Rosetta Won't Support x86 Virtualization Apps Running Windows

[Maximara]

not further problem than the 2 years of Binary 2 apps, catalyst mess ETC.

Seriously, no computer in 5 years ahead will be sold with Intel inside.
massive consumer computing is mobile, is ARM, so computers would use ARM, also even this probe to be inferior (I dont think so)

mobile/tablet market weight is getting so big, in some years it wont be possible to compite with it, consumer PCs need to adapt, last try to survive

There are just people who are happy with the status quo and complain about any change. To anyone playing attention mobile/tablet market is where it is and that is effectively down to two players: Apple and Android. And Apple is doing a healthy 25% in that market. It makes sense to cross pollinate a market they are going reasonable in with a market where they have always struggled. I would like to point out that back in the 68k days the Mac was also around 10% marketshare.

 

[icymountain]

The problem I see is that free VM solutions being nonexistent on Apple silicon. Think Virtualbox or QEMU. I hate to say it but I knew this day would come where it’s not a possibility.

My concern exactly...
I guess we have to wait and see for what comes up...

 

[dukebound85]

I wish apple would offer both arm and intel simultaneously

 

[cmaier]

I wish apple would offer both arm and intel simultaneously

That would hold the entire ecosystem back. Unless Arm is the one-and-only future, there’s little incentive for developers to target Arm, and then little incentive for consumers to buy Arm. What you end up with is Microsoft’s half-assed Arm strategy.

 

[pankajdoharey]

It’s possible Microsoft will make a retail version of Windows on ARM available, but unlikely, at least for now. For starters, a MacBook Pro would likely run it faster than Microsoft’s own Surface X. Second, it probably isn’t a big enough market. Apple sells about 16-20 million Macs a year. Maybe 10% (probably far less) run Windows in any form.

That said, if Microsoft thinks making Windows on Arm available for Apple users will improve its popularity overall, it might release it eventually.

Ah Well Microsoft has made Office Sweet for Mac since the PowerPC days, there was even a Internet Explorer if you remember. It doesn't cost much for Microsoft to do a port to Mac because of their architecture, and also they dont want to seed that space to someone else so why wouldn't they optimise it? It just makes sense for them mostly because they want their office/enterprise customers to continue using their apps.

 

[cmaier]

Ah Well Microsoft has made Office Sweet for Mac since the PowerPC days, there was even a Internet Explorer if you remember. It doesn't cost much for Microsoft to do a port to Mac because of their architecture, and also they dont want to seed that space to someone else so why wouldn't they optimise it? It just makes sense for them mostly because they want their office/enterprise customers to continue using their apps.

Office is fine, but i wouldn’t call it ”sweet.” ;-) The current version is pretty nice, but it doesn’t quite have feature parity with windows. What I’d really like is the real thing on iPad - i can do a lot of my editing on the fly using ipad, but tables of contents, and that sort of stuff can’t be inserted, etc.

 

[Detnator]

I doubt that very much. With its current generation ARM CPU, Apple has closed the gap, but is still slower (sometimes dramatically) than Intel in most single-core tests (see e.g. here). They will be able to benefit from very tight software/hardware customization for specific tasks, and maybe keep an advantage in the performance/watt ratio thanks to TSMC's better process, but that is less important for computers than it is for phones and tablets, and may be short-lived anyway as Intel's "Tiger Lake" with their new process comes out later this year (which supposedly also has a much improved iGPU).

In any case, leaving artificial benchmarks aside, I very much doubt that the average user will perceive a performance difference due to the CPU in practice. CPU performance has simply not been a problem for most types of consumer applications for several years now (except games and a handful of niche things).

Your chart compares 2018's iPad Pro to 2018's MBP 13". Fair enough. But compare 2020's top MBMP13" with 2019's iPhone 11 Pro Max. I can't seem to figure out from Geekbench's website how to do a comparison, but here's the numbers according to Everymac.com:

2019 iPhone 11 Pro Max:

iPhone 11 Pro Max (US/CA/A2161) 64, 256, 512 GB Specs (A2161*, MWFC2LL/A*, 3306*, iPhone12,5): EveryiPhone.com

Specs and features for the iPhone 11 Pro Max (US/CA/A2161) 64, 256, 512 GB (A2161*). Dates sold, capacity, battery life, networks, size, price and more.

everymac.com

GB5 single core: 1326
GB5 multi core: 3393

2020 MacBook Pro 13:

MacBook Pro "Core i7" 2.3 13" 2020 4 TB 3 Specs (2020 13" (4 TB 3), BTO/CTO, MacBookPro16,2, A2251, 3348): EveryMac.com

Technical specifications for the MacBook Pro "Core i7" 2.3 13" 2020 4 TB 3. Dates sold, processor type, memory info, hard drive details, price and more.

everymac.com

GB5 single core: 1226
GB5 multi core: 4507


If we go back in time and compare iPhones/iPads with Intel Macs, the difference has been shrinking and this year, by at least one metric Apple's chip is now faster. The MBP is faster multi-core but the iPhone is faster single core. A phone is faster by at least one metric than the latest Mac laptop. And that's with the Apple A13 - not even the (not yet released A13X or A13Z that the iPad Pro would typically have). This year's processors will be A14's, and while I'm curious why this year's iPads didn't have an up to date Z or X chip, I can't imagine Apple's putting Macs on their silicon if they've hit roadblocks with those chips.

Apple's rate of improvement is far greater than Intel's and that's exactly been Apple's issue with Intel. They're improving too slow, while Apple's been improving things dramatically every year.

The take-away here is that unless Apple hits some roadblock they're not anticipating and/or Intel pulls something magic out of the bag suddenly, then tomorrow's Apple A chips will be a LOT faster than tomorrow's Intel chips. And "tomorrow" starts near the end of this year when the first "real" Apple Silicon Mac is released. And of course if Apple's and Intel's rates of change continue as they have been, next year, the year after, and beyond, will see Apple way ahead.

I'm pretty sure that's what Apple's betting on, and again, I can't imagine they're forging ahead with this not having prototype A15s and maybe A16s already in development that aren't proving very promising already. They're not making this change with vague guesses as to their roadmap at least a few years ahead.

All that said, you're absolutely right, that the average user doesn't care, as long as it's fast *enough*, and so even if Apple and Intel keep up with each other, Apple will still produce better products (for their target market, not for everyone) simply because of the tighter integration.

 

[Rigby]

Your chart compares 2018's iPad Pro to 2018's MBP 13". Fair enough. But compare 2020's top MBMP13" with 2019's iPhone 11 Pro Max. I can't seem to figure out from Geekbench's website how to do a comparison, but here's the numbers according to Everymac.com:

2019 iPhone 11 Pro Max:

iPhone 11 Pro Max (US/CA/A2161) 64, 256, 512 GB Specs (A2161*, MWFC2LL/A*, 3306*, iPhone12,5): EveryiPhone.com

Specs and features for the iPhone 11 Pro Max (US/CA/A2161) 64, 256, 512 GB (A2161*). Dates sold, capacity, battery life, networks, size, price and more.

everymac.com

GB5 single core: 1326
GB5 multi core: 3393

2020 MacBook Pro 13:

MacBook Pro "Core i7" 2.3 13" 2020 4 TB 3 Specs (2020 13" (4 TB 3), BTO/CTO, MacBookPro16,2, A2251, 3348): EveryMac.com

Technical specifications for the MacBook Pro "Core i7" 2.3 13" 2020 4 TB 3. Dates sold, processor type, memory info, hard drive details, price and more.

everymac.com

GB5 single core: 1226
GB5 multi core: 4507

I'll just point out that the Macbook number includes both the i5 and i7 versions. It stands to reason that the i7 version would get a higher score than the i5/i7 average that you quoted. It's also a good idea to look at the individual results rather than the overall score, since phone CPUs are designed with different priorities than laptop CPUs.

Apple's rate of improvement is far greater than Intel's and that's exactly been Apple's issue with Intel. They're improving too slow, while Apple's been improving things dramatically every year.

Of course it has been faster, since Apple started at a much lower level. A few years ago their CPUs weren't in the same ballpark as desktop CPUs. As mentioned earlier, they have closed the gap compared to low-power x86 CPUs (which is great), but that was also in a time period where Intel had fab-related problems, which they seem to have mostly solved now. And we have not seen yet how well Apple's CPUs scale outside of compact mobile devices.

I'm pretty sure that's what Apple's betting on, and again, I can't imagine they're forging ahead with this not having prototype A15s and maybe A16s already in development that aren't proving very promising already. They're not making this change with vague guesses as to their roadmap at least a few years ahead.

Of course not. But I don't think performance is the main factor that's driving them to make this move. Money is. I'd just caution people not to expect miracles, and not to forget the significant downsides of moving away from the industry standard compute platform.

 

[cmaier]

I'll just point out that the Macbook number includes both the i5 and i7 versions. It stands to reason that the i7 version would get a higher score than the i5/i7 average that you quoted.
Of course it has been faster, since Apple started at a much lower level. A few years ago their CPUs weren't in the same ballpark as desktop CPUs. As mentioned earlier, they have closed the gap compared to low-power x86 CPUs (which is great), but that was also in a time period where Intel had fab-related problems, which they seem to have mostly solved now. And we have not seen yet how well Apple's CPUs scale outside of compact mobile devices.
Of course not. But I don't think performance is the main factor that's driving them to make this move. Money is. I'd just caution people not to expect miracles, and not to forget the significant downsides of moving away from the industry standard compute platform.

The apple silicon will be significantly faster than intel’s chips for the same TPD/market segment.

 

[newfoundglory]

there are positives and negatives to changes like this. I totally understand why there is concern about the move to arm. Ideally, this would be faster macs using much less power, and costing less money for us to buy. bonus: you can also run ios/ipados apps on your mac. reality: they will probably be priced the same.

Microsoft and google have and do release arm-based products, and Apple's switch might trigger industry change to use more arm-based cpu's. speculation: apple releases a 3GHz+ arm processor which is fast and low power, the industry is left in shock. industry begins its own switch to arm for consumer devices. assume mostly windows10 on arm. this might be good if you have an apple silicon mac, but that sort of industry change could be 3-5 years away.

And not forgetting, the rest of the industry doesn't benefit from the same massive Apple app store of arm-based apps.

The 2005 WWDC is on youtube for those interested, Steve Jobs announcing the PPC->Intel switch. Its an interesting watch which stands true even today. He talks about how Apple envisioned products which it didnt know how to build with the future PowerPC roadmap: He talks about how they couldnt put faster G5 processors in the PowerMac and how apple never managed a G5 powerbook. I bet you could have fried an egg on a PowerBook G5..........

 

[cmaier]

there are positives and negatives to changes like this. I totally understand why there is concern about the move to arm. Ideally, this would be faster macs using much less power, and costing less money for us to buy. bonus: you can also run ios/ipados apps on your mac. reality: they will probably be priced the same.

Microsoft and google have and do release arm-based products, and Apple's switch might trigger industry change to use more arm-based cpu's. speculation: apple releases a 3GHz+ arm processor which is fast and low power, the industry is left in shock. industry begins its own switch to arm for consumer devices. assume mostly windows10 on arm. this might be good if you have an apple silicon mac, but that sort of industry change could be 3-5 years away.

And not forgetting, the rest of the industry doesn't benefit from the same massive Apple app store of arm-based apps.

The 2005 WWDC is on youtube for those interested, Steve Jobs announcing the PPC->Intel switch. Its an interesting watch which stands true even today. He talks about how Apple envisioned products which it didnt know how to build with the future PowerPC roadmap: He talks about how they couldnt put faster G5 processors in the PowerMac and how apple never managed a G5 powerbook. I bet you could have fried an egg on a PowerBook G5..........

When apple released the first 64-bit A-series processor it shocked the whole industry. They were left flat-footed. Nobody had planned 64-bit for years. And they never caught up.

Same thing will happen again.

 

[1146331]

LOL.

Stop writing stupidities. If there is an OS that is productive, it’s macOS more than anything else. Especially for software development.

Parallels and vMWare will come out with native ARM64 virtualization of ARM64 guest OS.
Planet will recognize Intel and x86 must be dumped and will start shifting to ARM64.

Apple has ofter been the first to change paradigm. Drop legacy ports, drop audio jack in favor of wireless, and countless of other things. Now they want to drop legacy x86 support in favor of their own people are complaining 🤦‍♂️

I never said Intel doesn't need to be dumped and if you were somewhat intelligent with what you wrote, you'd know x86 has been mostly dead for years. It's all x64 now.

But x86 support is not ready to be dumped yet, at least for a company that desires "professional" [not "artistic"] customers. Anybody that actually works in IT for businesses will know that companies will choose the cheapest route to continue their business, even if that means patching a 40-year-old application to work on x86 versus redeveloping the application.

I would much rather give AMD a chance rather than ARM because I like to actually do useful things when I work, instead of fanboy for a company that doesn't care about me.

Apple has been the first to change the paradigm for phones and tablets, as well as OS's related to those. Other than that, they haven't done much except destroy the OS that Jobs made good over the past decade.

 

[Detnator]

I never said Intel doesn't need to be dumped and if you were somewhat intelligent with what you wrote, you'd know x86 has been mostly dead for years. It's all x64 now.

But x86 support is not ready to be dumped yet, at least for a company that desires "professional" [not "artistic"] customers. Anybody that actually works in IT for businesses will know that companies will choose the cheapest route to continue their business, even if that means patching a 40-year-old application to work on x86 versus redeveloping the application.

I would much rather give AMD a chance rather than ARM because I like to actually do useful things when I work, instead of fanboy for a company that doesn't care about me.

Apple has been the first to change the paradigm for phones and tablets, as well as OS's related to those. Other than that, they haven't done much except destroy the OS that Jobs made good over the past decade.

You’re right. Apple doesn’t care about you, nor the other people like you. Neither did Jobs. You’re clinging to a non-existent fantasy. macOS is better than its ever been - at least for the people it’s targeted at.

Apple cares about the people in their target market - that’s not you, nor is it most of those people working in “IT”.

For those people in Apple’s target market, Apple generally delivers on what they want/need. The millions and millions and millions of people in that market are happy, as evidenced by sales and constantly high customer satisfaction ratings.

Apple doesn’t give a rip about most IT people - never has. They’ve always been about making technology easy to use for normal people - the non-IT people. IT people are some of what Windows and Linux are for, and Apple’s perfectly happy with that.

 

[pldelisle]

I never said Intel doesn't need to be dumped and if you were somewhat intelligent with what you wrote, you'd know x86 has been mostly dead for years. KIt's all x64 now.

oh wow. That’s incredible. Thank you soooooooooo much for this information. It changed my life.
🤦‍♂️🤦‍♂️🤦‍♂️🤦‍♂️🤦‍♂️🤦‍♂️🤦‍♂️

x86 - Wikipedia

en.m.wikipedia.org

Bits	16-bit, 32-bit and 64-bit

[automerge]1593432477[/automerge]

Apple has been the first to change the paradigm for phones and tablets, as well as OS's related to those. Other than that, they haven't done much except destroy the OS that Jobs made good over the past decade.

another stupidity. MacOS has never been so great as it is today. It’s integration with iOS, iPadOS,and the whole ecosystem is working good. The introduction of Catalina hasn’t been good, but I got no bug since 10.15.3. Destroy the OS of Jobs ? Ahahahha. Yes because a company like Apple must stay in the past and not change anything, yeah right.
[automerge]1593432703[/automerge]

Apple doesn’t give a rip about most IT people - never has. They’ve always been about making technology easy to use for normal people - the non-IT people. IT people are some of what Windows and Linux are for, and Apple’s perfectly happy with that.

I’m in IT. Apple is currently the best platform ever for software development, and will still be the best one on ARM. Windows ARM will be available publicly, we are all gonna be able to virtualize it, it will soon run x86 64-bit softwares on it when ARM version isn’t available, so I don’t see any insurmountable problems for any professional using a mac.

I wouldn’t be surprised that with the great efficiency of the Apple GPUs we could see new softwares on the platform, especially in the 3-D segment, creation software and CAD.


Did I really have to say this ? Wow.

 

[von-uk]

That sucks, MS Project and MS Visio are essential to my work, plus two other old Windows apps. I've tried Mac alternatives, but I have huge plans and diagrams that fail to open in Mac equivalents e.g. Merlin, OmniPlan, Project Libre, Smart Draw, OmniGraffle, etc. I really hope I can stick with MacOS and still run my Windows Apps. I remember switching to Mac over 10 years ago on a whim and was more productive on the Mac a month later than I had ever been on Windows. I really don't want to have to go back to Windows once my current MBP needs replacing, but if I can't run my essential software I may have to!

 

[pldelisle]

That sucks, MS Project and MS Visio are essential to my work, plus two other old Windows apps. I've tried Mac alternatives, but I have huge plans and diagrams that fail to open in Mac equivalents e.g. Merlin, OmniPlan, Project Libre, Smart Draw, OmniGraffle, etc. I really hope I can stick with MacOS and still run my Windows Apps. I remember switching to Mac over 10 years ago on a whim and was more productive on the Mac a month later than I had ever been on Windows. I really don't want to have to go back to Windows once my current MBP needs replacing, but if I can't run my essential software I may have to!

Unless you totally switch to Apple’s equivalent, I don‘t think you will ever be able to make these softwares work again.

 

[luisfrocha]

From what I have read, aside from the technical hurdles of developing an emulation and/or virtualization solution, the next biggest blocker to running Windows on M1 machines is the fact that Microsoft doesn’t sell licenses of Windows to run on ARM machines (whether native, emulated, or virtualized) to the end user; it’s my understanding that only OEMs can get those. So once VMWare and Parallels deliver their solutions, we would still potentially need to wait for Microsoft to change their licensing.

 

[robinp]

From what I have read, aside from the technical hurdles of developing an emulation and/or virtualization solution, the next biggest blocker to running Windows on M1 machines is the fact that Microsoft doesn’t sell licenses of Windows to run on ARM machines (whether native, emulated, or virtualized) to the end user; it’s my understanding that only OEMs can get those. So once VMWare and Parallels deliver their solutions, we would still potentially need to wait for Microsoft to change their licensing.

If you're running Windows emulated, you would be buying an x86 version of windows and running it inside an emulation system. As Connectix did with their Virtual PC software for Macs using the PowerPC chips.

 

[blackcrayon]

From what I have read, aside from the technical hurdles of developing an emulation and/or virtualization solution, the next biggest blocker to running Windows on M1 machines is the fact that Microsoft doesn’t sell licenses of Windows to run on ARM machines (whether native, emulated, or virtualized) to the end user; it’s my understanding that only OEMs can get those. So once VMWare and Parallels deliver their solutions, we would still potentially need to wait for Microsoft to change their licensing.

AFAIK there's nothing in the Windows license that precludes running the off the shelf version of x86 Windows on any environment that it can run in. They might not be licensed to bundle Windows, already installed, with their emulator and sell it, but that's about it. I think this only applies to ARM native versions of Windows, which aren't sold separately- so of course there would be no license for installing in a virtual machine.

It would seem that Microsoft would be interested in selling extra copies of Windows though, so I'm expecting to see this happen with ARM. Since they have 3rd party hardware customers using it, why wouldn't they want Mac user sales as well.

 

[PinkyMacGodess]

AFAIK there's nothing in the Windows license that precludes running the off the shelf version of x86 Windows on any environment that it can run in. They might not be licensed to bundle Windows, already installed, with their emulator and sell it, but that's about it. I think this only applies to ARM native versions of Windows, which aren't sold separately- so of course there would be no license for installing in a virtual machine.

It would seem that Microsoft would be interested in selling extra copies of Windows though, so I'm expecting to see this happen with ARM. Since they have 3rd party hardware customers using it, why wouldn't they want Mac user sales as well.

I haven't seen any 'ready made' Windows VMs offered either, but I haven't looked for any. I'd imagine they won't want folks sharing them.

 

[Spock]

If you're running Windows emulated, you would be buying an x86 version of windows and running it inside an emulation system. As Connectix did with their Virtual PC software for Macs using the PowerPC chips.

Since Microsoft owns VirtualPC, it’s not totally impossible that they could offer a Windows on ARM solution for the Mac. Probably won’t happen, but it’s not impossible.

 

[biffuz]

It won't take long before we see yt videos on how to install Windows on Qemu on Arm Macs. It will be fun to see the benchmarks.

 

[cocoua]

unless x86 advances take off, most PC will be ARM in 2 years (which will pave the path for Chrome to evolve to a desktop OS, so Windows ARM will be broadly available).

I guess Bootcamp will come again in the short future.

 

[anshuvorty]

unless x86 advances take off, most PC will be ARM in 2 years (which will pave the path for Chrome to evolve to a desktop OS, so Windows ARM will be broadly available).

I guess Bootcamp will come again in the short future.

I wouldn't hold my breath on this. More likely this won't happen than will happen.

 

[cocoua]

yeah, I'm not sure about the Bootcamp thing, but as Windows will be ARM, some native virtualization will be possible (and legal ) of course nothing comparable to Bootcamp…