on ARM doesn't look like much. By December 2022 or 2023 however...
You know, I never thought of that. I always assumed Intel would be the primary CPU going forward still but what if ARM takes ahold of the industry? I wonder if it would usurp Intel. That would be very interesting.
 
Comment

jdb8167

macrumors 68020
Nov 17, 2008
2,251
1,483
Comment

EatinPonies

macrumors regular
Feb 15, 2016
108
204
18.85% is nowhere near to the 200% you would expect with a doubling of cores. I think that was the point.
True, but I've never seen performance double when you double the cores. In anything. Calculating a giant Excel sheet with array tables and what-not doesn't have this drastic of a performance boost either.
 
Comment

Gnattu

macrumors 6502a
Sep 18, 2020
624
686
The reason that many of the Windows apps don't run is that they are 32-bit Arm apps. For some reason known only to Microsoft, they compiled many apps as 32-bit and the M1 will not run 32-bit Arm binaries. Here is a list compiled by @Gnattu of all the 32-bit binaries. I'm guessing this will correlate to the apps that didn't run while testing for the article.

https://forums.macrumors.com/threads/parallels-tech-preview-for-m1.2275996/post-29412503
You CAN run them if you have to by replacing the internal one with the x86 version because we have "rosetta" to translate x86 to arm64, but we don't have such translation layer for arm32 to arm64. There is an example in the same thread.

Unfortunately there is no easy way to do such replacement and you need to manually update all the apps you are interested because the automatic update will update them to arm version and it will not work with M1.
 
Comment

chucker23n1

macrumors 603
Dec 7, 2014
5,925
7,516
When running macOS Big Sur, the M1 MacBook Pro with 8GB RAM has a single-core Geekbench score of 1719 and a multi-core score of 7384. When running Windows 10 through Parallels using the default two-core setting, the M1 Mac achieved a single-core score of 1491 and a multi-core score of 2753.

That's pretty similar to Microsoft's own Surface Pro X in terms of multi-core performance,

You're comparing two virtual cores achieving a score of 2753 to a 4+4 physical core setup achieving a similar score.

Yeah, the value is "pretty similar", but the meaning is quite different: it means the M1 is way faster.

With four cores activated, multi-core Geekbench scores were much better, hitting 5013. The single-core score didn't improve quite as much at 1518.

Why would the single-core score change from adding cores? (I'm actually surprised it doesn't decrease, but what we're seeing here is probably just jitter.)

You can enable eight cores too, but going from four to eight cores did little for performance (1524 single core and 5958 multi-core).

A 19% performance improvement is nothing to sneeze at, and given the M1's setup, it must come entirely from the efficiency (i.e., low-performance) cores.

 
Comment
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.