In light of the fact that we have very little substantiated information on exactly what Apple is doing, I'm just looking at the possibilites.
Good idea.
Apple has, of late, shown remarkable foresight in architecting OS X. They're designing for the future, not just the here and now.
The whole intel switch should bolster spirits. Making OS X hardware independent has apparently been a top priority. Not just in CPU choice but also with GPU choice (CI). Resolution independence also looks to be a major focus.
Considering this, I find it highly likely that apple has a team of 64bit code-monkeys going over their APIs with a fine tooth comb.
True, but that is not the only 64 bit architecture.
As the only architecture that Apple will be using is the x86 architecture with 64 bit extentions that AMD developed, that is the only relevent architecture to Apple, and thereby to us.
But at this point no one of authority has confirmed it.
Apple has given so little information on the its plans with Intel.
If it were going to follow the exact same path as the rest of the industry why hold out so much information? I don't see any strategic or business advantage in doing that.
If Apple is going to use EMT64, why not tell developers now, so they can prepare for it? This would empower developers to have 64 bit apps ready. That doesn't make sense to me.
At this point with an Apple/Intel collaboration. They pretty much have two choices.
They can follow what Microsoft dictates.
Or they can follow their own path.
There is little middle ground in this.
Apple has pretty much always gone its own way and has never directly followed MS policy.
Intel seems to be in a place where they have come to realize MS has more power to lead PC standards than they do. And Intel doesn't seem pleased with the situation.
From the evidence I've seen. I don't believe the Apple/Intel collaboration will follow the larger PC industry and its MS set standards.
I believe the point of this collaboration is to be the leaders that the larger industry will follow.
If Apple is going to use alternative technologies from Intel, they will be by nature fairly easier for the Wintel world to adopt, since ultimately Intel's goal will be to use Apple as a fulcrum of adoption for the Wintel word, as they were for USB. You can damn well bet Intel didn't forget that lesson. But due to the nature of the fulcrum Apple's hardware differences will be in support chipsets such as BIOS, RAM/bus interfaces including the more mundane peripheral interfaces, form factor/motherboard designs and non-PC computer devices such as home electronics, tablets, etc.
Why? Because if Apple popularizes new markets, makes Media PCs more than 5% of the population gives a shit about, and ignites interest in high powered computing applications that not only allows Intel to sell more high margin product to Apple but after the laggard Wintel market adapts their products to more closely match Apple's then Intel also sells more high margin CPUs and fewer Celeron class low margin CPUs in the popularity spillover.
That is why Intel wanted Apple so bad. Intel sees the Wintel market as a commoditized market without an agent for growth and the coming game consoles as stealing part of its precious high margin gaming market, what AMD hasn't already seized, and one of the few higher margin areas left with good volume.
But Apple doesn't want custom CPUs or 64 bit architectures. That would kill the advantage of their deal with Intel. And Intel doesn't want to give them to Apple yet, not until they've gained market momentum, because Apple's fulcrum doesn't have enough leverage to move Wintel that much yet. You can take that to the bank and smoke it.
But at this point no one of authority has confirmed it.
Apple has given so little information on the its plans with Intel.
If it were going to follow the exact same path as the rest of the industry why hold out so much information? I don't see any strategic or business advantage in doing that.
If Apple is going to use EMT64, why not tell developers now, so they can prepare for it? This would empower developers to have 64 bit apps ready. That doesn't make sense to me.
With apple putting a P4 in the developer kits, we can assume that Apple is saying that x86 is the path they will be following. That's pretty obvious. The only path Apple can follow from there is the 64 bit path Intel is following. Intel is only following one 64 bit x86 path.
The only other 64 bit path Intel has is IA-64, and Apple certainly won't be following that one.
With apple putting a P4 in the developer kits, we can assume that Apple is saying that x86 is the path they will be following. That's pretty obvious. The only path Apple can follow from there is the 64 bit path Intel is following. Intel is only following one 64 bit x86 path.
The only other 64 bit path Intel has is IA-64, and Apple certainly won't be following that one.
I'm not so sure.
That Apple are following 32bit x86 for now is a definite.
But Intel haven't decided themselves on their 64bit path. So far it's been determined by Microsoft and AMD and Intel have reluctantly followed whilst complaining (rightly or wrongly) along the way that it's not a good solution.
There's still a lot of technology to flesh out in their 'next generation' architecture and that includes virtualization and who knows what there. Intel must be looking at salvaging something from Itanium. They've a customer now with a PowerPC codebase. The ultra low power chips are getting down to StrongARM territory. Maybe this is the point where Intel's chips become ISA agnostic and the 64bit path Intel takes is PowerPC for Apple computers, AMD64 for Windows, ARM for low power.
Oh, it's just speculation. But in the absence of any facts, that's all we've got - speculation.
Part of it is just my geekness being very very sad that the Intel X86 architecture has won out over every good architecture that has come along over the last 30 years including numerous attempts at dethroning it by Intel themselves.
Part of it is wondering what else is behind Intel's curtain as the IDF raised more questions than answers. And surely the future isn't what they've announced so far?
It does amaze me that many of the Apple 'faithful' were so quick to embrace Intel and the legacy X86 strain at that and immediately dismiss PowerPC. It's difficult to hold that position without being seen as a PowerPC fanboi.
I'm just not prepared to get laid by the fat girl even if that's the only girl left and you know she puts out. Intel is that fat girl and even though she's on a diet now, you know she's never going to be Elle McPherson.
Merom, Conroe and Woodcrest will be EM64T chips, much like the Prescott P4 chips are. There really shouldn't be any question that Apple will be using the x86 ISA, will be using EM64T extensions (a clone of AMD's x64 extensions), will be using MMX/SSE for SIMD, and whatever else technology Intel will implement on the platform which would be common to all Intel vendors.
As for 64-bit support in OS X/Intel, I don't think much can be based off of the transition kit yet. At minimum, it will be like it is now, 4+ GB memory support and 64-bit CLI process support. Anything more will require a bit of time.
Oh, it's just speculation. But in the absence of any facts, that's all we've got - speculation.
Part of it is just my geekness being very very sad that the Intel X86 architecture has won out over every good architecture that has come along over the last 30 years including numerous attempts at dethroning it by Intel themselves.
Part of it is wondering what else is behind Intel's curtain as the IDF raised more questions than answers. And surely the future isn't what they've announced so far?
It does amaze me that many of the Apple 'faithful' were so quick to embrace Intel and the legacy X86 strain at that and immediately dismiss PowerPC. It's difficult to hold that position without being seen as a PowerPC fanboi.
I'm just not prepared to get laid by the fat girl even if that's the only girl left and you know she puts out. Intel is that fat girl and even though she's on a diet now, you know she's never going to be Elle McPherson.
There isn't any speculation here. It's over. Intel has their roadmap laid out. AMD has theirs. There is no way that Intel would be allowed to make PPC's.They wouldn't even if they were allowed to. If by some odd and impossible deal, they, and Apple were going that way, it would have been announced at the convention in June, and the machines sent out would have reflected that.
Really guys, I know this is fun and all that, but it's over. Really, it is.
Apple will be using x86 chips. 32 bit now, and 64 bit when they're ready.
I'm sorry that some still don't want to believe it. But it's true. There is no where else for Apple to go. Apple would not have produced those dev. machines for a joke. Do you really think that they send out thousands of those machines because they were going to use a different chip? That's absurd. The developers would drop Apple like a stone, and you know it.
Our trip through Fantasyland is over. It's time for everyone to get off.
I've never said Intel would be making PowerPC chips, melgross, I theorised that the future held instruction set agnosticism and that something else was afoot.
x86 *WILL* run out of steam. The architecture is fundamentally flawed. It has too much baggage. It's hit many walls in it's history and we're approaching the point where new process technologies just can't keep up. It's already essentially a big legacy emulation engine attached to a RISC core. Wouldn't it be much better to bypass that legacy? That is partly why many people are confused with Apple's move to 32bit intel, not 64bit. There's no legacy to support, why start that now?
Intel have tried to shift many times only to be held back by Microsoft. They tried it with i860 and then Itanium. They aren't happy about EMT64. They've got technology in spades that is underused or not used by the worlds most popular OS. They've teams of very talented engineers that used to work on very advanced designs like Alpha, ARM and VLIW now working on squeezing performance out of X86.
I may be in fantasy land but I reckon this Intel transition marks the end of both x86 and PowerPC and just one step on the road to something else. Intel wants to shed it's legacy and Apple are the only ones that can provide a mainstream OS to do it. AMD can't provide the technology.
Otherwise I fail to see any compelling reason to switch from PowerPC to Intel. I'm sure you're going to argue 'What about the PowerBook' but it's going to take Apple a year to 18 months to get an Intel laptop out and IBM and Apple would have to be seriously incompetent to not be able to get low power PPC out in that timeframe. And with 970MP workstations around the corner they're going to have to go some on the Intel front to beat those too. So the next year to 18 months isn't the end game IMHO, it's a step down the road to platform independence.
The Developer kit isn't about getting people on to Intel, it's about getting them on to XCode, Cocoa and platform independent code. In order to create universal binaries, it's not just the intel code that has to be clean, there are restrictions on which PowerPC technologies you can use also. Then the next transition to whatever Apple and Intel have up their sleeve is an easy one. If, as I theorise, we're heading to agnosticism then the Intel code will just work anyway and Apple have pulled a fast one on us to get the codebase clean and into a state where they've killed off the OS9 cruft.
Call me mad but I just think there's something else going on. I've no idea what but I don't think the future lies in X86 or PowerPC, technology from 30 years ago.
I've never said Intel would be making PowerPC chips, melgross, I theorised that the future held instruction set agnosticism and that something else was afoot.
x86 *WILL* run out of steam. The architecture is fundamentally flawed. It has too much baggage. It's hit many walls in it's history and we're approaching the point where new process technologies just can't keep up. It's already essentially a big legacy emulation engine attached to a RISC core. Wouldn't it be much better to bypass that legacy? That is partly why many people are confused with Apple's move to 32bit intel, not 64bit. There's no legacy to support, why start that now?
Intel have tried to shift many times only to be held back by Microsoft. They tried it with i860 and then Itanium. They aren't happy about EMT64. They've got technology in spades that is underused or not used by the worlds most popular OS. They've teams of very talented engineers that used to work on very advanced designs like Alpha, ARM and VLIW now working on squeezing performance out of X86.
I may be in fantasy land but I reckon this Intel transition marks the end of both x86 and PowerPC and just one step on the road to something else. Intel wants to shed it's legacy and Apple are the only ones that can provide a mainstream OS to do it. AMD can't provide the technology.
Otherwise I fail to see any compelling reason to switch from PowerPC to Intel. I'm sure you're going to argue 'What about the PowerBook' but it's going to take Apple a year to 18 months to get an Intel laptop out and IBM and Apple would have to be seriously incompetent to not be able to get low power PPC out in that timeframe. And with 970MP workstations around the corner they're going to have to go some on the Intel front to beat those too. So the next year to 18 months isn't the end game IMHO, it's a step down the road to platform independence.
The Developer kit isn't about getting people on to Intel, it's about getting them on to XCode, Cocoa and platform independent code. In order to create universal binaries, it's not just the intel code that has to be clean, there are restrictions on which PowerPC technologies you can use also. Then the next transition to whatever Apple and Intel have up their sleeve is an easy one. If, as I theorise, we're heading to agnosticism then the Intel code will just work anyway and Apple have pulled a fast one on us to get the codebase clean and into a state where they've killed off the OS9 cruft.
Call me mad but I just think there's something else going on. I've no idea what but I don't think the future lies in X86 or PowerPC, technology from 30 years ago.
Ok, you're mad!
I don't know how old you are but your prediction of the demise of the x86 is as old hat as is the prediction that Apple is going out of business.
In 1988 the x86 was pronounced as being obsolete. In 1993 every PC magazine had cover stories about how the PPC was going to take over the PC industry, and that the old CISC x86 was doomed. Even in the late '90's when Macs outperformed the PC the x86 was thought to have run out of steam.
You might not notice that it's still here. Not only that, but in areas where it was thought to be hopelessly outclassed, in the medium and large server business, it has been pushing the high end chips out. Look at what has happened to Sun, SGI, and others. They are all using x86 chips more and more.
What you and others here aren't considering is what I said earlier. The 64 bit OS's aren't using much of that legacy microcode and hardware on those chips. If you look at MS's Win 64, you will notice that it's exclusive, you can't go back. In 64 bit mode it's MUCH more efficient. Why is that? Because these chips work very well in 64 bit mode.
After a while, when everyone is well into 64 bit hardware and software, Intel will start removing old legacy functionality from their x86 chips. As the chips will be MUCH more powerful that they were before, that functionality will be able to be emulated in software. After a while that too will disappear. All will be run in 64 or clean 32 bit mode - just the way it is done now on the G5.
The x86 has plenty more life left. They will continue to evolve, just as they are RISC at the core, they will take on whatever they need to compete. In doing that they will slowly change, as does everything.
In eight or ten years, just about the time when the advances in die shrinks are expected to have reached their end, the x86 line will be almost unrecognizable. Just as todays new x86 is is almost unrecognizable to those of ten years ago.
Then we will be on to something else. Nanobars, molecular computing, Spintronics, possibly even Quantum computers.
But I'll bet that it still will use the latest x86 ISA.
I don't think there was a compelling reason to change from PowerPC either - in fact, IMHO the x86 has always been a kludge, from the first 8086 all the way - but for whatever reason, the decision has been made, and as Mel says, deal with it.
Apple's not going to get any chips early. They're not going to get any better treatment than any other OEM. Mac hardware is going to be no better or worse than mainstream PCs. The quest for hardware excellence is over.
Apple's going for market share, and that's the way to do it. We may even see Macs selling for less than PCs, because Apple can pay less for Mac OS X than Dell pays for XP Home. Apple's advantage over Wal-Mart/Lindows is that there will be many more mainstream apps available.
Ultimately we will all benefit as Apple's market share soars (hopefully). Years from now, I imagine some new architecture will supplant x86, possibly by emulating it at sufficient speed. But in the interim, as Gerald Weinberg says, things are the way they are because they got that way.
I don't think there was a compelling reason to change from PowerPC either - in fact, IMHO the x86 has always been a kludge, from the first 8086 all the way - but for whatever reason, the decision has been made, and as Mel says, deal with it.
Apple's not going to get any chips early. They're not going to get any better treatment than any other OEM. Mac hardware is going to be no better or worse than mainstream PCs. The quest for hardware excellence is over.
Apple's going for market share, and that's the way to do it. We may even see Macs selling for less than PCs, because Apple can pay less for Mac OS X than Dell pays for XP Home. Apple's advantage over Wal-Mart/Lindows is that there will be many more mainstream apps available.
Ultimately we will all benefit as Apple's market share soars (hopefully). Years from now, I imagine some new architecture will supplant x86, possibly by emulating it at sufficient speed. But in the interim, as Gerald Weinberg says, things are the way they are because they got that way.
There is a very compelling reason. The G4 is totally obsolete, and becoming more so every day.
The G5 is behind the latest Xenon's, and even further behind the latest Opterons.
There is almost no development dollars going into either line.
If you go to Freescale's site, you will see that Apple's chips, the 7447a, and the as yet unreleased 7448, are listed not under "desktop" , or "workstation", or "Mobile". All indicating a continued development for Apple, but under "embedded", indicating automobiles, machinery, etc. Apple is using chips for their computers that are being developed for CARS!
IBM is no better, even though the G5 isn't relegated to the embedded category, IBM's interest in appealing to Apple's needs have waned.
IBM has more experience in designing and building multi-core chips than ANYONE else. It was expected that they would have a dual core chip for Apple over a year ago. There is even some vague mention of development work in IBM's documentation. So what happened? IBM has had major problems at its G5 chip plants, it has spread development out over several years instead of rushing to compete with AMD and Intel. Why? Because it isn't worth it to them to spend all that money at once. IBM can afford to do what it wants at its own pace. If that pace doesn't meet Apple's needs, so be it.
Remember that Apple's leaving IBM will result in less than a 5% loss in it's chip division, which itself is only about 15% of the whole company. IBM won't even notice the loss. And if IBM was losing money on it, it will benefit from Apple leaving.
Nah, there are VERY good reasons for Apple leaving.
Comments
I don't use or know too many people who use 64 bit apps.
In light of the fact that we have very little substantiated information on exactly what Apple is doing, I'm just looking at the possibilites.
Looking at how Intel really feels about x64 its possible Apple will not go in that direction.
In case I was being misunderstood. I was not saying Apple will not have a 64 bit OS.
I was saying Apple may not use the AMD64 architecture.
Originally posted by TenoBell
In case I was being misunderstood. I was not saying Apple will not have a 64 bit OS.
I was saying Apple may not use the AMD64 architecture.
There is only one x86 bit architecture. The one developed by AMD and adopted by Intel.
True, but that is not the only 64 bit architecture.
Originally posted by TenoBell
In light of the fact that we have very little substantiated information on exactly what Apple is doing, I'm just looking at the possibilites.
Good idea.
Apple has, of late, shown remarkable foresight in architecting OS X. They're designing for the future, not just the here and now.
The whole intel switch should bolster spirits. Making OS X hardware independent has apparently been a top priority. Not just in CPU choice but also with GPU choice (CI). Resolution independence also looks to be a major focus.
Considering this, I find it highly likely that apple has a team of 64bit code-monkeys going over their APIs with a fine tooth comb.
Originally posted by TenoBell
Did you mean there is only one x64 architecture?
True, but that is not the only 64 bit architecture.
As the only architecture that Apple will be using is the x86 architecture with 64 bit extentions that AMD developed, that is the only relevent architecture to Apple, and thereby to us.
But at this point no one of authority has confirmed it.
Apple has given so little information on the its plans with Intel.
If it were going to follow the exact same path as the rest of the industry why hold out so much information? I don't see any strategic or business advantage in doing that.
If Apple is going to use EMT64, why not tell developers now, so they can prepare for it? This would empower developers to have 64 bit apps ready. That doesn't make sense to me.
They can follow what Microsoft dictates.
Or they can follow their own path.
There is little middle ground in this.
Apple has pretty much always gone its own way and has never directly followed MS policy.
Intel seems to be in a place where they have come to realize MS has more power to lead PC standards than they do. And Intel doesn't seem pleased with the situation.
From the evidence I've seen. I don't believe the Apple/Intel collaboration will follow the larger PC industry and its MS set standards.
I believe the point of this collaboration is to be the leaders that the larger industry will follow.
Why? Because if Apple popularizes new markets, makes Media PCs more than 5% of the population gives a shit about, and ignites interest in high powered computing applications that not only allows Intel to sell more high margin product to Apple but after the laggard Wintel market adapts their products to more closely match Apple's then Intel also sells more high margin CPUs and fewer Celeron class low margin CPUs in the popularity spillover.
That is why Intel wanted Apple so bad. Intel sees the Wintel market as a commoditized market without an agent for growth and the coming game consoles as stealing part of its precious high margin gaming market, what AMD hasn't already seized, and one of the few higher margin areas left with good volume.
But Apple doesn't want custom CPUs or 64 bit architectures. That would kill the advantage of their deal with Intel. And Intel doesn't want to give them to Apple yet, not until they've gained market momentum, because Apple's fulcrum doesn't have enough leverage to move Wintel that much yet. You can take that to the bank and smoke it.
Originally posted by TenoBell
That may come true.
But at this point no one of authority has confirmed it.
Apple has given so little information on the its plans with Intel.
If it were going to follow the exact same path as the rest of the industry why hold out so much information? I don't see any strategic or business advantage in doing that.
If Apple is going to use EMT64, why not tell developers now, so they can prepare for it? This would empower developers to have 64 bit apps ready. That doesn't make sense to me.
With apple putting a P4 in the developer kits, we can assume that Apple is saying that x86 is the path they will be following. That's pretty obvious. The only path Apple can follow from there is the 64 bit path Intel is following. Intel is only following one 64 bit x86 path.
The only other 64 bit path Intel has is IA-64, and Apple certainly won't be following that one.
Originally posted by melgross
With apple putting a P4 in the developer kits, we can assume that Apple is saying that x86 is the path they will be following. That's pretty obvious. The only path Apple can follow from there is the 64 bit path Intel is following. Intel is only following one 64 bit x86 path.
The only other 64 bit path Intel has is IA-64, and Apple certainly won't be following that one.
I'm not so sure.
That Apple are following 32bit x86 for now is a definite.
But Intel haven't decided themselves on their 64bit path. So far it's been determined by Microsoft and AMD and Intel have reluctantly followed whilst complaining (rightly or wrongly) along the way that it's not a good solution.
There's still a lot of technology to flesh out in their 'next generation' architecture and that includes virtualization and who knows what there. Intel must be looking at salvaging something from Itanium. They've a customer now with a PowerPC codebase. The ultra low power chips are getting down to StrongARM territory. Maybe this is the point where Intel's chips become ISA agnostic and the 64bit path Intel takes is PowerPC for Apple computers, AMD64 for Windows, ARM for low power.
Originally posted by ChevalierMalFet
You are kidding right? I'm at a loss.
Oh, it's just speculation. But in the absence of any facts, that's all we've got - speculation.
Part of it is just my geekness being very very sad that the Intel X86 architecture has won out over every good architecture that has come along over the last 30 years including numerous attempts at dethroning it by Intel themselves.
Part of it is wondering what else is behind Intel's curtain as the IDF raised more questions than answers. And surely the future isn't what they've announced so far?
It does amaze me that many of the Apple 'faithful' were so quick to embrace Intel and the legacy X86 strain at that and immediately dismiss PowerPC. It's difficult to hold that position without being seen as a PowerPC fanboi.
I'm just not prepared to get laid by the fat girl even if that's the only girl left and you know she puts out. Intel is that fat girl and even though she's on a diet now, you know she's never going to be Elle McPherson.
As for 64-bit support in OS X/Intel, I don't think much can be based off of the transition kit yet. At minimum, it will be like it is now, 4+ GB memory support and 64-bit CLI process support. Anything more will require a bit of time.
Originally posted by aegisdesign
Oh, it's just speculation. But in the absence of any facts, that's all we've got - speculation.
Part of it is just my geekness being very very sad that the Intel X86 architecture has won out over every good architecture that has come along over the last 30 years including numerous attempts at dethroning it by Intel themselves.
Part of it is wondering what else is behind Intel's curtain as the IDF raised more questions than answers. And surely the future isn't what they've announced so far?
It does amaze me that many of the Apple 'faithful' were so quick to embrace Intel and the legacy X86 strain at that and immediately dismiss PowerPC. It's difficult to hold that position without being seen as a PowerPC fanboi.
I'm just not prepared to get laid by the fat girl even if that's the only girl left and you know she puts out. Intel is that fat girl and even though she's on a diet now, you know she's never going to be Elle McPherson.
There isn't any speculation here. It's over. Intel has their roadmap laid out. AMD has theirs. There is no way that Intel would be allowed to make PPC's.They wouldn't even if they were allowed to. If by some odd and impossible deal, they, and Apple were going that way, it would have been announced at the convention in June, and the machines sent out would have reflected that.
Really guys, I know this is fun and all that, but it's over. Really, it is.
Apple will be using x86 chips. 32 bit now, and 64 bit when they're ready.
I'm sorry that some still don't want to believe it. But it's true. There is no where else for Apple to go. Apple would not have produced those dev. machines for a joke. Do you really think that they send out thousands of those machines because they were going to use a different chip? That's absurd. The developers would drop Apple like a stone, and you know it.
Our trip through Fantasyland is over. It's time for everyone to get off.
I've never said Intel would be making PowerPC chips, melgross, I theorised that the future held instruction set agnosticism and that something else was afoot.
x86 *WILL* run out of steam. The architecture is fundamentally flawed. It has too much baggage. It's hit many walls in it's history and we're approaching the point where new process technologies just can't keep up. It's already essentially a big legacy emulation engine attached to a RISC core. Wouldn't it be much better to bypass that legacy? That is partly why many people are confused with Apple's move to 32bit intel, not 64bit. There's no legacy to support, why start that now?
Intel have tried to shift many times only to be held back by Microsoft. They tried it with i860 and then Itanium. They aren't happy about EMT64. They've got technology in spades that is underused or not used by the worlds most popular OS. They've teams of very talented engineers that used to work on very advanced designs like Alpha, ARM and VLIW now working on squeezing performance out of X86.
I may be in fantasy land but I reckon this Intel transition marks the end of both x86 and PowerPC and just one step on the road to something else. Intel wants to shed it's legacy and Apple are the only ones that can provide a mainstream OS to do it. AMD can't provide the technology.
Otherwise I fail to see any compelling reason to switch from PowerPC to Intel. I'm sure you're going to argue 'What about the PowerBook' but it's going to take Apple a year to 18 months to get an Intel laptop out and IBM and Apple would have to be seriously incompetent to not be able to get low power PPC out in that timeframe. And with 970MP workstations around the corner they're going to have to go some on the Intel front to beat those too. So the next year to 18 months isn't the end game IMHO, it's a step down the road to platform independence.
The Developer kit isn't about getting people on to Intel, it's about getting them on to XCode, Cocoa and platform independent code. In order to create universal binaries, it's not just the intel code that has to be clean, there are restrictions on which PowerPC technologies you can use also. Then the next transition to whatever Apple and Intel have up their sleeve is an easy one. If, as I theorise, we're heading to agnosticism then the Intel code will just work anyway and Apple have pulled a fast one on us to get the codebase clean and into a state where they've killed off the OS9 cruft.
Call me mad but I just think there's something else going on. I've no idea what but I don't think the future lies in X86 or PowerPC, technology from 30 years ago.
Originally posted by aegisdesign
Nope. I'm still not convinced.
I've never said Intel would be making PowerPC chips, melgross, I theorised that the future held instruction set agnosticism and that something else was afoot.
x86 *WILL* run out of steam. The architecture is fundamentally flawed. It has too much baggage. It's hit many walls in it's history and we're approaching the point where new process technologies just can't keep up. It's already essentially a big legacy emulation engine attached to a RISC core. Wouldn't it be much better to bypass that legacy? That is partly why many people are confused with Apple's move to 32bit intel, not 64bit. There's no legacy to support, why start that now?
Intel have tried to shift many times only to be held back by Microsoft. They tried it with i860 and then Itanium. They aren't happy about EMT64. They've got technology in spades that is underused or not used by the worlds most popular OS. They've teams of very talented engineers that used to work on very advanced designs like Alpha, ARM and VLIW now working on squeezing performance out of X86.
I may be in fantasy land but I reckon this Intel transition marks the end of both x86 and PowerPC and just one step on the road to something else. Intel wants to shed it's legacy and Apple are the only ones that can provide a mainstream OS to do it. AMD can't provide the technology.
Otherwise I fail to see any compelling reason to switch from PowerPC to Intel. I'm sure you're going to argue 'What about the PowerBook' but it's going to take Apple a year to 18 months to get an Intel laptop out and IBM and Apple would have to be seriously incompetent to not be able to get low power PPC out in that timeframe. And with 970MP workstations around the corner they're going to have to go some on the Intel front to beat those too. So the next year to 18 months isn't the end game IMHO, it's a step down the road to platform independence.
The Developer kit isn't about getting people on to Intel, it's about getting them on to XCode, Cocoa and platform independent code. In order to create universal binaries, it's not just the intel code that has to be clean, there are restrictions on which PowerPC technologies you can use also. Then the next transition to whatever Apple and Intel have up their sleeve is an easy one. If, as I theorise, we're heading to agnosticism then the Intel code will just work anyway and Apple have pulled a fast one on us to get the codebase clean and into a state where they've killed off the OS9 cruft.
Call me mad but I just think there's something else going on. I've no idea what but I don't think the future lies in X86 or PowerPC, technology from 30 years ago.
Ok, you're mad!
I don't know how old you are but your prediction of the demise of the x86 is as old hat as is the prediction that Apple is going out of business.
In 1988 the x86 was pronounced as being obsolete. In 1993 every PC magazine had cover stories about how the PPC was going to take over the PC industry, and that the old CISC x86 was doomed. Even in the late '90's when Macs outperformed the PC the x86 was thought to have run out of steam.
You might not notice that it's still here. Not only that, but in areas where it was thought to be hopelessly outclassed, in the medium and large server business, it has been pushing the high end chips out. Look at what has happened to Sun, SGI, and others. They are all using x86 chips more and more.
What you and others here aren't considering is what I said earlier. The 64 bit OS's aren't using much of that legacy microcode and hardware on those chips. If you look at MS's Win 64, you will notice that it's exclusive, you can't go back. In 64 bit mode it's MUCH more efficient. Why is that? Because these chips work very well in 64 bit mode.
After a while, when everyone is well into 64 bit hardware and software, Intel will start removing old legacy functionality from their x86 chips. As the chips will be MUCH more powerful that they were before, that functionality will be able to be emulated in software. After a while that too will disappear. All will be run in 64 or clean 32 bit mode - just the way it is done now on the G5.
The x86 has plenty more life left. They will continue to evolve, just as they are RISC at the core, they will take on whatever they need to compete. In doing that they will slowly change, as does everything.
In eight or ten years, just about the time when the advances in die shrinks are expected to have reached their end, the x86 line will be almost unrecognizable. Just as todays new x86 is is almost unrecognizable to those of ten years ago.
Then we will be on to something else. Nanobars, molecular computing, Spintronics, possibly even Quantum computers.
But I'll bet that it still will use the latest x86 ISA.
Apple's not going to get any chips early. They're not going to get any better treatment than any other OEM. Mac hardware is going to be no better or worse than mainstream PCs. The quest for hardware excellence is over.
Apple's going for market share, and that's the way to do it. We may even see Macs selling for less than PCs, because Apple can pay less for Mac OS X than Dell pays for XP Home. Apple's advantage over Wal-Mart/Lindows is that there will be many more mainstream apps available.
Ultimately we will all benefit as Apple's market share soars (hopefully). Years from now, I imagine some new architecture will supplant x86, possibly by emulating it at sufficient speed. But in the interim, as Gerald Weinberg says, things are the way they are because they got that way.
Originally posted by cubist
I don't think there was a compelling reason to change from PowerPC either - in fact, IMHO the x86 has always been a kludge, from the first 8086 all the way - but for whatever reason, the decision has been made, and as Mel says, deal with it.
Apple's not going to get any chips early. They're not going to get any better treatment than any other OEM. Mac hardware is going to be no better or worse than mainstream PCs. The quest for hardware excellence is over.
Apple's going for market share, and that's the way to do it. We may even see Macs selling for less than PCs, because Apple can pay less for Mac OS X than Dell pays for XP Home. Apple's advantage over Wal-Mart/Lindows is that there will be many more mainstream apps available.
Ultimately we will all benefit as Apple's market share soars (hopefully). Years from now, I imagine some new architecture will supplant x86, possibly by emulating it at sufficient speed. But in the interim, as Gerald Weinberg says, things are the way they are because they got that way.
There is a very compelling reason. The G4 is totally obsolete, and becoming more so every day.
The G5 is behind the latest Xenon's, and even further behind the latest Opterons.
There is almost no development dollars going into either line.
If you go to Freescale's site, you will see that Apple's chips, the 7447a, and the as yet unreleased 7448, are listed not under "desktop" , or "workstation", or "Mobile". All indicating a continued development for Apple, but under "embedded", indicating automobiles, machinery, etc. Apple is using chips for their computers that are being developed for CARS!
IBM is no better, even though the G5 isn't relegated to the embedded category, IBM's interest in appealing to Apple's needs have waned.
IBM has more experience in designing and building multi-core chips than ANYONE else. It was expected that they would have a dual core chip for Apple over a year ago. There is even some vague mention of development work in IBM's documentation. So what happened? IBM has had major problems at its G5 chip plants, it has spread development out over several years instead of rushing to compete with AMD and Intel. Why? Because it isn't worth it to them to spend all that money at once. IBM can afford to do what it wants at its own pace. If that pace doesn't meet Apple's needs, so be it.
Remember that Apple's leaving IBM will result in less than a 5% loss in it's chip division, which itself is only about 15% of the whole company. IBM won't even notice the loss. And if IBM was losing money on it, it will benefit from Apple leaving.
Nah, there are VERY good reasons for Apple leaving.
Originally posted by cubist
....., as Gerald Weinberg says, things are the way they are because they got that way.
Gerald Weinberg related to Lawrence Peter "Yogi" Berra?