I feel like I'm living in a crazytown where nobody sees what's so plainly obvious. Doesn't anyone remember when IBM wasn't able to deliver chips to Apple on the schedule they demanded, and customers were complaining, and they were taking a drubbing in the press for it? And Apple let their existing product line stagnate for a little while, and then we got the Intel Macs?
Well, this time it's Intel and not IBM, except Apple owns their own chip designs and has 200 billion dollars laying around.
What others are seeing that you aren't is that
(a) Unlike the PowerPC days when the Mac gave Apple over 90% of it's profits, it now probably accounts for closer to 10% of Apple's profits
(b) The Intel excuse seriously does not fly with the Mac Pro.
(c) The CEO has openly wondered why anyone would buy a PC. Some have argued he meant Windows computers, but it doesn't inspire confidence (especially since he has stated he uses an iPad pro for most of his work)
(d) Clearly Apple's top priorities are making things smaller and lighter. This is out of step with a lot of "Pro" users' needs. Compare the new Mac Pro to its previous incarnation, and the changes in design priority could not be more obvious.
Apple has indeed some great chip designs available. But one should not forget that the switch to Intel not only gave Apple nicer CPUs but also made Apple compatible to the rest of the PC world. You can run any x86 application on a Mac via VMware or Parallels. Moving away from Intel would kill this.
I've been wondering, is it practically possible to have a dual arch hardware + OS?
The kernel would run on an ARM processor, along with all the supported software. Any x86 binary would be offloaded to the intel cpu. Sharing RAM and even the remaining hardware would be nearly impossible I guess. Not sure though. This way it can be like the integrated / discrete gpu design and switch can be done a lot less painfully.
Note that many operating systems today are already multiple architecture. It may not seem like it, but for example x86 and AMD64 are different architectures, just as 32 bit ARM and 64 bit ARM are different. The operating system has to define system call interfaces for them, have copies of libraries for each architecture, ensure the processor(s) are in appropriate modes when running the applications, provide inter-process interoperability even when different architectures etc.
Apple could make the equivalent of https://en.wikipedia.org/wiki/Rosetta_(software) to run x86 apps on non-Intel Macs. But the thing is, Apple has no good reason to make a different type of computer, considering computers are not that important, at least to Apple.
Rosetta worked, because the Intel chips were a lot faster than the Power PC processors they replaced, and it ran on single applications, not whole virtual machines. Considering how much Apple wants to optimize power usage for the laptop lines, emulating Intel seems to be counter productive to that.
Intel can certainly deliver chips to Apple to make better MacPros. I've got a hackintosh that I built last year which will easily outdo an Apple MacPro. Apple could decide to not be clever and just ship a pretty standard looking desktop computer with the latest chipsets and CPUs, spacious memory expansion and a couple expansion slots for GPUs and crank out a new rev every year and it would keep the people buying the MacPros happy and Apple wouldn't have to invest a ton of money overdesigning the thing -- and Intel would certainly be able to keep up with the demand easily.
The territory that they'd concede though is that expandability and people buying CPUs, RAM, drives and GPUs from non-Apple vendors. But that would buy them continuing to tie professionals down to macOS and then selling them laptops and building hype. Instead I think they're going to ditch macOS and switch to iOS on desktops and laptops and you'll see their laptop and desktop sales shrink and they'll be become entirely an iphone/ipad company. I don't think this has anything to do with supply chain and this is entirely self-inflicted damage because they're prioritizing their walled garden strategy above all else.
>Well, this time it's Intel and not IBM, except Apple owns their own chip designs and has 200 billion dollars laying around.
And? If any of the handful of people on the entire planet capable of building an ARM capable of rivaling Intel's current x86 designs were hired by Apple, it would be all over the tech industry news sites instantly. 200 billion doesn't mean a thing if it isn't applied properly.
This time Apple's explicitly saying the iPad Pro is their next personal computer so I wouldn't bet on that transition to ARM including macOS!
The only piece missing is creating iOS apps on iOS devices, for which they've already created a language and an app to learn programming, then everything we do and know and learn just to keep our computers running is kind of legacy stuff.
I think the Mac will be around for a while, but I strongly suspect that it will be running iOS soon. It won't be difficult for Apple to add Windowing to iOS, and mouse pointer and keyboard support, and replace MacOS with iOS.
The only issue will be major apps like Adobe CS and Office, but even those have iOS incarnations that can potentially reach parity with their MacOS counterparts.
I suspect a Mac running a KBM and Windowing friendly iOS would make a really nice machine.
What's wrong with Intel? They release new processors every year. Apple could easily update every mac every year with new processor, they just don't care.
Intel releases have become stretched out a bit, and they do no longer renew the whole processor lineup in parallel any more. It took Intel almost a year to finish the transition to Skylake - the late 2015 iMac already has Skylake processors, but the models needed for MB Pro only became available not too long ago. The 32g ram support for laptops seems to depend on the Kaby Lake earliest - while the ultra low power versions have been released, the more powerful versions are going to be released all over 2017.
Well, this time it's Intel and not IBM, except Apple owns their own chip designs and has 200 billion dollars laying around.