or Connect
AppleInsider › Forums › Mac Hardware › Future Apple Hardware › ARM seen challenging Intel's notebook chip dominance by 2013
New Posts  All Forums:Forum Nav:

ARM seen challenging Intel's notebook chip dominance by 2013

post #1 of 89
Thread Starter 
ARM, the company whose reference designs are used in the chips that power Apple's iPhone and iPad, is expected by industry insiders to carve out a significant presence in the traditional notebook computer market by 2013.

While ARM's raw processor performance is not expected to catch up with PC chipmaker Intel in the next two or three years, ARM CPUs are seen becoming powerful enough to power more traditional computers, rather than primarily smartphones and tablets. Citing sources in the notebook PC market, DigiTimes reported Friday that industry insiders expect that Intel will face its "biggest crisis yet" from ARM in 2013.

Anticipation of success from ARM in the notebook PC market stems partially from the fact that Microsoft's next-generation operating system, Windows 8, will offer compatibility with processors based on ARM's reference designs. Windows-based notebook makers, hoping to counter the success Apple has had with its Mac lineup, reportedly hope that competition between Intel and ARM will help the industry and could drive chip prices lower.

"The sources pointed out that Intel is currently focusing on improving the power consumption of its processors, while ARM is enhancing its performance," the report said. "Since ARM is cooperating with Microsoft, and Intel with Google, multiple choices of platform will mean a single platform will no longer dominate the market, and competition will also force the upstream giants to provide a wider range of options to their downstream partners."

Starting with the first-generation iPad in 2010, Apple began crafting their own custom-designed ARM-based systems-on-a-chip with both CPU and graphics processing included. Earlier this year, one rumor claimed that Apple was internally testing a new MacBook Air model powered by the same A5 processor as the iPad 2 and iPhone 4S, and that Apple was impressed with the results.



ARM CPUs are popular in devices like smartphones and tablets because they offer lower power consumption than traditional chips, like the Intel processors that power most PCs as well as Apple's entire Mac lineup. Apple originally attempted to develop the iPad based on Intel's low-power Atom chips, but eventually went with ARM designs. Apple also warned Intel that it would end its partnership with the chipmaker for Mac CPUs if it did not improve power consumption, a threat that Intel executives admitted was a "real wake-up call" for the company.

If ARM does enter the notebook PC business, its most likely entrance point would be thin-and-light laptops that sacrifice power for portability, like Apple's popular MacBook Air lineup. Intel has attempted to counter the MacBook Air with its own "Ultrabook" reference design for ultraportable laptops, but that strategy, while still in its infancy, has struggled thus far.
post #2 of 89
It won't be until the A8 and equivalents that ARM will be able to compete in that manner.

At least, I can't imagine it being any earlier.

Originally Posted by Slurpy

There's just a TINY chance that Apple will also be able to figure out payments. Oh wait, they did already… …and you’re already f*ed.

 

Reply

Originally Posted by Slurpy

There's just a TINY chance that Apple will also be able to figure out payments. Oh wait, they did already… …and you’re already f*ed.

 

Reply
post #3 of 89
I think it is still too soon to say. x86 will have its place and won't be displaced any time soon.

iPod nano 5th Gen 8GB Orange, iPad 3rd Gen WiFi 32GB White
MacBook Pro 15" Core i7 2.66GHz 8GB RAM 120GB Intel 320M
Mac mini Core 2 Duo 2.4GHz 8GB RAM, iPhone 5 32GB Black

Reply

iPod nano 5th Gen 8GB Orange, iPad 3rd Gen WiFi 32GB White
MacBook Pro 15" Core i7 2.66GHz 8GB RAM 120GB Intel 320M
Mac mini Core 2 Duo 2.4GHz 8GB RAM, iPhone 5 32GB Black

Reply
post #4 of 89
With the 8 core Mali chip coming out, SGX graphics cores being scalable up to 12, and the Cortex A15's to come, I can see it competing with low end Atom processors. Still far away from even ULV processors, but I can see ARM getting their foot in the PC space with Windows 8 being ARM compatible.
post #5 of 89
Interesting...

ARM's OPEN system may beat Intel's CLOSED system.

Between Apple's eco-system to include it's supply-chain and ARM's reference designs which they sell to manufactures, these two are turning the WINTEL model into a real disavantage.

The only thing in the WINTEL model that was really open with Microsoft controlling the OS and Intel controlling the CPU, was the commoditized PC - the least profitable link in the chain.
post #6 of 89
I see them challenging AMD before they take much of Intel's market.
post #7 of 89
Quote:
Originally Posted by studentx View Post

ARM's OPEN system may beat Intel's CLOSED system.

Wow people got so bias with open/closed system. ARM is not open it is licensed
post #8 of 89
I bet this happens. Many home users have modest processing requirements, which ARM can handle fine. The only thing is app compatibility. I think we will soon see the return of Universal Binaries, only this time the will be x86/ARM, and it will be a permanent arrangement, not a transition. ARM will probably never approach Intel in raw performance, and those machines will continue to be needed by many.
post #9 of 89
ARM efficiency is their biggest punch against Intel.

Of course ARM performance is nowhere near Intel processor, the i7 is roughly 10x more powerful but an Intel Core i7 in a MacBook need 30 times more power than the A5.

Performance per watts ARM is unbeatable right now.
post #10 of 89
"Win8 compatible" needs to be in quotes.

Because being able to run an OS with essentially no actual compatible software isn't exactly going to leverage the installed base.

The same applies to Apple if they were to release an ARM laptop. What exactly would you run on it for software?

Neither company will be transitioning to ARM, so how would it really make sense to have a predominately x86 ecosystem with a few portables running ARM, with ostensibly a "compatible" OS that actually is NOT compatible with the software base.

That is just a mess of confusion. I can see Microsoft doing it, but not Apple.

Before anyone trots out Apples transitions, remember that those were complete transitions of the product line. That is a very clear strategy. Not simply having mixed CPU architecture releases in the market confusing everything, and the transitions included CPU emulation software, which ARM will be incapable of running adequately.
post #11 of 89
Quote:
Originally Posted by studentx View Post

ARM's OPEN system may beat Intel's CLOSED system.

ARM is to CPUs as nVidia's graphics card designs are to GPUs.

That is to say, not open, just licensed.

Originally Posted by Slurpy

There's just a TINY chance that Apple will also be able to figure out payments. Oh wait, they did already… …and you’re already f*ed.

 

Reply

Originally Posted by Slurpy

There's just a TINY chance that Apple will also be able to figure out payments. Oh wait, they did already… …and you’re already f*ed.

 

Reply
post #12 of 89
PowerVR version 6 is possibly already included in the A6 processor.
This GPU is 10 times faster than the already incredible GPU of the A5 processor and comparable to the performance of the PlayStation 3.
This blows away the intel CPU GPU combination of the MacBookAir and will be more than enough to give exceptionally good performance for the next generation of Airs.
The point is that Mac OS X (*) relies heavily on GPU performance for most tasks and can offload even more to the GPU (via openCL) if needed. Of course a blazing 4 or 8 core A15 processor with a fifth to a tenth of the power consumption of a comparable Intel processor won't hurt.

J.

(*) and iOS
post #13 of 89
Performance wise, sure, I imagine they'll be a strong competitor to Intel soon. The problem is that pesky x86 instruction set. Apple pulled a magic rabbit out of the proverbial hat with their transition to x86 by way of Rosetta. They could do it again should they choose to move to ARM for MacOSX. I don't think they will...

It's all about the apps. iOS apps are not the same as MacOSX apps regardless of similar tools. Even if Apple were to do this, I don't think you'd find any love whatsoever in the massive Windows market. There is no Rosetta for Windows.

Does anyone really think hundreds of thousands of developers are going to rewrite all their Windows apps to run on ARM or maintain additional code bases to run on both ARM and x86? I think not. They already have to maintain multiple code bases for different platforms (the devs that aren't Windows only) and they're not going to exponentially increase that effort.

This will also be the achilles heel for Windows 8. Corporate America is still having problems getting it's IT infrastructure off of the IE6 crack. Does anyone think they'd buy new versions of all their apps to support ARM even if the developers wrote them? Yea, I don't think so either.
post #14 of 89
deleted
post #15 of 89
Quote:
Originally Posted by mytdave View Post

Performance wise, sure, I imagine they'll be a strong competitor to Intel soon. The problem is that pesky x86 instruction set. Apple pulled a magic rabbit out of the proverbial hat with their transition to x86 by way of Rosetta. They could do it again should they choose to move to ARM for MacOSX. I don't think they will...

I don't think they CAN. They were able to make Rosetta work because the Intel chips were at least as powerful as PPC. The loss in performance caused by translation was overcome by the fact that most people were replacing older systems. ARM starts out significantly less expensive than x86. Adding on the performance penalty from Rosetta and you'd have a major problem.

Quote:
Originally Posted by Snowdog65 View Post

"Win8 compatible" needs to be in quotes.

Because being able to run an OS with essentially no actual compatible software isn't exactly going to leverage the installed base.

The same applies to Apple if they were to release an ARM laptop. What exactly would you run on it for software?

Neither company will be transitioning to ARM, so how would it really make sense to have a predominately x86 ecosystem with a few portables running ARM, with ostensibly a "compatible" OS that actually is NOT compatible with the software base.

However, software availability is overrated. If Microsoft were to port Windows Office to ARM, that (plus IE and a few other lesser apps) would be all that many low end laptop users need. I don't see them running Adobe Creative Suite on this type of machine.

Quote:
Originally Posted by Snowdog65 View Post

That is just a mess of confusion. I can see Microsoft doing it, but not Apple.

Before anyone trots out Apples transitions, remember that those were complete transitions of the product line. That is a very clear strategy. Not simply having mixed CPU architecture releases in the market confusing everything, and the transitions included CPU emulation software, which ARM will be incapable of running adequately.

I agree, I don't think it's LIKELY for Apple to do it, but I don't see it as impossible. For example, I could see an 'iPad Pro' which has a 11-13" screen, keyboard, and iOS. Preload it with Pages, Numbers, and a few other things and it would meet a lot of people's computing needs.
"I'm way over my head when it comes to technical issues like this"
Gatorguy 5/31/13
Reply
"I'm way over my head when it comes to technical issues like this"
Gatorguy 5/31/13
Reply
post #16 of 89
deleted
post #17 of 89
Quote:
Originally Posted by jragosta View Post

I don't think they CAN. They were able to make Rosetta work because the Intel chips were at least as powerful as PPC. The loss in performance caused by translation was overcome by the fact that most people were replacing older systems. ARM starts out significantly less expensive than x86. Adding on the performance penalty from Rosetta and you'd have a major problem. <snip>

You are probably right that they could not turn around and run the existing Rosetta on ARM tomorrow. However, ARM performance is increasing, and one would presume that if they did do it again, it wouldn't be your grandfather's Rosetta - I would presume they'd tweak Rosetta to run in conjunction with LLVM and Klang.
post #18 of 89
Quote:
Originally Posted by MacRulez View Post

Long live ARM! Long live Intel!

Indeed. ARM will finally give Intel the competition it has so greatly needed for so long.

"But AMD!"

AMD is the bargain bin of CPUs. ATI's great, but AMD is… well.

Anyway, Haswell's supposed to kick what we know about power draw squarely in the genitals, and I, for one, hope that increased pressure from ARM will speed along Haswell's release.

And hey, they might even start work on Skylake/Skymont early because of ARM. Hopefully. I want them to hit the physical limit for traditional transistors before my hair goes completely white…

I know a guy at Intel. (You'll never believe me because I can't prove this, but whatever.) He says that Intel's greatest concern is ARM, even taking into account their own post-Haswell roadmap.

Originally Posted by Slurpy

There's just a TINY chance that Apple will also be able to figure out payments. Oh wait, they did already… …and you’re already f*ed.

 

Reply

Originally Posted by Slurpy

There's just a TINY chance that Apple will also be able to figure out payments. Oh wait, they did already… …and you’re already f*ed.

 

Reply
post #19 of 89
Quote:
Originally Posted by jragosta View Post

However, software availability is overrated. If Microsoft were to port Windows Office to ARM, that (plus IE and a few other lesser apps) would be all that many low end laptop users need.

That covers 80 or 90 percent of the software the average person needs; however, that still leaves 1 or 2 apps missing which is likely to be a deal breaker. Unfortunately, those final 1 or 2 apps will vary considerable between different people so you actually need to port a lot of apps to cover 100% of the average users needs.
post #20 of 89
Quote:
Originally Posted by Orlando View Post

That covers 80 or 90 percent of the software the average person needs; however, that still leaves 1 or 2 apps missing which is likely to be a deal breaker. Unfortunately, those final 1 or 2 apps will vary considerable between different people so you actually need to port a lot of apps to cover 100% of the average users needs.

Only if you expect this device to be sold to everyone.

As I said, there are many millions of people who don't need anything more than an email client, web browser, and MS Office. A low end portable device that ran those things could have potentially a large market.

You are correct that you'd need more than that to sell to EVERYONE, but since that would be a foolish thing to try to do with a cheap, low end device, I don't see how it's relevant.
"I'm way over my head when it comes to technical issues like this"
Gatorguy 5/31/13
Reply
"I'm way over my head when it comes to technical issues like this"
Gatorguy 5/31/13
Reply
post #21 of 89
Quote:
Originally Posted by mytdave View Post

You are probably right that they could not turn around and run the existing Rosetta on ARM tomorrow. However, ARM performance is increasing, and one would presume that if they did do it again, it wouldn't be your grandfather's Rosetta - I would presume they'd tweak Rosetta to run in conjunction with LLVM and Klang.

x86 performance is also increasing.

I don't think anyone sees ARM catching up to Intel any time soon, so ARM will be at a performance disadvantage for many years. That means that you'd be running Rosetta on a machine which is slower to start with - which is a formula for disaster. Rosetta really only makes sense if the machine is at least as fast as the one it's emulating.
"I'm way over my head when it comes to technical issues like this"
Gatorguy 5/31/13
Reply
"I'm way over my head when it comes to technical issues like this"
Gatorguy 5/31/13
Reply
post #22 of 89
Quote:
Originally Posted by tipoo View Post

With the 8 core Mali chip coming out, SGX graphics cores being scalable up to 12, and the Cortex A15's to come, I can see it competing with low end Atom processors. Still far away from even ULV processors, but I can see ARM getting their foot in the PC space with Windows 8 being ARM compatible.

The ImgTec PowerVR SGX Series5XT supports up to 16 cores, presently.

http://www.scribd.com/doc/61938492/P...ore-Family-3-1

The next jump most likely will be 32 cores.
post #23 of 89
Quote:
Originally Posted by mytdave View Post

You are probably right that they could not turn around and run the existing Rosetta on ARM tomorrow. However, ARM performance is increasing, and one would presume that if they did do it again, it wouldn't be your grandfather's Rosetta - I would presume they'd tweak Rosetta to run in conjunction with LLVM and Klang.

Klang?
post #24 of 89
Quote:
Originally Posted by jragosta View Post

x86 performance is also increasing.

I don't think anyone sees ARM catching up to Intel any time soon, so ARM will be at a performance disadvantage for many years. That means that you'd be running Rosetta on a machine which is slower to start with - which is a formula for disaster. Rosetta really only makes sense if the machine is at least as fast as the one it's emulating.

For applications written in Cocoa, a Rosetta type environment wouldn't be necessary. Recompile for ARM and redeploy. Rosetta would only be required for old binaries. When Apple switched from PPC to Intel, very few applications had been written in Cocoa and so could not be retargeted easily. Times have changed.
post #25 of 89
Quote:
Originally Posted by bsimpsen View Post

When Apple switched from PPC to Intel, very few applications had been written in Cocoa and so could not be retargeted easily. Times have changed.

"Very few"? Cocoa was around a DECADE before the introduction or even announcement of the transition to Intel.

Now, granted, this isn't at all the Cocoa we know and love today, but it's the psychological foundation for what Cocoa would become.

Cocoa as we know it came about around, what, aught three? Developers had (and have had) plenty of time to change.

Also, does that kid remind anyone else of Steve Jobs? In presentation style, I mean.

Originally Posted by Slurpy

There's just a TINY chance that Apple will also be able to figure out payments. Oh wait, they did already… …and you’re already f*ed.

 

Reply

Originally Posted by Slurpy

There's just a TINY chance that Apple will also be able to figure out payments. Oh wait, they did already… …and you’re already f*ed.

 

Reply
post #26 of 89
While I understand the focus of the article, the event is the market for a traditional PC is dying much quicker than anyone wants to admit. ARM will succeed because its in the right place at the right time. Not at all so for Intel.

I look at my iMac and realize it's become nothing more than a boat anchor. It syncs iDevices, so can our media server if we choose to dump the iMac. Then there's the 2nd gen Air. Haven't seen it in months.
post #27 of 89
deleted
post #28 of 89
Quote:
Originally Posted by jragosta View Post

x86 performance is also increasing.

I don't think anyone sees ARM catching up to Intel any time soon, so ARM will be at a performance disadvantage for many years. That means that you'd be running Rosetta on a machine which is slower to start with - which is a formula for disaster. Rosetta really only makes sense if the machine is at least as fast as the one it's emulating.


ARM doesn't need to catch up to Intel's CPU that's coming in a few years. ARM only need to catch up to what Intel's CPU can do right now, or even a few years ago.

CPU of today are grossly overpowered for what people normally do. If ARM CPU can have equivalent performance of an early generation Core2Duo, it will have ample performance to run any of the today's common PC applications.
post #29 of 89
Quote:
Originally Posted by bsimpsen View Post

For applications written in Cocoa, a Rosetta type environment wouldn't be necessary. Recompile for ARM and redeploy. Rosetta would only be required for old binaries. When Apple switched from PPC to Intel, very few applications had been written in Cocoa and so could not be retargeted easily. Times have changed.

In theory. It rarely works out that simply ("just recompile and deploy"). You are right, though, that the job SHOULD be easier than it was for the PPC-x86 transition. Still, I don't think it matters that much. An iPad Pro with nothing more than Office (or Pages, Keynote, and Numbers), web browser, quicktime, and email would satisfy a TON of people. And since it would run iOS, all the existing iOS apps would work, too.
"I'm way over my head when it comes to technical issues like this"
Gatorguy 5/31/13
Reply
"I'm way over my head when it comes to technical issues like this"
Gatorguy 5/31/13
Reply
post #30 of 89
deleted
post #31 of 89
Quote:
Originally Posted by mytdave View Post

Performance wise, sure, I imagine they'll be a strong competitor to Intel soon. The problem is that pesky x86 instruction set. Apple pulled a magic rabbit out of the proverbial hat with their transition to x86 by way of Rosetta. They could do it again should they choose to move to ARM for MacOSX. I don't think they will...

It's all about the apps. iOS apps are not the same as MacOSX apps regardless of similar tools. Even if Apple were to do this, I don't think you'd find any love whatsoever in the massive Windows market. There is no Rosetta for Windows.

Does anyone really think hundreds of thousands of developers are going to rewrite all their Windows apps to run on ARM or maintain additional code bases to run on both ARM and x86? I think not. They already have to maintain multiple code bases for different platforms (the devs that aren't Windows only) and they're not going to exponentially increase that effort.

This will also be the achilles heel for Windows 8. Corporate America is still having problems getting it's IT infrastructure off of the IE6 crack. Does anyone think they'd buy new versions of all their apps to support ARM even if the developers wrote them? Yea, I don't think so either.

Rosetta isn't needed. It's a matter of recompilation and fat binaries.
Problem solved (for Mac OS X)

J
post #32 of 89
I can't see ARM replacing Intel x86 architecture in 'traditional' clients (laptops/desktops) anytime soon. Intel (and AMD) design not just the CPU, but the entire systems architecture that supports it (memory, I/O, etc). Most likely, Apple is prodding Intel for lower-power designs for their traditional clients (especially MacBook Air) instead of focusing Atom vs. ARM in mobile devices. The irony is that several years ago Intel licensed ARM which they subsequently abandoned in favor of x86 (Atom). My sense is that the market for traditional desktops will continue to be dominated by Intel while mobile devices will be dominated by ARM designs...
post #33 of 89
the cheapest laptops can be bought for $299. what's the point of buying ARM vs an older Intel CPU? once you add in all the other parts that go into a laptop the savings vanish
post #34 of 89
Quote:
Originally Posted by al_bundy View Post

the cheapest laptops can be bought for $299. what's the point of buying ARM vs an older Intel CPU? once you add in all the other parts that go into a laptop the savings vanish

Size/weight and power consumption. It's not just about price.
Do you realize that fluoridation is the most monstrously conceived and dangerous Communist plot we have ever had to face? - Jack D. Ripper
Reply
Do you realize that fluoridation is the most monstrously conceived and dangerous Communist plot we have ever had to face? - Jack D. Ripper
Reply
post #35 of 89
by 2013 you will be able to buy super thin and power efficient Intel based laptops for $500 or less. and there won't be an issue of only running dumbed down apps
post #36 of 89
The smart move would be a laptop running iOS. For Apple this reduces confusiOn that would result in ARM binaries running on Mac OS devices.
post #37 of 89
The problem is that the next processors after ivy-bridge is said to be highly influenced by apple.

The question is when arm processors are as powerfull as intel processors will intels processors be as power efficient.
post #38 of 89
Quote:
Originally Posted by wizard69 View Post

The smart move would be a laptop running iOS. For Apple this reduces confusiOn that would result in ARM binaries running on Mac OS devices.

iOS doesn't work in a laptop setting. OS X doesn't work in a tablet setting.

Originally Posted by Slurpy

There's just a TINY chance that Apple will also be able to figure out payments. Oh wait, they did already… …and you’re already f*ed.

 

Reply

Originally Posted by Slurpy

There's just a TINY chance that Apple will also be able to figure out payments. Oh wait, they did already… …and you’re already f*ed.

 

Reply
post #39 of 89
Of-course ARM will eat Intel's lunch. It's a given.

Intel might be able to produce chips 10 times as fast, but most people don't care about 10 times as fast, all they wan't is enough speed to do every day mundane things.
Once ARM approaches that point (which is any moment now), say goodbye to a large part of your market Intel.

The iPad is already having a good feast on it right now.
post #40 of 89
Quote:
Originally Posted by MacRulez View Post

Would you run your Mac with an Intel Atom?

http://www.osnews.com/story/22704/In..._ARM_Cortex-A9


Only because Apple don't have a Mac that uses it.

Since Atom based systems runs Windows 7 and office applications with acceptable performance, I don't think such a system would have any real trouble running OSX. So as long as ARM CPU can get to a stage that's similar, it would become a real viable option.
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Future Apple Hardware
AppleInsider › Forums › Mac Hardware › Future Apple Hardware › ARM seen challenging Intel's notebook chip dominance by 2013