Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
Implementations of the new chip designs, which will deliver performance in excess of 2 GHz while drawing less than 0.25 W of power per CPU, will be available in the fourth quarter of this year.

This should be reworded:

The new core designs, which will deliver performance in excess of 2 GHz while drawing less than 0.25 W of power per CPU, will be to available to chip designers in the fourth quarter of this year. Licensees can then design chips incorporating the new core.
 
If your cell phone was 2GHz, you wouldn't use the maximum processing ability of the processor most of the time. The phone could come with a cooling fan (with a door that electrically closes when not running) that only runs when the processor is being used heavy-duty. For the rest of the time, the fan would not be needed. :apple:
 
Just as we lost MacOS when Apple switched from Motorola 68040 processors to PowerPC. And then we lost MacOS X again when Apple switched from PowerPC to Intel. No, wait a second, actually we didn't! MacOS X runs just fine on PowerPC and on Intel. And an ARM processor is much more similar to the Intel x86 processors than PowerPC is.

Now putting an ARM processor into an iMac doesn't make sense because the low power advantage doesn't count for much on a desktop processor, but if Apple wanted to create an ultra portable laptop with tremendous battery life by using ARM processors, there would be nothing stopping them.



Let me see...

Start XCode, open your project, open "Targets", click on your Target, Cmd-I for Info, and under the "Architecture" setting you add "arm". That's it. MacOS X is portable. For years when MacOS X ran on a PowerPC only, Apple had a secret project to make it work on Intel processors as well, and any Apple developer who wrote non-portable code got whacked over the head unless all code was portable. And all the external developers have learned how to write portable code as well. Today, any code that is written by a Macintosh developer with half a brain will work on any processor.

Hopefully Apple will make an ULTRA-LIGHT Mac (400 g or less; either a tablet, with sliding keyboard or clamshell) with video-out and USB ports, and Microsoft has half a brain for Office.
 
I still think the reason Apple kicked PPC support out of SL is so they can bring in ARM support. x86/ARM fat binaries. You read it here first...

It makes a lot of sense when you put the pieces together. Why would they spend so much time overhauling the inside of SL to be more efficient (to make sure it compiles and runs well on ARM).

You and carmenodie have proven once again why there needs to be the death penalty for people who say stupid Scheiße. Please, make a contribution but don't say stupid Scheiße that is so moronic it makes my head hurt.
 
The Cortex-A9 speed-optimized hard macro implementation will provide system designers with an industry standard ARM processor incorporating aggressive low-power techniques to further extend ARM's performance leadership into high-margin consumer and enterprise devices within the power envelope necessary for compact, high-density and thermally constrained environments. This hard macro implementation operates in excess of 2GHz when selected from typical silicon and represents an ideal solution for high-margin performance-oriented applications.

And now in English? Who writes this babble?
 
Can someone explain to me why an ARM processor is so power efficient, while my Intel uses ~20 watts, and my desktop uses closer to ~100 watts?

A decades worth of multiple architectural and implementation design decisions.

For desktop CPUs, if some design trick (more functional units, more pipelining, more instruction reorder buffers, cache, etc.) could make the CPU run 1% faster, and the CPU didn't melt (didn't exceed an expensive package's TPD with a monster heat sink), that made sense for companies competing for the fastest desktop and server CPUs.

Whereas ARM, relegated to the embedded market, only added a design improvements that made the CPU 1% faster if it used less than 1% more power. Thus fewer performance tricks could be added, only the most power efficient ones.

Repeat that decision hundreds of times over the years as the processor families went from 10's of MHz to more than a GHz, and the performance/power differential became huge.

Acorn also started out with a cleaner RISC-like architecture that required fewer layers of tricks and additions to make it run fast, whereas the x86 architecture was already a layer on top of some ease of 8080 code porting requirements.
 
Some people may dismiss the idea of Apple putting ARM chips in their desktop Macs, but you should never say never when it comes to Apple :)

As some of you will know, ARM was developed by Acorn for their Archimedes line of desktop computers in the mid to late 80's and they could easily give a 386/486 machine a run for its money.

While today the primary use of ARM chips is in mobile and embedded platforms (the only recent "classic" desktop ARM based PC is the now discontinued Iyonix) I believe with computers getting more and more compact and portable, plus a growing desire for more efficient systems, we could see the ARM platform make some headway in the non embedded market.
 
ARM in desktop

I think the biggest problem with an ARM chip in anything but a device is the lack of 64-bit.
 
I think the biggest problem with an ARM chip in anything but a device is the lack of 64-bit.

I think that the biggest problem would be the lack of x86 compatibility.

Look at a Windows netbook - just about any Windows software will install and run on it. I have the standard Photoshop Elements on my Win7 Atom-based eeePC. Runs fine (a bit slower than my Core i7, but fine). Anything that I've wanted to install on the netbook installs and runs fine.

Think about an Apple netbook/tablet with an ARM processor. Won't run any Mac software. Fail. It would be a big Ipod Touch, not a small Mac.
 
I think that the biggest problem would be the lack of x86 compatibility.

Look at a Windows netbook - just about any Windows software will install and run on it. I have the standard Photoshop Elements on my Win7 Atom-based eeePC. Runs fine (a bit slower than my Core i7, but fine). Anything that I've wanted to install on the netbook installs and runs fine.

Think about an Apple netbook/tablet with an ARM processor. Won't run any Mac software. Fail. It would be a big Ipod Touch, not a small Mac.

Maybe that is the aim - to sell an over sized iPod Touch as a tablet - it would be sufficiently different from the rest as not to raise the same level of expectations people have from a standard laptop and it could tap into the iPhone OS software ecosystem - they might need to tweak some aspects of the interface due to the larger screen size.
 
I think that the biggest problem would be the lack of x86 compatibility.

Look at a Windows netbook - just about any Windows software will install and run on it. I have the standard Photoshop Elements on my Win7 Atom-based eeePC. Runs fine (a bit slower than my Core i7, but fine). Anything that I've wanted to install on the netbook installs and runs fine.

Think about an Apple netbook/tablet with an ARM processor. Won't run any Mac software. Fail. It would be a big Ipod Touch, not a small Mac.

For the tablet, most of the software will come up from the iPhone not down from the Mac. The adaption to a touch interface has already been done for the iPhone, and Mac software wouldn't be a good fit.

I think developers know how to recompile their software to run on a hot, new Apple platform. At this point, even Adobe and Microsoft are getting the hint that Carbon is going to go away (heck, Lightroom is 64-bits and Cocoa). I just think that the lack of 64-bit on the ARM is a serious problem for non-devices such as iPhone / iPod touch / iTablet.
 
For the tablet, most of the software will come up from the iPhone not down from the Mac. The adaption to a touch interface has already been done for the iPhone, and Mac software wouldn't be a good fit.

As I said, it would be a big Ipod, not a small Mac. None of your existing programs will work, you'd have to buy new apps (and the control freaks at Apple would probably make it so that you had to buy them from Apple).

One of the main advantages of a Windows netbook is that it runs almost any app that you already have. No need to buy a new app, and no need to learn a new app because the one you use isn't available for the Maxi-Pod.
 
CPU choices

There could be many interesting strategies:

-A9 ARM running iphone apps as widgets and a recompiled ARM OS X (new INTEL/ARM universal binaries, new XCode)

-ARM CPU running iphone apps only in a light desktop version of iPhone OS

-A9 ARM +X86 processor "cocktail"... X86 idle most of the time, ARM running iphone apps. When an X86 OS X app is running the X86 kicks in. OpenCL (not limited to GPU) used for ARM co-processing whenever possible to lower the X86 charge.

-X86+ GPU, running iphone apps in an emulator: it could work but it'd draw a lot of power from an Atom CPU to do that.


-vaporware: nothing at all :)

A tablet needs to stay cool to be usable, you also need a very good battery life, ideally making having to carry a charger with you optional.

You also need a nice-looking GUI. You won't just have fullscreen zoomed-in iphone apps. Maybe not the full OSX but OSX is not unusable for a touch interface with a few tweaks. It hasn't stopped MS launching TabletPC versions of the full version of XP. Innovation is needed here.

The Archos 9 proves an Atom tablet is possible but drains the battery.

Let's see!
 
Seems to me that regardless of iPhone, Mac, mythical tablet, what Apple needs to control in order to have complete control of the Hardware is not the CPU or GPU but IOhub (southbridge, chipset,..., would a rose by any other name still do the same job).

If they control how information moves between the parts then they can do anything they like with the hardware. Remember one of the things PA Semi where really good at is low power PCI routers and similar IO switching systems.

They could build Apple a SOC that worked for both the iPhone and Mac. The iPhone one might bundle additional ARM Cores and GPUs that the Mac one doesn't. Both would handle all IO, like the custom trackpad/touchscreen engine they use already, also things like wifi and 3G.

The interesting thing this might open up in conjunction with 64bit memory space is a flash memory bus as part of the system memory. So instead of adding SSD constrained by the SATA interface. They could have flash chips on a memory bus on a module similar to the current DIMMs, sure not going help much with the write speed much but read speed would be huge.

Some thing like the Air (or Mythical Tablet) would be a good place to debut the system.
 
Can't wait until I get my first iPhone next year after my current contract is up. This will be very exciting with this new processor. :D

Hopefully they make the screen as high quality as the new HD Zune but this could result into poor battery life. Battery life is a big key though. I'd rather see more battery juice in the next generation. I'd also like to see the phone slimmer and lighter if possible.
 
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.