The bigger story here actually has to do with Atom. The biggest gains Intel is showing are at very low voltages, exactly what will benefit ultra mobile SoCs. Atom has had a tough time getting into smartphones and while we may see limited success at 32nm, the real future is what happens at 22nm. Atom is due for a new microprocessor architecture in 2012, if Intel goes the risky route and combines it with its 22nm process it could have a knockout on its hands.? http://www.anandtech.com/show/4313/i...ing-in-2h-2011
The bigger story here actually has to do with Atom. The biggest gains Intel is showing are at very low voltages, exactly what will benefit ultra mobile SoCs. Atom has had a tough time getting into smartphones and while we may see limited success at 32nm, the real future is what happens at 22nm. Atom is due for a new microprocessor architecture in 2012, if Intel goes the risky route and combines it with its 22nm process it could have a knockout on its hands.• http://www.anandtech.com/show/4313/i...ing-in-2h-2011
But let?s return to the rumor, from SemiAccurate, that the Mac and Intel will soon be ?arm-in-ARM.? (That bad pun isn?t mine.)First, let?s consider the name of the website.Second, what will Apple do at the high-end, for media creation and editing? What about Photoshop, FinalCut, and other applications, including CAD where the Mac is getting back in the game? There?s no roadmap for ARM chips to beat Intel in these computationally intensive areas.Today, going ARM is technically feasible on entry-level Macs. Tomorrow, newer multicore ARM chips might work for middle-of-the-line Macintosh products. But will Apple abandon the faster x86 processors at the high end just to avoid the kind of forking that awaits Windows in its own move to ARM? If not, we?ll again see Universal applications (a.k.a. fat binaries?two versions inside the same container), just as we did with the PowerPC to x86 transition. Microsoft is doing it because it must; Apple did it because the PowerPC didn?t have a future. But now?? http://www.mondaynote.com/2011/05/08...-why-and-when/
That said, Apple could include the ARM architecture in their computers in an advantageous way. In fact, there were rumors two or three years ago that may have been just premature.
The trackpad on Mac laptops (and the transition to trackpads on the iMac has begun as well) is already basically an iDevice screen with no display behind it. Well, put a display and an A5 processor behind it. How much extra would that cost? As much as $50?
You could run any iOS apps on just the trackpad. They could be mirrored on the big screen, or not. (Imagine the battery life if the main screen were blanked!) You could always switch back and forth, and the tighter integration between iOS and OS X in Lion would make that easier.
It's even possible that the main CPU could do some of the computational heavy-lifting?mainly generic number-crunching, probably. Maybe someday?many, many years down the road?the two architectures would be more or less merged.
Right now, a lot of PC users are also buying iPhones, iPod Touches, and iPads. Suppose they're looking for a new PC. Why wouldn't they buy one that could run their iOS apps as well?not to mention Windows, too.
The writing is on the wall.
Lemon Bon Bon.
It's just that each of us seems to see it saying different things! What I see is they will drop the towers, have iOS running on every piece of hardware AND use their "A" series chips as CPUs.
I like the idea of that...
But how would that work -- Isn't the A5 RAM part of the PoP SoC?
Just guessing the reason why apple would turn to such a low power junk chip . I guess intel would do a fabless production for apple and or samsung doing the same . P A semi must being creating something. maybe apple will make zombie machines with low power and have all of humans live in the SERVER farm clouds.
I think you?re talking about two different options here.
The one you say seems ridiculous is ridiculous because of inherent performance loss and lack of need for excessive power savings for a desktop system. The one you say isn?t so crazy isn?t so crazy if you look at the Motorola Atrix and and Chrome OS which are both ?desktop OSes? running on ARM. With the former I can see the potential for Apple to leverage their iOS/OS X/Darwin foundation with CocoaTouch UI for the included touchscreen but when connected to applicable peripherals it instantly switches to the Mac OS Aqua UI like any other app.
Within this Mac OS Aqua UI you can launch any series of windowed or fullscreen apps from LaunchPad, Finder, Dock or Spotlight. App Store apps will be have an option in Xcode by then to be Universal for iOS for iPhone/Touch, iOS iPad, IOS for Aqua UI, and Mac OS. with a possible inclusion for AppleTV despite it?s unique UI.
For example, this would allow you to use your iPhone in the same manner for accessing webpages but a simple dock setup with a display and keyboard and mouse would turn off the iPhone?s display (or use as a Magic Trackpad) that is now displaying a the same Safari window you are used to on Mac OS X. You finish what you?re doing, charging your device or simply need to take off and the phone kicks back to CocoaTouch when you disconnect the device.
If anyone can do this it?s Apple. They have all the Doozers and all the candy radishes (obscure reference) to make it work. Motorola had to start from scratch to give their Atrix a UI and it fails because of it.
That is the only way I can see Apple bringing the Mac OS UI to iOS in a sensible way.
Could all this ARM chip talk be about apple making a nano phone and lining up supply before it starts a production ??
90% of the people calling "bogus" here are a) not looking far enough ahead, b) haven't checked the evidence (see the link above - real companies are working on this) and c) as in the next post below blows away are simply making straight-line projections that the future will just be more of the present.
Thanks to Lemon Bon Bon (interesting handle) for saving me a lot of writing. To concentrate on a point below.....
Lots of venerable companies are in trouble. Including both halves of the long-dominant WinTel duopoly.
Office and its file formats are the only things tethering many 10's of millions of MS's customers to its products. Even if Win 8 and all its variants all hit their marks and then some, MS has years of fighting to reclaim position in smart phones and tablets (which may have been supplanted by new device classes by then, introduced at least partially by guess who?). MS's biggest OEM, HP, is trying to re-invent itself in the Apple mode, controlling its own hardware and software stack with Web OS even as we talk about this.
As for Intel, neither MS nor Apple has the slightest interest in Atom - that point where the x86' basic architectural flaws - a kludged design dating back to the original IBM PC - have finally become the limiting factor they always were destined to be - the computing power/power-for-speed roadblock that is keeping Atom from being a serious rival to ARM. And iDevices are the growth story for CPU's for the next decade most likely.
At this point then, x-86 volumes aside Apple is arguably already Intel's most important customer, both for their imprimatur (consider how openly Intel partnered with Apple on Thunderbolt even tho' it hopes the volume will eventually go into non-Apple PC's and peripherals) and the volume they could throw Intel's way with their mobile business.
So.... ....Intel wants to fab for Apple. Apple wants a fab with companies it's in less litigation with. Intel and Apple seem to be working fairly well together and there are few areas where they're in competition unlike Apple's former friend Google. And Intel's realizing Atom is likely to stay a bomb. And both Win 8 and iOS are Arm systems.
But Intel doesn't want to be reduced to a commodity chip vendor to stay in the game....
So.... ....Apple's ARM implementation via the chops of their recent acquisitions is arguably the best ARM around or close to it. Intel has gobs of talent, the most modern fabs and that new 3D tech noted above.
Another collaboration - much bigger than Thunderbolt - suggests itself. An Apple-Intel leveraged ARM design as suggested by Jameskatt2 could easily stand out in the commoditized ARM market.
Intel would be rescued from its Atom box canyon.
Apple would - by virtue of its IP contributions - get the best deal by far on the chips - and - collect royalties on every one of the chips sold to anyone else (i.e., Apple would be making money on the internals of the most advanced Android and Windows smart phones and tablets on the planet as well as the highest margins on its own). They would also ensure that Intel's future iterations would be designed with their specific new device designs in mind.
All cotton candy thinking for the moment, but Intel really needs to make some kind of out of the box move. Atom's not cutting the mustard.
That 3D New Tech has been developed at AMD several years ago. They will be introducing it in upcoming CPUs. Intel didn't lead in this category anymore than they lead in x64.
As for Intel, neither MS nor Apple has the slightest interest in Atom - that point where the x86' basic architectural flaws - a kludged design dating back to the original IBM PC - have finally become the limiting factor they always were destined to be - the computing power/power-for-speed roadblock that is keeping Atom from being a serious rival to ARM.
All cotton candy thinking for the moment, but Intel really needs to make some kind of out of the box move. Atom's not cutting the mustard.
I don?t disagree with your comments about Atom, but Anand seems to think 2012 Atom will go nuclear.
I don’t disagree with your comments about Atom, but Anand seems to think 2012 Atom will go nuclear.
Mebbe -- and mebbe not -- but I agree fully they don't have a shot before the 22nm gen (and one article below is already talking about the 14nm shift to come):
Interesting article in general: http://www.computerworlduk.com/in-de...or-technology/ including [emphasis added]:
"Much like it has been near impossible to displace either Intel or Microsoft in the PC space, it is likely to be equally hard to displace ARM in these new device classes because ARM is entrenched and the eco-system around it is becoming more robust by the day," Enderle said.
He said Intel desperately needs some very large design wins to get moving in the tablet/smartphone markets. And this 3D chip win is a strong one to start off with.
"Intel has virtually nothing compelling in either the tablet or smartphone space in terms of a shipping product," he added. "[The new technology] is core, no pun intended. This is generally how they advance. But Intel needs to address technology and marketing requirements and, so far, they are only effective on technology."
(Note: Enderle is often both a) wrong and b) a jerk btw, at least according to the Macalope - and see quote from PC World below about "barriers to entry" not being the same in the tab/phone space)
Where the 32nm Atoms are going:
The new 32-nanometer Atom processors are expected to be launched on the 1st of August this year while the two i7 models will hit the market in the later end of the month of June. In addition to this, to enhance the speed of the complete i-series processors, the company is also launching a new motherboard chip this month. The release of the new Atom series processors is breaking news in the field. Low power consumption processors, Atom D2500 and D2700 are said to be codenamed as Cedarview which will enter the market in between 1st August to 31st October.
Speaking more about the new Atom processors, they are expected to have their applications in low-power consumption laptops i.e., the shrinking netbook market, set-top boxes and home-theatre systems. They are also expected to have support for the new technology, Blu-ray 2.0, to have their own media engine and to produce much lesser heat during the service.
I didn't read "phones" or "tablets" in that list....
PC World chimes in:
as the market for devices like tablets and smartphones grows, Intel still faces pretty big challenges, even with its new "3D" chips. Here some issues Intel will have to address if it wants an Apple, a RIM, or a Motorola to pick its x86-based Atom processors over the ARM chips those device makers have so far chosen to power their mobile devices.
Developers, Developers, Developers
Intel says it's going to have a three-year jump on the competition with its 22-nanometer Tri-Gate process technology. But when you consider that the Atom product line won't be moved to 22nm as quickly as the company's PC and server processor families, that lead isn't quite as long. Other semiconductor companies should be adopting Tri-Gate by the time they ramp the 14nm process node in a few years time.
What this means is that mobile OS developers will have to consider if it's worth it to pour a lot of resources into developing for Intel's x86 architecture when they've already invested quite a bit towards getting their operating systems to run on ARM. If Intel's advantages with Tri-Gate are seen as only temporary, could some potential customers decide to just wait it out until the ARM chips catch up?
The good news for Intel is that cross-architecture software development isn't nearly as difficult as it was even a few years ago. For example, Dalvik, part of Google's Android OS, provides a platform-independent programming environment, meaning individual app creators don't have to make the x86/ARM decision.
And it's not like Intel doesn't have a pretty robust developer ecosystem for the x86 architecture already. They don't call it "Wintel" for nothing.
But here's the thing. Intel pushes a lot of semiconductor innovation forward, as it is now doing with Tri-Gate a generation ahead of its competitors. But those competitors aren't exactly standing still. Companies licensing the ARM architecture will have '3D' transistors on their chips eventually, they'll get to 22nm even sooner and they are moving into multi-core designs. Nvidia isn't just hoping to get into tablets—Tegra's already there. And guess what? ARM's got working 22nm silicon.
Finally, a number of posters have noted that AMD's in the 3D transistor game too, and announced first, but the Register (UK) notes:
AMD is ramping up 32 nanometer processes and a new chip design, but Intel is already gearing up to start production on its 22 nanometer process and its related Tri-Gate 3D transistor, the combination of which looks to give Intel a greater advantage in performance and power efficiency than many had been expecting from the chip giant's shrink to 22 nanometers.
On the Atom front, the Intel has been quiet, but obviously with the Tri-Gate transistors and the 22 nanometer shrink, Intel should be able to field more power-efficient, low-end Atom parts that are more suitable for smartphones, tablets, and consumer electronics. It should also be able to create Atoms that have more oomph in the same power envelope – perhaps a lot more, like twice as many cores or a significant jump in single-thread performance.
It remains to be seen if these future Atoms based on Tri-Gate/22 nanometer technology will be able to take on the ARM collective, however. And even if the future Atoms don't hold out well against the Cortex-A15 derivatives due in late 2012 or early 2013 for tiny computers, they may nonetheless be a boon for micro servers.
The sum of this probably augurs against my speculative notion of an Apple-Intel partnership on a super-ARM chip, but whatever, things just keep getting more interesting.
The ARM A-15 looks like a good chip, up-to 4-cores, 2.5GHz, up-to 1TB memory - that would easily beat up an Atom. Could it beat up an i7? Of course not. Could it beat up a Core2Duo? Of course not. I wouldn't be surprised if it could hold its own with a Celeron Dual Core. Why would Apple ditch the x86-64 platform for a processor who's only advantage is power consumption when everything else about the current crop of chips, and the ARMv7 architecture itself, is far behind x86-64? With the feverish pace Intel is moving at, I honestly cannot see Apple leaving x86-64 for ARM in full-sized computers.
You are thinking too narrowly. If ARM is coming to Macs, it will be destined for lightweight portables, such as MacBook Air.
Power efficient processors aren't just about lowering power bills, but about increasing battery life for portables, as well as (1) allowing silent fan-free computing and (2) smaller and lighter form factor.
I believe ARM-based Mac OS X is at least in R&D mode, especially since iOS is a subset of Mac OS X with additional components and layers.
Why can't Apple let it rest and keep Intel and stop farting around!
And the fact that Apple/NeXT has ported this OS from 68K to Intel (92), SPARC, PA-RISC, back to PowerPC, and then to Intel again, as well as built iOS/ARM as a subset of the current OSX offering.
Which means nothing,
No one buys mac for it's emulation power... they buy it for the ability to run Mac apps. iOS is proving that it's the APPS that sell devices... and if you have some 200Million iOS users, wouldn't it be nice to have a way for those app developers and users to work with a more 'universal' binary?
For many of us Windows is an app, an app that runs under emulation. Your statement clearly indicates that you don't know why many of us switched to the Mac and why we see moving away from i86 as stupid. Believe me it would be a very stupid move indeed.
With MS porting to ARM, Apple's timing makes sense, in a performance/price curve.
ARM chips are sold on a performance per watt basis and little else.
If high end ARM chips are demanded by 300Million devices a year... the price per chipset drops... at best putting pressure on Intel/AMD to improve it's pricing/delivery, and at worst, driving the inevitable bottom up migration from phones to pads to laptops to desktops to servers to mainframes.
ARM pricing is a lot more difficult than that. It is all making niche products cost effective right now, that is done with SoC construction. This is only effective when you can get by with ARMs extremely limited performance.
So with this lead time, Apple is likely just firing a shot over the bow of Intel: 'Want the 3rd largest 'pc' maker to continue using Intel chips.... Make them to our specs, at our price, and at our timeline... or run the risk of losing our business in 2 years [and remember, this is like commodities... the pricing/capacity has to be planned out 2 years in advance.... the timing of this makes total sense]'.
It makes no sense at all because ARM can't effectively compete with i86 on a performance basis. In two years time you are likely to get chips from intel with 6 to 8 cores running at 4GHz. Possible even more cores than that. These are real 64 bit cores which are very important to apples pro users. ARM is a long way from having a viable 64 bit core.
All this being said I fully expect Apple to expand it's iOS line up. It only makes sense as they have a winner on their hands and the current line up has serious holes. However I don't see them marketing a direct MBP replacement running iOS or even a direct replacement for the AIR. Even on the AIR an ARM chip would be a horrible step backwards and extremely limiting for users.
If there already exists Windows OS code in native ARM binary format, then it should be a lot easier to virtualize. Otherwise you would need to create a whole virtual x86 emulator - a much larger task.
People mis the whole applications space problem that forces many of us to use Windows. There is much out there that requires i86 that will never be ported to another architecture. Emulation really isn't acceptable here.
I really hope that Apple realizes how many Macs are being sold right now because the support legacy software needs. It would be a blunder of epic proportions to drop i86 in the laptop line up.
Because multiple core gets you dimishing returns. A dual-core chip doesn't double your performance. A quad-core chip doesn't quadruple your performance. Except for highly parallel functions (such as video processing), adding more cores eventually gets to the point that it doesn't speed up the types of tasks that most people perform.
Apple is doing a lot of work to make it easier for programmers to take advantage of multiple cores and the GPU, and that helps. But if you have a series of calculations where each step is dependent on the previous steps being completed first, the work can't be distrubuted to multiple cores.
Right now, 4 cores seems to be a practical limit for a typical user. In the previous iMacs where Apple offered faster dual-core or slower quad-core CPUs as options, many typical users were better off with the faster dual-core iMac. I'm sure we'll eventually get to the point where adding more than 4 cores will be of practical benefit to the average consumer, but that's likely more than a year or so away still.
The transition to dual core was extremely quick because people saw an immediate benefit with respect to system responsiveness. It was almost always a positive user experience. Going quad core like wise enhances system availability especially in the age of constant network connections. The important thing to realize is that modern OS's like Mac OS/X have a lot happening in background these days that users aren't aware of. Those cores are being put to use! The problem is many people look at core use on an individual app basis, this is foolish as some apps will never use all cores and others depend upon the user demand put on the app. It also assumes only one app is in use at a time.
Even iPhone iOS was multitasking apps from day one with E-Mail and other services running all the time. On Mac OS it is not uncommon to see people running multiple apps at one time, these are average users that might have E-Mail, Safari and other apps open at once. Safari can easily consume a considerable amount of processor power all on it's own especially if Flash is involved. With Flash off in it's own process and an enhanced Webkit, simple web browsing will be able to leverage most of the cores in today's machines. Remember though that Safari is almost never running alone on these machines.
I agree the Macbook Air form factor is the future of computing, because once you use one for some time, you realize how comfortable it is to have a computer like that, and when you go back to a weightier computer you feel it as a step backwards... you miss the Air. So, yes, the market will demand the Air form factor soon, no matter how hard MBP users try to think otherwise
That is like saying Ford only needs to build cars on one chassis. I agree that the new AIRs are a great achievement but they are not perfect for everybody. As such Apple will continue to offer up additional models to serve the varied user needs.
Given the right configuration I could see myself moving to a 15" AIR like machine. Mostly for the larger screen, but it would need to have a few more features than the current AIRs. Frankly I expect those features in the next AIR rev, but will likely hold off for Ivy Bridge. However Apple does need to address performance.
However, this form factor is only useful if it has software, and I'm afraid a transition from Intel to Arm would be a severe problem in lack of software (the best known Mac developers would support Arm soon, but other developers would take years to support Arm, and that would be a problem).
Exactly! Legacy software is a huge problem. Further it can be a nightmare for developers of advanced or complex software to test on multiple architectures. I can see many developers giving Apple the finger and simply ignoring an ARM based Mac. It simply isn't worth the trouble to support low performance architectures for many developers, this can even include old i86 hardware.
I use my (4GB RAM, 256GB SSD, 2.13GHz) Air for serious work (finite element analysis, CAD, OpenGL graphics, and photo retouching) (yes, the Air is not a netbook, and can be used for such tasks, even if a lot of people are completely wrong believing the opposite), but if it had an Arm processor I could only use it as a netbook or as an ipad... I wouldn't buy a netbook, I don't need one.
Many people are almost spastic in their belief that an ARM based AIR would be competitive. It is like they have no concept of the performance delta between ARM and current i86 chips. I can see Apple making another iOS device that might pass for a notebook of sorts, but I can't see them marketing them as Macs. It would be a total embarrassment.
Actually the first move here buy Apple might be an iPad with a slide out keyboard. It would make for a tablet that touch typist could better leverage and still maintain all the usability qualities of an iPad.
As a side note I've been playing around with CAD viewers on my iPad. There is great potential here but let's face it Touch and CAD don't really mix. So even if there is a massive jump in ARM performance I don't think you would be doing CAD or other engineering tasks on it. This highlights another issue where I see iOS and Mac OS traveling their own development paths, Touch simply does not work for some work flows.
The future is the power of the MBP on the form factor of the MBA. Anybody who believes the opposite will be proven wrong by the market evolution in the coming years. Period. But that needs software, and Arm would be a problem.
I don't buy the above either. Let's face it you are happy with the AIR performance in your specific engineering capacity. However many would not be so happy. The fact is Apple will always be able to pack more performance into a MBP and some will go that route because of that. It might not work for you but many are not bothered by the size of a MBP. If Apple never makes a 15" AIR I likely would never buy one (old people and tiny screens don't mix).
Plus you dismiss the fact that the old plastic Mac Book still sells in massive volume. It isn't a one size fits all world.