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.
ARM based MacBook Pro discussion [merged]
- Thread starter Omega Mac
- Start date
-
- Tags
- arm intel macbook pro macos sierra
- Sort by reaction score
I see. Makes sense. I use plenty of VMs myself but also commercial software so no way out at the moment ?
Yeah! I do that all the time!
Precisely what I imagine an Apple II user with Visicalc told one of the first Mac buyers
And they'd be absolutely correct. My workplaces gives us systems that are one to two years old and doesn't allow us to run on the latest operating systems until our security and other software are validated. If the new system doesn't run your application, then it's useless.
Agreed, if the new system doesn’t run your application then it’s useless to you. If other people can use it to make a living, well, good for them, but still useless for those that can’t.
This is Jean-Louis Gassee's take on ARM based Mac: https://mondaynote.com/arm-ed-mac-we-have-an-answer-545a20419a46
Enjoy.
Enjoy.
Honestly they better ditch Intel. The 4800 and 4900 CPUs from AMD are destroying Intel laptops in performance and consumption, requiring much less cooling... Just to put in perspective, (IIRC) the 4900HS is beating the 9900K and 3700X in some situations. It would be amazing to have MB + MBA ARM first years (so software devs start compiling apps, and everything gets compatible unless Catalyst works flawless from day 0) and AMD MBPs and iMacs. Won't happen tho, Apple won't mess with getting different x86 CPUs when in just 2-3 years probably the whole Macbook line will be ARM.
I don't really get his point, he basically comes to say that since Apple is making iPad a real computer there's no reason to move Macs from Intel CPUs. Tha makes no sense because Apple wants to sell both iPads and Macs, not only iPads that cost way less than laptops. That also contradicts Ming Chi Kuo rumors (who usually hits the nail) of ARM Macs. And to end with, having mouse support doesn't make it a computer with a full desktop OS, that will be left to Macs and iPads will be left as an occasional substitute and mostly as a complementary tool.
Last edited:
J-L Gassee in general says that Apple won't move Macs from x86, because it is completely unfeasible from investment perspective.
The biggest problem that people do not understand is ARM is fast ONLY in lightly threaded workloads. When you get something that has to scale, something heavily threaded, it chokes performance. Latest Graviton v2 ARM Chip, from Amazon is faster than anything in Single Threaded workloads, or lightly threaded, but those 32 ARM cores, clocked at 2.5 GHz completely choke when they have to scale.
They are 40% of performance of AMD Naples, which is last generation Server chip from AMD, also with 32 cores.
And in single threaded workloads, Graviton v2 is as fast as Apple A12! The same problem affects every, single ARM chip, out there, because its the nature of the architecture.
The main difference between Mobile stuff, and Desktop stuff is the software complexity, and how lightly or heavily threaded software is.
You cannot have BOTH worlds on Mobile. The reason why iOS is so fast is because majority of workloads are lightly threaded(for battery life purposes). Switching Macs to ARM would require MASSIVE redesign of Apple Arm architecture for higher scalability, which is not a given, or massive redesign of the software on Mac, to fit lightly threaded workloads, for more performance, which is also not a given, because it requires time, and money, and not all workloads can be redesigned to fit specific software design guidelines. There is a reason why x86 is considered General Purpose architecture. Because it has to fit ever, single glove, out there. And redesigning ARM for higher scalability with heavier threading means more power consumption. And here is where all of the benefits of ARM go away. So what is the point?
Could you technically explain exactly what in the ARM architecture prevents it from scaling? Beause stating that without any argumentation or logic behind is like saying x86 scales horribly because Intel +8core CPUs are monolithic toasters that make its power consumption unbearable for laptops (that'd be cherripicking to prove something, so you get my point). One specific case doesn't make a law. AMD has overcome the scaling problem with Infinity Fabric, it's not like there aren't solutions/alternatives for architecture bottlenecks.
You also seem to forget 2 important things:
1. If you design your own chip you can add specific hardware modules (like NPU for AI, or VPU/IPUs for video and image processing) that are way more powerful at that specific task than general purpose cores. So you don't need to have a big amount of cores, specially in laptops/tablets.
2. We are far yet to see ARM desktops, ARM will be used in laptops wich aren't intended to substitute powerful workstations where high core and threat counts matter. The scaling problem becomes negligible on laptops the moment an iPad pro is already equally or more powerful than a MBP.
Sorry but what comes after that phrase is just wording and opinions, not facts. Ironically the new AMD laptop chips (the x86 that you defend) proves you wrong on your quote: the new 4900HS performs like a desktop (among 3700X and 9900K) having a power envelope of a mobile chip, that's why it's put on a laptop.
For what you say next: all your points are that you need lots of money to redesign an ARM arch, that it would add more power consumption in computers, and that it makes no sense. Those are really easy to answer: Apple has all the money they want to put into R&D. Not only that, they already spent it on a full ARM design which gave great results and that (responding your 2nd point) performs like a Mac consuming way less. Taking the main workload of designing a chip that already does like laptop x86 competitors and improving it is a breeze compared of what Apple CPU designers have already done. The main point of doing it is not depending on any 3rd party vendor, and specially having more benefit margins.
You also seem to be implying x86 is the only general purpose architecture beause it "has to fit ever, single glove, out there". Well there's news for you: almost any actual type of processor has all the units needed to compute any task, but that doesn't mean it will be good at it. That's why GPUs, NPUs, and specific hardware in SoCs exists even you can play Crysis only using the CPU (in the AMD 64c Rome for example). People seem not to realize that ARM and x86 are basically the same, the main difference is one is full RISC and the other just has a CISC decoder that translates complex instructions into reduced ones, but the rest of the pipeline and operations are RISC. And all the "ARM isn't good enough for heavy loads" myth comes because until now it was only designed and used in mobile stuff, but there's nothing in it's architecture preventing an ARM chip to be designed for high performance computation.
Also keep in mind Apple might as well go ARM only for low-end, low-perf MacBook / MacBook Air while maintaining its x86 architecture on higher-end Pros. That would be potentially messy though.
I tend to favour that hypothesis, too. Prosser says pro apps support may be coming to the iPad.
Looks like "laptopisation of iPads" is more likely than "iPadisation" of the MB lineup.
Looks like "laptopisation of iPads" is more likely than "iPadisation" of the MB lineup.
When MacOS runs on ARM, who don't they just put it on an iPad, as soon as it is connected to the Magic Keyboard?
Why not go the other way, make macOS more like iOS and in many respects that's been occurring. macOS is even more locked down, it has the ability to run iOS apps.
to me MacOS on iPad makes more sense at this point, because the MacBooks do not have touch. I do not think Apple will merge the two anytime soon, so having the choice on an iPad depending on the situation would be awesome.
That means Apple would need either to make MacOS work on both ARM and x86 systems or to develop two different versions for each. Not sure they would want to do that.
Technically iOS is macOS on some level. I'm not sure we need macOS on iPad as much as using the interface that best suits the device. Having XCode is a game changer.
Apologies if this has already been asked but will moving to ARM based processors remove bootcamp as an option?
There's already an ARM Win10, the only reason to remove bootcamp option would be if Windows remained x86 exclusive.
Look, from Apple's perspective is really easy: "we make computers and tablets, tablets complement computers so we can sell both instead of only one". Thus putting macOS or a full desktop OS in an iPad Pro won't happen. You won't see people using iPad to code apps for iPad/iPhone, they will make you use a Mac so you have to buy a Mac.
That's how business work, and that's why they intend (with Catalyst project): to allow Macs to natively execute x86 and ARM, thus being a real computer and then having iPad Pro as a tool/complement (even tho you can occasionally use it to avoid using the laptop/tower). You as a user really want it so bad, it would be amazing of course, but it makes no sense for Apple to allow you to do everything you can do in a Mac in an iPad so you can spend only 800 bucks instead of 800 + 1300 or whatever you pick (usually more) so it won't happen.
Yes. Unquestionably. If Macs move to ARM it is the effective end of Bootcamp as well as VMware fusion and Parallels. Docker Desktop as well.
The fact that there is a Windows for ARM of some sort solves precisely 0% of people’s need for the tools listed above. Nobody uses boot camp because they want to run Windows. They use boot camp because they want to run a Windows application. And those applications that people want to run are x86. That’s not changing any time soon.
So, I google "docker desktop arm" and I found a billion websites documenting how to cross-compile or build multi-arch docker desktop images. The harder Apple hits the message home that ARM is here to stay, the faster the number of images for ARM will ramp up.
As for virtualisation software, it of course depends. There seems to be Xen for ARM, but others would have to be rewritten, which is a big task. Qemu is there as well. But nothing serious.
That is the problem indeed. There's no commercial software for the platform. But docker and other FOSS software? It's just a matter of time and it'll be all there.
It's a nice thought, but doesn't seem realistic to me. Eventually, probably, someday the arm container ecosystem will achieve parity or even eclipse the x86 container ecosystem. But for now ARM is just a fringe platform for containers and cloud infrastructure. People develop, test, and use containers on macOS only to the degree that they need to interoperate with existing infrastructure that is x86. In this area it's important for macOS users to be compatible with the world, not for the world to be compatible with macOS. Apple have no leverage to move the ecosystem away towards ARM. If that's going to happen it's going to come almost entirely from Google, Microsoft, and Amazon. It's going to be influenced strongly by the body of virtual private server providers like IBM/Softlayer and Linode and others in that market. It's going to be minority (but loudly) influenced by players like Synology and QNAP and tons of geeks who want to run a Plex server somewhere. Apple are on the wrong end of the equation to move the needle here.
For me, personally I can't exactly tell my employer that they need to move all our Kubernetes clusters to ARM just because I want to buy a new MacBook.
You don't seem to realize that compiling for x86 or ARM is just a matter of using one toolchain or another, your IDE does everything for you. The only difference for a developer is which primitives he'll use in case he goes for a high optimization level (most developers won't), i.e: using AVX/SSE or using NEON/SVE. And even then, if that was the most minimum hassle for a developer, there are many libraries and frameworks that already wrap those primitives for you so you don't even have to specifically code for an specific architecture/OS, you just use "SIMD" wrapper and the framework does the rest of the job.
And for the sake for it, again, Apple won't kill x86 from one year to another, they are working on Catalyst (a Rosetta like, but expcting it to be better) to execute both x86 and ARM instructions in the same OS.
With all respect, you are not understanding what I'm saying. I'm not talking about local desktop application development. Whether or not an application can be compiled for one or both platforms doesn't change the fact that a hypothetical ARM macOS machine will completely lose the ability to virtualize x86 containers and operating systems. That's fatal to the workflow for me and many thousands of other developers.
If I can't run and produce "regular" docker containers locally I can't do my work.
Emulating x86 on ARM (a rosetta-style solution) is a pipe dream. There's no way to achieve reasonable performance. The only reason Rosetta was usable is because the incoming Intel CPUs were significantly faster than the PPC chips they were replacing, so users could afford the inefficiency without it being too painful. ARM and Intel are much closer for performance, so the overhead of emulation isn't tenable. Even if you could manage to emulate x86 on ARM in a performant way, you still haven't solved the virtualization problem.
edit to add: Catalyst is not a rosetta-like solution. It doesn't involve transcoding or emulation at all. It's just a set of hybrid macOS/iOS libraries so that application developers can target both operating systems with a single application/UI. When a developer is using catalyst they're compiling independently for both platforms and producing fat binaries or separate binaries. Think of it as a UI unification toolkit, not an emulator. it does not enable an ARM machine to run x86 binaries or the opposite.
Last edited:
