That would also be the same Microsoft that restructured last week around the cloud and is de-emphasizing Windows? For many people their next PC will be a Chromebook. That’s already the case in education. My guess is that Apple has a pretty good sense of the percentage of their users who need Windows support. I don’t think the transition will be as quick as from PPC to Intel. I suspect iMac and Mac Pro will keep Intel the longest. MacBook is likely to be the first to go ARM. The Cannon Lake delay hits the MacBook line the hardest since Apple has historically released each version about six months after the 5W Y chips became available. To date there hasn’t been an official announcement for the Cannon Lake Y series.
Got a tip for us?
Let us know
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
Apple Plans to Ditch Intel and Use Custom Mac Chips Starting in 2020
- Thread starter MacRumors
- Start date
- Sort by reaction score
I think the best plan for Apple would be to over it as a no-cost option for someone buying a Mac. If they want the Universality of the Intel chip, then thats all good. If someone wants the optimization and efficiency of the Apple chips, then thats all good too.
I am sure they will build an emulator to support x86 instruction set. They will just have to license that from Intel. Maybe that will be a build-to-order feature. When the speed is fast enough and Windows is phasing out, inefficient instruction set level emulation is not that bad.
This will not be an issue.
Apple is going back to the 80s...
Custom computers with own peripherals (e.g printer) made to work only with Apple computers.
We all know how this ended...
That was the past and the average consumer market is larger than the pro so why should they focus on such a tiny segment?
The pros are also never ever happy it’s almost as if they want a new computer every day. I’ll be happy when that modular pro is finally released so that the crying can finally stop.
Actually, I have a vested interest in Apple doing well. I own Apple stock, not Intel.
I know all about Fujitsu supercomputers and using ARM and also Qualcomm has a server class ARM processor.
In the case of Fujitsu; they also have their own wafer Fab and have had wafer fabs for 30 years.
They were partnered with Amdahl doing Sparc V9 processors. But is you read the article closer, it's a supercomputer that execute the ARMv8 instruction set. The architecture is unknown.
Anyway, Fujitsu has been doing supercomputer design for 30 years.
Apple was an initial investor in ARM. ARM was in previous Apple devices. Research Apple Newton.
Apple also participated in AIM; Apple, IBM, Motorola for PPC processors.
Intel in the silicon industry is like Apple in the mobile industry.
While the A series processors do great in mobile devices, the desktop/laptop is a different animal.
You need the the ability to access 256GB of real memory on a high speed bus using multiple channels of DDR.
You need DDR controller and scheduler IP.
You need the ability to access multiple lanes (16-32) of PCIe as a root complex.
You need cache coherency that is much different than ARM and it's ACE or CHI.
You need really good floating point performance.
The ARM CPU in an iPhone is a far cry from a dual chip Xeon system.
And a dual chip or 4 chip Xeon system is what they have to compete with.
it make absolutely no sense to segment the Mac offerings with two CPUs and the developer porting with "fat" binaries like the past.
We'll see in 2020.
But these same folks said the same thing in 2012, 2015 and now 2018.
Let me know when Apple has a FAB, that's the only way they will ever replace Intel.
I hope this doesn’t happen. It’s going to be terrible if it does, beginning with incompatibility and slow apps running through emulation layers, and it’s going to end up with apps written for macOS probably working fine, but virtualization of other operating systems is going to be terrible, and I have my doubts about speciality software like X-Plane running with full performance in the first couple of years.
I can see it happening for the most portable Macs, due to ARM being suitable for low energy devices, but Apple has already changed architectures three times...
I can see it happening for the most portable Macs, due to ARM being suitable for low energy devices, but Apple has already changed architectures three times...
It very often is not simple. It can take days to get some stuff running, and sometimes it just won't or only with a number of issues. We are also talking about a completely different architecture. You most certainly need tweak the compilation procedure to get stuff running.
So what is the issue they are trying to solve? It does not make sense to spend billions trying to design a CPU when you already have a large supply of cheap cpus that are quite powerful. not like we are CPU bound anymore, most cpus sit idle.
I do not see what the point of all the effort will bring other than to lose focus on what they do best, which is focus on good software and the rest of the system (ie display,etc..)
I do not see what the point of all the effort will bring other than to lose focus on what they do best, which is focus on good software and the rest of the system (ie display,etc..)
Oh wow, thousands of doubts from someone who is feeling the need to replace its own Core2Duo MBP pretty soon...
Do I get a 2018 or 2019 (last Intel) MacBook Pro? Or just try to wait until 2020? I don't usually trust in first gen Apple hardware, so it would be 2021... Just too much for my old -and only machine- MacBook Pro.
Also, software compatibility with Intel CPUs will be great, but I bet the desktop Apple CPUs will be truly awesome: 7 or 5nm process build, more than 4 cores, huge energy efficiency... I jus't don't know what to do. Maybe buy, and wait until ARMs are mainstream in Apples computers.
Do I get a 2018 or 2019 (last Intel) MacBook Pro? Or just try to wait until 2020? I don't usually trust in first gen Apple hardware, so it would be 2021... Just too much for my old -and only machine- MacBook Pro.
Also, software compatibility with Intel CPUs will be great, but I bet the desktop Apple CPUs will be truly awesome: 7 or 5nm process build, more than 4 cores, huge energy efficiency... I jus't don't know what to do. Maybe buy, and wait until ARMs are mainstream in Apples computers.
I used an emulator back when Apple was on the PPC, and it was horrible.
I'll wait until we get more confirmation, but I'd probably just buy non-apple computer if this turns out to be true.
Worse. Software compatibility is tossed out the window. Every emulator I ever saw Apple produce worked but slowly.
What's the benefit here? "It's more like an iPhone and iPad" - paraphrasing their own words. If we wanted iPads or iPhones we would have bought one of those (many of us did). I honestly don't want my mac to operate more similarly to my iPad or iPhone.
And this would be the end of me ever buying an Apple computer again.
but everytime there's a new chip sighted with Apple's logo on it, someone runs this story. it's dubious at best
but everytime there's a new chip sighted with Apple's logo on it, someone runs this story. it's dubious at best
Windows supports ARM nowadays. What's the big deal as ARM Windows can run desktop x86 apps. Everybody is blowing this out of proportions as the underlying CPU simply doesn't matter like it did back in 2005. Apple now makes the most power efficient chips in the world. The x86 is fast but it is the modern day equivalent to the PowerPC. Also, why give Intel all the $$$ for buying their chips when Apple has its own more than capable chips. Maybe if users need x86 then Apple could load it on a sort of CPU daughterboard to allow those who need it to use it.
Me too. Let’s hope that whatever they are doing, they execute well. (Unintentional pun!)
I'm very skeptical at this.
As a developer, the best part about the x86 architecture is I can build for both Windows/Mac/Linux reliably.
If Apple goes the way I think they are doing (e.g. PowerPC days), it will be 2 steps backwards.
Any 1st party Apple software solutions for porting will be complete trash. Some people on here seem to have completely forgotten what it was like in the PowerPC days, and WHY Apple went Intel in the first place.
As a developer, the best part about the x86 architecture is I can build for both Windows/Mac/Linux reliably.
If Apple goes the way I think they are doing (e.g. PowerPC days), it will be 2 steps backwards.
Any 1st party Apple software solutions for porting will be complete trash. Some people on here seem to have completely forgotten what it was like in the PowerPC days, and WHY Apple went Intel in the first place.
It's a good way to reduce their market share even more. I would have to go back to windows/linux. I need to run those as VM's anyway. I choose to run MacOs. I don't have to.
Ditching Intel is a minor issue if you compare it to all the things Apple is doing:
- Ignoring standards that help developers (OpenGL and OpenCL are treated not even as second, but as third-class citizens, while Vulkan is completely ignored).
- Pushing users to depend on iCloud, and try to force them to update even if they don’t want to.
- MacOS updates are now several gigabytes, take ages to update, and these years key components became broken (PDFKit, Disk Utility, just to name a couple).
- Nvidia unsupported, not only currently (no way to install a Titan in a Mac unless it’s a Hackintosh), but also in the future, according to the documents made public yesterday.
- No true Mac Pro (“true” here means what the Mac Pro always meant, not including the cylinder, nor the joke called “iMac Pro” which has obviously been an attempt at trying to kill interest in any future Mac Pro).
- Lots of other annoyances (minor, but they add up to the final sum: the nonsense of versions and mobile-like autosave, the obsession of turning MacOS into another iOS, getting rid of great inventions like the MagSafe, Macbooks with only one port, and lots of annoyances more).
Now, if you tell me that ditching Intel is bad... well, I wouldn’t know what to reply, honestly.
- Ignoring standards that help developers (OpenGL and OpenCL are treated not even as second, but as third-class citizens, while Vulkan is completely ignored).
- Pushing users to depend on iCloud, and try to force them to update even if they don’t want to.
- MacOS updates are now several gigabytes, take ages to update, and these years key components became broken (PDFKit, Disk Utility, just to name a couple).
- Nvidia unsupported, not only currently (no way to install a Titan in a Mac unless it’s a Hackintosh), but also in the future, according to the documents made public yesterday.
- No true Mac Pro (“true” here means what the Mac Pro always meant, not including the cylinder, nor the joke called “iMac Pro” which has obviously been an attempt at trying to kill interest in any future Mac Pro).
- Lots of other annoyances (minor, but they add up to the final sum: the nonsense of versions and mobile-like autosave, the obsession of turning MacOS into another iOS, getting rid of great inventions like the MagSafe, Macbooks with only one port, and lots of annoyances more).
Now, if you tell me that ditching Intel is bad... well, I wouldn’t know what to reply, honestly.
this gets said time and time again and proven wrong. Not because the logic is entirely faulty, because it doesn't take into account that gaming PC's tend to be a moving target. Sure in 2020, a thumbstick PC might be equal to today's mid range PC's. But mid range PC's in 2020 are going to have advanced by 2 years as well. And for those who want bleeding edge gaming (maybe not a majority), those thumbstick PC's in 2020 are still going to be completely inadequate.
We've seen it with NUC's. Today's NUCs are actually quite powerful. if they were out 5 years ago, they'd be trouncing many of the computers people used. However, we're now 5 years on and Today's NUC's aren't up to modern gaming standards for today.
this isn't a bad thing. who the hell wants computers / technology to stagnate just so the "small" stuff catches up
Apple probably commands enough of the market that neither Adobe or Microsoft would abandon MacOS development if Apple moved away from Intel chips. Some smaller developers would probably have to eliminate cross-platform support, and I assume Bootcamp and Windows virtualization would no longer be practical options.
This would certainly impact users like me, who only could make the switch to Mac due to the benefits of the Intel chips (I rely heavily on cross platform statistical software and the occasional Windows app). If support for these products disappeared, I would be forced to move back to a Windows PC.
I have no data to make a claim, but my best guess is that tighter integration between iOS and MacOS, as well as control over their product release cycle, would enable Apple to increase upgrade rates and probably new Mac adoption, and that these sales would be greater than those lost by people like me.
This would certainly impact users like me, who only could make the switch to Mac due to the benefits of the Intel chips (I rely heavily on cross platform statistical software and the occasional Windows app). If support for these products disappeared, I would be forced to move back to a Windows PC.
I have no data to make a claim, but my best guess is that tighter integration between iOS and MacOS, as well as control over their product release cycle, would enable Apple to increase upgrade rates and probably new Mac adoption, and that these sales would be greater than those lost by people like me.
On a laptop raw performance isn’t an useful as balanced and consistent performance.
As to ARM, if anything, expect them to adopt RISC-V (https://riscv.org)...
I have no evidence for this expect that it’s what I would do if I were to jump head-first into fully custom silicon that far ahead... in any case, exciting!