Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
This would only occur if the price were low enough to justify the poor performance, which I don't see coming from Apple. Or an amazing solution coming out in WWDC.
 
Wirelessly posted (Mozilla/5.0 (iPhone; U; CPU iPhone OS 4_3_3 like Mac OS X; en-us) AppleWebKit/533.17.9 (KHTML, like Gecko) Version/5.0.2 Mobile/8J2 Safari/6533.18.5)

blow45 said:
Err... GCD can efficiently handle as many cores you can throw at it. Take a look at the dual processor 12-core Mac Pro.

yeah but how many os x software support it?:confused:

Almost all apps use GCD to one extent or another. The number of cores an app can leverage depends upon how the code is written and what the data models can support. Some apps are inherently single threaded but most can find some level of concurrancy.

I'm not sure where all your negativity with respect to GCD comes from. GCD provides one approach to leveraging modern hardware that works in hand with other approaches.
 
Wirelessly posted (Mozilla/5.0 (iPhone; U; CPU iPhone OS 4_3_3 like Mac OS X; en-us) AppleWebKit/533.17.9 (KHTML, like Gecko) Version/5.0.2 Mobile/8J2 Safari/6533.18.5)

blow45 said:
Only QuickTime X to my knowledge as of right now but Final Cut Pro X will support it so I'd expect Apple's other apps to follow suit soon.

oh ok, as I expected. Actually I thought it was a bit better and that adobe where on the gcd bandwagon too (yeah right...). If the pro apps are switching to GCD now I think one can safely assume that most non pro apple apps will go to gcd in a couple of years, if ever, cause of course, I wouldnt think that adding e.g. gcd to mail.app would be something really productive and apple will be up for it.

My mind is totally blown! Do you guys actuall know what GCD is?

As to that Mail app you are whining about it got a fairly significant speed up when Snow Leopard was released. That was directly due to GCD being enabled in SL. Many apps that had used NSOperation prior to SL got a speed up with zero to little effort on the developers part when SL arrived. That again due to high level abstractions making use of GCD.

You guys seem to have jumped on this GCD is shlt bandwagon without having a clue as to it's usage. You really need to read up on things a bit.
 
This sounds like it would resemble the performance of PC "Netbooks", which Apple has explicitly stated they don't want anything to do with. I agree 100% with those who say that, if this rumor is true, it was for purely experimental purposes and no implementation will be possible until ARM performance comes close to that of Intel with significantly lower power usage... otherwise, what's the draw (no pun intended).

Apple refused to do netbooks because they know that a poorly performing Apple branded device (i.e. poor/choppy video playback, constant "rainbow wheels", etc) is only going to hurt their reputation... they would gain nothing from this in the long run.
 
This rumor is about MBA specifically and jumping to wild conclusions is your own problem.

Pardon? I think you're getting confused (much like in that discussion regarding games). I only said I wouldn't buy a MBA with an A5. What conclusions were I jumping to?
 
Wirelessly posted (Mozilla/5.0 (iPhone; U; CPU iPhone OS 4_3_3 like Mac OS X; en-us) AppleWebKit/533.17.9 (KHTML, like Gecko) Version/5.0.2 Mobile/8J2 Safari/6533.18.5)

FvL said:
I guess that starting with Lion, Apple will force Apps submitted to the Mac App Store to be compiled with LLVM and distributed in LLVM-IR format. Then the OS will JIT compile to the target processor. This would not only make the Apps mostly processor-independent, but also automatically make use of the most efficient instruction set for the target processor.

This might very well be where they are going. LLVM is one of the most interesting open source projects going with a lot of great brains behind it.

I do wonder if it is ready for prime time though. It would be a big jump for Apple. That and the lack of rumors or leaks has me thinking Lion won't get it. The other issue is with developers some whom might object to the distribution of IR code. This would eliminate the emulation problem though. Of course optimizing for a platform won't be easy this way.
 
...

I called this a year ago, and yes Apple will eventually do it. It will run iOS not OSX its just a matter of getting iOS to be close enough to a desktop OS that it is acceptable...I think that will only take a year or two more. Then it will basically be an ipad with a real keyboard and a small mirrored LCD touchscreen instead of a trackpad so you don't have to touch on the vertical surface. In a laptop that thing would have days and days of battery life (but still no flash lol)
 
Ahh yes, the external keyboard for an iPad... How well will that work on somebodies lap on a train? Sitting on a bench in a park?

An author would be a good example of someone who would use an A5 Air yet find an iPad to be cumbersome. Especially if an A5 Air costs noticeably less than an Intel Air.

One size doesn't fit all.

Correct. I'm a journalist and author and while I love the iPad for taking notes (yes, typing on the screen -- which isn't nearly as bad as everyone says it is) for long form writing I'd prefer a physical keyboard.

Perhaps these A5 MBA's are just a testing platform for a whole new class of notebook positioned below the Air. They're just using Air hardware -- or something that resembles Air hardware.
 
This sounds like it would resemble the performance of PC "Netbooks", which Apple has explicitly stated they don't want anything to do with. I agree 100% with those who say that, if this rumor is true, it was for purely experimental purposes and no implementation will be possible until ARM performance comes close to that of Intel with significantly lower power usage... otherwise, what's the draw (no pun intended).

Apple refused to do netbooks because they know that a poorly performing Apple branded device (i.e. poor/choppy video playback, constant "rainbow wheels", etc) is only going to hurt their reputation... they would gain nothing from this in the long run.

There's a big difference between poor performance and low power, though. I wouldn't say the iPad has poor performance, but it certainly does last a long time on a single charge. Those Atom netbooks couldn't even play YouTube videos and were cheaply made. A theoretical ARM Air would almost be slotted north of $500.
 
Nope, just pointing out that a SSD is only an advantage in a disk I/O heavy workflow. ;) Which was my initial and still is my point really. You went off on a tangeant about main computer components (again, network is as much a main component in today's boxes as HDDs) and assumptions about my own workflow.

We're not talking about you, we're talking about the OP who I quoted and you chimmed in on. So by talking about your own personal needs, who went off on a tangeant?
 
how come they 've not done it already, I am wondering...maybe they were waiting to establish a developers market place under their control first so they could enforce programming directives better.
Apple needs to get their ducks in a row first. Mac OS X and iOS have to share the same core functionality, the hardware has to be powerful enough, and Apple needs to get their apps (iLife, iWork) running on anything from an iPad to a dual, 8-core Mac Pro with hyper-threading. The minimum requirements to make this happens are:
  1. iOS 5 and OS X "Lion"
  2. 1.5GHz, quad-core processor (at least for the MacBook Air)
More than likely, iOS and OS X "Lion" are going to be used for "testing" and we will not see an ARM processor on the Mac until the next OS update. Look what Apple has already done with/to iCal. To change the day/week/month view you used a slider, but now you press a button. UI elements could change based on the device it is on, so when an app is installed on a Mac you press buttons, but if it is installed on an i<device> then you use a slider. I can even see the same thing happening with input devices (fingers on screen, trackpad, keyboard, mouse, etc).

To make this a seamless as possible, Apple might combine iOS and OS X into one OS which would need Xcode 5 to create the binaries.
 
Those Atom netbooks couldn't even play YouTube videos and were cheaply made.
I sprinkled a bit of Apple magic on my Dell Mini 9 netbook, allowing it to run Mac OS X. Not only can it play YouTube videos, it can play them full screen without dropping frames. Running Windows XP on the same netbook causes it to take about 3x longer to boot and YouTube videos drop frames like crazy.

It's all about the software. The hardware existed for a long time. I think Apple could do some amazing things with the combination of ARM and Mac OS X. Hell, if Microsoft can get Windows running on ARM (albeit poorly from the reports I heard), Apple can make OS X sing on ARM.
 
Not excited about this at all. If they end up going this way for the MBA, they'll never have me as a customer. I'd stick with the MBP as long as it exists.
 
Wirelessly posted (Mozilla/5.0 (iPhone; U; CPU iPhone OS 4_3_3 like Mac OS X; en-us) AppleWebKit/533.17.9 (KHTML, like Gecko) Version/5.0.2 Mobile/8J2 Safari/6533.18.5)

KnightWRX said:
I don't know all the technical challenges, but exactly how many "power users" buy a MacBook Air?

*Raises hand*. It's also my primary computer. Not all power users are about sheer CPU power. In my case, I use the shell and the Unix tools a lot. The MBA is plenty of computer for that.

I dabble some in graphics, but only when I make some websites/iOS stuff, I could do these graphics on my Pentium 2 333 mhz in Gimp if I really needed to (so it's not sweat for the MBA).

You aren't the only one! I've seen many posts from people using AIRs in advanced or demanding ways.

Personally I don't have an AIR (yet) but I'm tempted to replace my old 15" MBP with one. Being an old fart the biggest thing keeping me from doing so is the small screen followed by the lack of storage. These are things Apple could easily address in the future.

Sometimes people mix up the concepts here. A power user in my mind is anyone conversant with his machine and the tools required for whatever he is working on. Frankly there are not many power users out there with most people having a casual relationship with their machine.

Back in the day there where many HP calculator power users out there. That is a tiny device compared to the AIR but the same concept applies. For every one HP calculator power user there where tens of more casual users.

In the end the new AIRs are very capable. With a Sandy Bridge implementation they will simply attract more users.
 
There is no reason why ARM can't blow the doors off of X86. There is no reason why Intel might not be thinking of moving to an ARM based architecture.

X86 is getting old. It needs to die someday. Now is as good a time as any.
 
Checkmate to Intel

I think it will take 2 to 3 years to go ARM in any laptop with considerable performance and battery life.

Intel CPU is the expensive part in $999 MacBook Air (almost $250 to $300), I was surprised Apple gotta good discount on LV/ULV CPUs and making profits on the MBA (35% margin).

so this rumor is doing checkmate to Intel and see how the industry plays out in next 24 to 36 months.

of course these are prototypes are on the work to see how ARM performs in a laptop form

remember there was a rumor Samsung working on Desktop class ARM CPUs

give it a few years we will be there, not now :D
 
Think outside the box

Many people are speculating that the Macbook Air will ONLY run on the A5. But what if they add an A5 and offload things like audio/video playback or iOS apps to it? This could lower the Intel chip usage and increase battery life...
 
Many people are speculating that the Macbook Air will ONLY run on the A5. But what if they add an A5 and offload things like audio/video playback or iOS apps to it? This could lower the Intel chip usage and increase battery life...

Interesting point.
 
Or think 2 boxes...

Many people are speculating that the Macbook Air will ONLY run on the A5. But what if they add an A5 and offload things like audio/video playback or iOS apps to it? This could lower the Intel chip usage and increase battery life...
Or Apple could release an A5 Air as a new machine and not a replacement for the Intel Air.

The 2 different Air lines would be for different classes of users.

I do think it would need a name differentiation if they did have both.
 
They were testing Intel processors since basically the first release of OS X, Apple wants to keep it's options open. If they made the switch to ARM it wouldn't be for another 2 years most likely
 
There is no reason why ARM can't blow the doors off of X86.
Yes there is a reason why intel blows the door off of ARM based processors. They're orders of magnitude faster. ARM is great low powered mobile devices, not for desktop computing needs. How many cores does the typical ARM processor currently have 2? How many does the core i7 have with hyper-threading 8. Additionally Sandy Bridge has shown to be incredibly faster over the prior edition, never mind ARM based products.
 
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.