Apple Testing ARM Based Mac Prototypes with Large Magic Trackpad?

[sehix]

when was the last time Apple came out with a truly low cost machine that was significantly cheaper than its regular models or competed in the low end market for computers?

The closest to that is the iPhone 5c which hasn't exactly been a huge success nor particularly cheap for that matter.

Yeah, it's only, what, the #3-selling smartphone in the U.S. right now.

----------

Motorola... Let me think...

They spun off their hardware business in 2004. Called Freescale. Then they split into Motorola Solutions and Motorola Mobility. Motorola Mobility was bought by Google. Motorola Solutions is what remains. A shadow of its former self.

Motorola in 2004 was not Motorola 1997. Which was most of the (missed) point (by the OP).

----------

PowerPC was developed by multiple companies (Motorola and IBM) which both were _not_ interested in the market that Apple was interested in. If you haven't noticed, Apple is actually developing ARM processors itself and is ahead of everyone else.

The AIM consortium developed PPC. Apple-IBM-Motorola.

They *all* had input into design, two manufactured, one didn't.

 

[deconstruct60]

There's no forking in browser-based Javascript!

so if open a new window/tab and start a another. For a multiple window/tab browser being able to "walk and chew gum at the same time" would be highly beneficial.

Multiple instances running is multiple paths of execution running.

The classic 'fork' in Unix got you a whole new, heavyweight, execution process. A new window/tab isn't all that far off from a that.

 

[sehix]

Change processors again and I'll never buy another Mac again. The downgrade from PowerPC to Intel was enough for me.

Having recently used one of my old G4 PowerBooks ... the difference between then and now is like night and day. The current rMPB is hugely better than my old 1GHz TiBook in every way imaginable.

 

[MrCheesecake]

Hilarious. These are RUMORS and no one other than Apple knows the details. I find it incredibly hard to believe that Apple would abandon x86 architecture altogether. This will probably be an addition to the product line in a way none of us really anticipates right now.

x86 wasnt an addition to ppc, it was a replacement. Apple's recent activity coincides with them just changing things without warning. Like removing sync services in Mavericks. Then after people bitching through 10.9 10.9.1 and 10.9.2, they finally brought it back in 10.9.3 after realizing thats not what people want.

Are they going to do the same thing with hardware too?

And the recent rumors have been true for the most part. The iOS redesign, the new features in Mavericks, the ipad mini, retina macbooks, fcpx being torn apart and destroyed, and the 4.7" iphone 6 already has a design frame released to case makers thats pretty similar to the renderings from late last year.

The days of ultra secrecy in Apple are dead. We know what they're releasing long before they announce it. If they're testing it, we'll probably see it soon. I just hope they don't switch everything over to ARM. If they make it another product line, awesome, I'll probably get one, but as a replacement, I really hope they don't switch to ARM.

 

[Goftrey]

I for one am totally digging this. Have never been a fan of the Intel Macs, heck I still use a G4 based PowerBook everyday. Getting developers on board shouldn't be too much of an issue, especially given the fact the Mac as a platform is now a massive player, especially in audio and visual design. Apple needs to do something bold in order to 1. kick start a movement, or 2. simply to get Intel's arses in to gear. IMO they've become far too complacent these last 7/8 years.

 

[SlCKB0Y]

Fact: Intel CPU performance has been stagnating for many years. Apple A series performance has skyrocketed. If this trend continues it will be criminal NEGLIGENCE to continue using Intel instead of ARM.

1. Regardless of performance, ARM is still a very specialised architecture, a far cry from the general purpose x86_64 processors and not suitable for desktop use.

2. ARM is still nowhere near the performance of the desktop or enterprise class Intel CPU's. People who claim it is are almost always comparing it to Atom, which is not "desktop class" or at the extreme low end of desktop class.

3. To be comparable, ARM will have to add many more extensions to graduate from a mobile CPU to a more generalised design and since it will need to virtualise x86, it needs to be as powerful as x86 and have a lot to spare (since we can no longer paravirtualise Windows and have to fully virtualise it)

4. x86 hasn't progressed as fast in performance as it otherwise would have because they refocussed on power efficiency (no doubt at the urging of Apple).

5. Apple would have the figures on the proportion of people using bootcamp and they would realise that each one of those people will potentially be lost in a move to ARM.

6. At the time that Apple abandoned PPC, x86 was outperforming it both in terms of raw power AND power efficiency, to the point that the latest gen offerings from IBM could not even be used in portables. x86 is not even close to this situation.

7. Apple is known for hedging their bets (eg. Compiling every pre-Tiger OS against x86), so this news is hardly surprising, but most likely doesn't mean much (if its even true to begin with).

For f***s sake, x86 is a 40 year old crap architecture that should have died decades ago. How many more years do you want to keep it alive? 10 years? 100 years? LET IT GO!

ARM is nearly 30 years old. How is this relevant at all? OS X has it's roots in the late 1960's, how is this relevant?

Why is x86 a "crap" architecture?

 

[wizard]

This is what they're moving towards by slowly eliminating upgrade functions in their laptops...soldered ram etc.

Soldered in RAM has nothing to do with a laptop being an iPad. I see no connection here at all. If anything soldering in your RAM makes a laptop more reliable. As for upgrades well we are coming upon a new generation of RAM that won't work in your old machine.

 

[Frost7]

2014 (or 2006 for that matter) is not 1997. There was a period of several years during which PPC outperformed Intel fairly well with the G3 and G4 families. Then IBM failed to deliver with the G5, and Motorola wandered off where the volumes/profits lay at the lower end, selling PPC processors for automotive applications. Switching to Intel made sense, particularly since Intel had made huge advances in performance/watt compared to just a few years earlier.

As for "a small company and IBM"... PPC was developed by the AIM consortium, which meant Apple, IBM and Motorola. Apple didn't make the chips, but shared in their design. And at the time, both IBM and Motorola not "small companies" by any industry measure.

A little bit of homework can put public embarrassment in abeyance, you know.

Not to mention Power architecture's seeming stall was not very long-lived. Not very long after the switch, IBM got a lot of their electrical/thermal issues under control, and Power has been consistently crushing everything else, especially Intel, in the server and high-performance computer market, especially when the Power7 arrived on the scene.

I would argue it's still a superior platform. If they really wanted to make competitive desktop CPUs, they could. There's just no reason because there's so much inertia behind x86. Heck, Microsoft almost went with a power efficient experimental 16-core Power CPU on the XBone, but ultimately went with the AMD Jaguar for cost.

Apple leaving at the time they did made sense though, and I don't begrudge them that. At the time there was no indication IBM would have most of their issues solved inside of two years. You can't go around betting the company on hope.

 

[jlc1978]

Originally Posted by jlc1978
when was the last time Apple came out with a truly low cost machine that was significantly cheaper than its regular models or competed in the low end market for computers?

The closest to that is the iPhone 5c which hasn't exactly been a huge success nor particularly cheap for that matter.

Yeah, it's only, what, the #3-selling smartphone in the U.S. right now.

However, Apple did not get the expected sales mix for the 5c which is why I said it wasn't a huge success. In addition, it isn't significantly cheaper than a 5S (which no doubt played a part in the sales mix) which gets back to my point that Apple doesn't do low cost devices. The sales results for the 5c bear out the correctness of Apple's strategy and why I doubt some low cost ARM based macbook is in the works; if anything I'd expect it, if ides exist, to be priced comp;arable to today's Intel based models.

 

[sehix]

Not to mention Power architecture's seeming stall was not very long-lived. Not very long after the switch, IBM got a lot of their electrical/thermal issues under control, and Power has been consistently crushing everything else, especially Intel, in the server and high-performance computer market, especially when the Power7 arrived on the scene.

I would argue it's still a superior platform. If they really wanted to make competitive desktop CPUs, they could. There's just no reason because there's so much inertia behind x86.

Apple leaving at the time they did made sense though, and I don't begrudge them that. At the time there was no indication IBM would have most of their issues solved inside of two years. You can't go around betting the company on hope.

Power is great in the server space, and probably would still be competitive for workstations/desktop. Not so sure it would be near to being competitive for laptops at all.

For consumers, and a lot of software developers, though, movement is very much away from desktops towards laptops.

Still sort of wish IBM had sorted out the issues that killed any G5 or later laptop-appropriate products back then.

 

[SlCKB0Y]

An ARM Mac would be a welcome change. I'm pretty sure within a couple of years every single mac outside of the mac pro will be using ARM processors. Most people don't need a metric ton of power for the things they do. I'm still using my 2009 C2D macbook and it runs everything just fine, and I'm pretty sure ARM can at least match that performance. I can't imagine why I need more power. You know what I would welcome though? Much better battery life and cooler operation.

Some of you "power users" really need to get out of your bubble.

By "power users", you must mean anyone who uses virtualisation, manipulates large images, video or audio, anyone who compiles software or a dozen other things.

Without those "power users" in their bubble, you wouldn't have much software to run on your Mac or phone and forget about music, movies or pretty websites.

Or I know, maybe those people should just spend half their day waiting for their computer to catch up with them because for you, a Macbook Air which will go 12 hours on a single charge whilst you facebook just isn't enough...

Even better, how about you buy a Mac which isn't 4 years old and you too can have much increased performance AND battery life so that the entire collective of "power users" doesn't have to change CPU architecture. The current x86 CPU processors run much cooler, for much longer whilst being much more powerful.

 

[wizard]

You sound like you know absolutely nothing about economics. It's not just a question about IF an ARM processor can become fast enough. It's about changing an entire software platform over to a new system and the CPU penalties and huge loss of developers that will be incurred in the process and all to accomplish WHAT, exactly? I haven't seen ONE thing mentioned that is better about ARM other than it runs cooler and I question that 100% given how flipping HOT my iPod 4G gets (and I mean HOT).

Performance per watt is really good with ARMs offerings. This is only part of the story though.

As for the software issue you are going way overboard here with your negativity. For most developers it is no problem at all. It is even less of a problem when people realize that they will be making gobs of money with this change.

I can see why Apple might be interested in it. They bring control of CPU updates in-house for one thing. Apple has always been a total control freak (well under Jobs anyway). But other than that, I see ZERO benefit.

You aren't looking hard enough. It isn't so much control of updates but rather control of the design. I've. Mentioned several times in this thread but the primary motivator here is that silicon is the new printed circuit board, the entire functionality of a computer, except for storage, will soon be in a single piece of silicon. Without access to that silicon or a vendor willing to work with you, you have no room to innovate. Right now the only vendor showing any willingness to work with customers in this manner is AMD.

If anything, they should be pushing Intel to get their CPUs more mobile-friendly so they can DITCH ARM on iOS instead. Then the platforms can properly unite.

That doesn't solve Apple access and engineering problem. Here is reality, unless Intel changes its way that processor will be a generic device overloaded with features for the general market. Apple would have zero ability to tailor that chip to their needs. This in a nut shell is the big problem with staying i86.

It would be simple to go from ARM to Intel because everyone is already using Apple's developer platform anyway. The Mac is OPEN and therefore people can develop whatever they want, however they want with the constraints of the OS (even make X11 apps instead if they desire).

You assume here that an ARM based Mac wouldn't be open! Apple could go either way here. However I think it is in their best interest to keep the platform open to the point that Linux can be loaded if wanted. In a literal sense they could set a new standard in PC hardware.

 

[Frost7]

Power is great in the server space, and probably would still be competitive for workstations/desktop. Not so sure it would be near to being competitive for laptops at all.

For consumers, and a lot of software developers, though, movement is very much away from desktops towards laptops.

True. Supposedly, software is a lot easier to move between ARM and PPC than between x86 and PPC/ARM, though... I suspect ARM-powered laptops with PPC-powered desktops could potentially have been a thing today if PPC hadn't gone off the rails for a few years a decade ago. Or maybe investment in low-power PPC would have made that a moot point. We'll never know.

Still sort of wish IBM had sorted out the issues that killed any G5 or later laptop-appropriate products back then.

You and me both, the CPU switch was a HUGE headache for me since I had a not-insignificant amount invested in PPC workstations. Hell, the Quad with the Quadro is still in service.

 

[deconstruct60]

Fact: Intel CPU performance has been stagnating for many years. Apple A series performance has skyrocketed. If this trend continues it will be criminal NEGLIGENCE to continue using Intel instead of ARM

Non fact. Last several tick tock iterations Intel CPUs have seen 8-15% general increases. It is not some revolutionary increase but it is far from stagnation. PC sales are stagnate (or negative), most vendors, including Apple, would kill now for a 10% yearly increase.


CPUs didn't radically shoot up consistently in past years either. Intel is managing the incremental roll outs but much of the work now is on expanding performance zones and making other trade-offs. It isn't just a crank clock rates and CPU drag racing results.

Fact: Intel engineering is way way way overrated. The only thing that keeps them afloat is their fabrication technology and market inertia.

Another non fact.

Tweaker overclocking and CPU drag racing are about as overrated as this notion of Intel engineering. Intel has a ton of engineering. All of it isn't better than everyone else. That doesn't mean Intel doesn't have a substantial set of world class folks. Head-to-head best best against best Intel competes with anyone else. Intel has broad product portfolio. Get into the variant products meant to add breath to product listing page and yes can get to match ups that are not clearly better and far more mundane.

Intel is suffering right now not so much from engineering shortfalls but a lack in scope of what they are making. They have also been suffering from a lack of competition ( AMD shooting themselves in the foot multiple times).
Intel can drift into lack of focus if there nobody to be paranoid about. (frankly Apple has a sometimes exhibited a very similar problem. )

For f***s sake, x86 is a 40 year old crap architecture that should have died decades ago. How many more years do you want to keep it alive? 10 years? 100 years? LET IT GO!

The latest large jump in ARM was it letting go of its legacy 32 bit ( and smaller) baggage. ARM isn't "new" and not without baggage. Apple's move to ARM v8 architecture was not for the 64-bits but to get to the new instruction set.

There isn't another major transition for x86 to make. ARM has some headroom largely because it is lagging; not more progressive.

By in large, x86 is all decoded into something else before actually executing. The current implementations have the first stab at caching those encodings. Those caches are only going to get bigger and more effective. The transistor budget of the x86-to-micro-op decoder is only going to get relatively smaller to total CPU package budget. Exposing the micro ops wouldn't particularly make execution go up a revolutionary amount.


Market inertia? At this point ARM has very similar baggage as x86.

 

[gnurf]

But when you think about what people do with their computers, it's hard to see why they need an i7/i5/i3. Facebook, mail, youtube, etc - how much power do you need?

Quite a lot, actually. If Ableton Live and Logic is your main tool, you need a beefy CPU for instruments and effects. I have instruments I couldn't really bother with before I went from C2D to i7.

 

[the-msa]

You assume here that an ARM based Mac wouldn't be open! Apple could go either way here. However I think it is in their best interest to keep the platform open to the point that Linux can be loaded if wanted. In a literal sense they could set a new standard in PC hardware.

You actually think anyone would buy a Mac to put a proper Linux on it? If this rumour is true, apple is just trying to ipadize their macs. Close the system so they have total control over it.

 

[SlCKB0Y]

It is true about A7. Without the limitation of the small battery, they can crank up the Mz and add a few more cores to handle graphics to make it a desktop class machine.

No it isn't "true" and no one knows how ARM will perform. These things don't scale linearly.

Maybe if they "crank up the Mz and add a few more cores", power consumption increases at a much greater rate and eliminates the benefit.

 

[Frost7]

Intel is suffering right now not so much from engineering shortfalls but a lack in scope of what they are making. They have also been suffering from a lack of competition ( AMD shooting themselves in the foot multiple times).
Intel can drift into lack of focus if there nobody to be paranoid about. (frankly Apple has a sometimes exhibited a very similar problem. )

This. Intel seems to start growing problems at a rapid rate when competition isn't holding their feet to the fire. Kind of the same with Nvidia. Both have been experiencing laziness and lack of focus from having to do nothing but keep plugging along slowly while AMD keeps tripping over their own feet.

Intel's nonstop Broadwell delays and Nvidia getting stupidly cocky with that overpriced Titan Z and stretching out their GPU roadmap unnecessarily accentuate the effect lately.

Look at the "improvements" in Intel's upcoming Devil's Canyon CPUs. In another time when AMD was roughly at parity, almost everything Intel did there would have been in the higher-end Haswells from the get go.

 

[SlCKB0Y]

You actually think anyone would buy a Mac to put a proper Linux on it?

Why not? I know at least a few that have..

 

[wizard]

And how do you tap all that "untapped" power?

1. Faster RAM. This could mean a wider path to RAM or it could mean new technology.
2. Faster secondary storage. Right now the interface to flash isn't exactly fast so a new interface would do wonders.
3. Double the clock rate. You have to remember that A7 only runs at 1.3 GHz, so their is plenty of room to move up. We don't even know where the current A7 tops out at.

Note that those three things only address the changes required outside of the chips main core. These are also three things that have a noticeable impact on the user experience.

 

[apolloa]

My point is that ARM CPU has not been yet to be designed to handle such tasks.

The PS3 used CELL to compute graphics...a dedicated nvidia card has been used for menus and such. CELL was never intended to be a desktop CPU but it was ahead of its time back then. It was difficult to code for that is true but that is mainly IBM/Sony's fault.
Apple knows how to develop things. They know how to attract developers etc.
They are not HP,IBM,Sony or any other of this fail rotten corps. It is time to change and we no longer need to be dependent on the OS/ architecture...that is not what matters anymore.

The CELL was used to do some graphics processing but the majority was done by the NVIDIA GPU, don't know where you got the idea from that the CELL did all the graphics, have you ever seen the size of the NVIDIA GPU in a PS3? It's bigger than the CELL. And if Apple did this it would prove nothing more than they know how to ruin their business.
I would never buy a Mac with an ARM CPU, I would always think why do I have a machine less capable than a cheap Windows desktop, and developers won't want to support it, why should they write programmes to work on two totally different platforms? I imagine the OS is bad enough, but of they have to code for the different hardware platforms too than I think many would not bother, they will look at the user base of Mac and not do it because Windows is so much bigger.

 

[wizard]

You actually think anyone would buy a Mac to put a proper Linux on it? If this rumour is true, apple is just trying to ipadize their macs. Close the system so they have total control over it.

Yes it happens. In fact more than a few people have done so for the same reason some people dual boot to Windows. Me, I run Linux in a VM on my MBP right now. However if i buy a new Mac soon you might see me wiping the old machine to install Linux. Why? Because of specific software I might want to run.

 

[SlCKB0Y]

Not to mention Power architecture's seeming stall was not very long-lived. Not very long after the switch, IBM got a lot of their electrical/thermal issues under control, and Power has been consistently crushing everything else, especially Intel, in the server and high-performance computer market, especially when the Power7 arrived on the scene.

That "stall" lasted for years. Maybe we should just stick Power7 in a MBA - i'm sure that would work well.

 

[flux73]

I really hope that this isn't true.

How do you know you don't want it when you haven't seen it, touched it, or know know how well it works? Is there something about ARM architecture that inherently limits it compared to one with x86 architecture?

 

[SlCKB0Y]

I could see a lower end ARM based OS X device (similar to a MacBook Air) if there wasn't a noticeable performance hit, it would be much cheaper than the current MacBook Air (US$899)

No, it wouldn't. The proportion of the total cost of a MBA which goes towards the CPU is small to the point of being irrelevant.