And the 7460 is supposed to go to a .13 process, but the roadmap says the G4 has " .15µ copper process for initial G4 product (migrating to SOI)," and then the G5 has ".13µ process with SOI initial G5 product." Doesn't that suggest that the transition from .15 to .13 is they key to the transition from G4 to G5?
Hypothesis - the 7460 is released in January and called the G5.
I heard the G5 "will" come out in March, not January. And I am currently downloading 10.2 which is 64-bit. And Mac OS 10.2 is due out in March as I heard as well. :cool:
<strong>I heard the G5 "will" come out in March, not January. And I am currently downloading 10.2 which is 64-bit. And Mac OS 10.2 is due out in March as I heard as well. :cool: </strong><hr></blockquote>
I really find it VERY hard to believe that Apple would make very single other processor out there obsolete with a 64bit compiled version of 10.2, it makes no sense. Apple has been pushing this product, and now you can't use it?? C'mon, get real.
10 will more than likely stay at 32 bit. Maybe forever. That might be a OS 11 (or is it XI) version, but NOT 10.2. Apple just would be that stupid.
Ok I may gone overboard . I will ask tomorrow or Monday about when is G5's released date just to be extra certain. And I do know the person who knows what we don't know, and need to know . As for the OS, I may have exaggerated a bit, who knows, it probably wont be released until 2 years from now (the whole 64-bit thing), because if it was released next year, it would suck for all Mac Users including me. So you're probably right. Maybe OS 11 or something, but not tomorrow.
I'm not supporting claims of the next 10 release being 64bit, I think it's way too soon and little benefit unless you're playing with a lot of data.
But the claims of a 64bit version of 10 making all other processors out of date are wrong. 10/HFS allows you to combine code for different platforms into a single FAT binary. For example I could give you a FAT version of Photoshop with both PPC and x86 code and the operating system would pick the appropriate segment to load.
It's not as simple as I make it sound. Apple would have to do some tricks with the linker for 32/64bit library calls and various other components that cross the boundary. But to sum it up it's very possible for Apple to deliver a 32/64bit OS.
The other thing that is interesting is that the speeds indicated are exactly what the Register's source said: Debut at 1.2, 1.4, 1.6 and then move to 2ghz by summer.
That Geek.com info was there before The Register and MOSR got all their insider reports on the G5 too.
I think they should announce the G5 if it can ship by March. Kinda sucks that it wouldn't ship until Febuary or March but I guess I could wait a little longer.
What they could do with a 10.2 is make it a 64-bit OS but make it 32-bit backwards compatible so it would basically run as a 32-bit OS on any computer that doesn't have the G5. Or maybe it will just be 32-bit and the G5 will run software 32-bit until Apple brings out a 64-bit version of X.
If the G5 is to be in the 85XX product line, as a lot of people have been posting, I thought it was interesting that the MPC8540 won't be even sampling until " the second half of 2002."
The MPC8540 is an anounced chip with the specifications fairly well outlined. The MPC85XX - G5 has not been announced and information on it is relatively speculative.
My guess is that the G5 will come out after the MPC8540, hence, would be sampling later in the 2nd half of 2002 and introduced at the end of the year. Hope I'm wrong, but .......
Think about it. When Motorola puts an 8 infront of the processor numeration, it means it will be a SOC type embedded chip. This must mean that the G5 was changed from 7500 to 8500 because it will have built-in features that were normally external to the CPU core like memory controller, ethernet, and PCI controller. This should reduce latency and provide much more bandwidth.
<strong>Product numbers are just so much marketing, and marketing departments can play hell with company or even industry conventions. 8500, 7500, neither really means anything until you can slap one on a board.</strong><hr></blockquote>
Agreed. That reference above means nothing. It could be a freakin embedded chip for all we know.
I think Apple isn't going to release this kick-ass new chip with a 20 year old OS as a default system.
They'll surely ship it with Mac OS X preinstalled and set for default, and we all know this won't happen before March. I say we'll see G5 around 1400MHz with GeForce 4 cards at MWNY.
And forget about that Raycer thing, I doubt there will ever be a "product" based on that. Some minor chips sure, but not a graphics card or something like that.
"Agreed. That reference above means nothing. It could be a freakin embedded chip for all we know."<hr></blockquote>
The MPC8540 is an embedded chip. But so is the G4. Below is the directory for finding the G4 on Motorola's web site.
Motorola : Semiconductors : Product Catalog : 32-Bit Embedded Processors : PowerPC ISA : MPC7XXX, MPC7XX and MPC6XX Host Processors
Point is the G5 is supposed to be in the 85XX family of processors and a much less complicated chip, 8540, won't be sampling until the 2nd half of 2002.
Maybe the G5 development has been secretly accelerated beyond the 8540 and is near final design, but.........well...... this seems to be wishful thinking.
Hope I'm totally wrong, especially since I will be buying a new low end tower sometime in Jan.
[quote]All we need from Apples raycer acquisition is that tiny little Quartz 2D chip... Let the big boys worry about 3D.<hr></blockquote>
3D acceleration is vector acceleration. Quartz is vector-based. 2D acceleration affects bitmaps.
Apple could design a chip that would accelerate Quartz, but it would be redundant. They'd be better off finding a way to rewrite Quartz to take advantage of OpenGL, and using whatever 3D acceleration the video card provides.
I have a hunch that that optimization would have less of an impact on overall performance than a few rounds of tweaking Carbon (and PowerPlant, and Finder). Quartz is already pretty efficient. On the other hand, some Carbon apps seem to spam the OS with events (AppleWorks 6.2 reportedly posts 8,000 events per second when it's idling!) and that's a performance killer right there.
Don't forget that Apple doesn't like Mot releasing info about new chips, Apple likes to be the one to introduce them in their machines. If Mot released info (legit info as in Press Release) about the G5 we would all be able to see everything it would have and figure a release date.
My guess is that the 8540 (which I've looked at for a good guess of what will be in the G5) is Mots way of saying this is what we have in the pipe without giving away the G5 release for Apple.
Comments
And the 7460 is supposed to go to a .13 process, but the roadmap says the G4 has " .15µ copper process for initial G4 product (migrating to SOI)," and then the G5 has ".13µ process with SOI initial G5 product." Doesn't that suggest that the transition from .15 to .13 is they key to the transition from G4 to G5?
Hypothesis - the 7460 is released in January and called the G5.
<strong>I heard the G5 "will" come out in March, not January. And I am currently downloading 10.2 which is 64-bit. And Mac OS 10.2 is due out in March as I heard as well. :cool: </strong><hr></blockquote>
I really find it VERY hard to believe that Apple would make very single other processor out there obsolete with a 64bit compiled version of 10.2, it makes no sense. Apple has been pushing this product, and now you can't use it?? C'mon, get real.
10 will more than likely stay at 32 bit. Maybe forever. That might be a OS 11 (or is it XI) version, but NOT 10.2. Apple just would be that stupid.
[ 11-25-2001: Message edited by: artenman ]
[ 11-25-2001: Message edited by: artenman ]</p>
But the claims of a 64bit version of 10 making all other processors out of date are wrong. 10/HFS allows you to combine code for different platforms into a single FAT binary. For example I could give you a FAT version of Photoshop with both PPC and x86 code and the operating system would pick the appropriate segment to load.
It's not as simple as I make it sound. Apple would have to do some tricks with the linker for 32/64bit library calls and various other components that cross the boundary. But to sum it up it's very possible for Apple to deliver a 32/64bit OS.
-Bill
<strong>I just thought I'd get this topic going with some info I found -
Moto intros the G5 <a href="http://maccentral.macworld.com/news/0009/27.mot.shtml"; target="_blank">here</a>
Basic feature set of the G5 <a href="http://www.geek.com/procspec/apple/g5.htm"; target="_blank">here</a>
Just some more info/juice to breath some life back into this thread
This seems to be the most credible evdience yet for a G5 release in Jan....possibly available in March.
I am starting to get giddy.
I think they should announce the G5 if it can ship by March. Kinda sucks that it wouldn't ship until Febuary or March but I guess I could wait a little longer.
What they could do with a 10.2 is make it a 64-bit OS but make it 32-bit backwards compatible so it would basically run as a 32-bit OS on any computer that doesn't have the G5. Or maybe it will just be 32-bit and the G5 will run software 32-bit until Apple brings out a 64-bit version of X.
<strong>Oh, and I am still downloading this so called 10.2 64-bit OS. Interesting.</strong><hr></blockquote>
From where?
If the G5 is to be in the 85XX product line, as a lot of people have been posting, I thought it was interesting that the MPC8540 won't be even sampling until " the second half of 2002."
The MPC8540 is an anounced chip with the specifications fairly well outlined. The MPC85XX - G5 has not been announced and information on it is relatively speculative.
My guess is that the G5 will come out after the MPC8540, hence, would be sampling later in the 2nd half of 2002 and introduced at the end of the year. Hope I'm wrong, but .......
<strong>Product numbers are just so much marketing, and marketing departments can play hell with company or even industry conventions. 8500, 7500, neither really means anything until you can slap one on a board.</strong><hr></blockquote>
Agreed. That reference above means nothing. It could be a freakin embedded chip for all we know.
I think Apple isn't going to release this kick-ass new chip with a 20 year old OS as a default system.
They'll surely ship it with Mac OS X preinstalled and set for default, and we all know this won't happen before March. I say we'll see G5 around 1400MHz with GeForce 4 cards at MWNY.
And forget about that Raycer thing, I doubt there will ever be a "product" based on that. Some minor chips sure, but not a graphics card or something like that.
G-news
"Agreed. That reference above means nothing. It could be a freakin embedded chip for all we know."<hr></blockquote>
The MPC8540 is an embedded chip. But so is the G4. Below is the directory for finding the G4 on Motorola's web site.
Motorola : Semiconductors : Product Catalog : 32-Bit Embedded Processors : PowerPC ISA : MPC7XXX, MPC7XX and MPC6XX Host Processors
Point is the G5 is supposed to be in the 85XX family of processors and a much less complicated chip, 8540, won't be sampling until the 2nd half of 2002.
Maybe the G5 development has been secretly accelerated beyond the 8540 and is near final design, but.........well...... this seems to be wishful thinking.
Hope I'm totally wrong, especially since I will be buying a new low end tower sometime in Jan.
[quote]All we need from Apples raycer acquisition is that tiny little Quartz 2D chip... Let the big boys worry about 3D.<hr></blockquote>
3D acceleration is vector acceleration. Quartz is vector-based. 2D acceleration affects bitmaps.
Apple could design a chip that would accelerate Quartz, but it would be redundant. They'd be better off finding a way to rewrite Quartz to take advantage of OpenGL, and using whatever 3D acceleration the video card provides.
I have a hunch that that optimization would have less of an impact on overall performance than a few rounds of tweaking Carbon (and PowerPlant, and Finder). Quartz is already pretty efficient. On the other hand, some Carbon apps seem to spam the OS with events (AppleWorks 6.2 reportedly posts 8,000 events per second when it's idling!) and that's a performance killer right there.
My guess is that the 8540 (which I've looked at for a good guess of what will be in the G5) is Mots way of saying this is what we have in the pipe without giving away the G5 release for Apple.