After Mojave

[Lifeisabeach]

I'd wager the the possibility of that happening is about zero. Intel owns the rights to X86. They aren't going to just easily give a license to a huge customer to allow them to become a huge competitor instead.

From what I vaguely recall, there’s something about the history of x86 that forces Intel to license it to anyone who wants a license. That’s one reason they were pushing the Itanium architecture early on... they wouldn't have to license it out. I could be mis-remembering this. Maybe it’s just that AMD has a perpetual license to x86. Actually in just reading around now, that latter thought appears to be the case, more or less. Anyway, there are other licensees of x86 (yeah, I was surprised to see that there are), so there is precedence. Plus, Intel’s CEO openly regrets rejecting Apple’s inquiries to use them as a foundry for their ARM chips. Maybe an x86 license could be part of a deal to let Intel take on that business? Unlikely, probably, but it’s a thought to explore.

Also, no one on the planet has the experience with X86 that Intel has. It seems highly unlikely to me that anyone is going to do it better. AMD has been chasing this goal for ages and still hasn't.

Could the same have been said for ARM? Apple is at the top of that architecture despite coming in late. Of course they had to buy the talent, but hey, maybe there are some Intel guys looking for a change.

Also, going ARM-only coupled with traditional emulation software to provide backwards compatibility with old Mac software would likely be unacceptably slow. Transitive had very unique just-in-time technology that made Rosetta viable, but Transitive no longer exists; it was gobbled up by IBM. Once again, it might be an impossible licensing issue.

I agree completely. Emulating hardware is very processor intensive. It wasn’t that long ago when emulating arcade machines from the 80’s wasn’t even a good experience. Even less long ago before early Nintendos could be emulated.

 

[MacInWin]

But for most cases, I would propose, speed of emulation is not the critical factor. Just being able to emulate so that the job can be done would suffice, as long as the speed was not incredibly slow. I needed, for years, to be able to run Windows 7 because I had a specialty printer that had no Mac support. I would create the files (decorative labels for my wife's business) in macOS, then run Parallels to have a virtual Win7 device to send the print job to the printer. That was the ONLY reason for Windows on my Mac, and I didn't have any reason for it to have to be that swift.

So if a Rosetta-like emulator came along, just to let me do what I want/need to do until the vendors can shift and recompile in an ARM environment, that would suffice, particularly if the speed of the ARM was such that the emulation was anywhere close to the speed on the older Intel box.

And Apple could still say that the Mac can run Windows, just through this emulation. Wouldn't have the same meaning as now, but could still be true, particularly if they provided the emulation at no cost to the buyer.