Apple Testing ARM Based Mac Prototypes with Large Magic Trackpad?

[maflynn]

This is not going to be about the CPU. Yes, intel will be faster for some time. But they are not giving apple what they want.
Apple wants a fast cpu, with an amazing GPU, low power consumption, for a good price. Out of the 4, intel can deliver only 1.

Apple will probably want to do a retina macbook in a fanless design, with great battery life and competitive price. Intel chips don't allow for this design(afaik).

I think intel is delivery more then one on your list.
Fast CPU, amazing GPU, low power consumption (MBA getting 8 hours of battery)

They're probably charging a lot (or more then what apple wants to pay) so I'd say intel is delivering 3 out of the 4.

Where as if apple changed platforms they'd only get 2 out of the 4.
low power and a great price. You said it yourself intel is producing faster CPUs and their Iris Pro GPU is surprisingly very good as well.

It would seem to me they'd be going backwards using that logic.

 

[leman]

Apple wants a fast cpu, with an amazing GPU, low power consumption, for a good price. Out of the 4, intel can deliver only 1.

Intel can deliver fast CPU and amazing GPU, its also fairly low-power. I also think that ARM CPU performance is less of a concern here — if Apple can up the clock of their A7/A8 by 50%, it should be 'enough' for a consumer laptop. GPU is a real problem though. The PowerVR GPUs are simply no where near Intel solution right now and won't be any time soon...

 

[Steve121178]

This is not going to be about the CPU. Yes, intel will be faster for some time. But they are not giving apple what they want.
Apple wants a fast cpu, with an amazing GPU, low power consumption, for a good price. Out of the 4, intel can deliver only 1.

Apple will probably want to do a retina macbook in a fanless design, with great battery life and competitive price. Intel chips don't allow for this design(afaik).

Intel deliver all four. Anyone wanting more than what currently exists has unrealistic expectations.

Regarding the fanless design, this is what the low power Broadwell CPU's are for. These Broadwell CPU's are ideally suited to professional tablets (Surface) as well as thin & light notebooks, such as the MacBook Air and Ultrabooks.

 

[maflynn]

Another issue of completely switching platforms (as opposed to just introducing a low end model) is the ability to manufacture the chipset in quantity. Right now they're churning out a lot of A7s for their iPhone and iPads, can whom ever they contracted with handle an increased load?

One (out of many) reason apple dropped Motorola from PPC platform was their inability to meet the demand. They were unable to provide CPUs in quantities for apple. IBM fell into that issue early on with their G5s as well, but that I think was more of a teething issue as rolled out the new PPC design for Apple. Either way Apple was frustrated by the chip makers inability to provide CPUs in quantities. Intel's fabrication ability is astounding and they'll be moving away from that (assuming a complete platform shift)

 

[antonis]

For convincing s/w makers. Take a look at WP/Android and iOS. How apple convinced developers to write SW for their platform ? Well...because the apple ecosystem is the best for monetizing their product. They will do it again.

I don't think these 2 things are comparable. Intel has established an ecosystem long long ago in the desktop / workstation market. If apple isolate themselves (again) they are in big trouble (again). Mobile market was at a dead-end before iOS and iPhone. But PC market is nothing like that. Software houses are not going to be so nice with apple, porting their applications from intel (that they are going to develop on, anyway, because of PCs) just for apple's shake.

Apple might be an insanely rich company, but market-wise they are just a small spot on the map.

 

[quackers82]

If this is true please do not stop making the Intel range. I only moved to Mac in 2011 and i still come across problems of not being able to use some PPC programs, that were never available on Intel Mac and have to get the Windows version and use WINE or Pararalls another platform change would be a nightmare.

 

[gnasher729]

Why do so many people think Intel really cares about Apple? Apple Worldwide market share in computers is just 5%. Companies like HP, Dell, Lenovo, Asus etc. buy a lot more processors from Intel than Apple.

Google for "Intel pricelist". Then check what the processors in Apple laptops and desktops cost. Hint: They are not cheap. You won't find them in $300 laptops or desktops. Apple's percentage of Intel revenue, and more importantly of Intel profit, is a lot higher than five percent. And Macs are regarded as quality computers. Losing Mac processor sales would make a real dent into Intel's numbers and it would be an awful PR blow.

 

[jscooper22]

When they dropped the XServes they basically said "we're focusing on the consumer and have no real incentive -- or interest -- in keeping our business customers happy." If they drop Intel and therefore vmware/parallels/bootcamp, they're basically saying "go jump in a lake" to everyone but kitschy sole-proprietors.

The year they make this switch -- if they do -- is the year we'll need to, too. Unfortunate. We use Macs for a reason, but reality is what it is. It's bad enough we need to use Win for the few business-critical program we do. Going 100% Windows is going to suck, but will be necessary.

 

[gnasher729]

Yeah .. if somehow Apple stayed with PPC, we would have an awesome MacPro. The latest 12 core POWER8 CPU is ~2X faster than the Xeon in the current top end 12 core MacPro and is better at performance/watt as well. POWER8 is 190W TDP.

"IBM said that a base POWER Systems S-Class server would cost $7,973, but did not say what configuration or what specific system that cost was tied to.". So that's about $8,000 for the cheapest POWER 8 system that you can buy.

----------

Well, all you said is right, but only from a theoretical perspective.

I cannot write a 250-page document on my iPad. No iPad software supports cross-references or other advanced features. Yes, Word and even Pages may improve a million times and become even more powerful than their counterparts available for PC/Mac. But they may never evolve as well. It is a matter of software, but the software has to be written, and it takes several years of development for a piece of software to achieve such levels. There is no reference manager for the iPad as well, nor integration with Word. iOS can evolve and be capable of it. Of course it is a possibility. Theoretically, everything is possible.

You are confusing ARM and iOS. They are not the same thing. If I had a computer with ARM processor running MacOS X, the development effort to make my x86 software run on that ARM processor would be exactly ZERO.

----------

1. Regardless of performance, ARM is still a very specialised architecture, a far cry from the general purpose x86_64 processors and not suitable for desktop use.

2. ARM is still nowhere near the performance of the desktop or enterprise class Intel CPU's. People who claim it is are almost always comparing it to Atom, which is not "desktop class" or at the extreme low end of desktop class.

3. To be comparable, ARM will have to add many more extensions to graduate from a mobile CPU to a more generalised design and since it will need to virtualise x86, it needs to be as powerful as x86 and have a lot to spare (since we can no longer paravirtualise Windows and have to fully virtualise it)

1. Is nonsense.
2. ARM doesn't use 50+ Watt. However, it's no problem to build an 8 core ARM processor running at 3 GHz, eating the same energy as a desktop Intel processor, and being competitive with it. It defeats the purpose, but it's possible. On the other hand, it is highly unlikely that is something Apple wants to do.
3. Nonsense. And asking for virtualising x86 is absolute nonsense.
4. There are serious estimates that ARM will take about 20% of the server market by 2016. Not Apple's market, but just to give you something to think about.

 

[theluggage]

Another issue of completely switching platforms (as opposed to just introducing a low end model) is the ability to manufacture the chipset in quantity. Right now they're churning out a lot of A7s for their iPhone and iPads, can whom ever they contracted with handle an increased load?

So, Apple just pull a few wads of cash out from under their mattress and hire, buy or (if all else fails) build more chip fab capacity.

This is possibly the most significant way that ARM differs from Intel/IBM/Motorola - Apple can license processor components from ARM on a pick'n'mix basis, incorporate them into its own processor design (they've already bought in the expertise for this - the A7 is the result) and pay whoever the hell they like to make them. Heck, Intel would probably build A7s for them for the right price.

Plus, Apple have more cash and buying clout now than they did in 2005.

 

[lewismayell]

Yep trackpads on Windows laptops are almost universally terrible. In fact, before I switched to Mac I would never use a trackpad for anything for this season. I would also plug in a small travel USB mouse.

Yeah, I never expected I would actually use the Mac trackpad because of PCs, but sometimes I find myself using it over a mouse

 

[driftless]

I have read this thread with interest, along with a prior thread speculating about the rumored 12" retina, fanless Mac. I don't understand all of gnashing of teeth and threatening to leave the Apple ecosystem every time that Apple does what Apple does best: change the paradigm. I have a couple of predictions, based on nothing but a gut feel. They are:

1. The new Mac will be a premium product and priced accordingly. It will not be a glorified iPad.
2. It will be useful from day 1.​

In the past I have avoided MBA's because their screen and lack of power. I am hoping that this will be a great ultra-portable Mac.

I appreciate the technical posts in this thread. Thank you for the discussion.

 

[theluggage]

I don't think these 2 things are comparable. Intel has established an ecosystem long long ago in the desktop / workstation market.

...an ecosystem which is now shrinking, as mobile devices make consumers finally realise that they don't need MS Office to write letters or Photoshop to manage their snapshot collection, and with the increasing use of platform-independent development and high level languages.

Even at the workstation end, technologies like OpenCL (which Apple is pushing actively with the new Mac Pro) and high-level languages based on virtual machine bytecodes or just-in-time compilation mean that there is less and less need to write your high-performance code in lovingly hand-crafted assembler or use processor-specific code in C.

Photoshop 2020 could easily be a Javascript/HTML5 frontend with an OpenCL backend and an absolute minimum of native-code 'glue'.

If you want to support two or more of Windows, OS X, iOS, Android and Linux then differences in processor architecture are the absolute least of your worries...

 

[mateus]

Yeah, I never expected I would actually use the Mac trackpad because of PCs, but sometimes I find myself using it over a mouse

This is one of the (many) joys of using a Mac. I think most Mac users opt for the trackpad over a mouse, even for Photoshop work. And the reason the Magic Trackpad exists.

However, I've heard a number of people claim that a Windows 8 upgrade can turn trackpads on PC laptops into a Mac-like experience. Apparently all Windows systems prior to W8 had a serious flaw effecting trackpad input which has been totally redesigned for W8.

 

[n8mac]

You gotta love these threads, it really brings out the hypocrites. People say they want competition, then turn around and say "I will only compute if it runs on an intel with the option of Windows/x86", without realizing how much Microsoft and Intel have stifled computing in the last several decades. A lack of competition meant that Intel was able to get away with simply increasing the mhz of Pentiums fooling many people into thinking they were better than PPCs. A lack of competition allowed Microsoft to get away with releasing Vista having people thinking that their only option was to run XP until Win7 was released.

This idea that there is only one way to compute needs to die. I applaud apple for having the balls to take on such stale and backwards thinking.

 

[thunng8]

"IBM said that a base POWER Systems S-Class server would cost $7,973, but did not say what configuration or what specific system that cost was tied to.". So that's about $8,000 for the cheapest POWER 8 system that you can buy.

That is if you buy the server from IBM. There will hopefully much cheaper servers from third parties in the future. With Nvidia onboard as well, there will be systems that will incorporate multiple Nvidia gpus with power8 processors.

----------

Intel can deliver fast CPU and amazing GPU, its also fairly low-power. I also think that ARM CPU performance is less of a concern here — if Apple can up the clock of their A7/A8 by 50%, it should be 'enough' for a consumer laptop. GPU is a real problem though. The PowerVR GPUs are simply no where near Intel solution right now and won't be any time soon...

I am fairly sure that Apple is designing their custom Gpu at this very moment and may incorporate it in the A8. They have been hiring a huge amount of gpu designers in the last few years.

 

[antonis]

...an ecosystem which is now shrinking, as mobile devices make consumers finally realise that they don't need MS Office to write letters or Photoshop to manage their snapshot collection, and with the increasing use of platform-independent development and high level languages.

Even at the workstation end, technologies like OpenCL (which Apple is pushing actively with the new Mac Pro) and high-level languages based on virtual machine bytecodes or just-in-time compilation mean that there is less and less need to write your high-performance code in lovingly hand-crafted assembler or use processor-specific code in C.

Photoshop 2020 could easily be a Javascript/HTML5 frontend with an OpenCL backend and an absolute minimum of native-code 'glue'.

If you want to support two or more of Windows, OS X, iOS, Android and Linux then differences in processor architecture are the absolute least of your worries...

Yup, that's a most probable future snapshot of the development platforms (unless something else comes up out of nowhere in the meanwhile). But it's not there yet, and it'll take some more time to get there. Withdrawing now an entire desktop line from Intel and AMD it will definitely isolate apple.

Just saying that, as things are now, Apple is not so big in the computer market to trigger such changes and expect anyone to follow.

 

[holydude]

That is if you buy the server from IBM. There will hopefully much cheaper servers from third parties in the future. With Nvidia onboard as well, there will be systems that will incorporate multiple Nvidia gpus with power8 processors.

----------



I am fairly sure that Apple is designing their custom Gpu at this very moment and may incorporate it in the A8. They have been hiring a huge amount of gpu designers in the last few years.

Sorry but sparc/power market is dead. IBM and Oracle killed them with short term solutions and decisions.

 

[roadbloc]

Which of course is a joke on your part because I don't think you are that stupid to believe that Intel processors of the time where slower. The fact of the matter is that most users reported very good performance out of the box for those new machines. Why? Because of the vastly superior integer performance the Intel chips of the time hard.

Of course you can find a benchmark here or there to prove any point, but for most people Intel was an immediate win and frankly Intel went on a performance binge shortly after.

By the way most of the arguments here against ARM are directly or indirectly related to integer performance. This argument is only of partial value anymore though.

This isn't just about performance. Apple have stung me once with ditching older app compatibility (when Lion axed Rosetta for PowerPC apps). After slowly rebuilding my software collection for Intel processors, they're just going to turn around and do the same thing to me again, expecting to rebuy all my old software yet again? Even though the older software is perfect, it just doesn't run on their newer machines?

Nope. Say what you want about Windows, at least it is compatible with older apps.

 

[thunng8]

Sorry but sparc/power market is dead. IBM and Oracle killed them with short term solutions and decisions.

Power marketshare might be down, but at least IBM are trying with openpower. And technically and performance wise the latest power8 absolutely slaughters high end Xeons.

----------

This is one of the (many) joys of using a Mac. I think most Mac users opt for the trackpad over a mouse, even for Photoshop work. And the reason the Magic Trackpad exists.

However, I've heard a number of people claim that a Windows 8 upgrade can turn trackpads on PC laptops into a Mac-like experience. Apparently all Windows systems prior to W8 had a serious flaw effecting trackpad input which has been totally redesigned for W8.

I doubt that. I've used trackpads with windows 8 and they seem as bad as they were on windows 7.

 

[Rocketman]

This seems aimed at Chrome books, where the user does not have an attachment to an ecosystem. Apple could offer the relatively low cost of Chrome books without reducing its margins, and with an optional ecosystem in place for media and of course a new ecosystem for ARM based productivity apps that are keyboard and track pad aware.

Certainly a 2-4 chip ARM system would have both higher benchmark performance and also lower chip costs than a low power Intel chip. Intel is pricey.

I for one would like to see Apple continue to differentiate UI with OS's. OSX(1-2xX86), iOS (1xARM), AppleTV (1xARM), iOS in the car (1xARM), now ARM laptop (2-4xARM). It follows Apple's statements that they do not intend to converge the OS code bases.

Rocketman

 

[Val-kyrie]

I remember the Intel transition well. Back then, I was excited for the transition because I was coming from Windows and the ability to emulate and/or dual-boot Windows made the Mac appealing.

The appeal of the Intel transition was what it offered to Windows users, some of whom switched. Unless MS announces a switch to an ARM-based OS in the near future, Apple will have to give serious consideration to the impact of this transition on their customer base.

Another problem with a transition to ARM is in what it will offer to consumers. Better performance per watt without a noticeable performance hit? lower cost? Is this significantly better than Intel? behind Intel? on par?

The other problem for Apple will be software segmentation and consumer confusion. As others have pointed out, the Surface RT was a dismal failure compared to the Surface Pro. How will Apple explain software segmentation to the average consumer? What happens when someone buys an ARM-based system and finds out post-purchase that they cannot run a particular program because they did not purchase an Intel-based machine?

Is this move being contemplated because of AMD's Skybridge project (i.e., the co-development of pin-compatible ARM and X86 chips)?

Apple is nowhere near the level of performance Intel is and it never will be. Why? Apple's focus is on hardware. Intel's focus is on CPU's and chipsets. Ya you forgot about chipsets. There's not a chipset in the world that will support the ARM processor. You will be stupid in thinking that Intel's chipset will support ARM.

Computers have chipsets because it's a lot more complicated then tablets. It has multiple IO's, a BIOS, Multiple PCI-e lanes, and CPU to support where the tablet doesn't.

The real reason is because there's no chipset that supports Apple's ARM CPU. Apple is stuck with Intel.

You raise a good point about chipsets and I/O for ARM-based Macs. I don't know enough about chipset design to comment, but I am curious how Apple would handle the need for I/O on ARM-based Macs. What about USB 3 or 3.1? HDMI? Before someone chimes in about TB, there needs to be an affordable TB docking station! For the cost of a TB docking station at $200-$300 (none of which do it all), Apple might as well stay with Intel.

I think the real motivation to do this is because Intel doesn't have enough competition from AMD to release new chips on time. They keep delaying them and Apple is sick of it. The Broadwell architecture is coming in December at the latest. It'll be about a year late. Delayed at least twice. Apple's going to miss the back to school season because Intel knows people will buy their chips anyway because AMD isn't releasing anything competitive yet. That must make them mad. On top of that, the basic math is that Apple can crank up clock speed, core count and redesign their architecture enough to soundly beat Intel chips at the same power consumption. It's probably not about power consumption. Just performance and Intel hurting Apple's bottom line. Nobody should be happy that their Mac costs $300-$400 more because intel charges an arm and a leg (no punn intended). Compare that to maybe $40-$50 for an ARM chip.

Hopefully any decision regarding ARM-based Macs will be made based on a logical response to the overall market picture and not on an emotional reaction to delayed chips.

I wonder if Intel's statement that it is willing to fab ARM chips has anything to do with the rumblings in this rumor?

 

[Ezhik]

When they dropped the XServes they basically said "we're focusing on the consumer and have no real incentive -- or interest -- in keeping our business customers happy." If they drop Intel and therefore vmware/parallels/bootcamp, they're basically saying "go jump in a lake" to everyone but kitschy sole-proprietors.

The year they make this switch -- if they do -- is the year we'll need to, too. Unfortunate. We use Macs for a reason, but reality is what it is. It's bad enough we need to use Win for the few business-critical program we do. Going 100% Windows is going to suck, but will be necessary.

This switch does not seem likely to me. Apple would create a schism between ARM and x86 Macs - and this confusion is definitely not something that Apple does.

So if they do go to ARM, they will pretty much lose their pro markets - nobody wants an ARM Mac Pro, that I'm sure of.

 

[lewismayell]

This is one of the (many) joys of using a Mac. I think most Mac users opt for the trackpad over a mouse, even for Photoshop work. And the reason the Magic Trackpad exists.

However, I've heard a number of people claim that a Windows 8 upgrade can turn trackpads on PC laptops into a Mac-like experience. Apparently all Windows systems prior to W8 had a serious flaw effecting trackpad input which has been totally redesigned for W8.

Would you say, that for MacBook on a stand used a Magic Trackpad would be a good idea or a waste of money, I already have again mouse, unlike a few other people I can't really flaw it, works fine for me, but I'm wondering how much more, if at all, intuitive the trackpad would be

 

[2457282]

It may happen in the future, but we aren't close to have similar performance between Apple's A series iOS chips and Intel's mainline i3-i7 chips - it'd be too big a gulf to bridge at this point.

Give it another 5 - 10 years though…at this point I think it'd be way too much of a sacrifice in performance to work for the Mac brand. JMHO..

In previous rumors, I predicted that it would be possible either with the A8 or A9 chip. The issue is not going to be performance of the CPU, but the issue of the OS. If they use iOS, it must allow for multiple simultaneous windows. Frankly if they do this then the iPad maxi (or whatever they call it) with a keyboard cover, it replaces the laptop. In addition, since the mac line of laptop is a small percentage of the business for Apple and shrinking, I see this as a real possibility.

Having said all that, my guess is that apple would not do this at all. Instead, they will keep the Mac Intel based and then just continue to develop the iPad in terms of size and in terms of the OS. Over time, the mac sales will continue to dwindle and the iPad sales take over. This also gives what few software titles that are not on the iPad time to convert. It also gives apple time to improve the cloud integration further, which is the only other major issue other than multiple windows. Finally, the cell providers are finally getting there with LTE and plans that give enough data volumes to allow this to all happen. Therefore, I stick with my prediction that by the A9, there is absolutely no reason, IMO. That we cannot have fully capable iPads replacing macs, even in the workplace – it requires iOS changes for multiple windows, and better cloud integration. If apple can do this and keep the cost under $1000, I can see corporations expanding the BYOD policy to include computers that will save them significant capital investment and would be a huge issues for Lenovo and Dell and HP which are still primarily Desktop and Laptop based.