"In theory, there's not a whole lot that Linux can do that you can't do with OSX."
Sure there is:
Binary compatibility with binaries provided for Linux and not MacOS
Source compatibility -- even if most software can be ported, it's not close enough that I can do my Linux development, and there exists software that hasn't been ported.
Linux is by far a better performing kernel, and this can be an asset for some jobs.
I'm sure there's other things as well, but those would be my reasons if I were to do it.
"Most of the ppl that got into Linux so they'd look l337 have already jumped onto the OSX bandwagon anyways."
Sure, and some people that got into Linux did it had a technical motivation.
"I think the cost involved here is the Cost of admitting that despite announcing the shift to Intel over 6 months ago, Intel still doesn't have a suitable 64-bit chip available is too high. So they push these hobbled machines out now with Apple logos"
All of their options would have had a cost. This one had the smallest.
"hoping nobody makes a big thing about the return to 32-bit computing in the newest machines."
The laptops were never 64-bit, and iMacs were never capable of 64-bit computing in a way helpful to desktop users. The apps that would have mattered (Photoshop etc) were waiting for Apple to release 64-bit versions of their graphical libraries, so except for a few niche products that also had 32-bit support, there's really no benefit.
At the end of the day, Apple would not have been able to launch with a 64-bit OS even if they had 64-bit hardware because it simply wasn't ready. They would have been stuck with legacy support regardless. Even after they move to 64-bit, most applications will stay 32-bit if only to support computers running Tiger. No one is denying that it's a 32-bit system or that that's not ideal, but Apple could not have escaped that fate except by sacrificing sales that are much more important to them.
They're a for-profit company. Don't expect them to give up that much money over a feature people are obviously not that worked up about.
They didn't have a 64-bit OS ready, so even if they launched on 64-bit hardware they'd still have to support the applications for a long time. Later versions of the OS could upgrade to 64-bit, but they wouldn't have been able to avoid the legacy support for applications.
Also, G4s were completely outclassed as it was. Another 9 months of that when everyone else had just gotten another core would have killed their laptop sales. The cost of dealing with 32-bit will probably be less than the hit to their hardware sales would have been.
Most software doesn't need to be 64-bit, and 64-bit machines can offer backwards compatibility.
"My guess is x86-32 on Apple will be VERY short and buying one of these Macbooks will in the long run a pretty bad choice if you intend to run Macintosh software on it."
There's no way they'll have a 64-bit version of the OS out any time soon, so even if they're shipping 64-bit hardware it's not going to be taken advantage of. The earliest 64-bit OS I can imagine would be an interim release of 10.5 like they did for 10.2, and it probably won't be until 10.6. Given that kind of timeframe, the current MacBooks will be nearing the end of their 3 year support window by the time anything happens.
Yes, it means Apple is behind on 64-bit support... but we already knew that, even on the G5s.
"So? What the hell has that got to do with SMP servers? RISC chips had been "optimised for 32-bit" for a decade already, and most had long movedo on to 64-bit."
Merely pointing out an example of where they picked server/workstation performance over traditional PC performance. Pentium Pros were adequate for 1 and 2-way systems, and they cost significantly less than the RISC systems you would have needed to beat them.
"They sucked and still do for Unix on SMP boxen.
In a previous life, I used to build software (many gigabytes daily) on 64-bit SMP RISC and 32-bit intel Xeon boxes. The cheap (and hot and unreliable) 4-way 1.9GHz Pentium IV Xeon Dell boxes were fast with one user, compared to the 64-bit RISC boxen."
And here I thought we were talking about Pentium Pros. Pentium Pros have very little to do with the Netburst chips you are now criticizing, it's unfair to treat the two as similar as you are doing.
The Yonah chips that have just been released are decended from Pentium Pros, and while I hesitate to make SMP comparisons because there is not yet any Xeon version of the chip, they can keep up with the closest comparable RISC chip with a fraction of the power (eg a dual-core Yonah at 2 ghz uses about the same amount of power as a single-core PowerPC 970FX at the same clock speed, while either of the cores can beat the FX at almost anything other than FP performance).
"My take on history is that Apple have chosen the right processor architecture at any given moment taking account everything that was known at the time."
Even if PowerPC was the right choice in the early 90s, it's been the wrong one for 5+ years. IBM and Moto/Freescale don't care about desktop chips, and the time has passed when other chips can be pushed into service against AMD and Intel's best and brightest.
It's hard to take you seriously when you say that.
"They were never intended for anything other than PeeCees. Sever and workstation applications were an afterthought, as anyone with any experience of SMP systems will tell you."
Actually, they optimized for 32-bit protected mode performance at the expense of real-mode performance. It hurt them because there was still a lot of real-mode software being used, but they were fine for UNIXes and NT.
"Also, a security issue if you live in close-quarters (apartments, office buildings) because people could sniff the "packets" between the hub and device. They could watch you on your webcam, when you have your webcam software off."
Even if there's strong encryption, this is an issue. There are attacks that can figure out what you're typing from the timing of the keystrokes alone, and wireless USB would provide very accurate timing information to any attackers.
"Yeah, especially since (thouse of us old enough should remember;-) ) one of the original goals of PPC consortium (IBM, HP, Motorola)"
Wasn't it Apple, IBM, and Moto? The acronym everyone uses is AIM.
"I guess ISA simplicity goes both ways and not it makes it easy to emulate PPC on x86..."
Not really. It's actually pretty difficult to emulate PowerPC on x86, and Rosetta isn't that fast. It's impressive because it doesn't completely suck, but I wouldn't call it fast.
"The Core Duo iMac, with two x86 cores, is a very small improvement over the single PPC970. Compared to a two-and-a-half-year-old dual-CPU Power Mac G5, the Core Duo iMac is pathetic. And this is running Intel-native applications like QuickTime and iTunes."
The thread on the Ars article goes into detail about why the benchmarks weren't done very well, and you're ignoring Macintouch's own caveats with their numbers. It's clear Apple still has a lot of optimization to do before they reach the level of refinement they had on PowerPC, and plenty of the stuff where people are complaining it's not twice as fast is I/O bound.
I've noticed three things with the benchmarks: First, single-threaded benchmarks still show an advantage for the Intel machine. It's small, but it's still there meaning the cores are fast individually. Next, I/O bound benchmarks still show an improvement, even though the hard drive is essentially the same. This is probably a result of Intel's chipset having a faster SATA controller, and having the chipset provided is another clear advantage of the switch.
Finally, all the reviews have made a point of saying how responsive the machine feels. As these things are desktop machines, that's probably more important than any of the other performance measures. It's faster in other ways, but responsiveness is what you'll notice day to day.
"Yes it is, because MB means 1000 Kilobytes, and MiB means 1024 Kibibytes. Usually you're off by a factor of about 0.91 or 1.1, which can mean a lot. Please don't insult the functionality of accuracy.:D"
MB means 1048576 bytes either way, unless it's coming from a storage vendor.
I've read that chips like PowerPC 970 that have an extremely small die can be hard to cool, and also that cache is a good way to increase the size of the die without increasing power usage that much.
If AMD can provide a reasonable cache that only requires a fraction of the die space, might this become a problem?
DRAM uses 1 transistor plus a capacitor per bit. SRAM uses 6 transistors. ZRAM is denser than both, as it's just one transistor.
I don't think it's a drop-in replacement for current types of memory, but it may become popular in the future. The advantage isn't as big as with SRAM, so there's less of a rush.
"Yes, but you don't have to emulate the whole system, just user mode down to libraries. For GUI refresh (especially a compositing GUI), you can get that done in native code, if done right."
True. And that makes it efficient. But it doesn't make it not an emulator because the translation to local libraries happens in addition, not in lieu.
"Now, one predicted trend for the future is for all architectures to move to simple, cheap, in-order cores, and put a lot of them on the chip to give increases in TLP without using a hugely complicated, expensive, lots-of-power-and-chip-area out-of-order core."
While TLP will undoubtedly become more important in the future, I think the single-thread performance is going to remain important in the future, at least for the desktop. It's already hard to get significant amounts of parallelism out of most applications (even though most have several threads, most threads sleep most of the time), and the developer effort required to optimize for multiple cores increases dramatically with each one you add. Ultimately, I think this means we'll be stuck with chip designs that maximize ILP, to the detriment of the core count.
There are applications for which TLP is not a problem, for example the servers that will run Niagara chips. But they have more sources of work (many clients making requests). I don't think desktops will ever have it that easy.
"I find is odd that Intel keeps backtracking to its 20 year old Pentium Pro design. Both of their recent high-budget designs, the P4 and the Itanium proved to be a flop to some extent, while the P6/Pentium Pro/PII/PIII/Centrino/Banias architecture has scaled amazingly well since its humble 200 MHz beginnings."
We hit the wall in single-threaded performance that was remarkably similar for all the various architechtures. I imagine the Pentium Pro architechture appeared at a time when it was possible to incorporate most of the features we've come to know and love in something approaching their current form.
"But Transmeta was the last memorable innovation, and since then everyone seems to be exclusively focused on cache megabytes and transistor sizes."
I'd say Cell and Niagara are pretty interesting. They don't have the general-purpose utility of other chips, but they're very good at what they're specialized for, and they're certainly interesting.
"Well not really. Apple isn't really emulating PPC on the x86...since they WROTE the operating system that is running on both they have a lot more control over what's running and can do all kinds of cool translations and on the fly stuff for what the program is doing, making it not really emulation. More like how WINE is not really an emulator."
That is not correct.
Rosetta runs binaries of a different architechture. The fact that does this with translation makes it fast, but it doesn't make it not an emulator.
WINE makes it possible to run binaries that conform to different standards but use the same architechture. Some people call what WINE does emulation too (acronym notwithstanding). I don't know if I'd go that far, but Rosetta is certainly an emulator.
Just what we need. More uninformed Cell zealots that don't understand the Cell's weaknesses. Even better, they'll be management types that can't be reasoned with because they don't understand anything past the word "parallelism".
"If I'm going to be forced to use a garbage architecture like x86 in the future,"
You get better performance and a longer battery life. The only way you would be able to tell the difference between it and a PowerPC without digging around in the low level stuff is that it's faster and the battery lasts longer. PowerPC is clean in ways that don't benefit the user, while stagnating in the desktop/laptop areas. For better or worse, x86 has consolidated the desktop/laptop by getting better results than anyone else.
I think it's unreasonable to expect others to avoid a clearly better choice because it's not aesthetically pleasing for you.
"then one would have to kill 12 times the amount of people now, to have a comparable affect."
I disagree. The impact of a death is highly variable, in ways not necessarily related to the global population. In, say, 0th century Rome, or in WWII, a single death would not have attracted much attention unless it was someone important. These days, you can't disconnect a feeding tube without everyone going nuts.
"One of the things Apple has enjoyed a long time before people on Intel/Windows have is the 64bit "inside"."
Intel took a while to catch up, but AMD had 64-bit chips before Apple. Windows took a while to get support, but Linux supported it almost immediately and OS X still doesn't support 64-bit GUI apps. It was only an Apple first if you define the criteria of the comparison to exclude the ways in which x86 was first without excluding Apple for lacking GUI support.
"I understand (correct me if Im wrong) that PowerPC has been 64bit since the G4's or at least the G5's and OSX is 64bit ready."
G4s aren't 64-bit, OS X doesn't support 64-bit GUI apps. Also, OS X doesn't have any 64-bit at all on x86, as it's significantly more complicated there.
"My question or concern is what is going to happen to that? Is Apple switching from 64bit to 32bit Yonah's or has intel made a 64bit Yonah just for apple.
that is the only factor from this whole "switch" to intel thing that has had me worried. What do you guys think or know of this?"
Yonahs are 32-bit, and if Apple uses them they're going to be stuck with it. Even if they had 64-bit hardware right now, the OS wouldn't be ready for it so they'd have to run in 32-bit mode. It'll be a problem for the future since they'll have to support the older chips for some time, but all the alternatives are worse so it's worth it.
"Eh? Firstly, the new G5's available from IBM are much lower in power consumption."
Those are available at much lower clock speeds, Pentium Ms have very good performance clock for clock (better than G5s) as well as significantly higher clock speeds and more cores (IBM's low power G5s are still single core).
Even if Apple had released a PowerBook based on G5s, it would be barely competitive with the low end of Intel's offerings. It couldn't touch the faster chips, and even if that had been enough, it wouldn't have done any good at all against the dual-core Yonahs Intel is releasing.
"Secondly that Pentium M you are touting has more in common with that Pentium II you just trashed than any other processor design."
That's irrelevant. Whatever its heritage, it's been shipping for years, and the fact that it's fast and low power can be confirmed by anyone. While it's related to older chips, it should be judged on its merits because it has also seen plenty of new development. Its merits are very strong clock for clock performance and very low power usage, which make it better suited to laptop use than anything IBM has.
"All of the current top3 Supercomputers in the world are POWER5 IBM designs. The intel transition was a mistake IMO."
Desktops/laptops require good performance with individual threads more than they need lots of cores/threads. While desktops have many threads, it's rare for more than one of them to be runnable at any given time, the vast majority are sleeping waiting for something. x86 chips have started getting more than one core now, but only because everyone (including IBM) has hit a wall in individual thread performance. Niagara and POWER5+ both require many threads to reach their potential, and they are therefore not well suited to desktop systems.
"Intel doesn't have any solid plans for the future of x86, they're just going to keep using process advances to cram more PII cores on a chip, until they end up with an x86 based NIAGRA clone"
You say that like IBM isn't adding cores to their POWER chips. More cores doesn't make it a Niagara clone. Niagara uses many hardware threads per core to allow the chip to mitigate the penalty of things that would otherwise cause a stall. Intel does completely different things, like branch prediction, and you'll note they've dumped hyperthreading now that they're moving away from Netburst.
Intel's priorities are appropriate for desktop/laptop machines, so they're the best choice. IBM's priorities lead them to make good server and game chips, but these are both specialized in a way that's not well suited to desktops/laptops. The switch is the right choice.
Slashdot Mirror
Sure there is:
I'm sure there's other things as well, but those would be my reasons if I were to do it.
"Most of the ppl that got into Linux so they'd look l337 have already jumped onto the OSX bandwagon anyways."
Sure, and some people that got into Linux did it had a technical motivation.
"I think the cost involved here is the Cost of admitting that despite announcing the shift to Intel over 6 months ago, Intel still doesn't have a suitable 64-bit chip available is too high. So they push these hobbled machines out now with Apple logos"
All of their options would have had a cost. This one had the smallest.
"hoping nobody makes a big thing about the return to 32-bit computing in the newest machines."
The laptops were never 64-bit, and iMacs were never capable of 64-bit computing in a way helpful to desktop users. The apps that would have mattered (Photoshop etc) were waiting for Apple to release 64-bit versions of their graphical libraries, so except for a few niche products that also had 32-bit support, there's really no benefit.
At the end of the day, Apple would not have been able to launch with a 64-bit OS even if they had 64-bit hardware because it simply wasn't ready. They would have been stuck with legacy support regardless. Even after they move to 64-bit, most applications will stay 32-bit if only to support computers running Tiger. No one is denying that it's a 32-bit system or that that's not ideal, but Apple could not have escaped that fate except by sacrificing sales that are much more important to them.
They're a for-profit company. Don't expect them to give up that much money over a feature people are obviously not that worked up about.
They didn't have a choice.
They didn't have a 64-bit OS ready, so even if they launched on 64-bit hardware they'd still have to support the applications for a long time. Later versions of the OS could upgrade to 64-bit, but they wouldn't have been able to avoid the legacy support for applications.
Also, G4s were completely outclassed as it was. Another 9 months of that when everyone else had just gotten another core would have killed their laptop sales. The cost of dealing with 32-bit will probably be less than the hit to their hardware sales would have been.
Most software doesn't need to be 64-bit, and 64-bit machines can offer backwards compatibility.
"My guess is x86-32 on Apple will be VERY short and buying one of these Macbooks will in the long run a pretty bad choice if you intend to run Macintosh software on it."
There's no way they'll have a 64-bit version of the OS out any time soon, so even if they're shipping 64-bit hardware it's not going to be taken advantage of. The earliest 64-bit OS I can imagine would be an interim release of 10.5 like they did for 10.2, and it probably won't be until 10.6. Given that kind of timeframe, the current MacBooks will be nearing the end of their 3 year support window by the time anything happens.
Yes, it means Apple is behind on 64-bit support... but we already knew that, even on the G5s.
"And at least my shiny new (1 year old) power book doesn't quite feel outdated just yet ..."
With a G4 in it, it was outdated the moment you opened the box.
"So? What the hell has that got to do with SMP servers? RISC chips had been "optimised for 32-bit" for a decade already, and most had long movedo on to 64-bit."
Merely pointing out an example of where they picked server/workstation performance over traditional PC performance. Pentium Pros were adequate for 1 and 2-way systems, and they cost significantly less than the RISC systems you would have needed to beat them.
"They sucked and still do for Unix on SMP boxen.
In a previous life, I used to build software (many gigabytes daily) on 64-bit SMP RISC and 32-bit intel Xeon boxes. The cheap (and hot and unreliable) 4-way 1.9GHz Pentium IV Xeon Dell boxes were fast with one user, compared to the 64-bit RISC boxen."
And here I thought we were talking about Pentium Pros. Pentium Pros have very little to do with the Netburst chips you are now criticizing, it's unfair to treat the two as similar as you are doing.
The Yonah chips that have just been released are decended from Pentium Pros, and while I hesitate to make SMP comparisons because there is not yet any Xeon version of the chip, they can keep up with the closest comparable RISC chip with a fraction of the power (eg a dual-core Yonah at 2 ghz uses about the same amount of power as a single-core PowerPC 970FX at the same clock speed, while either of the cores can beat the FX at almost anything other than FP performance).
"My take on history is that Apple have chosen the right processor architecture at any given moment taking account everything that was known at the time."
Even if PowerPC was the right choice in the early 90s, it's been the wrong one for 5+ years. IBM and Moto/Freescale don't care about desktop chips, and the time has passed when other chips can be pushed into service against AMD and Intel's best and brightest.
"PeeCee"
It's hard to take you seriously when you say that.
"They were never intended for anything other than PeeCees. Sever and workstation applications were an afterthought, as anyone with any experience of SMP systems will tell you."
Actually, they optimized for 32-bit protected mode performance at the expense of real-mode performance. It hurt them because there was still a lot of real-mode software being used, but they were fine for UNIXes and NT.
lucky != okay
"Also, a security issue if you live in close-quarters (apartments, office buildings) because people could sniff the "packets" between the hub and device. They could watch you on your webcam, when you have your webcam software off."
Even if there's strong encryption, this is an issue. There are attacks that can figure out what you're typing from the timing of the keystrokes alone, and wireless USB would provide very accurate timing information to any attackers.
"Yeah, especially since (thouse of us old enough should remember ;-) ) one of the original goals of PPC consortium (IBM, HP, Motorola)"
Wasn't it Apple, IBM, and Moto? The acronym everyone uses is AIM.
"I guess ISA simplicity goes both ways and not it makes it easy to emulate PPC on x86..."
Not really. It's actually pretty difficult to emulate PowerPC on x86, and Rosetta isn't that fast. It's impressive because it doesn't completely suck, but I wouldn't call it fast.
"The Core Duo iMac, with two x86 cores, is a very small improvement over the single PPC970. Compared to a two-and-a-half-year-old dual-CPU Power Mac G5, the Core Duo iMac is pathetic. And this is running Intel-native applications like QuickTime and iTunes."
The thread on the Ars article goes into detail about why the benchmarks weren't done very well, and you're ignoring Macintouch's own caveats with their numbers. It's clear Apple still has a lot of optimization to do before they reach the level of refinement they had on PowerPC, and plenty of the stuff where people are complaining it's not twice as fast is I/O bound.
I've noticed three things with the benchmarks: First, single-threaded benchmarks still show an advantage for the Intel machine. It's small, but it's still there meaning the cores are fast individually. Next, I/O bound benchmarks still show an improvement, even though the hard drive is essentially the same. This is probably a result of Intel's chipset having a faster SATA controller, and having the chipset provided is another clear advantage of the switch.
Finally, all the reviews have made a point of saying how responsive the machine feels. As these things are desktop machines, that's probably more important than any of the other performance measures. It's faster in other ways, but responsiveness is what you'll notice day to day.
"Yes it is, because MB means 1000 Kilobytes, and MiB means 1024 Kibibytes. Usually you're off by a factor of about 0.91 or 1.1, which can mean a lot. Please don't insult the functionality of accuracy. :D"
MB means 1048576 bytes either way, unless it's coming from a storage vendor.
I've read that chips like PowerPC 970 that have an extremely small die can be hard to cool, and also that cache is a good way to increase the size of the die without increasing power usage that much.
If AMD can provide a reasonable cache that only requires a fraction of the die space, might this become a problem?
DRAM uses 1 transistor plus a capacitor per bit. SRAM uses 6 transistors. ZRAM is denser than both, as it's just one transistor.
I don't think it's a drop-in replacement for current types of memory, but it may become popular in the future. The advantage isn't as big as with SRAM, so there's less of a rush.
"Yes, but you don't have to emulate the whole system, just user mode down to libraries. For GUI refresh (especially a compositing GUI), you can get that done in native code, if done right."
True. And that makes it efficient. But it doesn't make it not an emulator because the translation to local libraries happens in addition, not in lieu.
Nice post. I have one thing to add:
"Now, one predicted trend for the future is for all architectures to move to simple, cheap, in-order cores, and put a lot of them on the chip to give increases in TLP without using a hugely complicated, expensive, lots-of-power-and-chip-area out-of-order core."
While TLP will undoubtedly become more important in the future, I think the single-thread performance is going to remain important in the future, at least for the desktop. It's already hard to get significant amounts of parallelism out of most applications (even though most have several threads, most threads sleep most of the time), and the developer effort required to optimize for multiple cores increases dramatically with each one you add. Ultimately, I think this means we'll be stuck with chip designs that maximize ILP, to the detriment of the core count.
There are applications for which TLP is not a problem, for example the servers that will run Niagara chips. But they have more sources of work (many clients making requests). I don't think desktops will ever have it that easy.
"I find is odd that Intel keeps backtracking to its 20 year old Pentium Pro design. Both of their recent high-budget designs, the P4 and the Itanium proved to be a flop to some extent, while the P6/Pentium Pro/PII/PIII/Centrino/Banias architecture has scaled amazingly well since its humble 200 MHz beginnings."
We hit the wall in single-threaded performance that was remarkably similar for all the various architechtures. I imagine the Pentium Pro architechture appeared at a time when it was possible to incorporate most of the features we've come to know and love in something approaching their current form.
"But Transmeta was the last memorable innovation, and since then everyone seems to be exclusively focused on cache megabytes and transistor sizes."
I'd say Cell and Niagara are pretty interesting. They don't have the general-purpose utility of other chips, but they're very good at what they're specialized for, and they're certainly interesting.
"Well not really. Apple isn't really emulating PPC on the x86...since they WROTE the operating system that is running on both they have a lot more control over what's running and can do all kinds of cool translations and on the fly stuff for what the program is doing, making it not really emulation. More like how WINE is not really an emulator."
That is not correct.
Rosetta runs binaries of a different architechture. The fact that does this with translation makes it fast, but it doesn't make it not an emulator.
WINE makes it possible to run binaries that conform to different standards but use the same architechture. Some people call what WINE does emulation too (acronym notwithstanding). I don't know if I'd go that far, but Rosetta is certainly an emulator.
"So what you are saying is that Apple bought Intel's marketing hype?"
That, or they saw that the chips Intel already had were fast enough to justify the switch.
Just what we need. More uninformed Cell zealots that don't understand the Cell's weaknesses. Even better, they'll be management types that can't be reasoned with because they don't understand anything past the word "parallelism".
"If I'm going to be forced to use a garbage architecture like x86 in the future,"
You get better performance and a longer battery life. The only way you would be able to tell the difference between it and a PowerPC without digging around in the low level stuff is that it's faster and the battery lasts longer. PowerPC is clean in ways that don't benefit the user, while stagnating in the desktop/laptop areas. For better or worse, x86 has consolidated the desktop/laptop by getting better results than anyone else.
I think it's unreasonable to expect others to avoid a clearly better choice because it's not aesthetically pleasing for you.
"then one would have to kill 12 times the amount of people now, to have a comparable affect."
I disagree. The impact of a death is highly variable, in ways not necessarily related to the global population. In, say, 0th century Rome, or in WWII, a single death would not have attracted much attention unless it was someone important. These days, you can't disconnect a feeding tube without everyone going nuts.
"One of the things Apple has enjoyed a long time before people on Intel/Windows have is the 64bit "inside"."
Intel took a while to catch up, but AMD had 64-bit chips before Apple. Windows took a while to get support, but Linux supported it almost immediately and OS X still doesn't support 64-bit GUI apps. It was only an Apple first if you define the criteria of the comparison to exclude the ways in which x86 was first without excluding Apple for lacking GUI support.
"I understand (correct me if Im wrong) that PowerPC has been 64bit since the G4's or at least the G5's and OSX is 64bit ready."
G4s aren't 64-bit, OS X doesn't support 64-bit GUI apps. Also, OS X doesn't have any 64-bit at all on x86, as it's significantly more complicated there.
"My question or concern is what is going to happen to that? Is Apple switching from 64bit to 32bit Yonah's or has intel made a 64bit Yonah just for apple.
that is the only factor from this whole "switch" to intel thing that has had me worried. What do you guys think or know of this?"
Yonahs are 32-bit, and if Apple uses them they're going to be stuck with it. Even if they had 64-bit hardware right now, the OS wouldn't be ready for it so they'd have to run in 32-bit mode. It'll be a problem for the future since they'll have to support the older chips for some time, but all the alternatives are worse so it's worth it.
"Eh? Firstly, the new G5's available from IBM are much lower in power consumption."
Those are available at much lower clock speeds, Pentium Ms have very good performance clock for clock (better than G5s) as well as significantly higher clock speeds and more cores (IBM's low power G5s are still single core).
Even if Apple had released a PowerBook based on G5s, it would be barely competitive with the low end of Intel's offerings. It couldn't touch the faster chips, and even if that had been enough, it wouldn't have done any good at all against the dual-core Yonahs Intel is releasing.
"Secondly that Pentium M you are touting has more in common with that Pentium II you just trashed than any other processor design."
That's irrelevant. Whatever its heritage, it's been shipping for years, and the fact that it's fast and low power can be confirmed by anyone. While it's related to older chips, it should be judged on its merits because it has also seen plenty of new development. Its merits are very strong clock for clock performance and very low power usage, which make it better suited to laptop use than anything IBM has.
"All of the current top3 Supercomputers in the world are POWER5 IBM designs. The intel transition was a mistake IMO."
Desktops/laptops require good performance with individual threads more than they need lots of cores/threads. While desktops have many threads, it's rare for more than one of them to be runnable at any given time, the vast majority are sleeping waiting for something. x86 chips have started getting more than one core now, but only because everyone (including IBM) has hit a wall in individual thread performance. Niagara and POWER5+ both require many threads to reach their potential, and they are therefore not well suited to desktop systems.
"Intel doesn't have any solid plans for the future of x86, they're just going to keep using process advances to cram more PII cores on a chip, until they end up with an x86 based NIAGRA clone"
You say that like IBM isn't adding cores to their POWER chips. More cores doesn't make it a Niagara clone. Niagara uses many hardware threads per core to allow the chip to mitigate the penalty of things that would otherwise cause a stall. Intel does completely different things, like branch prediction, and you'll note they've dumped hyperthreading now that they're moving away from Netburst.
Intel's priorities are appropriate for desktop/laptop machines, so they're the best choice. IBM's priorities lead them to make good server and game chips, but these are both specialized in a way that's not well suited to desktops/laptops. The switch is the right choice.