Intel Confirms: Macs to switch to ARM by 2020.

[Mooshi]

Would you spend 2.5K on an ipad with an integrated keyboard though?

[Ashleyyyy]

Just now, VegetableStu said:

question would be how damned custom would Apple's ARM chip be o_o or how much hardware Apple will be opening up to windows

probably enough for Windows to function. 

[Mira Yurizaki]

I'm going to chime in on the whole instruction set architecture (ISA) thing. Which ARM and RISC V are.

 

The ISA has little to no bearing in whether or not a processor performs better. It depends on the implementation of the processor. Just because it's ARM doesn't mean it performs better or worse than something else. Otherwise it doesn't make sense that the Vortex core in Apple's A12 beats the snot out of a Cortex-A76 when they both use ARM. And we can look at x86 examples too, like how the Athlon/Athlon XP beat the Pentium 4 or how the second-gen Core processors beat the FX series.

 

The ISA is merely a blueprint on how to interpret the software bits and bytes. And while I wish there was more apples to apples comparison and data, considering benchmarks between the iPad Pro and Apple's laptops, I'm willing to believe Apple's implementation of ARM can go toe-to-toe with Intel's and AMD's implementation of x86. We just haven't seen Apple's architecture really bare its fangs in a high performance setting

 

The only thing that would make this transition hairy is anyone who doesn't have an iOS compatible version of their app. I don't think Apple could make an x86 emulator like they did making a PPC emulator during the Intel transition considering Intel tried to sue Microsoft for doing it. I mean, Apple has a lot of money but my armchair business sense would rather put that money to helping others transition only rather than fight a court battle and make an emulator on top of the transition

[Donut417]

I wonder if they are moving this direction to make it easier to add LTE connections to their computers. Because to my understanding ARM chips can have the radios built in making it easier to have this capability. Plus you dont see too many Intel or AMD devices with LTE connections currently. People want all of their devices connected to the internet all the time and be able to take them everywhere. 

 

 

16 hours ago, VegetableStu said:

wonder if they're still going to support booting windows on an ARM Mac by then ._.

Microsoft is working on a version of Windows that runs on Qualcomm chips, so maybe.

 

16 hours ago, VegetableStu said:

also don't think they'll be replacing the entire mac lineup

I have to agree. I have a feeling the lower end/cost devices might get ARM. I have a feeling their more "Professional" line will still have Intel chips in them. 

 

16 hours ago, Blademaster91 said:

Unless there would be some kind of emulation to allow programming for x86,though that would just be slower than actual x86 hardware

Microsoft has been experimenting with this with their ARM version of Windows 10. Plus they also do a bit of emulation with the Xbox One, allowing people to play 360 games. So its all possible some form of emulation can be used. 

[Mira Yurizaki]

15 minutes ago, Donut417 said:

I wonder if they are moving this direction to make it easier to add LTE connections to their computers. Because to my understanding ARM chips can have the radios built in making it easier to have this capability.

There's nothing stopping anyone from integrating a radio into their chip. It's just for most applications of ARM are for embedded systems where integrating everything as much as possible is more economical than having the ability for system builders to add it on separately.

 

15 minutes ago, Donut417 said:

Plus they also do a bit of emulation with the Xbox One, allowing people to play 360 games. So its all possible some form of emulation can be used. 

I think in this case, they're recompiling the main binary to x86. The graphics and audio on the 360 uses the same API calls as the XB1 so there's no need to emulate those. The only thing that's emulated per se is the system behavior.

[Drak3]

After reading through the sources, I can't find anywhere (other than Tom's Hardware's clickbait title) that Intel confirmed anything. At best, just Bloomberg linking back to an article from last year where they claim that 'people familiar with Apple's plans' are confirming that some macs will be transitioning to ARM.

 

Which is basically the same rumors we've been hearing for years.

 

 

And the app integration to enable iOS apps to be Mac compatible was also known for about a year. The App store on OSX isn't exactly fruitful.

 

5 hours ago, captain_to_fire said:

macOS Photos > UWP Photos

I disagree. I'd rather be hit in the head with a sledge hammer than use OSX's photo app.

 

5 hours ago, captain_to_fire said:

HomeKit > does Windows 10 have any IoT control? 

Realistically speaking, is having a control panel for that set of glaring security holes worth it?

5 hours ago, captain_to_fire said:

System Preferences > Settings + Control Panel 

The common 'system preferences' menu shared amongst Unix and Linux is basically just the settings panel from Windows, except lacking more settings. I wouldn't constitute a win for OSX on this either.

10 hours ago, JoostinOnline said:

Are you sure about that? Thunderbolt isn't supported by any AMD products that I know of.

Probably because TB3 makes zero sense on most machines running Ryzen, and all machines running Threadripper or EPYC.

 

[Donut417]

1 minute ago, Drak3 said:

Probably because TB3 makes zero sense on most machines running Ryzen,

Or because Thunderbolt is an Intel Technology and Intel hasn't licensed it to AMD yet. 

[Drak3]

1 minute ago, Donut417 said:

Or because Thunderbolt is an Intel Technology and Intel hasn't licensed it to AMD yet. 

Thunderbolt has been royalty free for awhile now, and the only thing stopping AiCs is firmware.

[Hunter259]

2 hours ago, Mira Yurizaki said:

I'm going to chime in on the whole instruction set architecture (ISA) thing. Which ARM and RISC V are.

 

The ISA has little to no bearing in whether or not a processor performs better. It depends on the implementation of the processor. Just because it's ARM doesn't mean it performs better or worse than something else. Otherwise it doesn't make sense that the Vortex core in Apple's A12 beats the snot out of a Cortex-A76 when they both use ARM. And we can look at x86 examples too, like how the Athlon/Athlon XP beat the Pentium 4 or how the second-gen Core processors beat the FX series.

 

The ISA is merely a blueprint on how to interpret the software bits and bytes. And while I wish there was more apples to apples comparison and data, considering benchmarks between the iPad Pro and Apple's laptops, I'm willing to believe Apple's implementation of ARM can go toe-to-toe with Intel's and AMD's implementation of x86. We just haven't seen Apple's architecture really bare its fangs in a high performance setting

 

The only thing that would make this transition hairy is anyone who doesn't have an iOS compatible version of their app. I don't think Apple could make an x86 emulator like they did making a PPC emulator during the Intel transition considering Intel tried to sue Microsoft for doing it. I mean, Apple has a lot of money but my armchair business sense would rather put that money to helping others transition only rather than fight a court battle and make an emulator on top of the transition

Thank you. I was about to have to correct people on this again as every single post that involves ARM vs x86 devolves into "RISC vs CISC". The only advantages that might occur is that CISC would theoretically have better memory optimization while RISC doesn't need as long of a pipeline as modern x86 chips run what is essentially RISC instructions in the end.

[mr moose]

10 hours ago, ScratchCat said:

ARM's N1 and E1 platforms are approaching the performance of x86 systems. If this design launches in the next year or two there will not be time for x86 to leap forward in performance to maintain it's lead on ARM for long.

Source

 

Yes, there are already a few super computers using ARM,  I was not saying they don't exist, I said they will likely take a decade to to become a viable workstation/ server that will effect Intel or AMD's business.

 

9 hours ago, LAwLz said:

Depends on what platform you're targeting.

 

ARM computer = Good for developing ARM software on.

x86 computer = Good for developing x86 software on.

 

Linus Torvalds had a mail conversation about it and the problems "ARM powered cloud" will have, like a week ago and it's a really interesting read.

Here is a quote from it:

  Reveal hidden contents

Guys, do you really not understand why x86 took over the server market?

It wasn't just all price. It was literally this "develop at home" issue. Thousands of small companies ended up having random small internal workloads where it was easy to just get a random whitebox PC and run some silly small thing on it yourself. Then as the workload expanded, it became a "real server". And then once that thing expanded, suddenly it made a whole lot of sense to let somebody else manage the hardware and hosting, and the cloud took over.

Do you really not understand? This isn't rocket science. This isn't some made up story. This is literally what happened, and what killed all the RISC vendors, and made x86 be the undisputed king of the hill of servers, to the point where everybody else is just a rounding error. Something that sounded entirely fictional a couple of decades ago.

Without a development platform, ARM in the server space is never going to make it. Trying to sell a 64-bit "hyperscaling" model is idiotic, when you don't have customers and you don't have workloads because you never sold the small cheap box that got the whole market started in the first place.

The price advantage of ARM will never be there for ARM servers unless you get enough volume to make up for the absolutely huge advantage in server volume that Intel has right now. Being a smaller die with cheaper NRE doesn't matter one whit, when you can't make up for the development costs in volume. Look at every ARM server offering so far: they were not only slower, they were more expensive!

And the power advantage is still largely theoretical and doesn't show very much on a system level anyway, and is also entirely irrelevant if people end up willing to pay more for an x86 box simply because it's what they developed their load on.

Which leaves absolutely no real advantage to ARM.

This is basic economics.

And the only way that changes is if you end up saying "look, you can deploy more cheaply on an ARM box, and here's the development box you can do your work on".

 

 

I assume you mean hardware acceleration for video encoding since you brought up Intel's iGPU?

We already have hardware accelerated video encoding and decoding in ARM chips. Intel does not have exclusivity for that feature.

In fact, there aren't any Intel/AMD/Nvidia hardware acceleration features you won't find in ARM chips these days, apart from maybe ray tracing (which isn't widely used anyway).

I saw bits of that article and a response to it by some normy on youtube.  I have to say while his premise is sound I think it's backwards.  Even if they manage to get mainstream desktops all running on ARM, and end user software developed for ARM.  A lot of the corporate world that runs windows on x86 is not going to want to change just because they can, a really large incentive (along the lines of massive dollars) is going to be needed before companies change their back ends.  

 

8 hours ago, Lukyp said:

Depends on how old is that statement and what he meant by that, as in the past when PPC Macs were out and Linux was not something considerably decent on desktop, an x86 nix meant a lot even especially for programming, performance, and since x86 was the most wide architecture in the world... It was definitely a thing

 

Today since the arm importance becauz phonez, compiling for arm from x86 became a thing to but definitely this would be faster and easy to test on an arm native architecture, so not really useless for programming. You can still build x86 from arm but vice-versa you got the same issues

 

Compiling arm android for example takes an huge amount of time, and still since there are not any decent "arm Dev boxes" as Linus calls that can be even worse than x86, then we can only hope in the future for better implementations to be made, that's what I hope Apple does

Very recent statement by very experienced software engineers. 

4 hours ago, LAwLz said:

1) Gonna need a source on that. 

2) Even if that's true, super computers are a very, very VERY niche market and I don't think it's that relevant. More reasonably poweredservers are far more important. Servers that more than a handful of people use. 

There are a few but it's not major.  I mentioned Astra before which is at 205 in the top 500, however as you say, these are (as they are in the cloud services) niche and new, development and uptake is not that fast.   Also most of the articles I have read suggest that the software side is whats holding them back from mainstream adoption. 

 

 

[Guest]

1 hour ago, mr moose said:

Very recent statement by very experienced software engineers.

There will definitely be advantages on mobile development in that case and vice versa compromises on x86 deployment, nothing else I should particularly say here

Spoiler

and for other workloads depends on how support companies are gonna put on software support for anything else and if apple finds a way to support x86 apps on it other than the one already mentioned in this thread, the marzipan thingy looks like a bytecode and in theory is cool

 

[Guest]

1 minute ago, Schnoz said:

ARM? Why not Leg?

I'm crying...

[Lurick]

45 minutes ago, Schnoz said:

ARM? Why not Leg?

Because then it will run away >.>

[Curufinwe_wins]

21 hours ago, mr moose said:

I know quite a few software type people who like mac because it is both  a Unix like system and runs on x86.   I wonder if removing the x86 component makes them more or less useless for programing. 

This is pretty important.

 

Obviously it's a small user base, but the scientific HPC computing group would much rather drop MacOS than swap away from x86. Particularly with all the legacy fortran and older codes that are never going to be ported to arm, but are still being used and semi-supported.

[mr moose]

21 minutes ago, Curufinwe_wins said:

This is pretty important.

 

Obviously it's a small user base, but the scientific HPC computing group would much rather drop MacOS than swap away from x86. Particularly with all the legacy fortran and older codes that are never going to be ported to arm, but are still being used and semi-supported.

And from my limited understanding their are also quite a number of mainframes still running Unix type stuff from the very early days (thinking the banking sector) in large systems that cannot and likely will not be changed anytime in the foreseeable future. 

[Mira Yurizaki]

I'm sure the server/workstation market isn't going to care about this and will fight tooth and nail to keep x86 for a while. But the consumer likely isn't going to care as long as their daily apps and tasks can be done. Which doing this on say the normal MacBook first would make sense. Which if Apple can build up a base there means the rest may come.

[Chett_Manly]

I voted no, but then I wouldn't buy a Mac as it is as I am of the PC Master Race, and gaming is non negotiable for my daily driver.

 

That said I could see myself buying an ARM based laptop, where portability is a primary concern. So an ARM based MacBook Pro would be just as enticing to me as an Intel based MBP.

[Ashleyyyy]

38 minutes ago, Curufinwe_wins said:

Obviously it's a small user base, but the scientific HPC computing group would much rather drop MacOS than swap away from x86. Particularly with all the legacy fortran and older codes that are never going to be ported to arm, but are still being used and semi-supported.

maybe Apple will do a similar thing as they did when they moved from PowerPC to Intel, and use universal binaries. 

 

they might put ARM in the laptops, because portability, and have X86 in the desktops still. that would make sense. and use universal binaries so that both architectures are supported. 

[mr moose]

9 minutes ago, Mira Yurizaki said:

I'm sure the server/workstation market isn't going to care about this and will fight tooth and nail to keep x86 for a while. But the consumer likely isn't going to care as long as their daily apps and tasks can be done. Which doing this on say the normal MacBook first would make sense. Which if Apple can build up a base there means the rest may come.

 

I certainly can't see any company that runs x86 or relies on servers that are already x86 developing for ARM.  No one changes for the sake of change and the financial incentive has to be huge to take the risks usually involved in making such changes. 

[Guest]

20 minutes ago, mr moose said:

And from my limited understanding their are also quite a number of mainframes still running Unix type stuff from the very early days (thinking the banking sector) in large systems that cannot and likely will not be changed anytime in the foreseeable future. 

And even not every Unix in that era run on x86 processors, especially the IBM flavors

 

I know a person working for banks (I already told you this somewhere) and someone was using a an ibm unix on ppc, relatively old since they updated their architecture so the program could still run today, but their lineup never forced to replace x86

[mr moose]

7 minutes ago, Lukyp said:

And even not every Unix in that era run on x86 processors, especially the IBM flavors

 

I know a person working for banks (I already told you this somewhere) and someone was using a an ibm unix on ppc, relatively old since they updated their architecture so the program could still run today, but their lineup never forced to replace x86

All the more evidence pointing toward not seeing a major change from x86 to ARM in the near future.

[Guest]

11 minutes ago, mr moose said:

All the more evidence pointing toward not seeing a major change from x86 to ARM in the near future.

Imo it just doesn't make sense for everything currently

 

I don't think people would be getting that much improved battery life for example for using arm instead of x86, and the effort and money required to transitioning would be excessive considering how cheap and power efficient some chips are (and more will be in the future), but I still hope for Apple to make their move and see what happen, I may be wrong

 

Personally I'm willing to try it, also because I'm interested in arm development and would be better for me to test things

 

[descendency]

I use my 2016 MBP for virtualization tasks, so the best case scenario is losing a lot of performance in virtual workloads - I'd have to switch back to Windows for an upgrade. In the end, I think my use case is super niche, but it affects me personally, so - I'm not for them switching. Ultimately, I've been saying it would happen for a long time, though. It felt inevitable. 

[Uttamattamakin]

I wouldn't do this because I and other academics (I teach college and do theoretical physics research) have workflows that depend on a mix of really old software and really new (often custom coded that day in Matlab) software.  We know how X86 works, we know it's failure modes, we know how it is stable and unstable.  Furthermore, there is a HUGE installed base of such computers all over the place.  Macintosh had a certain Niche in the professional community.  It was a full blown X 86 Unix environment that would just work out of the box.  It allowed for natively running many X11 tools right on the same OS as a full blown version of office 365. To get that on PC takes having Linux and windows installed with one or the other virtualized.

 

Apple fans who only want to use their certified Apple products with eachother will be just fine. 

On the plus side... maybe now we'll get a proper Macintosh computer in a convertible tablet form factor.   IF that was all that an ARM based mac was that would be great.  Call it a Mac Pad or something.   I wouldn't trade a Surface Pro for it but it would give Mac Users a real option.  

[Ashleyyyy]

2 hours ago, descendency said:

I use my 2016 MBP for virtualization tasks, so the best case scenario is losing a lot of performance in virtual workloads - I'd have to switch back to Windows for an upgrade. In the end, I think my use case is super niche, but it affects me personally, so - I'm not for them switching. Ultimately, I've been saying it would happen for a long time, though. It felt inevitable. 

well if all of the OS's you virtualise support ARM then it shouldn't be an issue.