That's easy, I think that because it's always happened before on every odd machine I've ever owned.
Again, to clarify, I'm not saying that SOMEONE won't release an emulation based software for x86-64 VMs to run on Apple Silicon Macs that people will flock to. I'd say the odds of it happening are pretty decent. I do not, however, believe that VMware or Parallels will be the ones to do it. I've only been talking about Parallels and VMware this whole time.
That said, it's foolhardy to think that just because something has always happened that it will continue to happen, especially in an industry where change is constant.
They haven't, but they'd really like to.
No, you're actually 10000% wrong on this one. Microsoft is a cloud services company now. Windows merely a vehicle for them to sell Microsoft 365 and Azure which is where the core of their business is now. They don't have any interest in becoming a hardware company in the way that Apple is a hardware company. Yes, they sell Surface computers and XBoxes and that IS hardware, but it's not what drives their business and certainly not to the degree that it is for Apple.
That's been their model to date, but like I said, they'd like more. As for Microsoft wanting WOA to succeed, I just don't see it. They've had on and off support of ARM processors for a long time and they still don't have a viable product. If it were important, it would have happened long ago.
Again, you're wrong here. Microsoft cares more about Azure and Microsoft 365 (which Windows 10 Enterprise is technically a component of on the Business/Enterprise subscriptions). That is their primary focus. Devices are important, but not even remotely to the degree that Azure and Microsoft 365 are. I do IT for a living and I practically live on the Microsoft site and regularly attend their many webinars. Trust me when I say that I know this stuff for a fact.
They care more about you running Windows in the form of Azure hosted Windows Virtual Desktop than they do about you running Windows on someone's ARM64 based system or not.
The only reason why Windows 10 for ARM64 isn't a viable product is that Microsoft CAN'T force developers and users off of x86-64 systems the way that Apple is currently trying to. It would be stupid of them to even try. So, they need to try to convince users and developers that ARM64 is just as good. But there's no underlying incentive for anyone to make the move on either side. That's why innovation here has been stalled. Not because Microsoft doesn't care. Because they can't and won't hold the gun to everyone's head like Apple is famous for doing and CURRENTLY IS doing.
Thinking from our end of the question, yes, that would be logical, but from Microsoft's end, I'm not so sure. They haven't shown that yet.
Again, like you've said, we don't know what Microsoft is thinking or planning behind closed doors. In either case, there's a lot more that needs to happen to make Windows 10 for ARM64 function on Apple Silicon (either in direct boot or virtualization scenarios) than was needed to bring x86 and x86-64 Windows to Intel Macs. Updating all ARM32 code present in the OS to be 64-bit is a key part of it as, that will certainly make the user experience confusing at best and suck at worst (as Apple Silicon hasn't had 32-bit ARM instruction sets since A10/A10X Fusion and all derivatives therein such as T2 and Windows 10 for ARM64 DOES have a bunch of ARM32 code still in it). But that's just one of many things that needs to be done. The licensing element isn't tricky, but it's more than what needed to be done back in 2006 when Boot Camp for Intel Macs first arrived. Similarly, Apple is throwing many more hurdles in the way of a native boot solution that will require much more than simply enabling CSM support in UEFI and writing drivers for components that don't already have pre-existing drivers due to being in other commercially available Windows PCs. Nothing insurmountable, but certainly much more work and collaboration will need to happen between the two companies than was done in 2006.
So far as what we have seen in public, and I think it's idiotic of them if that's their plan. All it'll do is make sure those of us that need x86 compatibility are pushed away from Apple altogether. Shrinking your market intentionally isn't a very good idea. They are also assuming WOA will ever be viable and licensable. I'm not so sure about that.
ARM is expanding in commonality. For those that HAVE to use x86 or x86-64, there will be emulators or the option of buying another machine. I'm not saying I'm particularly happy about it, because I'm really not.
Windows 10 for ARM64 will be viable and it would be actually stupid on Microsoft's part to not help it find its way to an Apple Silicon Mac, even if only for a VM.
Those that are not served by that (us included) will be incentivized to buy a PC or just not have a Mac. Again, as I've said, those that need to run x86-64 OSes on their Macs are not a big enough group of people, otherwise Apple wouldn't be gambling. They want MORE Mac users, not fewer.
What I'm describing is nowhere near the price of a full PC. Check them out, you'd be surprised. And I'm thinking more as a portable thing anyway. I already have Windows PC's at home to fill that role...
I have checked them out. The good ones are pricey enough to effectively be comparable to an actual PC. The ones that suck REALLY suck and are not even worth it even if they are nowhere near the price of a full PC. This is why people don't use them that much.
It is for me right now too though I hope that will change in the future. For now though, I bought a 2020 27" iMac with an i9. My thinking is that by the time it starts to feel slow and outdated, x86 virtualization will not be that important to me.
Worst case scenario, x86 virtualization won't suck on a PC. It'll just suck if what you want to do is virtualize x86-64 releases of macOS. But yeah, certainly if you're looking for a Mac that will virtualize x86-64 Mac, Windows, and Linux operating systems, a 2020 27" iMac is probably going to be as good as it gets short of going with a Mac Pro.
Add me to that boat as well. But since I mostly need a couple VMs for server use, rather than development, I can run the Mac Pro headless for that. By the time the Mac Pro is too old for the job, I’ll have a better idea of what to replace it with.
Unless you're talking about macOS boxes running the macOS Server app, I'd say that you're better off getting a PC tower. Maybe an HP Z workstation or a Dell Precision or, maybe an actual server, throwing either Hyper-V Server or ESXi onto it. I know that's not the cool thing to say around here, but your options will be more flexible on average. Plus you can get something adequate for said VMs for WAY less than a Mac Pro would cost you (that is, unless you need all 28 cores split between that small handful of VMs).
I wonder how much of the Apple market is in this boat. How many licenses do VMWare and Parallels sell to us? VMWare in particular has seemingly seen the macOS version as an interesting side project that pays for itself. They’ve been bringing parity features to the Mac, but very slowly. I expect we’re a minority in both the Apple and VM user markets, but I wonder how much of one.
It's to the point that VMware was publicly on the fence about even continuing with an Apple Silicon version of Fusion at all! So, I'd imagine it's not so big that they're worried, but big enough that it swung in favor of continuity of Fusion into the post-Intel era of Macintosh. Parallels has much more of their business rooted in Mac support, so it makes sense that they're down to continue into the Apple Silicon era, let alone to the degree that they're featured in the WWDC 2020 keynote video as the flagship partner for virtualization on Apple Silicon Macs.
That said, both of these products are virtualization products. They’ve never been emulators, so I do think it is a bit of a stretch to expect them to bulk up an emulator team just to be able to “not suck as bad as QEMU”. And to get to that point would require a good amount of expertise, and time. Not just a year or two either. I’d half expect 5 years is not a bad guess as to when emulation starts getting to the “not suck as bad as QEMU” state.
If either do, then Parallels is most likely to do it, with so much of their product line focused on the Mac.
I completely agree with this assessment. I don't think Parallels will do it because I think the narrative is still "ARM is going to become more commonplace as it expands and eventually x86 will be less relevant/essential" and that they're buying into it just as Apple is. But, again, for those of us serious about x86 virtualization, Apple is still selling us their high end Intel Macs and we'll have x86 PCs that are more than adequate to serve those needs (save for running supported x86-64 macOS guests) for years to come.
I don't see myself ever not needing x86, at least until I retire. (not all that far off, 6-10 years..)
It will become less and less relevant as time goes on, if current computing trends are to be given weight. Though, certainly it will be relevant for 6-10 years easily. Get an Intel Mac though. Any Intel Mac you buy today has at least that long left in it in terms of support. Otherwise, there are tons of great PCs out there that will be more than enough to handle your x86 virtualization needs. And again, I say this as someone who is in the same boat as you in terms of needing x86 support specifically.
