“Apple’s move from Intel to ARM processors may be closer than you think,” Mark Reschke writes for T-GAAP.
“The highly rumored iPad 3 is said to be doubling the current iPad 2 pixel resolution, moving from 1024×768 to 2048×1536. The move effectively increase the pixel count by over 4x, which is stunning for a device this size,” Reschke writes. “Apple is likely to upgrade to a quad-processor to help things along, but a massive boost in graphics processing power would certainly be required to push that many pixels.”
“Apple has a knack for great timing. On Tuesday, Imagination Technologies publicly announced it’s newest line of PowerVR, which may boast up to 20 times more of the performance vs their current core of GPU cores,” Reschke writes. “Such power integrated into a quad-core A6 processor would be able to make the iPad 3 roar, but it may deliver enough power to justify landing into Apple’s Mac product lineup, starting with the MacBook Air.”
Read more in the full article here.
MacDailyNews Take: Things can go south for Intel very, very quickly and they seem to know it.
If Intel’s MacBook Air wannabe “ultrabooks” fail to catch on, Apple’s leverage over Intel will ratchet up to tremendous levels. Intel will become Apple’s lapdog.
As I’m only familiar with the name more so than the tech, what is/are the advantages of an ARM processor over the Intel chip? I’m guessing power management and such but can someone give more detail?
Technically, it’s the low voltage and low current operation that makes ARM processors more successful in mobile devices. I believe there is a business angle now too, since Intel refused to address Apple’s needs for so long (and apparently still haven’t delivered a satisfactory solution). A bit like Adobe and Flash – Apple gives you a chance, but if you ignore them you have to earn your way back into their future.
Sheesh, you’d think Intel would know better. The reason Apple switched Macs from PowerPC to Intel was because Motorola and IBM dragged their feet on improving the G3 and G4 processors.
——RM
Wait, that was G4 and G5. Damn my lousy memory of history!
——RM
The G3 (PPC 750) series of processors is alive and well today in embedded applications and specialty applications, such as spaceflight (requires greater radiation tolerance). It is still valued for its combination of good performance/power characteristics and low cost.
It was a decent general purpose CPU for a while, especially for laptops, but lacked the math coprocessor functions and acceleration provided in the G4 and G5. It is still a useful device, however, especially with industry attempting to add “intelligence” to many common devices.
Particularly power management and low temperature; ARM chips are designed for mobile devices, so an ARM quad-core processor with this new GPU, along with the tech supplied by the new Israeli acquisition could well produce chips capable of running laptops, as well as the lower end iMacs. The high-end ones use desktop processors, but the cheaper ones use laptop tech.
No more Boot Camp then…?
Well why not? How else would you expect to run the new version of Windows 8 that is written for ARM (not that you’d really want to though).
Imagination Technologies did not publish timeline for the claimed 20x perfect increase. (can be a decade 🙂 ).
Intel demos ivy bridge on ces and announces next year’s Haswell.
Only rumors and hot air for Imagination Technologies … and Apple is not commenting.
ARM in MBA won’t happen this or next year.
All “expert’s” Apple HW related over projections are laughable since the iP4S fiasco. MDN included 🙂
Only Apple knows what they are going to use infuture products but one is sure: they are going to use the best … and ARM is far from able to compete for MBA CPU.
It’s also not much likely Intel to be able to compete for CPU in smart phones this year or next …. So it will be really interesting in 2014-2015 but not sooner.
My 2c.
Apple already owns 9.5% of Imagination Technologies; I wonder if they’ll just buy them outright in the near future.
I am definitely excited about ARM making its way to the Mac platform (I have nothing against Intel though)!
Still skeptical.
Just sayin’.
If it happens, it will probably be timed with the next major release of the Mac OS (after Lion). If Lion has a two-year run, that would make it 2013.
However, I can see something similar to the PowerPC to Intel transition happening. Back then, Apple kept a secret build of Mac OS X for Intel in development from Day One. Tiger was actually TWO major releases. The first Tiger release was for PowerPC. Then, Apple released an Intel build of Tiger, along with each new Intel Mac model. There never was a “retail” version of Tiger for Intel. Leopard was “universal.”
So, if Apple repeats that process, there is a secret ARM build of Mac OS X Lion right now. During Lion’s run, Apple will release a build of Lion for ARM along with each new Mac that uses an ARM processor. If this happens, Lion’s run as “the king” could be longer, just as Tiger’s run was much longer than previous releases.
I don’t think Apple will completely transition away from Intel (or it will take a lot longer than the PowerPC to Intel transition). It will obviously start with MacBook Air.
That “secret build of OS X Lion for ARM” is called “iOS 5” and you can download it from Apple’s website.
There just isn’t any desktop-hardware beyond the iPad to run it on.
While I believe that a future “Mac OS” will be “iOS for Mac,” it’s certainly NOT there yet with iOS 5. 😉
There’s no user-accessible file system. There’s no windowing, so every app takes over the whole screen. There’s no pointer (cursor) on the screen. The only available screen resolution is 1024×768. There are no drivers for the diversity of internal and external components used by a Mac. The list of things iOS 5 cannot do that is needed for the Mac’s OS is endless.
“Mac OS X Lion for ARM,” if such a thing currently exists in secret, is NOT the same thing. It’s the current Mac OS X Lion, with all of Lion’s features and capabilities, that happens to run on ARM-based processors. It’s equivalent to Tiger having a PowerPC build and a separate Intel build.
I don’t think “iPad 3” will have a quad-core processor or a double-resolution display, if the release date is March/April.
If the release date is late Summer or Fall, maybe… But going directly from 1024×768 to 2048×1536 still seems impractical. Yes, Apple doubled the iPhone’s screen resolution, but that’s a MUCH smaller screen. The iPad’s screen is like 8 or 9 times larger. The rumor mongers will always go with whatever is technically possible, without regard to viability as a product, because no one wants to read a boring rumor. All they care about is hits to their web site.
If the next major release of iOS is resolution independent, then Apple could go from 1024×768 to something more reasonable, like 1600×1200.
That elusive resolution independence is the key. To date, Apple has not demonstrated that capability and the iPad 2 still has a sub-HD display. Therefore, the reasonable assumption is that if Apple intends to step up the iPad display resolution to compete with Android offerings, then it will once again double the number of horizontal and vertical pixels.
The current 1024×768 display isn’t even sufficient for 720p (1280×720), much less 1080p (1920×1080). Doubling to 2048 x 1536 would admittedly be a bit of overkill, but the resulting ~264 ppi is not unreasonable, and well below the 326 ppi of the iPhone 4S. It is undoubtedly challenging to manufacture a 9.7″ display with 3,145,728 pixels in large quantities and high yield at a low enough cost to keep the iPad 3 cost within reason. But Apple can make that happen, if anyone can.
LOL… just because 264 PPI is “well below” 326 PPI (on a 3.5-inch screen), that does not make it reasonable. It is WELL ABOVE any existing LCD screen of that size or larger used in any existing product. The MacBook Air’s display is very nice (with tiny pixels) and it’s only about 130 PPI. You’re talking about TWICE the density (4x the pixels per square inch).
Is 2048×1536 technically feasible? Probably. Is it commercially viable? NO. In my opinion, not in 2012 (especially early 2012). Instead, I think Apple finds a way to move up to “reasonable” level, such as 1600×1200, while ingeniously minimizing impact on the iPad platform and developers.
I think we forget that iPad was Apple’s answer to “netbooks,” the low-end laptop. Instead of an ultra-cheap MacBook, Apple gave us the $499 iPad. Well, iPad has quickly become the “luxury” high-end tablet in perception, and the typical speculation about “iPad 3” seems oblivious to the need to keep iPad (all models) within a reasonable price range.
The sarcasm embedded in your LOL is unwarranted. I caveated my input sufficiently. About the only thing that you did was change my “not unreasonable” to your “technically feasible,” which I admit is a better characterization.
As I stated:
While I do believe that resolution independence is the better long term answer across the board, it must also be difficult to implement. Otherwise Apple would have done that a decade ago during the early steps of Mac OS X. As I stated when comparing those options (resolution independence versus 2048×1536) previously:
> While I do believe that resolution independence is the better long term answer across the board, it must also be difficult to implement.
Not really. It already exists in a limited way in Mac OS X.
I was at an Apple Store the other day playing with a 27-inch iMac. I tried changing the resolution setting (in System Preferences Displays pane) to lower settings, such as 1920×1080. In the past, not using the native resolution resulted in unacceptable “fuzziness.” However, on these iMacs, the pixels are already small enough so that it looks quite good. And that’s “only” at about 109 PPI. The current problem with doing this on Macs is that the choices are too incremental; true “resolution independence” would allow the user to select ANY resolution setting.
But that problem does not exist if the same method was used on an iPad. If the next iPad has a 1600×1200 display, Apple could make iOS scale the current iOS to fit on that screen (just like that 27-inch iMac scaled 1920×1080 to fit). At 9.7 inches, that would be about 206 PPI. The user would not be able to see a difference, because the pixels are too tiny. If the user holds an iPad 50% further away from eyes compared to any iPhone, 206 PPI is already “retina.” But Apple can selectively use the full 1600×1200 resolution, such as in playing movies or displaying pictures, and allow developers to do the same for their apps. Call it “HD mode.”
Therefore, once the pixels become tiny enough, it doesn’t really matter if Apple “doubles” the resolution to 2048×1536 or uses something like 1600×1200. Resolution independence is achieved and native resolution becomes mostly irrelevant.
Apparently, to watch a movie, the iPad can be rotated to landscape mode and 720p could quite nicely display all 720 lines of resolution on a display with 768 pixel rows.
Did you know that you can rotate the iPad to landscape mode?
1024 is less than 1280, so your point does not “fit” in the horizontal direction.
The article says some people prefer their Air over their iPad. Doesn’t that make the case that Intel’s chipsets (in the Air) outperform the ARM CPUs (iPad)?
Intel has some pretty nice x86 processors on the market now, and even better ones in the pipeline.
The only company that could probably deliver a great ARM+GPU solution is NVIDIA.
That argument involves a lot of assumptions. The MBA does offer several similarities to the iPad because of its solid state memory architecture – in some ways it is an iPad with a keyboard, but with extra OS X versatility. But there are many possible reasons that a person may have for preferring an MBA to an iPad. Your assumption that the CPU is the key is just that – an assumption. It certainly does not “make the case.”
If I was Apple, I would do everything I could (without endangering my company) to completely SCREW INTEL over their helping the Winblows world compete with Apple.
Personally I don’t care what processor is in my MacBook Air. Personally I would like to see Intel hurt really, really bad.
… and where they please. We all know that.
Going to ARM from Intel would be a huge move, though. The two architectures are not even close, the speeds are not even close, they’d have to change all their hardware AND all their software. And the ARM processors would have to speed up by a factor of three – just to be where Intel is today. And would have to boost the cores on the chip. And would have to boost from 16-bit(?) to 64 bit. And Apple would have to re-write OSX and/or iOS. And then there’s all the programs that run on OSX!
Sounds like a WHOLE lot of WORK!
So ARM wouldn’t give Apple a leg up…?
Not if ARM can’t go toe to toe with Intel, no.
Well two big thumbs down, then….
Apple is preparin’ a knuckle sandwich for the boys at Intel. Can’t say as I blame ’em after gettin’ the finger from ’em before.
A knuckle sandwich?
You’re pullin’ my leg…..
Apple’s A series processors are 32-bit machines. 64-bit versions of ARM do exist, but I don’t know of anyone selling laptops based on them currently.
It would not surprise me to see Apple eventually go from x86 to ARM, but I wouldn’t ever expect Apple to go back to a 32-bit Mac.
-jcr
… an alternative that amounts to “sort of”.
While it would be quite expensive and time consuming to replace the Intel CPUs with ARM CPUs in the OSX boxes, how hard would it be to “grow” the iOS family?
Think about the Next Generation of the MBA – and only that model – going iOS and ARM. Now we’re talking about bigger screens, a real keyboard, real ports, greater structural integrity, longer battery life and the like (vs iPad). We are ALSO talking about a slight reduction in power and fewer powerful apps (vs MBA) and greater weight and cost (vs iPad).
Thought I would muscle in on this conversation by announcing a knees-up.
If apple move Mac to arm they would just go ios and kill osx as we know love it. the writing is on wall. someday we will not have powerful Macs but fully control appliance ios devices from apple
Apple’s Development system, Xcode, already is able to generate code for both ARM and Intel, it is able to work with iOS and OSX GUI libs, even memory management is converging, in the sense that automatic reference count is about to replace garbage collection. I am sure there may be some remaining issues, but if the performance and the economics are right, the transition may be far from difficult.
Won’t happen. Windows needs to continue to run on Macs for many more years.
I doubt very seriously that we will see ARM in Apple laptops this year, or next. The A6 – even with a fabulous new GPU – will have only a fraction of the processing power of intel’s Ivy Bridge. The impact on the user experience is still too great to switch.
Is it completely out of the question to build a hybrid? Have intel for high performance that can shut off when the horsepower is not needed and ARM the rest of the time. Running a movie could be ARM GPU and running a word processor might run on ARM okay.
We’re more likely to see more integration of iOS and OS X first. Only after that will a switch to ARM chips.
Also have to remember that OS X has to perform a lot more functions – multiple buses (USB, FW, Thunderbolt) and run more complex (and poorly written) apps like Office.
What I like about iOS is the app response and clean memory management. A lot more is happening in the background in a Mac but I would like to see the same responsiveness in OS X.
Apple doesn’t need to move from Intel to ARM. The Mac only accounts for 17% of Apple sales. Eventually in a few years the iPad will replace the MacBook as we see larger iPads thus reducing the Mac percentage further. All without a painful and unnecessary transition – just a natural move from Intel MacBooks to ARM iPads. Eventually the iMac will be the only Mac left. Maybe then they will go all ARM.
good news: arm is great, nothing new here
bad news: intel chips in the MB Air, MB Pro, iMac, etc. are MUCH more powerful and faster than the small device ARM processors discussed in this article.
So, in short, an ARM such as those contemplated can not replace the Intel chips at this point. Sorry everyone but actually this is a good thing. Apple switching to Intel and embracing USB helped convert me to Mac. Ease of running virtual machines expecting to see an Intel chip is important for many of us long time Windows sufferers in switching over to Mac so we can use Boot Camp or virtual software as a crutch to continue to run some Windows software until we finally realize its not necessary. But in the meantime, Intel has a good track record and I am glad intel chips are in Macs. Not sure if I would continue with Apple if they started using something else again (like the G3/4 and PowerPC days of yore – arghh)
Intel has confirmed that Apple is likely (etechmagdotcom) to use Medfield chip in its iPhone. This will leave us to the question whether Apple will quit using A5 chips or Medfield will be a partial choice.
@Zeet, when you make a definitive statement such as “Intel has confirmed..” how about providing a link from the Intel or Apple press office rather than a rumour site?
Dear Billy it is not a rumor, this thing is discussed at Consumer Electronics Show going on in Las Vegas, and many online newspapers have headlined the news. Official statements are there by Intel’s representative who have initiated talks with Apple over the matter we’re talking about.