I wonder how this will effect OS's I have to imagine that there will be a different code structure for ARM as opposed to Intel? No?
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 Expected to Announce Arm-Based Mac Plans at WWDC, Transition Away From Intel to Begin in 2021
- Thread starter MacRumors
- Start date
- Sort by reaction score
First - dual boot is a "don't know" until more information is released. There's no fundamental reason why an ARM Mac shouldn't be able to boot the ARM version of Windows 10 - but it's impossible to know without more hardware and firmware details... but it will be Windows 10 for ARM so it's usefulness would depend on how support for that develops in the near future - but it does have x86 emulation so most windows software will run, but with unspectacular performance. Long-term, its usefulness depends on how Windows on ARM grows.
Second while MacOS on ARM will support "virtualization" (unless Apple are really holding it wrong - its a pretty essential part of modern operating systems) that will be ARM virtualisation so, again, you'll be talking about virtual machines running Windows for ARM, ARM Linux or Mac OS for ARM - still useful, but probably not for running old x86 Windows software - so you're back to betting on how well ARM Windows is going to take off. Ditto for ARM Linux - although Linux has supported ARM for years and most of the big open-source projects are already on ARM.
To run x86 Windows or x86 Linux you'll need an emulator that can simulate (using software) an x86 processor on an ARM. It's quite possible - that's what those Youtube vids showing MacOS running on an iPad are using - but it won't have anything like the sort of performance you get with Parallels/Fusion. (Anybody remember SoftWindows on PPC...? Right).
So, even in the unlikely event that you'll be able to buy an ARM Mac the day after WWDC, don't rush if your workflow depends on dual-booting or virtualising x86 operating systems, don't rush to the store.
If you're using a Mac to run high-end Windows software or build binaries for x86 Linux* this is probably the end of the road - although, frankly, if I were doing that I'd have bought a PC laptop with a decent spec and GPU years ago.
Apart from that, the world has changed a lot since 2006 when running Windows on a Mac was a killer feature - so I'm not sure that, 2-3 years down the line, it's going to be that big a deal.
(* ...apart from that, ARM Linux is pretty complete, Linux development is traditionally CPU-independent, most "Web development" is platform-independent Javascript/PHP/Java/Perl/Python... and if you absolutely need to do an x86 build you can spin up a Linux instance in the cloud, on demand, for next to nothing now)
And they does support that-- in an iPhone.
Since iPhone 6s the A9 already have a PCIe connection to use NVMe storage.
And I do not think Apple doesn't own a 7nm PCIe 4.0 blueprint/IP at this time.
I take the opposite view.
Buy an Intel Macbook now before being stuck with using the first version of a new architecture.
No.
Source code is usually 100% portable in todays environment.
User land applications are CPU independent for more than 2 decades
[automerge]1591736920[/automerge]
That my suggestion too.
If you fear of a transition then do not take a part in it.
Wait for the transition to finish.
And obviously you will lose any exciting feature/performance gain during the transition.
Ampere Altra is the first 80-core ARM-based server processor
Ampere has unveiled the industry's first 80-core ARM-based 64-bit server processor today in a bid to outdo Intel and AMD in datacenter chips.
Sorry, but, ARM is already on 80 cores and counting. Maybe the ARM per-core performance will lag behind AMD and Intel, but ARM has a built-in advantage when it comes to die size and power consumption if only because an ARM core doesn't need a bulky x86 instruction translation unit - enabling more cores and/or more space for GPU/vector processors/accelerator/codec etc.
Can't wait for the Photoshop speed comparisons.
That and editing speed comparisons, Premiere, Resolve, etc
AppleTV would make sense as a 'developer kit". All the more so if it is bumped up to a A12X/Z.
the touch screen on a iPad Pro is a determinant to testing macOS apps since Macs don't have touch screens. I highly doubt also Apple is up for tweaking the primary boot firmware on the iPad to do anything other than iPadOS.
All you need is a box that can connect to a bluetooth Keyboard , display out , and a decent amount of RAM/Disk to install on a handful of apps. If put a USB-C port on a new Apple TV then would have a physical port. Bring your own Display , Keyboard , and Mouse would keep the dev box costs relatively low. ( and pretty good chance would be a leased box , not a sale. ) .
Pretty good chance Apple is starting with a MacBook revival of the one port wonder they dropped because didn't have an Intel CPU+GPU package they liked. It would be thinner and lighter ( as the Bloomberg article points to Appel's desires being focused on. ). For apps for a device like that you don't need Apple's latest stab at a tower at all. (because towers are super hyper engineering things now at Apple). That Apple would throw a cheap case they found at Fry's at a developer kit along with some generic ATX motheborad they build I just don't see it. ( again just huge dicsonnect from the thinner lighter laptop primary focus).
P.S. If Apple build a AppleTV prototype back in 2018 with a A12X that would line up with trying it out a firmware tweaked board of of that as a macOS test run system there too. Or they threw it into a board that could fit a one port wonder MacBook. Both of those are more likely than an iPad.
( they skipped the AppleTV prototype then but probably just iterated on it till this year. )
There was a recent paper in Science:
There’s plenty of room at the Top: What will drive computer performance after Moore’s law?
Maybe someone subscribes, and can tell us if the register summarized it adequetely
Moore's Law is deader than corduroy bell bottoms. But with a bit of smart coding it's not the end of the road
Moore's law won't last forever. A move to ARM will delay the inevitable for a couple of years at most. If Mac programmers continue their undisciplined ways, macos will continue to be seen as the slower OS, regardless of how many chickens are available to plow the field.
Last edited:
Why? The rumor said Apple is designing its own GPU for this endeavor. And since most macs don't support changing the GPU...
And Apple likes its own flash controllers as well.
All apps will need to be re-compiled to ARM code or Bytecode. In many cases that just means the developer ticking the "ARM" box in XCode and hitting "build" (and then, obviously, doing some testing) and uploading the result. For most modern apps it should be far less effort than it took going from 32 to 64 bit and fixing other problems casued by the annual Mac OS update.
The problem will be the exceptions where the App has been "hand optimised" for x86 (which will typically only affect small parts of the code) in some way or relies on third-party libraries that haven't been re-built yet (likely the biggest cause of delay). Dropping 32-bit support in Catalina will have "killed or cured" a lot of those problems now.
“It has set Apple up for the next 20 years”
incredible foresight, but technology moves faster than even he could predict.
I think this is the most important bit. And, there’s the instruction length decoder that is not required on the A-series (all instructions being 64-bit). There’s a good amount of performance overhead on all x86 solutions.
More to the point, Thunderbolt/USB4 are based on PCIe... but PCIe already exists on ARM systems so there's no reason to think that a "Pro" ARM Mac wouldn't have PCIe, even if the first effort is an ultraportable based on a phone chip.
This is partly about performance increases going forward, but mostly about margins. A-series silicon costs Apple something like US$20-30 per CPU.
Article also says that mac has 10% of the market .... which it doesn't worldwide. Close to 10% USA but that isn't the world.
The article seems to indicate Apple may be living in a time bubble.
"...Inside Apple, tests of new Macs with the Arm-based chips have shown sizable improvements over Intel-powered versions, specifically in graphics performance and apps using artificial intelligence, the people said. Apple’s processors are also more power-efficient than Intel’s, which may mean thinner and lighter Mac laptops in the future. ..."
Versus what AMD and Intel will have rolled out by 2021.
Intel Gen12 Xe Graphics Leads AMD's Vega in Integrated GPU Performance
The tide has turned and roles have swapped, with AMD now delivering more CPU but less GPU.
Yes... Intel's 2018 iGPU performance probably will get whipped. but walking away from 2020 instances. That remains to be seen.
In the spectrum of thinner and lighter Mac laptops ...yes a A series derivative is probably a winner. But in terms of the overall Mac product line up. Apple has more than a few missng pieces. Their GPU get clobbers by most discrete GPUs. They have no facility for hooking those up. Nothing on multiple high speed I/O ports. etc.
The quote in the article is
"... Like it did then, the company plans to eventually transition the entire Mac lineup to its Arm-based processors, including the priciest desktop computers, the people said. ... "
'Eventually' doesn't seem to mean indicative of "soon". There other somewhat highly questionable assumption there is whether Intel and AMD are going to sit on their hands in the higher end space while Apple plans 'something". As long as Appel's custom ARM chips are anchored on iPhone workloads there is a pretty big disconnect with doing higher end stuff.
Apple could pull a future ARM N1 varaint ( N2 , N3 ) design and tweak it to Mac desktop usage but are they really going to put in that kind of effort for relatively low volume Macs models. They have been variable Rip van Winkle when it comes to acually dong work on the Mac Pro.... suddenly they are going to do super low volume chips just for that product? Quite questionable. if Apple goes ARM top to bottom on line up I'd expect the line up to shrink (or at least shift to more mid range targets. More laptops perhaps and less desktops. ).
Er... if you're testing Mac OS software then it's rather easier to not use the touchscreen that is there than, on an Apple TV, to test apps with (e.g.) the TouchID sensor, webcam etc. that aren't there. An iPad Pro could also emulate the touchbar (much as some people here would like it to go away, Apple clearly don't subscribe to that point of view)
The Intel dev system was, basically, the first Hackintosh, with a PC motherboard built into a G5 Tower case. If Apple build an ARM Mac into a basic mini-tower then developers can go hang because I WANT ONE!!!
...but, this time round it's probably going to need all the features of a MacBook Pro (maybe Thunderbolt or USB4, too) so they might actually need to build a MacBook-type developer machine. But an iPad Pro would probably come closest.
Perhaps not. But if you scale that down to the awhatever equivalent, is it better? Unlikely.
Now scale that iPad Pro chip up to be appropriate for your laptop... what happens?
Is that specific laptop the majority of laptops?
...because Intel always hit their launch date targets :->
True - and they only just launched the new Mac Pro. They can't "obsolete" that in the next year or two and retain any credibility (if they have any left after the trashcan...). The big "pro" audio and video suites are going to be the hardest to transition to ARM because of all their dependencies and plug-ins.
The CPU is probably, as you can that cpu from many makers including dell.
Was going to get the 16" but going to wait now.
Don't want a computer with a crappy resale after ARM comes out.
Don't want a computer with a crappy resale after ARM comes out.
So much drivel from one single poster.
No, but AMD has. And before that you could say the same thing about AMD. Intel will get its **** together, but creating integrated circuits has a lead time measured in years. But you would know that, mr. "I've designed CPUs".
Not really, no. Again, a facile and demonstrably wrong statement. Obviously ARM cannot run AMD64 code natively, but setting that aside, ARM cannot achieve performance for free. But you would know that, since you claim to have designed CPUs and are now posting on a Mac rumors forum. Like so many industry CPU engineers do.
ARM products, not computers. Garbage to take selfies and post Instagram stories, to consume and facilitate consumption, not to create. ARM SoCs are excellent for that, because they've been expressly designed to be low-power, low-performance chips. They're trash at GP computing, by design.
Could that be changed? Of course, ARM didn't start out as some low power RISC design, but a full fledged PC CPU architecture. But that doesn't come for free. In fact, calling ARM "RISC" today is pushing the definition farther than it can take, it's a CISC/RISC mixture.
If this rumor is true, and I'm sure it is, we're going back to the PPC era Macs - because neither AMD nor Intel are going to sit by and churn out 4-5% IPC increase per generation forever.
Your posts are a mixture of wishful thinking and blind fanboyism, mixed with an unhealthy dose of inflated self-importance. I've been a Mac user since 1991, I've been in this party before.
No, but AMD has. And before that you could say the same thing about AMD. Intel will get its **** together, but creating integrated circuits has a lead time measured in years. But you would know that, mr. "I've designed CPUs".
Not really, no. Again, a facile and demonstrably wrong statement. Obviously ARM cannot run AMD64 code natively, but setting that aside, ARM cannot achieve performance for free. But you would know that, since you claim to have designed CPUs and are now posting on a Mac rumors forum. Like so many industry CPU engineers do.
ARM products, not computers. Garbage to take selfies and post Instagram stories, to consume and facilitate consumption, not to create. ARM SoCs are excellent for that, because they've been expressly designed to be low-power, low-performance chips. They're trash at GP computing, by design.
Could that be changed? Of course, ARM didn't start out as some low power RISC design, but a full fledged PC CPU architecture. But that doesn't come for free. In fact, calling ARM "RISC" today is pushing the definition farther than it can take, it's a CISC/RISC mixture.
If this rumor is true, and I'm sure it is, we're going back to the PPC era Macs - because neither AMD nor Intel are going to sit by and churn out 4-5% IPC increase per generation forever.
Your posts are a mixture of wishful thinking and blind fanboyism, mixed with an unhealthy dose of inflated self-importance. I've been a Mac user since 1991, I've been in this party before.
Last edited:
Given the way the world is going, I don’t think there will ever be another processor architecture after ARM. In twenty years every chip will be ARM. We’re already 80% of the way there. It’s like how every operating system in world is now Unix except for the lone holdout of Windows.
See yea, have fun with Linux. I heard the same things when Apple switched to intel.