Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
I think while Apple may be experimenting with this idea, in the end because Intel made recent significant improvements in lowering the power consumption of CPU's and motherboard chipsets, Apple may continue to use Intel CPU's with the MacBook Air and MacBook Pro laptops. And because Apple buys so many Intel CPU's, they could get the latest desktop/laptop CPU's at reasonable prices.
 
I like the idea that Apple wont have to wait for Intel, its one more thing that Apple can control. I'd certainly buy one of these ARM based Macbooks.
I'm no expert, but most of my lecturers at university claim that ARM is a much better architecture, im sure somebody will probably argue this though #
 
Its a different version of windows because its on a different chipset. Today's applications for OSX would not work on an ARM based Mac because they're not compiled of run on a different CPU. OSX would need to an emulator, much like they did when they went to PPC to Intel.

Microsoft did not create a different windows version, but rather create a windows version to run on ARM. Apple is rumored to create an OSX version that will run on ARM

This is the exact same thing that MS is doing, running a copy of their OS on a different chipset, apps have to written/compiled for that chipset and it produces different binaries.

Well, did you read all the stuff I wrote above the part you quote? As it is, cross-compiling existing Cocoa applications to ARM should trivially work for most of the apps. You don't need an emulator. You also don't need a different version of the app for each platform, because OS X allows you to have native code for different platform in the same binary (the universal or fat binary format, which they have successfully used with the PPC and then with x32 to x64 transition).

And Windows RT is quite different from normal Windows — its a specialised OS with its own restricted API. I see no reason for Apple to do something similar (they already have their own specialised OS X with restricted API, its called iOS) — it makes much more sense for them to simply compile OS X for ARM and call it a day. An ARM version of OS X would be virtually indistinguishable from the Intel one.
 
I have to believe an ARM based Mac is set to be a low end product and they're not dropping Intel altogether - that would be insane.

That's my guess to ... It would be still interesting to soo how they diiferentiate the iPad from a low coat MBArm.

Plus my hope to see a Surface Pro/Helix type of Mac tablet is shrinking
 
And here we go back to Windows!

But Windows is rooted in the PC era.

The world is moving on to the post-PC era.

Does Apple need to accommodate people who are tethered to an increasingly anachronistic method of working? Or should Apple concentrate on the future?

Versions of Office and iWork already work on ARM processors...versions that are good enough for the vast majority of users, versions that can and will be improved in time.

Increasingly, people are moving to cloud based services and apps...where the processor is irrelevant.

If ARM processors can meet 99% of the needs of 99% of users, then the PC era really is dead.
 
How do the French get so many scoops?

With a 30 hour work week and a 40 week work year, where do they find the time?
 
There's been a few rumors of apple moving along to an ARM architecture, at least for one model. I don't understand it, look at the failure of the windows rt. I'd avoid it like the plague.

A difference is that Windows does not support fat binaries.
 
Sad to say but if Apple starts mainlining ARM I'm likely out of their Desktops. The switch off PowerPC was rough (still keep 10.6 around on an extra partition for Rosetta support) and at this point if my MacPro 3,1 goes down I'm likely going to go with hand assemble PC and run off dual booting Windows/Linux, and maybe Hackintosh it if I get nostalgic.

The reasons to switch from PowerPC to x86 were many. The reasons to switch from x86 to ARM are 1, mobile.

Mobile is what my iPad is for, a remote screen with just enough brains to operate indepently for minor tasks. Everything else gets run from the tower, either at the keyboard or by Remote access.

If Apple wants to up their game in remote hardware resource usage, I wouldn't necessarily say no to the idea of an ARM laptop that can run back to an x86 more desktopy home server for help with bigger programs. Its what SteamOS is doing.

However their latest batch of screwups with SSL does not fill me with confidence.
 
Well, did you read all the stuff I wrote above the part you quote? As it is, cross-compiling existing Cocoa applications to ARM should trivially work for most of the apps. You don't need an emulator. You also don't need a different version of the app for each platform, because OS X allows you to have native code for different platform in the same binary (the universal or fat binary format, which they have successfully used with the PPC and then with x32 to x64 transition).
I did read it, and I'm not sold that its a trivial move for developers since the cpu structure is so different. The same thing was said about PPC to Intel early on that developers only needed to cross compile and it would work, but that was not the case.

I'll concede the point about windows RT and its API, but still think its a valuable lesson for apple to observe so they don't repeat it.

I also think they're giving giving consumers reasons not to buy a product. Switching platforms (as opposed to rolling out a different model) is a scary proposition and consumers and pros want to be sure their apps can and will run on it now and the future. Moving off intel causes a lot of fear and anxiety that apple shouldn't need to address. This is one of many reasons why its sheer lunacy to move completely off the intel chipset and why I think it will be a low cost low end model they're releasing.
 
$1500 or $300?

+-------------------------+
+ At the end of 2015: +
+-------------------------+

Intel, Samsung, Global foundry will ALL have a 14nm FinFET

ARM CCN-508 : Available with 32 ARM cores + GPU + Very high speed memory controller + 'open' bus for 3rd party ALL on the SAME piece of silicon.

What would you rather have .... a $1500 Intel 16 core intel Xeon ... or a 32 core ARM chip for $300?

... this is why Apple will migrate to ARM ... so they can sell you portable supercomputer for under $1000.

AJ
 
I

I was actually watching the 2005 presentation where they announced the Intel transition the other day. I don't remember Steve's exact phrasing, but he said that Intel processors would see them through for at least another 10 years. That was 9 years ago.

In that same presentation, he said that OS X was built to be cross-platform, or at least platform-agnostic. Then there's the fact that they were prototyping Intel Macs for five years before that, which is consistent with this rumour.

In light of all this, I expect that we'll see an Intel->ARM transition, maybe not next year, but likely within 5 years. It will be a pain to recompile everything for ARM, but I daresay Apple would like to leave behind every developer who doesn't update their apps.

And let it be said that I am not in favour of this transition. It will be a sad day for cross-platform gaming (we hardly knew ye), and a sad day for people who run Windows on a Mac.
 
The 64bit Dual Core A7 running at 1.4Ghz gets around half the geekbench score of the dual core i5 in the MBP. If they use a quad core it would be on par, then if they are to use 4-8 of them, it will be 4-8X more powerful. That has to be enough extra power to real time translate Intel processor applications to run on ARM would it not?

----------

Most of your work is apparently still in the stone age.

haha

----------

But Windows is rooted in the PC era.

The world is moving on to the post-PC era.

Does Apple need to accommodate people who are tethered to an increasingly anachronistic method of working? Or should Apple concentrate on the future?

Versions of Office and iWork already work on ARM processors...versions that are good enough for the vast majority of users, versions that can and will be improved in time.

Increasingly, people are moving to cloud based services and apps...where the processor is irrelevant.

If ARM processors can meet 99% of the needs of 99% of users, then the PC era really is dead.

If they make a version of Final Cut Pro, and get Adobe to translate all of their apps, it could push devs.
 
If the processor helps to sell the machine, then yes, Apple should keep Intel.

I don't think its any coincidence that Apple Computers sales started to excelerate after Apple went Intel.

Still too early to move to Cloud services for apps etc - firstly, limited available ( how many of your favourite desktop apps have web versions? ), and secondly, too unreliable with security concerns.

Many companies aren't moving due to security and reliability issues. They'd rather keep their data local to their organization.

A long way to go until PC era is finished.


But Windows is rooted in the PC era.

The world is moving on to the post-PC era.

Does Apple need to accommodate people who are tethered to an increasingly anachronistic method of working? Or should Apple concentrate on the future?

Versions of Office and iWork already work on ARM processors...versions that are good enough for the vast majority of users, versions that can and will be improved in time.

Increasingly, people are moving to cloud based services and apps...where the processor is irrelevant.

If ARM processors can meet 99% of the needs of 99% of users, then the PC era really is dead.
 
The 64bit Dual Core A7 running at 1.4Ghz gets around half the geekbench score of the dual core i5 in the MBP. If they use a quad core it would be on par, then if they are to use 4-8 of them, it will be 4-8X more powerful. That has to be enough extra power to real time translate Intel processor applications to run on ARM would it not?

So if I understand your point, apples to apples, dual core A7 is 50% slower then a dual core i5 (in that specific benchmark), adding more cores will make it as fast as a dual core i5?

I'm not seeing the logic, throwing an quad core or octo core processor to compete against a dual core intel CPU doesn't seem to make sense. Plus the speed increase is not linear, more 2 more cores does not make it 2 times faster, 4 more cores makes it 4x faster.
 
So if I understand your point, apples to apples, dual core A7 is 50% slower then a dual core i5 (in that specific benchmark), adding more cores will make it as fast as a dual core i5?

I'm not seeing the logic, throwing an quad core or octo core processor to compete against a dual core intel CPU doesn't seem to make sense. Plus the speed increase is not linear, more 2 more cores does not make it 2 times faster, 4 more cores makes it 4x faster.

But I'm sure that by doubling the core count (while still having the single threaded performance so high), as well as having 8 of them running together will well outperform the dual core i5 in the 13" MBPr.

And if Apple still use fans, they could run these processors higher allowing for even more performance gains.
 
So if I understand your point, apples to apples, dual core A7 is 50% slower then a dual core i5 (in that specific benchmark), adding more cores will make it as fast as a dual core i5?

I'm not seeing the logic, throwing an quad core or octo core processor to compete against a dual core intel CPU doesn't seem to make sense. Plus the speed increase is not linear, more 2 more cores does not make it 2 times faster, 4 more cores makes it 4x faster.

Not to mention instruction sets vary, and even if there was 1-1 performance comparison, single core performance is still massive in computing.

If not, well AMD's 8 core behemoth 4-5Ghz chips would be all over and round intel's quality i7 quad cores.
 
Why don't they just buy AMD and bring all their chip design and manufacturing in house so they get total control.
 
Arm would be better for a high end application than budget/low power usage. A very large percentage of an X86 chip is dedicated to converting CISC instructions into microcode. A larger percentage is dedicated to branch prediction and speculative execution. A VLIW implementation of ARM gets rid of all of that.

If Apple were to build a VLIW ARM chip on the scale of a modern XEON, 16 times the performance of the top end Intel chip would be trivial and 256 would be a reasonable goal. There is a reason why both Intel and Microsoft are investing a great deal of effort into the ARM architecture. X86 should have died 10 years ago.

Note: Moving to VLIW arm would require a much smarter compiler. It would be a bit of a challenge for Apple or Microsoft, but not something out of their reach.

Edited to say, One big advantage of a VLIW implementation would be single threaded performance. There are many problems that can't be made parallel. VLIW RISC is perfect for those problems.
 
If the processor helps to sell the machine, then yes, Apple should keep Intel.

I don't think its any coincidence that Apple Computers sales started to excelerate after Apple went Intel.

Still too early to move to Cloud services for apps etc - firstly, limited available ( how many of your favourite desktop apps have web versions? ), and secondly, too unreliable with security concerns.

Many companies aren't moving due to security and reliability issues. They'd rather keep their data local to their organization.

A long way to go until PC era is finished.

Presumably they are not planning to ditch Intel in their Pro Macs but instead introduce a new range of cheaper consumer focussed Macs that run ARM and iOS and therefore can run iOS software. That would make sense.
 
But I'm sure that by doubling the core count (while still having the single threaded performance so high), as well as having 8 of them running together will well outperform the dual core i5 in the 13" MBPr.
Adding cores does not cause a linear increase in performance. The application needs to be able to written to take advantage of high cores and I don't most apps do. I think its overly simplistic to think you can double the core count to equal an intel cpu.

What you're saying is apple should put in a quad core to compete against Intel's dual core and octo core against Intel's quad core. That doesn't even take into account the Intel CPU's ability to hyper threading ability and virtualization technology in the cpu and lot of other technology that makes intel CPUs blazingly fast. Plus is Apple's GPU on par with the Iris Pro? Intel is really hitting the stride with their iGPUs and they're getting quite potent
 
Would be good if they do move. Then when everything goes wrong, then only have themselves to blame. Apple has been very good at passing the buck to someone else over the past 15 years.

Can't ever see it happening myself. Not the way Apple works. They will always need someone else to blame for the short comings. Would think there is more chance of the pro machines going back to PowerPC chips before Arm are in them.
 
I lol reading the comments about ARM based laptops running OS X. If they ever come they will come with some type of modified iOS that can show more than one app at the same time thus giving the user thousands of apps. It would be ridiculous for them to have to cut off iOS apps. They could just unite them both and just make one: ios

IMO this will be the end for apple. Even people that are not technical want an open Computer that can handle most task and iOS as of this day can't even handle older wifi printers/brands to just name one thing.

Instead, I see them releasing just one "Mac" with ARM, prob the 12"iPad Air w some type of removable keyboard.
 
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.