IBM G4's in 2003

[arn]

MacEdition posts (http://www.macedition.com/nmr/nmr_20020914.php) vague details and a report that IBM's 64-bit Power 4 Chip will make its way to the next generation of PowerMacs with a new Apple Processor Interconect bus.

While the new systems are reported to be in prototype form presently, are not expected for full production until later in 2003.

[vniow]

That report's kinda vaugue, nothing really specific about it and nothing that differs it from other rumors, except for the fact that it mentions something called the Apple Processor Interconnect bus and that it won't coincide with the official death of OS9.
Bah. :(

[rice_web]

first post! (maybe)

Anyway, This sounds promising. The general concensus seems to be that we'll see Power4 processors of some sort in 2003.

Most every site has ventured to guess that the Power4 is coming.

[Edit: no first post, dangit ]

[ibjoshua]

more fuel on the fire.
what everyone wants to hear etc etc.
we all hope it comes true - i just hope it's not wishful thinking. i pity anyone who has no choice but to buy a new machine in the next 6-9 months.

i_b_joshua

[ennerseed]

Sounds very good to me.
Has the gayblade ever been misleading?

[jaykk]

If this is true, it is one of the best thing for Apple. Its time to buy some apple stocks..

[QuiteSure]

ApplePI? Oh please.

[shadowfax0]

But will it still be the G4? Or is it also time for a name change....Also, does this mean Apple will be the first one to offer a 64-bit processor for the masses?

[ibjoshua]

Originally posted by shadowfax0
But will it still be the G4?

i believe the 'G' stands for 'generation' so i think if you wanted you could call it a G5

but that would probably confuse people so why not just call it a 'good idea'.

i_b_joshua

[reyesmac]

If Apple does have some chip that will replace the G4 next year, who in here thinks that Apple will have them go more than 30% faster than the previous generation? What has been the biggest speed bump that Apple has had compared to what it replaced? Will they use dual processors, or will they just put one that is a little faster than two? Hope we dont get the usual top two getting upgraded and the bottom getting Yikesed.

What will the low end look like? Would the low end going from dual 867 to a 1gig be enough? Has Apple ever given us a speedbump that has made the low end faster than the high end that it replaces? I have played with the new macs and they seem fast enough as long as you buy more ram for it but what size speedbumb would be enough to make PC users heads turn?

[bousozoku]

Processor Interconnect bus? Does this sound like the same business that's already running to allow the processor to interface with memory in interesting ways?

I'll wait to see but I would be skeptical as to whether these systems will ever come to light and, if so, whether they will have good or extraordinary performance.

[G4scott]

Originally posted by QuiteSure
ApplePI? Oh please.

I didn't notice it until you said this... I'm laughing so hard right now :D :D :D

What's MacEdition's track record on rumors? These rumors seem pretty good... We keep on hearing how far the G4 is behind the P4, but really, even with 3ghz clock speeds, I don't see them hurting Apple's market share any more than 2+ghz clock speeds...

I'm crossing my fingers...

Although, my next computer might be a dual 867... I can get $500 for my G4 400 at Comp USA...

[tjwett]

didn't anyone see their names? this is obviously fake. pretty funny though.

[DaveGee]

Originally posted by tjwett
didn't anyone see their names? this is obviously fake. pretty funny though.

Rule is with TGB (as it was with Mac The Knife) you usually gotta look PAST the 1st paragraph to get on to the TRUE meat of the story. Been that way for a long long time. TGB doesn't talk trash... You do see him talking about iWalks and QUAD Processor PPC based monsters and the day you seem him talking about em... well then that means we might really see em come true.

Biggest problem with TGB is he usually waits for REAL confirmation about the stuff he is working on and as such he doesn't report on nearly as much stuff as say SpyMac or MOSR does... (isn't about time MOSR dusts off that QUAD CPU rumor again?) :D

Dave

[Choppaface]

what is with that guy's writing style?? O_o

[soilchmst]

Originally posted by DaveGee

QUAD CPU rumor again?)

Did someone say quad processors ;) :D :D
I knew they aren't going to appear (any time in my youth, at least), and I would could never use that much power. BUT, I would like to say I have a quad processor Power 4 :cool:

[nuckinfutz]

If Apple does have some chip that will replace the G4 next year, who in here thinks that Apple will have them go more than 30% faster than the previous generation?

If Apple uses this new chips announced by IBM on Oct 15 then they won't have too many problems getting %30 more speed. This Proc supports up to 6.4GBps bandwith. Keep in mind today AMD/Intel systems are hovering around 2.6 to low 3.x Mem Bandwidth. Plus this Proc looks to be 8way Superscalar which is double the current G4 capabilities theoretically of course.

Processor Interconnect bus? Does this sound like the same business that's already running to allow the processor to interface with memory in interesting ways?


My guess is it would be a bus that supercedes Motorolas MPX bus. Proc to memory. I'm sure it will probably offer more features than MPX and will definitely support some nexgen tech. Wish NMR had more info on this.

didn't anyone see their names? this is obviously fake. pretty funny though.

I've been reading this guys stuff for what 5-6 years now when he wrote for Macweek Magazine. His stuff is legit. He was the first person I new that broke the rumors of Connectix VirtualPC back when the idea of Connectix shipping a PC Emulator was akin to Quark shipping an Audio app or something. Hearing this from him makes me feel MUCH more positive.



what is with that guy's writing style?? O_o

That's "The Gay Blades" style. I just cut to the chase and reade the bottom.

[dongmin]

Does this Apple PI thing mean they're not gonna use RapidI/O or Hypertransport? Or are they all separate things?

[Nipsy]

Originally posted by reyesmac
If Apple does have some chip that will replace the G4 next year, who in here thinks that Apple will have them go more than 30% faster than the previous generation? What has been the biggest speed bump that Apple has had compared to what it replaced?

Well, it will be MUCH faster. Much much faster, but it may still be a 1GHz chip.

The 1GHz Power4 I read about was blowing the doors off of everything in the consumer market by a factor of 5, so it is conceivable that a Power4 Lite would be twice as fast as the 3.2-3.5GHz Pentium 4 that will be shipping then, even if it is a 1GHz chip.

However, some people will still be concerned about the GHz disparity, regardless of processor power. These people deserve Windows!

[iwantanewmac]

Originally posted by jaykk
If this is true, it is one of the best thing for Apple. Its time to buy some apple stocks..

you might reconsider that..........:)

[barkmonster]

If Apple does have some chip that will replace the G4 next year, who in here thinks that Apple will have them go more than 30% faster than the previous generation?

46.6% is the largest speed bump apple have ever made on a whole lineup of powermacs. That was when they went from 500Mhz to 733Mhz. It wasn't too impressive in a lot of benchmarks because it was nowhere near 46% faster.

57.2% is the largest speed bump on 1 model when they went from 466Mhz digital audio to 733Mhz Quicksilver on the entry level model.

since the G4 jumped from 500Mhz to 733Mhz the fastest chip in the following range has always been not only predicted with pinpoint accuracy by overclocking attempts on the previous range but speed increases have been in 15 - 25% range :

500Mhz - 733Mhz - 46.6%
733Mhz - 867Mhz - 18.3%
867Mhz - 1Ghz - 15.4%
1Ghz - 1.25Ghz - 25%

If clockspeed actually equates to actual performance increases when you're comparing the same CPU family, the next power mac should be a dual model at 1.5 - 1.6Ghz. That's a 25 - 30% speed increase and I can't see someone buying a dual 1.25Ghz G4 this year only to buy a dual 1.6Ghz next year unless there's some serious change in the architechture that makes the actual perceived speed increase more like 50 - 60%.

[peterjhill]

Originally posted by Choppaface
what is with that guy's writing style?? O_o

It all goes back to MacWeek Magazine. MacWeek was like MOSR, but in print. I was sorry to see it get sucked up into (the sucky) maccentral. Mac the Knife always had the last page editorial/rumor page. The web version always had cool but strange links from the text. I hope that the MacEdition guy is the same person.

Anyone here have an official Mac the Knife coffee mug? That would be worth something.

[AmigaMac]

For one, Apple is not going to leave the PowerPC architecture! So that pretty much kills the specualtion of anything x86 (which would be dead on arrival anyways). The question is, who was going to deliver the 64bit incarnation of PowerPC to Apple first? Well it seems that the answer is laying right in our laps... IBM! I think Motorola/IBM might be switching roles in Apple's requirements where IBM will be delivering for the desktops and Motorola (if they get a 64bit G4/G5 of their own out there) will make its way to laptops because of their low power consumption while delivering high performance. Or IBM could take the whole ball and chain and Moto gets left behind. I figure IBM will be the best bet since Amiga has also adopted their offerings (though the AmigaOnes will get G4s in the near future).

Bottomline is that the PowerPC is here to stay folks!

[Telomar]

Originally posted by Nipsy


Well, it will be MUCH faster. Much much faster, but it may still be a 1GHz chip.

The 1GHz Power4 I read about was blowing the doors off of everything in the consumer market by a factor of 5, so it is conceivable that a Power4 Lite would be twice as fast as the 3.2-3.5GHz Pentium 4 that will be shipping then, even if it is a 1GHz chip.

However, some people will still be concerned about the GHz disparity, regardless of processor power. These people deserve Windows! For a start the POWER4 is a good chip but on single threaded programs it isn't computationally 5 times faster than current top desktop chips. It does have some aspects of its architecture that make it very attractive though.

Secondly I'd advise you not to get your hopes quite so high. I would expect Apple to handle the introduction of their next generation system in a similar way to the PIV introduction. That is to basically introduce it low even though you know it can scale and then scale it. I'd also expect a couple other things but you'll have to wait for those.

[G4scott]

Originally posted by AmigaMac
Bottomline is that PowerPC isn't going anywhere folks!

I hope you mean anywhere as in "Apple isn't going to leave the PPC architecture behind", because the way you say it, it sounds like the PowerPC is stuck where it is now, and can't be improved :eek:

I hope that's not what you meant ;)

[unclepain]

I'm one of those sorry saps who has to upgrade right now. Actually, I'm very pleased to be able to go from my original G4/400 with the Yikes MB to a new Dual 867 upgraded with a Superdrive (which is currently en route), but this IBM proc with improved MB interconnects could be an upgrade several orders of magnitude the likes of which we haven't seen in a loooooooong time. I know that there will ALWAYS be something better down the pipeline, but if this rumor is true, it will feel like I'm going from a KIA to a Lexus, when I could go from a KIA to a Ferrari for the same money if I had just waited a bit longer.

[eric_n_dfw]

Originally posted by Telomar
For a start the POWER4 is a good chip but on single threaded programs it isn't computationally 5 times faster than current top desktop chips. It does have some aspects of its architecture that make it very attractive though.
True, but most of the heavy hitter app's like FCP, Photoshop, Maya etc... should be saturated with multi-threaded code.

[Macmaniac]

This may sound like a dumb question but, will Apple have to develop a 64 bit version of OS X? Also will these new porcessors be compatible with older apps: Starcaft, Photoshop 6, and so on. Or will they just no be able to take advantage of 64 bit processing? It would be really cool if Apple is the first with a 64 bit processor for the masses, that would definantly win us some switchers. 2ghz would be nice:)

[Snowy_River]

Originally posted by Macmaniac
This may sound like a dumb question but, will Apple have to develop a 64 bit version of OS X? Also will these new porcessors be compatible with older apps: Starcaft, Photoshop 6, and so on. Or will they just no be able to take advantage of 64 bit processing? It would be really cool if Apple is the first with a 64 bit processor for the masses, that would definantly win us some switchers. 2ghz would be nice:)

As I understand it (and that's really not saying very much ;) ), OS X only needs to be recompiled to be fully 64bit native, no rewriting necessary (i.e. it's already written to be 64bit native).

As for older apps, they would simply not use the full capacity of the chip, using only the first 32 bits. Thus, they would effectively run the chip at only half capacity.

(Please don't hesitate to correct me if I've misspoken. It's been too long since I've read up on this aspect of architecture...)

[reyesmac]

If the G5 is to be 1ghz and it will only give you speed on certain kinds of apps, what will be left to run at the normal speed of a 1ghz chip? If you can play 5 high quality videos at the same time under transparent windows without frameskips but all your old programs will go at the same speed they where before unless they are updated, has Apple done anything to help us? Would this chip be like the G4, you had to buy new versions of your programs that used the velocity engine to see the maximum speed of the chip? If not, then we are in for a treat, but if it is only 1ghz, they are going to have to convince Mac users to buy, not just windows users. Not only that, but everyone who has a web page about macs have to agree that this chip does what it does for people to buy, just look at the powermacs we have now, if you took apples word for it, they are as fast as it gets when they are only marginally faster than what they replaced.

As for the person that bought a soon to be slow computer all I could say is:
Don't worry too much about it, Apple rarely makes the low end Powermac even chip-wise with the top two. Look at it now, it is like 45% slower than the dual gig because of the motherboard. The first generation of G5 will probably have a G4 1gig on a 166mhz mobo and the top two will go up in price and be 1 processor at the middle on the new mobo and dual 1 gig at the top with new mobo and faster graphics card. So for what you got and what you payed for, you did get a good deal. That is, if Apple does not keep its all dual powermac lineup strategy.

By the way, anyone here with a dual machine, how many 640x480 movie trailers can you play at the same time with sound and see no skipping? Does the system feel sluggish?

[nuckinfutz]

Originally posted by barkmonster


46.6% is the largest speed bump apple have ever made on a whole lineup of powermacs. That was when they went from 500Mhz to 733Mhz. It wasn't too impressive in a lot of benchmarks because it was nowhere near 46% faster.

57.2% is the largest speed bump on 1 model when they went from 466Mhz digital audio to 733Mhz Quicksilver on the entry level model.

since the G4 jumped from 500Mhz to 733Mhz the fastest chip in the following range has always been not only predicted with pinpoint accuracy by overclocking attempts on the previous range but speed increases have been in 15 - 25% range :

500Mhz - 733Mhz - 46.6%
733Mhz - 867Mhz - 18.3%
867Mhz - 1Ghz - 15.4%
1Ghz - 1.25Ghz - 25%

If clockspeed actually equates to actual performance increases when you're comparing the same CPU family, the next power mac should be a dual model at 1.5 - 1.6Ghz. That's a 25 - 30% speed increase and I can't see someone buying a dual 1.25Ghz G4 this year only to buy a dual 1.6Ghz next year unless there's some serious change in the architechture that makes the actual perceived speed increase more like 50 - 60%.

Anyone who actively follows processor developments new the jump from 500Mhz to 700Mhz G4+ wasn't going to yield a huge jump in speed.

The 500mhz G4 was a 4 stage pipleline Proc.

The 700mhz G4+ was a 7 stage.

Increase the piplelines in a proc and you ehance it's ability to be clocked at a higher rate but you lose efficiency generally. It's a design tradeoff that has propelled the P4 to great mythical heights.

The Power4 Lite proc is slated to have 10-15 Piplines I believe and will be no larger than a 130 Nano process...most likey 90 nano. So it's should clock at 2ghz relatively easy IMO.

It's frustrating to read so many Mac users who are disgruntled because Apple's computers aren't clocked as high as AMD/Intel when many do not realize the design tradeoffs involved.



If the G5 is to be 1ghz

Trust me the G5 is not going to be 1ghz. The current Power4 is a huge proc at 180 nano with multiple cores huge caching and much more. It's way too much for a Desktop.

The lite version will be made to clock and support SIMD.


Here's what you all can most likely expect next year in the Powermacs

Next update:

130 Nano G4+ Should go up to Dual 1.5-1.6 Q1 2003

130 or 90 Nano IBM G5. Apple could hit 1.8 2.0 and perhaps 2.2ghz. The ApplePI proc to memory controller is essential for taking advantage of the 6.4GBps memory bandwidth of this proc.

I think that this proc has the potential to be a large leap forward in speed but I'll have to temper my enthusiasm until we know more about the proc from IBM on Oct 15.

[TechLarry]

Originally posted by arn
MacEdition posts (http://www.macedition.com/nmr/nmr_20020914.php) vague details and a report that IBM's 64-bit Power 4 Chip will make its way to the next generation of PowerMacs with a new Apple Processor Interconect bus.

While the new systems are reported to be in prototype form presently, are not expected for full production until later in 2003.

I thought the whole "Power-4 on Mac" think was debunked in it's entirety a few weeks ago ?

TL

[chubakka]

one of those I heard from a guy who heard from an IBM engineer that... blah blah blah...

Only thing we do know as fact that IBM has said it is making 64-Bit PPC chips derived from Power4... for low end servers and desktops... with very Alitvec like features. And they plan to introduce them at around 2GHz.

[scem0]

I guess I am going to have to wait.................. and wait.................... and wait................... and wait................... and then it will finally come out............ and then I will wait........ and wait........... and wait........... and finally the price will drop 300-400 $'s......... and then I will buy one.................. and then I can die a happy man.............. but only after i have used it for at least a year.........

:D :D :D



if this rumor is even true at all :confused: :mad: :( :o :)

[tjwett]

Originally posted by DaveGee


Rule is with TGB (as it was with Mac The Knife) you usually gotta look PAST the 1st paragraph to get on to the TRUE meat of the story. Been that way for a long long time. TGB doesn't talk trash... You do see him talking about iWalks and QUAD Processor PPC based monsters and the day you seem him talking about em... well then that means we might really see em come true.

Biggest problem with TGB is he usually waits for REAL confirmation about the stuff he is working on and as such he doesn't report on nearly as much stuff as say SpyMac or MOSR does... (isn't about time MOSR dusts off that QUAD CPU rumor again?) :D

Dave

oh, i get it now. thanks for clearing that up. i haven't read this guys stuff before. at first glance it looks like something from crazyapplerumors.com.

[nuckinfutz]

http://www.plasma-online.de/index.html?content=http%3A//www.plasma-online.de/english/identify/picture/ibm_cpu.html

Power4-B (0.13 µm, Cu, 2GHz, 2003)

Look at the very last entry on the whole page.

http://www.midrangeserver.com/tfh/tfh030402-story02.html



Last paragraph states that Power4-II servers should be out sometime in October of 2002...

As for the timing of AIX partition support, I think it is much more likely that PASE in enhanced in OS/400 V5R2 and AIX partitions will become available around when OS/400 V5R3 and AIX 5L Release 5.3 ship on Power4-II servers. AIX 5L Release 5.2 and the Power4-II servers are due around October 2002, and AIX 5L 5.3 is expected sometime in the second half of 2003, probably on Power5 servers.

[sturm375]

Originally posted by shadowfax0
But will it still be the G4? Or is it also time for a name change....Also, does this mean Apple will be the first one to offer a 64-bit processor for the masses?

Nope.

1Q03, AMD will have their version of the x86-64 out for the masses. Rumor has it that it will start at the equivent of a P4 3 Ghz. Later in '03 AMD will be shipping 90 nm technology CPUs.

http://www.amd.com/us-en/Processors/ProductInformation/0,,30_118_608,00.html

[P-Worm]

Originally posted by reyesmac
By the way, anyone here with a dual machine, how many 640x480 movie trailers can you play at the same time with sound and see no skipping? Does the system feel sluggish?

I can't tell you how many 640x480 movies I can play on my older model dual 1ghz, but I can tell you that I have played 8 movies at one time, while ripping a cd, and saw vertually no hit in overall performace. I nearly died.

So if your question is wheather or not to get a dual proc. over a single, then my responce is a resounding yes!

P-Worm

[Mr T]

that believe Apple / IBM can put a g5 machine in the next 4 months. This is a long ways off . At least 8 months for Apple R&D develpment.

Get used to the dualies - you will be using them for a while .......

[G4scott]

Originally posted by unclepain
I'm one of those sorry saps who has to upgrade right now. Actually, I'm very pleased to be able to go from my original G4/400 with the Yikes MB to a new Dual 867 upgraded with a Superdrive (which is currently en route), but this IBM proc with improved MB interconnects could be an upgrade several orders of magnitude the likes of which we haven't seen in a loooooooong time. I know that there will ALWAYS be something better down the pipeline, but if this rumor is true, it will feel like I'm going from a KIA to a Lexus, when I could go from a KIA to a Ferrari for the same money if I had just waited a bit longer.

Sounds like me ;)

I might get my parents to trade in my 400mhz yikes G4 for $500 at CompUSA, and we might get the dual 867. I really see no need for the dual 1ghz for what my family uses it for. Maybe the super drive, but that's about it...

[Nipsy]

Originally posted by Mr T
that believe Apple / IBM can put a g5 machine in the next 4 months. This is a long ways off . At least 8 months for Apple R&D develpment.

Get used to the dualies - you will be using them for a while .......

What makes you think that this R & D phase will be publicly announced?

Could it have begun 8, 12, or 16 moths ago?

Could IBM possibly announce on the 15th that this chip has been designed, built, and tested, and wil now enter production?

Hmmmm...

[dongmin]

Originally posted by Mr T
that believe Apple / IBM can put a g5 machine in the next 4 months. This is a long ways off . At least 8 months for Apple R&D develpment.

Get used to the dualies - you will be using them for a while .......

if you had bothered to read the MacEdition report, it said that don't expect these IBM-powered Macs to be introduced this January. It also said that they had OS X running on the new CPU back in Nov. 2001. It's naive to think that any of us has a clue as to the development schedule of Apple and IBM. They could've been working on this for years, for all we know.

But if they're introducing the chip this October, I would have to think it'll take them at least another 6-9 months before they're producing in quantities. But again, I don't pretend to know Apple's development cycle anymore than the next guy.

[Dunepilot]

Originally posted by G4scott


Sounds like me ;)

I might get my parents to trade in my 400mhz yikes G4 for $500 at CompUSA, and we might get the dual 867. I really see no need for the dual 1ghz for what my family uses it for. Maybe the super drive, but that's about it...

I'll also be moving from a Mac of similar vintage, in my case a PBG3 333 Lombard. I've waited for almost 3 years to the day before getting a new Mac, and I'm looking forward to getting a dual G4 machine even if faster stuff comes out in the next 12 months. As I see it, OS X is finally up to an acceptable speed that it'll run well on this Mac, and I also want to be able to run OS 9-based music stuff if I want without Apple telling me I can't as they will in January.

At least these G4s are built for the OS they'll be running. I pity the poor fools who placed down money in the days of the dual 450 and dual 500 machines, before OS X was ready, and at a time when very very few apps could take advantage of the second processor. The time of the duals is ripe.:)

[shadowfax0]

I'm not a poor sap, right...? :(


:D

[danbirchall]

If you're one of the people saying "what is WITH that guy's writing style?" (as opposed to one of the people like me, who get the joke, and find it hilarious!) get over to Google and search for Nigerian 419 Spam.

All will be made clear. :)

[fluke]

[QUOTE]Originally posted by Dunepilot

I also want to be able to run OS 9-based music stuff if I want without Apple telling me I can't as they will in January.

-----------------

just a note regarding music software: pro tools will not run in classic under osX. but pro tools 6 should be shipping this month.

fluke

[primalman]

With the knowledge that the new 970/Power4 Lite has the AltiVec unit on it, it seems obvious to me that IBM has either:

A - Licensed the right to make AltiVec on their chips;
B - Reverse-engineered AV to work and be legal, or;
C - Apple has paid for IBM to use AV.

In any respects, does it seem likely at all to anyone else that IBM could use the license to make G4 chips for Apple, allowing them to split the supply with Motorola??

I mean, G3s for the iBook and and the G4+ and 970 for the PowerMac and [x]Mac lines from IBM, 7455 and 7457 chips for the eMac, iMac and maybe the some PowerBooks from Motorola. Basicly all G3 and DDR compliant G4s and such from IBM and everything else from M.

i dunno, just thinking out loud.

[ffakr]

Originally posted by reyesmac
If Apple does have some chip that will replace the G4 next year, who in here thinks that Apple will have them go more than 30% faster than the previous generation? What has been the biggest speed bump that Apple has had compared to what it replaced? Will they use dual processors, or will they just put one that is a little faster than two? Hope we dont get the usual top two getting upgraded and the bottom getting Yikesed.

What will the low end look like? Would the low end going from dual 867 to a 1gig be enough? Has Apple ever given us a speedbump that has made the low end faster than the high end that it replaces?
This is an entirely different family of processors. You can't base projected clock speed off of existing processors.
This will be a varient of the IBM Power4 processor, the PPC 970. It is a completely new processor (aside from the power4 roots).
IBM has already announced that they are targeting 1.4-1.8 GHz on a .13 micron process.
This is where you should expect initial shipments to fall.
It is a new processor though, so you have to remember that...
* it will be MUCH faster than a G4 at a given clock speed... IBM posted early SPEC scores for a 1.8GHz part that come close to the 3GHz P4 and the upcomming 1.8 GHz AMD clawhammer. PPC970 seems to run SPEC about twice as fast as the G4.
* IBM stated that the current SPECs are likely to go up... they are currently testing pre-production silicon
* IBM has recently stated that the original projections were conservative and they have hinted that they might debut at around 2GHz.
* IBM will produce these .13micron chips at a new plant that is supposed to ramp up a .09 micron process in the first quarter of this year... we may see the 970 come out as a cooler/faster .09 micron part later this year (rather than the .13 micron part they are testing now).

Anyway... I doubt that the entire Apple line would move to a new 64bit PPC at the end of this year. The 'Pro' line would go first... That would be xServe, towers, and the Powerbooks (when the chips were running cool enough)... the consumer machines probably won't follow until sometime in 2004... depending on cost, heat, yield... of the new chip.
The good news is, the cream of the G4 crop will be available for consumer machines when the 970 moves into the pro line... MOTO is _supposed_ to hit 1.8GHz with the G4 later this year...

[MacKid]

Originally posted by bousozoku
Processor Interconnect bus? Does this sound like the same business that's already running to allow the processor to interface with memory in interesting ways?

I'll wait to see but I would be skeptical as to whether these systems will ever come to light and, if so, whether they will have good or extraordinary performance.

Isn't there already something called the API (Apple Processor Interconnect) interface?:confused: Sounds familiar. . .

[MacKid]

Isn't there already something called the API (Apple Processor Interconnect) interface?:confused: Sounds familiar. . .

Sorry for the double post

[ffakr]

Originally posted by sturm375


Nope.

1Q03, AMD will have their version of the x86-64 out for the masses. Rumor has it that it will start at the equivent of a P4 3 Ghz. Later in '03 AMD will be shipping 90 nm technology CPUs.

http://www.amd.com/us-en/Processors/ProductInformation/0,,30_118_608,00.html


AMD has backed off from the 1st quarter talk.. they are officially stating first half. The release date is anything but sure. The last time anyone outside of an iron-clad NDA has seen a demo clawhammer, AMD was only running the clawhammer at 1.2GHz...
This was a month or so ago... and AMD is targeting 1.8 GHz for release.

It will be a good chip when it comes out, but I don't think anyone outside of AMD knows whether the public will see Clawhammer in March, or in June.

Personally, I think that Apple's R&D is moving in parallel with IBMs work. I hope that we might see a shipping PPC 970 product in the third quarter. The xServe is the likely candidate since the volumes are low, allowing Apple to get it shipped without a stockpile of processors.
I don't expect that we'll see 970 towers any time before Q4, 2003.

[MacKid]

Originally posted by Mr T
that believe Apple / IBM can put a g5 machine in the next 4 months. This is a long ways off . At least 8 months for Apple R&D develpment.

Get used to the dualies - you will be using them for a while .......

The rumor has always been "late 2003".

[locovaca]

Originally posted by ffakr


AMD has backed off from the 1st quarter talk.. they are officially stating first half. The release date is anything but sure. The last time anyone outside of an iron-clad NDA has seen a demo clawhammer, AMD was only running the clawhammer at 1.2GHz...
This was a month or so ago... and AMD is targeting 1.8 GHz for release.

It will be a good chip when it comes out, but I don't think anyone outside of AMD knows whether the public will see Clawhammer in March, or in June.

Personally, I think that Apple's R&D is moving in parallel with IBMs work. I hope that we might see a shipping PPC 970 product in the third quarter. The xServe is the likely candidate since the volumes are low, allowing Apple to get it shipped without a stockpile of processors.
I don't expect that we'll see 970 towers any time before Q4, 2003.

See here (http://www.2cpu.com/) Go to the news entry for January 7

[MacKid]

Originally posted by ffakr
Personally, I think that Apple's R&D is moving in parallel with IBMs work. I hope that we might see a shipping PPC 970 product in the third quarter. The xServe is the likely candidate since the volumes are low, allowing Apple to get it shipped without a stockpile of processors.
I don't expect that we'll see 970 towers any time before Q4, 2003.

I agree, and I think that the Xserves are pretty much due for an update, I mean things have been really quiet in that area lately. Something has got to change, since the PowerMacs now have faster processors than the servers. And I also think that people need to stop expecting Power 4's or 970's by July.

[locovaca]

Originally posted by MacKid


I agree, and I think that the Xserves are pretty much due for an update, I mean things have been really quiet in that area lately. Something has got to change, since the PowerMacs now have faster processors than the servers. And I also think that people need to stop expecting Power 4's or 970's by July.

I don't think it's necessary to bump the Xserves until the 970 comes into the arena. Servers are about stability before speed- that's why you see so many of them running older processors, not the latest and greatest. Heck, you don't see very many P4 processor servers, and it took a few years to get the Athlon into a server! Real high density servers (read: rack mounts) need to be stable, part of which comes the heat dissipation. If Apple wants to remain serious in the market I'd rather they took their time.

[josepht]

I've got an old G3 266 PowerMac. Needless to say, I need to upgrade. I'm planning to wait for the PPC 970s before I spend my family's money on a new PowerMac. The thing is: we're not rich. How much do any of you think the low-end PowerMac G5s will be?

Another thing: I'm probably not going to get the first G5s out of the shoot. Wouldn't it be wise to wait for the first update so some of the kinks could be worked out? I've waited this long. I can wait a bit longer.

Joseph

[idkew]

Originally posted by josepht
I've got an old G3 266 PowerMac. Needless to say, I need to upgrade. I'm planning to wait for the PPC 970s before I spend my family's money on a new PowerMac. The thing is: we're not rich. How much do any of you think the low-end PowerMac G5s will be?

Another thing: I'm probably not going to get the first G5s out of the shoot. Wouldn't it be wise to wait for the first update so some of the kinks could be worked out? I've waited this long. I can wait a bit longer.

Joseph

it will roughly cost the same as a low end powermac costs now. maybe cheaper. each top and bottom of the line costs about the same each generation...actually a little cheaper.

[ffakr]

Originally posted by locovaca


See here (http://www.2cpu.com/) Go to the news entry for January 7

Yes, it is a posting about a possible release of a dual opteron server. The subject ends with Feburary?
Yes, it could happen by Feburary but I want to point out a few things....

[list=1]
AMD promised the *hammer in the 4th Quarter of 2002, then the 1st Quarter of 2003, now the 1st half of 2003. Products slip and I'm not ripping on them, just saying that I wouldn't be betting on a product ship date in 2 weeks.
AFAIK, MS doesn't have a 64bit x86-64 Windows release ready to roll. If you want one, You'll be running a 32bit OS, or an essentially experimental Linux OS on it
A server company trying to make first to market is a far cry from the average consumer getting access to this chip. That might not happen for quite some time after the first production processor rolls out of Dresden (or whatever fab is making these)
The price on the dual CPU server is over $9000! Doh! That is for a 1U server that essentially has the performance of a dual, high end P4 rig. Sure the Opteron box will have much larger addresses, but a 1U will have real world constraints on how much memory/storage you can pop in.
[/list=1]

I might buy a Hammer myself next year... but I think it has a very long road to travel before it is a real option to the consumer. I think the hammer will directly compete with the 970 on the high end desktop because hammer will take a while to come down in price... especially compared to Athlon and even P4. Add in the uncertainty of the OS side of things and I really don't see hammer being a contender till later (2nd half) of this year.
Apple, on the other hand, will likely have a 970 box with a 64bit OS X ready to roll as soon as the processor is available in volume. I think Apple is going to give 970 an advantage by supporting it more whole heartedly than MS is supporting the x86-64 platform.

Unfortunately, it also appears that AMD is focusing its attention on the hammer line and not on the Athlon. The Athlon was stuck at around 2GHz for a long time due to errata, and it appears to be stuck again after the launch of a few new cores. At this rate, Intel will have 3.4GHz hyperthreading P4s out before we see a new Athlon :(

[ffakr]

Originally posted by locovaca
I don't think it's necessary to bump the Xserves until the 970 comes into the arena. Servers are about stability before speed- that's why you see so many of them running older processors, not the latest and greatest. Heck, you don't see very many P4 processor servers, and it took a few years to get the Athlon into a server! Real high density servers (read: rack mounts) need to be stable, part of which comes the heat dissipation. If Apple wants to remain serious in the market I'd rather they took their time.

Servers are about stability, redundancy, AND power.
I know people who are evaluating the xServe for Cluster use. Apple should definately update the xServe asap.

There are also more reasons than just raw Integer performance to put newer G4s in the xServe asap... the next processor [7457] is supposed to support a DDR processor bus. Servers are also about bandwidth. A dual CPU server, especailly one running at a Gig or higher needs a DDR bus (at least!).

Also, the low volume of the xServe would make it an obvious candidate for a new processor as volume is usually constrained at launch.

And finally, the xServe is targeted at education, research, AND the enterprise... they need to get the attention of the IT manager. :)

...as for the P4, the P4 wasn't a good candidate for the 1U rack server due to it's thermal characteristics. Why put in a 60+ watt P4 when you could put in two Tualitan P3s that were faster and shed less heat (combined) than the P4. The P4 is finally muscleing out the P3 in the 1U server market, because the heat/power ratio is getting better and the single P4s are faster than the Dual P3s... not to mention the fact that the hugely increased bandwidth of the P4 makes it more attractive in the server market.

..jmho... Ffakr.

[sturm375]

Originally posted by ffakr


Yes, it is a posting about a possible release of a dual opteron server. The subject ends with Feburary?
Yes, it could happen by Feburary but I want to point out a few things....

[list=1]
AMD promised the *hammer in the 4th Quarter of 2002, then the 1st Quarter of 2003, now the 1st half of 2003. Products slip and I'm not ripping on them, just saying that I wouldn't be betting on a product ship date in 2 weeks.
AFAIK, MS doesn't have a 64bit x86-64 Windows release ready to roll. If you want one, You'll be running a 32bit OS, or an essentially experimental Linux OS on it
A server company trying to make first to market is a far cry from the average consumer getting access to this chip. That might not happen for quite some time after the first production processor rolls out of Dresden (or whatever fab is making these)
The price on the dual CPU server is over $9000! Doh! That is for a 1U server that essentially has the performance of a dual, high end P4 rig. Sure the Opteron box will have much larger addresses, but a 1U will have real world constraints on how much memory/storage you can pop in.
[/list=1]

I might buy a Hammer myself next year... but I think it has a very long road to travel before it is a real option to the consumer.
Unfortunately, it appears that AMD is focusing its attention on the hammer line and not on the Athlon. The Athlon was stuck at around 2GHz for a long time due to errata, and it appears to be stuck again after the launch of a few new cores. At this rate, Intel will have 3.4GHz hyperthreading P4s out before we see a new Athlon :(

I think you might be wrong about the experimental linux OS on the "Hammer" processor. I have read in several articals that the "Hammer" would have been next to impossible to produce (at least the compiler) were it not for the fact tha AMD courted the open source/linux community. SuSE (http://www.suse.com) already has a full blown software package for the x86-64, the won't sell it because there is no hardware to put it on yet. At least not on the open market, I know that there are developmental rigs out there for this stuff to work on. Also in a recent convention in Los Vegas, AMD showed off their "Hammer" complete with a Windows version with 64-bit support.

I agree that consumers will not see any "Hammers" until late this year, but large OEM will see them much sooner than the rest of us. Also note that at the moment the current AMD Athlon XP 2800s out-perform the P4 3.06Ghz w/ hyperthreading in most respects. Only specific tasks that are coded to hyperthreading so minimal headway over the Athlons.

I will try to find links to the above references when I have more time.

[Kid Red]

Arn you know that IBM will not be making G4s. Tsk Tsk on the title.

[locovaca]

Originally posted by ffakr

[list=1]

AFAIK, MS doesn't have a 64bit x86-64 Windows release ready to roll. If you want one, You'll be running a 32bit OS, or an essentially experimental Linux OS on it

The price on the dual CPU server is over $9000! Doh! That is for a 1U server that essentially has the performance of a dual, high end P4 rig. Sure the Opteron box will have much larger addresses, but a 1U will have real world constraints on how much memory/storage you can pop in.
[/list=1]

I might buy a Hammer myself next year... but I think it has a very long road to travel before it is a real option to the consumer. I think the hammer will directly compete with the 970 on the high end desktop because hammer will take a while to come down in price... especially compared to Athlon and even P4. Add in the uncertainty of the OS side of things and I really don't see hammer being a contender till later (2nd half) of this year.
Apple, on the other hand, will likely have a 970 box with a 64bit OS X ready to roll as soon as the processor is available in volume. I think Apple is going to give 970 an advantage by supporting it more whole heartedly than MS is supporting the x86-64 platform.



Well, I don't necessarily disagree with you, but a few thoughts:

1. People who will really benefit from the 64 bit power (science companies, etc.) most likely use some kind of linux/unix base, so not having windows at launch I don't think will be that big of a problem. When the consumer chip comes out then the Windows will be a much bigger factor.

2. Linux in itself is one big experiment in my eyes. In fact, all software is one big experiment- development, testing, releasing, fixing, lather, rinse, repeat. There's no such thing as a finished software product. There will always be one more thing to add, one more thing to optimize, one more bug to fix.

3. When the Athlon 1 GHz was being presold it too had ridiculous prices like $6000 for a processor. That's not to say that this system is going to be the same, but I do not think that the $9000 pricetag will stick long. Opteron systems are supposed to compete with Xeon systems in terms of price.

4. Who's to say that the 970 is going to be in the consumer machines at launch? For all we know they may be only stuck in the Xserves. The 970 in powermacs are just speculation at this point- we know that Apple will bring out a machine that has the 970, but nobody knows which yet.

5. Once again, I reiterate that the people who will want 64 bit the most at it's launch- laboratories, renderers, etc. use linux as a base OS for the mean number crunching. Windows won't mean much to this market at this point (although they will eventually switch everything over to 64 bit as the Windows/X machines come out). Plus, do you really think that Microsoft will sit on their hands while Linux takes over the X86 server market?

[ffakr]

Originally posted by locovaca
2. Linux in itself is one big experiment in my eyes. In fact, all software is one big experiment- development, testing, releasing, fixing, lather, rinse, repeat. There's no such thing as a finished software product. There will always be one more thing to add, one more thing to optimize, one more bug to fix.
This is pretty much my take on it... There may be 'release quality' linux when hammer ships, but I'm sure there will be bugs to be worked out.

3. When the Athlon 1 GHz was being presold it too had ridiculous prices like $6000 for a processor.
This I don't recall. Why would anyone pay that much for a 1GHz Athlon when it was just an incremental upgrade to the existing processor line? The Athlon of that day was pretty friendly to overclocking too.
[/b]4. Who's to say that the 970 is going to be in the consumer machines at launch? For all we know they may be only stuck in the Xserves. [/B]
I actually think that the xServe will be the first to sport the 970 for a number of reasons. I think the desktop will follow very soon though... as soon as IBM production volume comes up enough to meet demand. The current desktop isn't significantly different from the xServer at a low level after all.

[josepht]

Originally posted by idkew


it will roughly cost the same as a low end powermac costs now. maybe cheaper. each top and bottom of the line costs about the same each generation...actually a little cheaper.

That's what I thought; but, I wanted someone else's opinion.

Thanks! :D

Joseph

[locovaca]

This I don't recall. Why would anyone pay that much for a 1GHz Athlon when it was just an incremental upgrade to the existing processor line? The Athlon of that day was pretty friendly to overclocking too.

This was before it was actually released, around February of 2000. It was some company over in Europe that was making these available for Pre-release orders. I'm not talking about the socket processors, this is the original, "First to 1 ghz" slot a processor. I don't remember if people actually bought them, but it was done. The slot A's weren't AS OC friendly because the cache wasn't integrated, and was only rated up to certain speeds, and unless you bought a nice "golden finger" device you couldn't change the cache divider unless you had some nice soldering skills.

[dongmin]

Originally posted by josepht


That's what I thought; but, I wanted someone else's opinion.

Thanks! :D

Joseph

Actually, it's possible that Apple will keep the old processor in the low-end Tower; they sometimes split processors even within one line of macs.

So, assuming that 7457s are coming in Feb and 970s are introduced in August, you might see the low-end tower retain the 7457 and the higher two models running the 970s.

Moral: Take all these rumors and prognostications with a grain of salt.

[MacKid]

Originally posted by josepht
I've got an old G3 266 PowerMac. Needless to say, I need to upgrade. I'm planning to wait for the PPC 970s before I spend my family's money on a new PowerMac. The thing is: we're not rich. How much do any of you think the low-end PowerMac G5s will be?

Another thing: I'm probably not going to get the first G5s out of the shoot. Wouldn't it be wise to wait for the first update so some of the kinks could be worked out? I've waited this long. I can wait a bit longer.

Joseph

I think the PowerMacs will cost in the same range as they do now, because whenever Apple upgrades, they usually keep the same general pricing system, and it would be up to you to decide whether you want to wait or not, because after a new computer comes out, you can expect a 3-5 month wait before they are updated with higher speeds.

[Centris 650]

I've got a question concerning the 970's that can only be answered in pure speculation but here it goes....

When I bought my first mac (a Centris 650) it had the 68040 processor in it. A few years after that the G3 processor came out. I waited to upgrade my C650 but a G3 upgrade never came out. I was told that due to differences in the motherboard my 68040 machine would never get a G3 upgrade.

MY QUESTION: Will this be the case for the G4s? If I buy a new machine and the 970s are released am I stuck with an unupgradeable computer? If so then I might just bite the bullet and buy an iMac and upgrade to a 970 Mac next time.

Thoughts?

[ffakr]

Originally posted by Centris 650
MY QUESTION: Will this be the case for the G4s? If I buy a new machine and the 970s are released am I stuck with an unupgradeable computer? If so then I might just bite the bullet and buy an iMac and upgrade to a 970 Mac next time.
Thoughts?
Since the 970 will use a processor-chipset bus that is unlike anything available on a current Mac, I'd bet that it'll be just about impossible to make a 970 upgrade for existing machines.
Sure, you can do just about anything with enough resources... but I think an 'adapter' would cost about as much as a motherboard chipset to produce.

[cubist]

Right, and you're going to 64 bit etc. ... I strongly doubt any upgrades will be possible.

However, if you buy, for example, the dual G4 now, there may be upgrade boards with 7457's which could get you up to 1.4GHz fairly soon.

I myself have a 450MHz Cube and there are upgrade boards with dual 1GHz. They won't be as fast as a dual 1G powermac because of the Cube's 100MHz bus, though.

Upgrades that don't change the whole MB are usually not cost-effective, because it won't be as fast as a machine designed around the newer processor. When I say 'cost effective', I mean that they don't add their price to the resale value of the machine. The Cube is a special case because there aren't going to be any new Cubes... or are there?:eek:

[GeneR]

Originally posted by nuckinfutz
http://www.plasma-online.de/index.html?content=http%3A//www.plasma-online.de/english/identify/picture/ibm_cpu.html

Power4-B (0.13 µm, Cu, 2GHz, 2003)

Look at the very last entry on the whole page.

http://www.midrangeserver.com/tfh/tfh030402-story02.html



Last paragraph states that Power4-II servers should be out sometime in October of 2002...

As for the timing of AIX partition support, I think it is much more likely that PASE in enhanced in OS/400 V5R2 and AIX partitions will become available around when OS/400 V5R3 and AIX 5L Release 5.3 ship on Power4-II servers. AIX 5L Release 5.2 and the Power4-II servers are due around October 2002, and AIX 5L 5.3 is expected sometime in the second half of 2003, probably on Power5 servers.

I'm sorry. Would you mind explaining this to me (the layman/lame man) I don't really understand. It's all greek to me. How much better is the Power4 than the G4? Sorry to bother. Thanks!

[barkmonster]

Anyone who actively follows processor developments new the jump from 500Mhz to 700Mhz G4+ wasn't going to yield a huge jump in speed.

The 500mhz G4 was a 4 stage pipleline Proc.

The 700mhz G4+ was a 7 stage...

...It's frustrating to read so many Mac users who are disgruntled because Apple's computers aren't clocked as high as AMD/Intel when many do not realize the design tradeoffs involved.

I agree there. I was just refering to the clockspeed increase, not the realworld performance increase. I know higher pipeline stages, a tiny L2 cache and a quarter clockspeed DDR L3 were done to increase the clockspeed at the expense of performance. I even remember reading benchmarks where the 533Mhz G4 was thrashing the 733Mhz model at quite a few tasks. I know we're getting a 512K L2 and a faster (200Mhz vs 167Mhz) bus speed on the 7457 G4s so that should improve things quite a bit.

I know it's no exactly the same comparison because of the pipeline going from 10 stages to 20 but when the original P4 had a 100Mhz x 4 bus speed and only 256K of L2 it was being thrashed by the P3 in the same way as the early 7450 G4s were thrashed by 7410 G4s. Once they increased the bus speed to 133Mhz x 4 and upgraded the L2 to 512K, the Pentium 4 became a lot more efficient. Now it only needs to have a 40% clockspeed advantage to match the speed of the Athlon.

With the knowledge that the new 970/Power4 Lite has the AltiVec unit on it, it seems obvious to me that IBM has either:

A - Licensed the right to make AltiVec on their chips;
B - Reverse-engineered AV to work and be legal, or;
C - Apple has paid for IBM to use AV.

Actually it was :

D - Desktop '98, A SiMD extension to be included in a future PowerPC design was originally refered to as VMX. Motorola called their implementation Altivec. It was a join project and neither IBM or Motorola have any more or less right to use the technology.

Incidentally, Desktop '99 is the codename for a 64 bit version of the PowerPC that was expected late 2002 (PowerPC 970). Info about proposed features of both the G4 and the GPUL was in a macworld article from '97 or something with quotes from IBM sources.

NO DDR on the 7457!!! (seriously, I read it in a Motorola PDF last year)

The G4 chip that's going to use DDR is only a proposed design and doesn't have a release date or even a name/product number. The 7457 L is going to be out Q2 '03 and scale up to 1.3Ghz. That is the only new version of the G4 with a definate release date that was featured in the document "PPCSALESFACT.PDF" available online sometime last year. After the L version comes the P version that can scale to 2Ghz+ but that still uses the 200Mhz bus and doesn't use DDR.

[zulgand04]

Everyone is talking about how Apple should put the new IBM chips in the PowerMacs but has anyone thought about the Xserve. Now would't it make sense for Apple to put this awsome chip in there servers before the powermacs. I mean it kinda is common sense, their not gonna put a IBM chip in a powermac and forget about there server hardware. So you never know Apple may put new chip in the Xserve before july and then in september put them in the powermacs.

just my theory

oky im not the only one with the theory

[trebblekicked]

Originally posted by zulgand04
Everyone is talking about how Apple should put the new IBM chips in the PowerMacs but has anyone thought about the Xserve. Now would't it make sense for Apple to put this awsome chip in there servers before the powermacs. I mean it kinda is common sense, their not gonna put a IBM chip in a powermac and forget about there server hardware. So you never know Apple may put new chip in the Xserve before july and then in september put them in the powermacs.

just my theory

oky im not the only one with the theory

as if powermac sales weren't bad enough, updating the xserve with a new, (hopefully) more competitive processor while leaving the overmatched desktops with the motochip would be like the G4 towers themselves drinking cyanide koolaid. apple knows how awful the towers are selling. i don't think they'd leave the towers out to dry like that...i do think that once the 970 is ready, it'll be in both the tower and the server. **unless, for some ungodly reason, the 970 is not cost-ready for the desktop sytstems. that would be a bummer.

[MacCoaster]

Originally posted by locovaca
4. Who's to say that the 970 is going to be in the consumer machines at launch? For all we know they may be only stuck in the Xserves. The 970 in powermacs are just speculation at this point- we know that Apple will bring out a machine that has the 970, but nobody knows which yet.
Au contraire. We actually don't know if Apple will ever use the PowerPC 970. There is nothing solid.

[Dont Hurt Me]

Originally posted by zulgand04
Everyone is talking about how Apple should put the new IBM chips in the PowerMacs but has anyone thought about the Xserve. Now would't it make sense for Apple to put this awsome chip in there servers before the powermacs. I mean it kinda is common sense, their not gonna put a IBM chip in a powermac and forget about there server hardware. So you never know Apple may put new chip in the Xserve before july and then in september put them in the powermacs.

just my theory

oky im not the only one with the theory This is a very good theory and I think thats what they might do,bump the powermacs with the 7457 and put the 970 in the xserves before filtering down to the powermacs. I wonder what the cost will be on these new chips because that will determine what apple will do with them dont you think!

[nighthawk]

Because the sales on the Xserve are so low, they might just do that at the next revision. IBM has had engineering samples since Oct 2002, and it may be possible for Apple to announce the new Xserve with 970 chips in about a month or so... with shipping 4 to 6 weeks later.

If I remember right, only 8000 Xserves were sold last quarter. That was out of about 750,000 macs sold.

[ffakr]

Originally posted by barkmonster

NO DDR on the 7457!!! (seriously, I read it in a Motorola PDF last year)

The G4 chip that's going to use DDR is only a proposed design and doesn't have a release date or even a name/product number. The 7457 L is going to be out Q2 '03 and scale up to 1.3Ghz. That is the only new version of the G4 with a definate release date that was featured in the document "PPCSALESFACT.PDF" available online sometime last year. After the L version comes the P version that can scale to 2Ghz+ but that still uses the 200Mhz bus and doesn't use DDR.

Moto also released a road map late last year that indicated two 7457 models were in the pipe. The first didn't have bus specifications in the roadmap, but the 7457 variant due later this year specifically mentioned a DDR memory controller on the actual processor die.

It is fair to say that Motorola hasn't specifically stated that the 7457 will support a DDR bus, but if the 7457+ (or what ever they called it) has an on chip DDR controller, it's likely DDR support (off chip) will be introduced in the 7457 family.

I simply can't believe that a company... even Motorola.. would release a redesigned processor in 2003 without a DDR memory bus. This type of support is important to Motorola not only for Mac sales, but also for the market they REALLY care about.. high bandwidth embedded router applications.

[Dont Hurt Me]

Originally posted by ffakr


Moto also released a road map late last year that indicated two 7457 models were in the pipe. The first didn't have bus specifications in the roadmap, but the 7457 variant due later this year specifically mentioned a DDR memory controller on the actual processor die.

It is fair to say that Motorola hasn't specifically stated that the 7457 will support a DDR bus, but if the 7457+ (or what ever they called it) has an on chip DDR controller, it's likely DDR support (off chip) will be introduced in the 7457 family.

I simply can't believe that a company... even Motorola.. would release a redesigned processor in 2003 without a DDR memory bus. This type of support is important to Motorola not only for Mac sales, but also for the market they REALLY care about.. high bandwidth embedded router applications. When it comes to motorola and their cpu history doing almost nothing is what i would believe!

[barkmonster]

I've been searching the motorola site for a link to a PowerPC Roadmap, I've had no luck finding one but it did give this tip on one of search result pages that you might find funny :

Tip: Separate unrelated proper names with a comma.

Example: Bill Gates, Steve Jobs

Someone at Motorola obviously has a sense of humour :D

Does anyone know how to find the roadmap, it used to be easy to find from the PowerPC ISA part of their site ?

[primalman]

Originally posted by barkmonster
I've been searching the motorola site for a link to a PowerPC Roadmap, I've had no luck finding one but it did give this tip on one of search result pages that you might find funny :

Tip: Separate unrelated proper names with a comma.

Example: Bill Gates, Steve Jobs

Someone at Motorola obviously has a sense of humour :D

Does anyone know how to find the roadmap, it used to be easy to find from the PowerPC ISA part of their site ?

All I could find...

http://e-www.motorola.com/webapp/sps/site/overview.jsp?nodeId=03M943030450467M983989030230

primalman

[Dont Hurt Me]

i saw it to it also said last modified was nov of 2001. i think that road map is more of a road block!

[MacKid]

Originally posted by barkmonster
I've been searching the motorola site for a link to a PowerPC Roadmap, I've had no luck finding one but it did give this tip on one of search result pages that you might find funny :

Tip: Separate unrelated proper names with a comma.

Example: Bill Gates, Steve Jobs

Someone at Motorola obviously has a sense of humour :D

Does anyone know how to find the roadmap, it used to be easy to find from the PowerPC ISA part of their site ? It's there. . .I found it once. . .Then threw it in the trash. The plans for the G5 are very vague, I remember.

[MacKid]

http://e-www.motorola.com/collateral/PPCRMAP.pdf

That's it right there.;)

[barkmonster]

I know about the 2001 PDF, it's very old and doesn't even mention the 7457 chips. I'm on about the one mentioned here :

Moto also released a road map late last year that indicated two 7457 models were in the pipe. The first didn't have bus specifications in the roadmap, but the 7457 variant due later this year specifically mentioned a DDR memory controller on the actual processor die.

I don't know where the roadmap above is supposed to be but I'm guessing it's one of 3 things :

1. The same PDF I've got that was dated 26/9/02, misinterpreted.

2. Fiction.

3. Being pulled off the motorola site because it shouldn't have been there.

The PDF I'm refering to is on the motorola site, Here (http://e-www.motorola.com/brdata/PDFDB/docs/PPCSALESFACT.pdf), unless it's being pulled.

The original Accelerate Your Mac! article linking to the PDF with a brief write up is Here (http://www.xlr8yourmac.com/archives/nov02/110402.html).

Here's 2 screen shots from the PDF, sorry about the quality, it's shrunk down so I don't bug modem users too much. The first one is the roadmap, clearly showing the DDR version is no more than an idea they're kicking around and it's not even being tested yet. The one below that is a section of the roadmap from a few pages down where it shows dates, bus speeds, clockspeeds etc... It also includes some info on current 74xx based chips for comparison.

[MacKid]

Here is the latest roadmap.

[technocoy]

guys, i dont know what other machines will be running, but i will tell you something from apple this year will have the new IBM chip... i've said it before and i'm saying it again... I KNOW IT WILL BE IN AN APPLE MACHINE. Desktop, or server i couldn't tell you, but the deal is done, and the new IBM powerPC will be used by apple. This is what i know, now, when it comes to the AMD rumors, i don't know... maybe apple's got something up their sleeve that is going to blow us all away... steve did say that it was going to be an amazing year from apple!!

trust me guys,
technocoy:D

[MacKid]

Originally posted by technocoy
guys, i dont know what other machines will be running, but i will tell you something from apple this year will have the new IBM chip... i've said it before and i'm saying it again... I KNOW IT WILL BE IN AN APPLE MACHINE. Desktop, or server i couldn't tell you, but the deal is done, and the new IBM powerPC will be used by apple. This is what i know, now, when it comes to the AMD rumors, i don't know... maybe apple's got something up their sleeve that is going to blow us all away... steve did say that it was going to be an amazing year from apple!!

trust me guys,
technocoy:D

I think we've pretty much all decided/agreed on that. . .:rolleyes:

[ffakr]

Originally posted by barkmonster

1. The same PDF I've got that was dated 26/9/02, misinterpreted.

2. Fiction.
Fiction? I say I've seen it and you say fiction?
I'm hurt, nay insulted.

The Moto processor road map WAS linked in this Architosh article.... http://www.architosh.com/news/2002-10/2002c-1023-mcp7457-rm1.phtml
The linked roadmap indicated two 7457 processors, the 7457 and the 7457-RM. The roadmap indicated that the RM would have an on chip DDR memory controller, but there was no indication what type of bus the 7457 would support. Architosh believed this was an indication that DDR support would emerge with the 7457-RM, but I believe that if you go by that roadmap, it is equally as likely that DDR support will be rolled out with the 7457 family and the controller will move to the die with the RM version (late this year).

3. Being pulled off the motorola site because it shouldn't have been there.This, at least, seem to be fairly certain. ;-)

[ffakr]

Originally posted by ffakr

Yes, it is a posting about a possible release of a dual opteron server. The subject ends with Feburary?
Yes, it could happen by Feburary but I want to point out a few things....

Ahem...
Feburary?

Just saw this...
An article a few days ago regarding how AMD still hasn't announced a release date for the Athlon 64:
http://www.extremetech.com/article2/0%2C3973%2C851347%2C00.asp

and this...
A story on Tom's hardware... linking eWeek...
"Chip maker Advanced Micro Devices Inc. will release its long-awaited Opteron 64-bit processor at an event April 22 in New York City. "

Guess that maybe the question mark behind Feburary in that article should have been larger.