PowerBook To Get Dual-Core G4?

[CmdrLaForge]

Why do you want a PPC970 in a PowerBook (a "G5 PB") ??

Do you need a 64-bit PB ?? Not really - unless you need 8 GiB of RAM in your PB 64-bit won't do much for you. (And besides, a PB with 8 GiB of RAM would be big, hot and heavy!!!)

It has been said many times, but I can repeat it: there are users who don't upgrade every other year, but like to keep their equipment for 5-7 years ! The G4 is old tech and 64 bit is the future for Apple products in 3 years there will be no 32 bit offering any more and the OS will go completly 64 bit.

Got my point ?

 

[G4-power]

Why do you want a PPC970 in a PowerBook (a "G5 PB") ??


You might want a *faster* PB, but a dual-core e600 with faster bus would certainly fit that bill.

Do you need a 64-bit PB ?? Not really - unless you need 8 GiB of RAM in your PB 64-bit won't do much for you. (And besides, a PB with 8 GiB of RAM would be big, hot and heavy!!!)

Maybe someday in the next year or two there will be enough 64-bit applications (absolutely none for OS X today, none) and memory will shrink so that 8 GiB will fit in 2 SO-DIMMs.

And, that someday in the next year or two, the dual-core e700 with 64-bit addressing will be available.
_________________________

Think about it - doesn't it make sense to continue to base the PB on chips from a low-wattage embedded heritage rather than from a heritage of bad-ass liquid cooled servers?

A G4 derivative with a fast bus, better memory, dual-core would be a great chip for a high end laptop.

You want performance, right? Does it matter which chip is used if the results are good?

I wonder if Apple will be smart enough to *never* put a G5 (or G6) in a PB....

What you say makes absolutely sense. My wording was not correct, so my point was not understood as I meant it to be. Anyway, I don't think in any way that it would be wise to put a 970 in a PB. But as someone earlier said, to the market nowadays they don't always look at the performance, but what they look at is "Another G4? G5 is newer, it'll smoke that." To answer your question "Does it matter which chip is used if the results are good?" To me, the chip inside is not important, if it just gives performance. The market looks at the number after the G, unfortunately.

I agree that Apple should keep on putting low-wattage chips, rather than server-chips in the PB's.

The future for Apple's partnership with FreeScale looks quite good too, if the (G4-derivative?) e600 and e700 are what they seem to be. I agree that today 64-bit is next to useless in PowerBooks, before we get higher capacity SO-DIMMs and a 64-bit system.

I guess that's all I had to say.

 

[G4-power]

It has been said many times, but I can repeat it: there are users who don't upgrade every other year, but like to keep their equipment for 5-7 years ! The G4 is old tech and 64 bit is the future for Apple products in 3 years there will be no 32 bit offering any more and the OS will go completly 64 bit.

Got my point ?

I know this post was addressed to AidenShaw, but I kinda agree with you.
There has been occasions when an Apple computer comes up with "Max memory something", like for example the iMac G4 (1GB). Then the 1GB DIMMs came and the max memory was bumped to 1.5 GB (2 GB?) Well, this could happen, but another point, that CmdrLaForge said. Some people keep the equipment for many many years. Like in the G5 intro "The computers of the future will be 64-bit. But the future is now at Apple" If OS X and new apps are "fully" 64-bit in 3-4 years, it will speed up the G5. But oh well, they may come up with a 32-bit chip for the next PB speed-bump, and it will be fast now (I hope), but not as fast as the G5-based would be.
We can see the same now with G3/G4. In non-altivec code G3 wins easily. But in AltiVec code a 350 MHz G4 smokes a 500 MHz G3 (well, 600 and 800 too). Take for example AltiVec Fractal Carbon. 500 Mhz G3 achieved 430 Mflops, 350 MHz G4 1220 Mflops. I know, all programs haven't gone the AltiVec way yet, but I'd say most. And this will happen for 64-bit too, I hope.

 

[CmdrLaForge]

I know this post was addressed to AidenShaw, but I kinda agree with you.
There has been occasions when an Apple computer comes up with "Max memory something", like for example the iMac G4 (1GB). Then the 1GB DIMMs came and the max memory was bumped to 1.5 GB (2 GB?) Well, this could happen, but another point, that CmdrLaForge said. Some people keep the equipment for many many years. Like in the G5 intro "The computers of the future will be 64-bit. But the future is now at Apple" If OS X and new apps are "fully" 64-bit in 3-4 years, it will speed up the G5. But oh well, they may come up with a 32-bit chip for the next PB speed-bump, and it will be fast now (I hope), but not as fast as the G5-based would be.
We can see the same now with G3/G4. In non-altivec code G3 wins easily. But in AltiVec code a 350 MHz G4 smokes a 500 MHz G3 (well, 600 and 800 too). Take for example AltiVec Fractal Carbon. 500 Mhz G3 achieved 430 Mflops, 350 MHz G4 1220 Mflops. I know, all programs haven't gone the AltiVec way yet, but I'd say most. And this will happen for 64-bit too, I hope.

Exactly, and thats the problem. E.g. iDVD. Comes with every Mac. But I can't use it on my G3 iBook even so it was delivered with the iBook as part of iLife.

And that will happen with many new apps (maybe Motion 2 ?) or even the OS itself.

For me - I would buy a dual core 2 GHz powerbook in a second if they release that - but many people just like to know that there equipment is future proof.

Cheers
LaForge

 

[AidenShaw]

It has been said many times, but I can repeat it: there are users who don't upgrade every other year, but like to keep their equipment for 5-7 years !

Good point, but Apple's track record has been to force more frequent hardware upgrades.

I would be surprised if the first generation G5 PowerBooks will support much larger SO-DIMMs than the max at introduction. Apple would prefer that you buy 2nd or 3rd generation systems, rather than take the effort (expense) of firmware upgrades for older systems to support new technology. Just like the iPod upgrades aren't released for older iPods, it's not in Apple's profit plan to keep upgrading old systems.

Too many times "future-proofing" means that you spend more than you needed to today, only to find that the newer technology won't work down the road. (What if those 4 GiB SO-DIMMs are only available in DDR3 format, and your 1st gen system only supports DDR?)

You can waste money over the long run and be stuck for years with slow CPUs and other components - when compared to a strategy of buying less expensive machines more often. This is especially true when prices are on a general downward trend, and a price/performance is on a steep downward trend.
____________

Plus, devices are inter-related today, so the "5 to 7 year" goal might be a masochistic approach.

For example, I replaced my Compaq laptop after about 2 1/2 years.

Was it not doing as well as it originally was doing? Of course not, it was still as fast as ever.

But, when I got the Compaq I had a 1 megapixel digital camera. I now have a 5 megapixel SLR-like model and a 4 megapixel pocketsize. The Compaq was too slow at processing these much larger pictures.

Especially in a laptop, no amount of "future-proofing" would have compensated for the fact that the CPU was just too slow. The CPU is in a socket, but the chipset couldn't handle a newer generation CPU. The chipset also supported SDRAM, so I couldn't add new, denser DDR SO-DIMMs.

So, I replaced it with a system with a CPU 4 times faster, 3.5 times the memory, faster bus and memory, GigE -- at half the price of the Compaq!

The G4 is old tech and 64 bit is the future for Apple products in 3 years there will be no 32 bit offering any more and the OS will go completly 64 bit.

I find this extremely unlikely - Apple's installed base is too large compared to their market share to so quickly abandon all the G3 and G4 systems out there. Can you imagine the class action lawsuits, let alone the bandwidth of whining on message boards? 😉 Apple also has to consider how much of their installed base would move to Longhorn instead of buying a G5....

Sure, there will be high-end *applications* that will be 64-bit only - but most apps have no use for 64-bit. Couple a huge 32-bit hardware base, and "no advantage to 64-bit" for most applications - I just don't see Apple taking such an arrogant step as moving to 64-bit only. (Look at the situation with the G3 today - still going strong after years of G4 and a year of G5.)

Note that Sun, which has no 32-bit SPARC hardware (AFAIK), still builds and sells 32-bit Solaris for SPARC. This is even though 64-bit Solaris runs the 32-bit Solaris applications natively.

 

[rareflares]

hey, any validity to this? if it's real, that really suggests "a supercomputer in your lap" as the next logical step.

is this real?

 

[AidenShaw]

hey, any validity to this? if it's real, that really suggests "a supercomputer in your lap" as the next logical step.

The logical step would be a reference to the G5 iMac.

Maybe "a supercomputer on your wall" if the rumours of a detachable, wall-mountable screen/system are true.


BTW, don't you find Apple's "supercomputer" hype to be a bit tiring, especially since most of the benchmarks show the G5 to be roughly the same speed as a Pentium/Xeon system (win some, lose some)?

 

[cloud 9]

The logical step would be a reference to the G5 iMac.

Maybe "a supercomputer on your wall" if the rumours of a detachable, wall-mountable screen/system are true.


BTW, don't you find Apple's "supercomputer" hype to be a bit tiring, especially since most of the benchmarks show the G5 to be roughly the same speed as a Pentium/Xeon system (win some, lose some)?

why do i want a computer on my wall? to look at? so i take a chair, sit down and stare at my wall? and a mouse and keyboard on my lap? 🙄

the logic step would be, something for at home and something for on the road 🙂 and not something for on the wall 😀 but that's my opinion

yes, i find the supercomputer hype also a bit over the top, but apple is very much into marketing, and super is better than mega, hyper, ultra, or an animal/tiger/fruit/... name for computer...imagine 'a megamac with tiger'...or a 'g5 gazelle' or something... 😀

 

[G4-power]

BTW, don't you find Apple's "supercomputer" hype to be a bit tiring, especially since most of the benchmarks show the G5 to be roughly the same speed as a Pentium/Xeon system (win some, lose some)?

At least I'm getting bored to the supercomputer hype. I think we all agree that the G5 is a much better chip, altogether, and clock-to-clock at least, compared to Pentium/Xeon. Just the thing that PPC is RISC-based rather than CISC makes it a lot faster.

And btw, the speed results Apple post showing that the G5 is superior to Wintel machines are a bit biased. Macs are generally slower on Photoshop because Adobe doesn't optimize it well enough for Macs. Apple chose the filters that work better on the Mac. But the biasing isn't as much that on the PC market, with MS claiming that Windows Server 2003 would be 500 % faster than Red Hat linux, yeah right.

 

[AidenShaw]

two things...

Just the thing that PPC is RISC-based rather than CISC makes it a lot faster.

1. PPC has many CISC-like instructions, and Pentiums and AMD chips have a RISC-like core, so the old distinctions are passée.

2. It's not a lot faster !! Sometimes it wins big, sometimes it loses big, usually it's in the same ballpark.

For example:

http://www.barefeats.com

 

[G4-power]

1. PPC has many CISC-like instructions, and Pentiums and AMD chips have a RISC-like core, so the old distinctions are passée.

2. It's not a lot faster !! Sometimes it wins big, sometimes it loses big, usually it's in the same ballpark.
http://www.barefeats.com

1. I know that nowadays the RISC/CISC battle is gone (x86 still sucks) but the x86 RISC-like-core and PPC CISC-like-instructions was news to me.

2.Yeah, I know. I wouldn't like to argue, but both Photoshop and Cinebench aren't very well optimized for Macs. But true, usually it's in the same ballpark.

 

[AidenShaw]

I know that nowadays the RISC/CISC battle is gone

It's gone because the two camps have basically merged, and are using the best features of both.

but the x86 RISC-like-core and PPC CISC-like-instructions was news to me.

The original RISC philosophy from 1974 was that one could get by with a small number of very simple instructions, which could be made blindingly fast.

Complex operations would need sequences of the simple instructions, but it would still be fast because the simple instructions were so fast. And, since much of the time only simple operations are needed, the overall program is also very fast.

Of course, the original RISC papers are from a time when top clock speeds were a few MHz and a few hundred thousand transistors on a chip were a lot.

Try to figure out how a chip with AltiVec could be called RISC by the early definition, or how it could have instructions like sqrt and the graphics instructions....
____________

Starting with the P6 architecture, Pentiums have used a RISC-like core that executes "micro-ops", and have a front end that translates the x86 instruction set into "micro-ops" for the core. Note that the P4 doesn't have an "instruction cache" in its specs, but instead lists a "trace cache". From Intel's site:

FAQ:
What does the Execution Trace Cache do?

Solution:
The Execution Trace Cache is an innovative way to implement a Level 1 instruction cache.

It caches decoded x86 instructions (micro-ops), thus removing the latency associated with the instruction decoder from the main execution loops.

In addition, the Execution Trace Cache stores these micro-ops in the path of program execution flow, where the results of branches in the code are integrated into the same cache line. This increases the instruction flow from the cache and makes better use of the overall cache storage space (12KB micro-ops) since the cache no longer stores instructions that are branched over and never executed. The result is a means to deliver a high volume of instructions to the processorâ??s execution units and a reduction in the overall time required to recover from branches that have been mispredicted.

http://www.intel.com/support/processors/sb/cs-001649-prd24.htm
___________


Here's another relevant quote (on a page with other interesting info):

Post-RISC
As the world enters the 21st century the CISC Vs. RISC arguments have been swept aside by the recognition that neither terms are accurate in their description.

The definition of 'Reduced' and 'Complex' instructions has begun to blur, RISC chips have increased in their complexity (compare the PPC 601 to the G4 as an example) and CISC chips have become more efficient.

The result are processors that are defined as RISC or CISC only by their ancestry.

The PowerPC 601, for example, supports more instructions than the Pentium. Yet the Pentium is a CISC chip, while the 601 is considered to be RISC.

CISC chips have also gained techniques associated with RISC processors. Intel describe the Pentium II as a CRISC processor, while AMD use a RISC architecture but remain compatible with the dominant x86 CISC processors.

Thus it is no longer important which camp the processor comes from, the emphasis has once-again been placed upon the operating system and the speed that it can execute instructions.

http://www.amigau.com/aig/riscisc.html

 

[jhu]

hmm... the risc/cisc debate still rages for some reason. i wish people would get it into their heads that there is no longer a risc/cisc distinction.

 

[G4-power]

Okay, thanks for the info AidenShaw.

 

[AidenShaw]

hmm... the risc/cisc debate still rages for some reason. i wish people would get it into their heads that there is no longer a risc/cisc distinction.

Agree, but first we have to get it into the heads of the marketing people !!

As long as "people" keep reading marketing hype about RISC, they'll think that it means something.


BTW, at the foot of the Amiga page that I quoted is a link to a long and thorough discussion of the topic at Ars Technica.

http://ars-technica.com/cpu/4q99/risc-cisc/rvc-1.html

In this paper, I'll argue the following points:

1. RISC was not a specific technology as much as it was a design strategy that developed in reaction to a particular school of thought in computer design. It was a rebellion against prevailing norms--norms that no longer prevail in today's world. Norms that I'll talk about.
   
2. "CISC" was invented retroactively as a catch-all term for the type of thinking against which RISC was a reaction.
   
3. We now live in a "post-RISC" world, where the terms RISC and CISC have lost their relevance (except to marketing departments and platform advocates). In a post-RISC world, each architecture and implementation must be judged on its own merits, and not in terms of a narrow, bipolar, compartmentalized worldview that tries to cram all designs into one of two "camps."

 

[CmdrLaForge]

Good point, but Apple's track record has been to force more frequent hardware upgrades.

I would be surprised if the first generation G5 PowerBooks will support much larger SO-DIMMs than the max at introduction. Apple would prefer that you buy 2nd or 3rd generation systems, rather than take the effort (expense) of firmware upgrades for older systems to support new technology. Just like the iPod upgrades aren't released for older iPods, it's not in Apple's profit plan to keep upgrading old systems.

Too many times "future-proofing" means that you spend more than you needed to today, only to find that the newer technology won't work down the road. (What if those 4 GiB SO-DIMMs are only available in DDR3 format, and your 1st gen system only supports DDR?)

You can waste money over the long run and be stuck for years with slow CPUs and other components - when compared to a strategy of buying less expensive machines more often. This is especially true when prices are on a general downward trend, and a price/performance is on a steep downward trend.
____________

Plus, devices are inter-related today, so the "5 to 7 year" goal might be a masochistic approach.

For example, I replaced my Compaq laptop after about 2 1/2 years.

Was it not doing as well as it originally was doing? Of course not, it was still as fast as ever.

But, when I got the Compaq I had a 1 megapixel digital camera. I now have a 5 megapixel SLR-like model and a 4 megapixel pocketsize. The Compaq was too slow at processing these much larger pictures.

Especially in a laptop, no amount of "future-proofing" would have compensated for the fact that the CPU was just too slow. The CPU is in a socket, but the chipset couldn't handle a newer generation CPU. The chipset also supported SDRAM, so I couldn't add new, denser DDR SO-DIMMs.

So, I replaced it with a system with a CPU 4 times faster, 3.5 times the memory, faster bus and memory, GigE -- at half the price of the Compaq!

I find this extremely unlikely - Apple's installed base is too large compared to their market share to so quickly abandon all the G3 and G4 systems out there. Can you imagine the class action lawsuits, let alone the bandwidth of whining on message boards? 😉 Apple also has to consider how much of their installed base would move to Longhorn instead of buying a G5....

Sure, there will be high-end *applications* that will be 64-bit only - but most apps have no use for 64-bit. Couple a huge 32-bit hardware base, and "no advantage to 64-bit" for most applications - I just don't see Apple taking such an arrogant step as moving to 64-bit only. (Look at the situation with the G3 today - still going strong after years of G4 and a year of G5.)

Note that Sun, which has no 32-bit SPARC hardware (AFAIK), still builds and sells 32-bit Solaris for SPARC. This is even though 64-bit Solaris runs the 32-bit Solaris applications natively.

Well, you have some valid points, but still, people want the G5 in the powerbook because they want a more future proof system. Period.

And yes, Apple will release a 64 Bit OS ONLY ! They will not really care about all those G3 and G4 owners. Do you remember what they did with OS X in comparison to 9 ? That was a even huger step and they made it. And its good. Its one of the reasons why Windows is soo much worse. Because they always keep the compatibility.

 

[AidenShaw]

And yes, Apple will release a 64 Bit OS ONLY ! ... Do you remember what they did with OS X in comparison to 9 ? That was a even huger step and they made it.

I'm sorry I wasn't more clear.

I was arguing that I found your timescale (3 years to 64-bit only) to be "extremely unlikely". I said that Apple wouldn't "so quickly abandon all the G3 and G4 systems". I was unclear with a later statement "moving to 64-bit only" - I should have repeated the "so quickly" qualification in some form on that line.


Surely there will come a time when Apple (and Microsoft) abandon the 32-bit systems. Of course, by that time we'll be expecting 4 GiB to 8 GiB of video RAM on our graphics cards, and the final PowerMac G4s will seem as quaint as an Apple [] .... 😉

The other comment is about the "hugeness" of the step. If you don't need more than 4 GiB of RAM per process, then a 64-bit operating system isn't even a step - it's of no benefit....

 

[MikeBike]

It has been said many times, but I can repeat it: there are users who don't upgrade every other year, but like to keep their equipment for 5-7 years ! The G4 is old tech and 64 bit is the future for Apple products in 3 years there will be no 32 bit offering any more and the OS will go completly 64 bit.

Got my point ?

You're not getting our point.
The G4, clock for clock, is Stronger then the G5.
So, the G5 must clock higher to give better performance.
Then, there's the DualCoreG4, which should give you from 80-100% better performance in 1 processor cycle.

But, you want a single core G5, HOPING for OS X to Optimize it into giving you More Than 100% increase in performance, In The Coming Years.
If Apple improves the performance of the OS by 20% a year, wouldn't it take you 4 YEARS or more, to equal the performance of the DualCoreG4, which you could enjoy for 4 years?!?

The Current G5 doesn't have a lot of tricks to increase it's performance relative to the current G4. In the future, rumor has it, IBM will improve altivec( in the Hardware ) and then there's SMT-symetric multi-threading coming in the G6. But, again, that's a hardware improvement.

Today, you're best bet to Future proof your machine would be to Hope you could buy a DualCoreG4.

But, that's just my two cents.

 

[AidenShaw]

you miss the point

You're not getting our point.
The G4, clock for clock, is Stronger then the G5.
So, the G5 must clock higher to give better performance.
Then, there's the DualCoreG4, which should give you from 80-100% better performance in 1 processor cycle.

You're missing the point.

"5" (as in "G5") is bigger than "4" (as in "G4"), therefore the "G5" is obviously better than "G4" - so we must have it.


[/sarcasm]


It doesn't matter if a dual-core G4 makes a lot more sense, marketing is pushing a "G5"....

 

[CmdrLaForge]

You're not getting our point.
The G4, clock for clock, is Stronger then the G5.
So, the G5 must clock higher to give better performance.
Then, there's the DualCoreG4, which should give you from 80-100% better performance in 1 processor cycle.

But, you want a single core G5, HOPING for OS X to Optimize it into giving you More Than 100% increase in performance, In The Coming Years.
If Apple improves the performance of the OS by 20% a year, wouldn't it take you 4 YEARS or more, to equal the performance of the DualCoreG4, which you could enjoy for 4 years?!?

The Current G5 doesn't have a lot of tricks to increase it's performance relative to the current G4. In the future, rumor has it, IBM will improve altivec( in the Hardware ) and then there's SMT-symetric multi-threading coming in the G6. But, again, that's a hardware improvement.

Today, you're best bet to Future proof your machine would be to Hope you could buy a DualCoreG4.

But, that's just my two cents.

I really got your point and as I earlier said, I would buy a dual core G4 powerbook in a second and would definitly prefer it a G5 system at the same clock speed.

But I was talking about many other users that just don't see it that way, wanting a more "future proof" system.

Cheers

 

[CholEoptera36]

I really got your point and as I earlier said, I would buy a dual core G4 powerbook in a second and would definitly prefer it a G5 system at the same clock speed.

Yeah I agree too CmdrLaForge, just like the majority of people on here can see that dual-core is the way of the future. A single dual-core processor can do what dual CPU chipsets can do.

Here's a good question though (if I over-read it from someone earlier then sorry), how many cores can you put on an IBM die?😕

-I know that by 2007 the quad-core AMD models are supposed to be released, and I've heard a couple different answeres as to how many cores can exist on a 939 pin chip like their current FX models. I've heard some numbers like 77, even 666 cores on the current chips. I can't seem to get a legit answer from articles so far. Would we just keep seeing more cores and more cores over and over again, much like how we expand more lanes for drivers in major cities? It definitely is a much better alternative than adding more CPUs like what IBM did with dual G5's, but what's the core limmit per chip?

-Also it's interesting to think that Apple would make a move toward AMD. If so, kudos to Apple for getting smarter. I know that IBM and AMD are currently in a joint development of multi-core technology, but AMD by far is the backbone of multi-core architecture as they designed it from the ground up.

-Hopefully by the time Apple releases a multi-core chip they won't have to worry about it being G4 still. Most likely all effort is going into the move to 100% 64bit CPUs first, then multi-core.

 

[thatwendigo]

Here's a good question though (if I over-read it from someone earlier then sorry), how many cores can you put on an IBM die?😕

It's mostly limited by transistor count, cache, and other components of the actual chip. The current Power5 is technically a dual-core design, but it's sold in four chip bundles that very nearly make it an octuple core. With SMT enabled, it presents itself to an operating system as sixteen logical cores.

-Also it's interesting to think that Apple would make a move toward AMD. If so, kudos to Apple for getting smarter. I know that IBM and AMD are currently in a joint development of multi-core technology, but AMD by far is the backbone of multi-core architecture as they designed it from the ground up.

Uh, no. I can understand why you'd want to believe that AMD is some kind of amazing leader in dual-core and multi-core technology, but the truth of the matter is that other companies have been doing it for a long, long time in the server market. IBM, in fact, had the dual-core Power4 quite some time ago (roughly 2001), and that's just the first dual core I can name off the top of my head.

Also, it wouldn't at ALL be "smarter" for Apple to move to an x86-based architecture, because it opens a Pandora's Box of problems. You could kiss the open usage model goodbye on Mac OS X, because they'd have to use copy protection or some kind of hardware detection to keep it from being pirated. People would have to replace their applications because they wouldn't run on the alternate processor and PC emulation of macs is less efficient than the other way around. More than anything, though, IBM owns the patents that other people are licensing to work on these designs of theirs... Apple's in a partnership with the company that files the most computer-related patents (and overall patents, I believe) in the world, every year.

-Hopefully by the time Apple releases a multi-core chip they won't have to worry about it being G4 still. Most likely all effort is going into the move to 100% 64bit CPUs first, then multi-core.

Why is any of that hopeful? The G4 is still a good chip, if a few changes were made to the basic systems. Guess what? The FreeScale e600 makes those changes! Everything that you like about AMD chips, other than the x86 instruction set, is going to be there - one die memory control, fast system busses, and so on.

People have latched onto the 64-bit thing way too quickly. It won't make a single bit of difference to home users for a while yet, because they're not going to have more than 4GB or RAM or be using 64-bit math. Basically, it's the same thing as what happened with the G5... Numbers got higher and it dazzles those who don't understand it.

 

[CholEoptera36]

It's mostly limited by transistor count, cache, and other components of the actual chip. The current Power5 is technically a dual-core design, but it's sold in four chip bundles that very nearly make it an octuple core. With SMT enabled, it presents itself to an operating system as sixteen logical cores.

Interesting, I'mma read up some more on it. But why does cache effect how many cores can be put on a die?

Also, it wouldn't at ALL be "smarter" for Apple to move to an x86-based architecture, because it opens a Pandora's Box of problems.

I never said move to x86 architecture, I'm not a moron. Are you suggesting AMD cannot produce CPU's for Apple? If Motorolla can do it, then what do you think...

Why is any of that hopeful? The G4 is still a good chip, if a few changes were made to the basic systems. Guess what? The FreeScale e600 makes those changes! Everything that you like about AMD chips, other than the x86 instruction set, is going to be there - one die memory control, fast system busses, and so on.

No one made it out that the G4 sucks. All I was saying is that I'm placing my bets that Apple will develop all chips in 64-bit before going dual-core. Why would they go dual-core on G4's when in a few years you might not even be able to use them on upcomming Operating System upgrades? And BTW AMD is an outstanding company with lots of proof in the short period of time they've come to play with the big dogs. And no one else is even close to multi-core anything in a personal computer, I wasn't talking about servers. Although everybody's gotta get technical now and then so my mistake for not saying "personal computer multi-core chips." 🙄 AMD's already got multi-core chips for personal computers. Where does IBM stand on that? What have they got in multi-core that isn't for servers? (this whole thread is about PowerBooks let me remind you, scratch the servers).

 

[AidenShaw]

32-bit CPUs can do 64-bit math, but 64-bit CPUs do 64-bit integers faster

People have latched onto the 64-bit thing way too quickly. It won't make a single bit of difference to home users for a while yet, because they're not going to have more than 4GB or RAM or be using 64-bit math.

Well, we're all using 64-bit floating point and 64-bit integer math, and have been for a very long time. 😉

Real native 64-bit floating point has been in the hardware in most modern machines (386 and later, all power/powerpc, alpha, vax, sparc, pa-risc,....).

On 32-bit machines, 64-bit integers are presented as native by the compilers - but executed as a sequence of 32-bit integer operations.


I think what you meant instead of "not be using 64-bit math" was something like "not be doing performance-critical operations on 64-bit integers".

All computers are doing 64-bit integer operations today, but if you're crunching a lot of them a 64-bit CPU with native 64-bit integer support will have a big advantage.

 

[CholEoptera36]

All computers are doing 64-bit integer operations today, but if you're crunching a lot of them a 64-bit CPU with native 64-bit integer support will have a big advantage.

I agree with that. It's known by now(or atleast you'd think so) that running a 64-bit OS on a 64-bit processor will increase performance no matter what machine it's on, no matter what the usage is for. Yeah, as of the moment in time... there are applications that would need a 64-bit processor to run 64-bit performance-critical applications. And this might as of the time, rarely apply to the home user. But it won't be too long before today's performance-critical 64-bit apps are tomorrows normal apps. I guess it just makes more sense to me for Apple to move all out 64-bit first and establish that much and get everyone on the same page. A lot of Apple's strategy with upgrades is to move everone at a certain pace to keep us all upgrading togather. The G4 is falling further and further behind as time goes by. Now only eMacs and the Apple notebooks are left to upgrade to G5s. I think that should come first before Apple thinks about having anything multi-core. JMO