I'd disagree. With Lion's iOS features and the general dumbing down of Apple's "pro" applications, I could really see Apple blowing up the MacBook and betting their fanbase/iPhone base won't blink. The only people that would mind would be the .01% who do more than facebooking and emailing, especially once iCloud gets the kinks worked out.
This ARM switcher talk is nothing more than naive ARM fan dreams with no basis in reality. ARM has no magic pixie dust, their CPUs are MUCH slower than state of the art desktops like Sandybridge.
The complexity/power use of modern desktop chips is a symptom of the performance level not the ISA (Instruction Set Architecure).
For any ARM chip to get anywhere near desktop levels of performance, ARM chips will need more complex OOO architectures, huge caches, high performance IO, high performance memory interfaces. ARM Power consumption would be similar at this point and the advantage of ARM would be what?
Every Apple CPU transition (68K-PPC, PPC->x86) had two factors in common:
1) The new architecture was already or projecting to eclipse the previous on performance.
2) Extensive CPU emulation of the previous architecture.
ARM is unsuitable from both perspectives, it is only projected to stay well behind x86 desktops on performance for the foreseeable future. Since it is so much slower it is completely unsuitable for emulation of the previous architecture.
Apple has already placed it's ARM bet and it is in iOS, if they want to create an ARM netbook, it will use iOS, not OSX.
This ARM switcher talk is nothing more than naive ARM fan dreams with no basis in reality. ARM has no magic pixie dust, their CPUs are MUCH slower than state of the art desktops like Sandybridge.
This is very true and for those of us looking for far more power in our next machine it kinda rules out ARM based products. However not everybody is concerned with performance or they are concerned with facilities that require special hardware like video decode.
Of course the problem with the low end is this, ARM won't be competing with Sandy Bridge but rather chips like AMDs Zacate and Ontario chips. Possibly also Intels Atom if they ever fix that family of chips. in this regard AMD has a rather huge lead over ARM giving good enough performance for some in laptop like machines.
Quote:
The complexity/power use of modern desktop chips is a symptom of the performance level not the ISA (Instruction Set Architecure).
Well again yes and no. I86 carries around a lot of extra baggage from way back. Current i86 processors are not what you would get from an attempt to do a ground up processor design today. Not all of that baggage is benign. Outside of the processor core though I have to agree, processor performance is dependent upon high performance gear supporting on and off chip.
Quote:
For any ARM chip to get anywhere near desktop levels of performance, ARM chips will need more complex OOO architectures, huge caches, high performance IO, high performance memory interfaces. ARM Power consumption would be similar at this point and the advantage of ARM would be what?
Well yes this is true in the sense of traditional architectures, however most organizations going the ARM route are also going the SoC route. The more hardware that can be put on that SoC the more you turn the power equation to your favor. The other thing is that these days cores count and so do specialized execution units.
Quote:
Every Apple CPU transition (68K-PPC, PPC->x86) had two factors in common:
1) The new architecture was already or projecting to eclipse the previous on performance.
2) Extensive CPU emulation of the previous architecture.
Yeah this one I've always had problems explaining to people that truly believed their G5 was oh so fast. G5 never was in a general sense.
Quote:
ARM is unsuitable from both perspectives, it is only projected to stay well behind x86 desktops on performance for the foreseeable future. Since it is so much slower it is completely unsuitable for emulation of the previous architecture.
Apple has already placed it's ARM bet and it is in iOS, if they want to create an ARM netbook, it will use iOS, not OSX.
This I agree with 100%. What they create might not even be a traditional net book/notebook. IOS gives them a great deal of freedom to experiment with other formats. For example if Apple could do a book format, that is a device that open like a book to produce a seamless screen they would. The technology isn't there yet but I'm sure Apple is looking into it. Another option would be a tablet with a slide out keyboard.
It is fair to say Apple has many options when it comes to ARM. I just don't see ARM running Mac OS/X as viable as it would lead to significant market confusion. This might all change in a couple of years, but right now there are to many negatives with respect to ARM driving a MBP like machine.
There's certainly a substantial delay between Steve Jobs vision on future laptops and what a majority of the mac user base thinks. This delay will perhaps be fixed when new MPBs begin to look like MBAs. I don't live in that delay because I use my MBA for demanding tasks (CAD, graphics, math), and yes, it's fast.
I see many people will be disappointed when the next 15'' MBP will have the MBA form factor while outperforming the current MBP offering. I see no reason to enjoy weight and size in a laptop, but there seems to be a lot of people who really wish laptops to be heavy.
To put it plainly if I'm about to shell out multi thousands of dollars I expect all the performance the technology will allow for at that time. There are certainly things that can be done to make a new MBP thinner and lighter, but it needs to stay in the 35 watt envelope for processor performance.
Another way to look at this is that I don't want to see Apple go back to marketing slow underpowered machines relative to the rest of the market. Fast is a relative thing and also reflects ones expectations but the current AIRs are not fast.
Remember, Apple is all about profit margins, and switching to ARM will significantly increase those margins!
That's not necessarily the case. Ipads and iphones are made in much higher volume. Scaling performance up would be quite a lot of work, and it's not a definitively great idea. Companies like Intel sell the same cpus to a number of oems, which brings down the cost significantly. For Apple this may not work out so well. Right now you're theorizing without any facts to back it up.
Quote:
Originally Posted by wizard69
Well again yes and no. I86 carries around a lot of extra baggage from way back. Current i86 processors are not what you would get from an attempt to do a ground up processor design today. Not all of that baggage is benign. Outside of the processor core though I have to agree, processor performance is dependent upon high performance gear supporting on and off chip.
If they were designing from the ground up, either architecture would look old at this point.
Quote:
Originally Posted by wizard69
It is fair to say Apple has many options when it comes to ARM. I just don't see ARM running Mac OS/X as viable as it would lead to significant market confusion. This might all change in a couple of years, but right now there are to many negatives with respect to ARM driving a MBP like machine.
I don't know that we'll ever see it on OSX. By the time it's a consideration, OSX may just be too old.
Quote:
Originally Posted by wizard69
To put it plainly if I'm about to shell out multi thousands of dollars I expect all the performance the technology will allow for at that time. There are certainly things that can be done to make a new MBP thinner and lighter, but it needs to stay in the 35 watt envelope for processor performance.
Another way to look at this is that I don't want to see Apple go back to marketing slow underpowered machines relative to the rest of the market. Fast is a relative thing and also reflects ones expectations but the current AIRs are not fast.
It's definitely skipped around a bit. Right now the Windows oems definitely charge less than you'd pay for a macbook pro with equivalent cpu, gpu, and ram. If you want a Mac, you still buy a Mac. I did a comparison on this a while ago. You're paying for possibly fit and finish, and OSX. I didn't compare to anything cheap, and I also left out the "mobile workstations" because those are just a totally different kind of design. Anyway regarding power and wattage, the 35W laptop cpu may soon be a thing of the past if you believe Intel. What would concern me more is that it's always possible Apple may throttle cpus again to keep them running cool enough (somehow I remember them doing that on the original air). Recall the dynamic tdp feature Intel mentioned? For Apple it may be an acceptable compromise even if the machine cannot run at maximum performance for an extended period of time.
Comments
?the general dumbing down of Apple's "pro" applications?
The only people that would mind would be the .01% who do more than facebooking and emailing,
And I'm accused of insane hyperbole.
Remember, Apple is all about profit margins, and switching to ARM will significantly increase those margins!
No. Apple is about experience. And ARM sucks eggs right now. It's nothing compared to Intel.
I'd disagree. With Lion's iOS features and the general dumbing down of Apple's "pro" applications, I could really see Apple blowing up the MacBook and betting their fanbase/iPhone base won't blink. The only people that would mind would be the .01% who do more than facebooking and emailing, especially once iCloud gets the kinks worked out.
This ARM switcher talk is nothing more than naive ARM fan dreams with no basis in reality. ARM has no magic pixie dust, their CPUs are MUCH slower than state of the art desktops like Sandybridge.
The complexity/power use of modern desktop chips is a symptom of the performance level not the ISA (Instruction Set Architecure).
For any ARM chip to get anywhere near desktop levels of performance, ARM chips will need more complex OOO architectures, huge caches, high performance IO, high performance memory interfaces. ARM Power consumption would be similar at this point and the advantage of ARM would be what?
Every Apple CPU transition (68K-PPC, PPC->x86) had two factors in common:
1) The new architecture was already or projecting to eclipse the previous on performance.
2) Extensive CPU emulation of the previous architecture.
ARM is unsuitable from both perspectives, it is only projected to stay well behind x86 desktops on performance for the foreseeable future. Since it is so much slower it is completely unsuitable for emulation of the previous architecture.
Apple has already placed it's ARM bet and it is in iOS, if they want to create an ARM netbook, it will use iOS, not OSX.
This ARM switcher talk is nothing more than naive ARM fan dreams with no basis in reality. ARM has no magic pixie dust, their CPUs are MUCH slower than state of the art desktops like Sandybridge.
This is very true and for those of us looking for far more power in our next machine it kinda rules out ARM based products. However not everybody is concerned with performance or they are concerned with facilities that require special hardware like video decode.
Of course the problem with the low end is this, ARM won't be competing with Sandy Bridge but rather chips like AMDs Zacate and Ontario chips. Possibly also Intels Atom if they ever fix that family of chips. in this regard AMD has a rather huge lead over ARM giving good enough performance for some in laptop like machines.
The complexity/power use of modern desktop chips is a symptom of the performance level not the ISA (Instruction Set Architecure).
Well again yes and no. I86 carries around a lot of extra baggage from way back. Current i86 processors are not what you would get from an attempt to do a ground up processor design today. Not all of that baggage is benign. Outside of the processor core though I have to agree, processor performance is dependent upon high performance gear supporting on and off chip.
For any ARM chip to get anywhere near desktop levels of performance, ARM chips will need more complex OOO architectures, huge caches, high performance IO, high performance memory interfaces. ARM Power consumption would be similar at this point and the advantage of ARM would be what?
Well yes this is true in the sense of traditional architectures, however most organizations going the ARM route are also going the SoC route. The more hardware that can be put on that SoC the more you turn the power equation to your favor. The other thing is that these days cores count and so do specialized execution units.
Every Apple CPU transition (68K-PPC, PPC->x86) had two factors in common:
1) The new architecture was already or projecting to eclipse the previous on performance.
2) Extensive CPU emulation of the previous architecture.
Yeah this one I've always had problems explaining to people that truly believed their G5 was oh so fast. G5 never was in a general sense.
ARM is unsuitable from both perspectives, it is only projected to stay well behind x86 desktops on performance for the foreseeable future. Since it is so much slower it is completely unsuitable for emulation of the previous architecture.
Apple has already placed it's ARM bet and it is in iOS, if they want to create an ARM netbook, it will use iOS, not OSX.
This I agree with 100%. What they create might not even be a traditional net book/notebook. IOS gives them a great deal of freedom to experiment with other formats. For example if Apple could do a book format, that is a device that open like a book to produce a seamless screen they would. The technology isn't there yet but I'm sure Apple is looking into it. Another option would be a tablet with a slide out keyboard.
It is fair to say Apple has many options when it comes to ARM. I just don't see ARM running Mac OS/X as viable as it would lead to significant market confusion. This might all change in a couple of years, but right now there are to many negatives with respect to ARM driving a MBP like machine.
There's certainly a substantial delay between Steve Jobs vision on future laptops and what a majority of the mac user base thinks. This delay will perhaps be fixed when new MPBs begin to look like MBAs. I don't live in that delay because I use my MBA for demanding tasks (CAD, graphics, math), and yes, it's fast.
I see many people will be disappointed when the next 15'' MBP will have the MBA form factor while outperforming the current MBP offering. I see no reason to enjoy weight and size in a laptop, but there seems to be a lot of people who really wish laptops to be heavy.
To put it plainly if I'm about to shell out multi thousands of dollars I expect all the performance the technology will allow for at that time. There are certainly things that can be done to make a new MBP thinner and lighter, but it needs to stay in the 35 watt envelope for processor performance.
Another way to look at this is that I don't want to see Apple go back to marketing slow underpowered machines relative to the rest of the market. Fast is a relative thing and also reflects ones expectations but the current AIRs are not fast.
Remember, Apple is all about profit margins, and switching to ARM will significantly increase those margins!
That's not necessarily the case. Ipads and iphones are made in much higher volume. Scaling performance up would be quite a lot of work, and it's not a definitively great idea. Companies like Intel sell the same cpus to a number of oems, which brings down the cost significantly. For Apple this may not work out so well. Right now you're theorizing without any facts to back it up.
Well again yes and no. I86 carries around a lot of extra baggage from way back. Current i86 processors are not what you would get from an attempt to do a ground up processor design today. Not all of that baggage is benign. Outside of the processor core though I have to agree, processor performance is dependent upon high performance gear supporting on and off chip.
If they were designing from the ground up, either architecture would look old at this point.
It is fair to say Apple has many options when it comes to ARM. I just don't see ARM running Mac OS/X as viable as it would lead to significant market confusion. This might all change in a couple of years, but right now there are to many negatives with respect to ARM driving a MBP like machine.
I don't know that we'll ever see it on OSX. By the time it's a consideration, OSX may just be too old.
To put it plainly if I'm about to shell out multi thousands of dollars I expect all the performance the technology will allow for at that time. There are certainly things that can be done to make a new MBP thinner and lighter, but it needs to stay in the 35 watt envelope for processor performance.
Another way to look at this is that I don't want to see Apple go back to marketing slow underpowered machines relative to the rest of the market. Fast is a relative thing and also reflects ones expectations but the current AIRs are not fast.
It's definitely skipped around a bit. Right now the Windows oems definitely charge less than you'd pay for a macbook pro with equivalent cpu, gpu, and ram. If you want a Mac, you still buy a Mac. I did a comparison on this a while ago. You're paying for possibly fit and finish, and OSX. I didn't compare to anything cheap, and I also left out the "mobile workstations" because those are just a totally different kind of design. Anyway regarding power and wattage, the 35W laptop cpu may soon be a thing of the past if you believe Intel. What would concern me more is that it's always possible Apple may throttle cpus again to keep them running cool enough (somehow I remember them doing that on the original air). Recall the dynamic tdp feature Intel mentioned? For Apple it may be an acceptable compromise even if the machine cannot run at maximum performance for an extended period of time.