all this talk of missed opportunity with this ibm thing. hey jobs knew of the path and potential of the ppc and is not just looking down 1 year....but future development. don't you think that SJ decribed exactly what he wants and where he wants to go with chips --intel, amd etc??? this i think is more for internal propaganda and to tell the other users of the ppc that ibm has a future with this chip so other manufacturers don't go away like apple. it's a defensive catch up mode.
all this talk of missed opportunity with this ibm thing. hey jobs knew of the path and potential of the ppc and is not just looking down 1 year....but future development. don't you think that SJ decribed exactly what he wants and where he wants to go with chips --intel, amd etc??? this i think is more for internal propaganda and to tell the other users of the ppc that ibm has a future with this chip so other manufacturers don't go away like apple. it's a defensive catch up mode.
I think Apple knew about these chips when IBM first proposed the designs.
I also think that Apple was already in the throes of deciding whether they would stay or leave.
Either IBM had these specs on hand then, and Apple decided months ago, but waited until the Dev Conf to make a decision, or IBM told Apple that the chips specs would be better than they finally came up with, and THAT straw broke the camel's back.
I don't believe that IBM intro'ed these chips to show Apple up. They aren't good enough for that.
I completely agree with melgross. The update process on Powerbooks is continuous at Apple. The Powerbook team is working on a new machine the day the last one goes into production, (at least after most of the production bugs have been worked out.) Apple has parts from IBM and has known the specs for a long time. They may even have faster parts than the official speed announced. I know for a fact that the fastest Powerbook is not always released. It depends on part availability.
The fact that these chips have been announced by IBM prior to Apple producing a computer using them is probably a bad sign. It means they are available for the highest bidder?
As for G5 Powerbooks with the low power 970FX, there could be all sorts of issues that held them up besides the G5's heat. Bridge chips, low power mode programming, sleep issues, power management, battery life. (Key employess leaving?) Steve is pissed (which I think has already been mentioned.)
I know the process is continuous as I knew an Apple employee who worked on the first Apple Portable with the black and white screen and worked there through Duo's, Wallstreets and Pismo, leaving just before the Tibook was released.
The first Portable wasn't released for months due to a design problem with the black and white screen. Everybody else was releasing portables where the mouse tracks ghosted on the screen and Apple wouldn't do it, they had to have dynamic refresh.
The fact that these chips have been announced by IBM prior to Apple producing a computer using them is probably a bad sign. It means they are available for the highest bidder?
It probably signals that the PowerBook G5 is in fact a finished product. I bet this presentation and information was set for release on that date before WWDC, with IBM's knowledge of the PowerBook G5's timetable and that it would have been announced by then.
Which newly announced processor has the faster front side bus?
-The new IBM 16W 1.6GHz G5
or
-The new Freescale 15W 1.67GHz G4
Also:
With the introduction of the new G5 chips, what would be the possibility of Apple putting the processors in the eMac and/or Mac-Mini, whilst putting the new G4 processors in the iBooks and/or PowerBooks? Would it make sense for Apple to have all their desktop computers powered by the G5 line of processors (single core for the eMac, Mac-Mini, and iMac; dual core for the PowerMac), and have all their mobile laptop computers powered by the G4 line of processors?
Which newly announced processor has the faster front side bus?
-The new IBM 16W 1.6GHz G5
or
-The new Freescale 15W 1.67GHz G4
Also:
With the introduction of the new G5 chips, what would be the possibility of Apple putting the processors in the eMac and/or Mac-Mini, whilst putting the new G4 processors in the iBooks and/or PowerBooks? Would it make sense for Apple to have all their desktop computers powered by the G5 line of processors (single core for the eMac, Mac-Mini, and iMac; dual core for the PowerMac), and have all their mobile laptop computers powered by the G4 line of processors?
The problem right now is that nobody knows. The specs aren't out.
Apple has been putting the mobile chips into the Mini, and into the EMac as well. It's a power, heat, price thing.
Which newly announced processor has the faster front side bus?
-The new IBM 16W 1.6GHz G5
or
-The new Freescale 15W 1.67GHz G4
Also:
With the introduction of the new G5 chips, what would be the possibility of Apple putting the processors in the eMac and/or Mac-Mini, whilst putting the new G4 processors in the iBooks and/or PowerBooks? Would it make sense for Apple to have all their desktop computers powered by the G5 line of processors (single core for the eMac, Mac-Mini, and iMac; dual core for the PowerMac), and have all their mobile laptop computers powered by the G4 line of processors?
The 970 has the option of 1/2, 1/3 or 1/4 FSB vis-a-vis clock speed. So a 1.8 Ghz G5 has the option of an FSB of 900, 600 or 450Mhz. The 7448 is a G4 with a 200Mhz Maxbus FSB. Therefore, the G5 will have the faster FSB.
However, that comparison is misleading. Others have a better memory than me, but I was under the impression that, because of the length and width of the G5 processing pipeline, the G5 was more bandwidth hungry than the G4. Hence, the G5 needs a comparatively larger FSB to keep all the execution units pumping (and minimising wasted CPU resources and heat expenditure). The speed battle between the G4 and G5 is a line ball when clockspeed between the two is close and the FSB is throttled back from 1/2 to the lower ratios.
The 970 has the option of 1/2, 1/3 or 1/4 FSB vis-a-vis clock speed. So a 1.8 Ghz G5 has the option of an FSB of 900, 600 or 450Mhz. The 7448 is a G4 with a 200Mhz Maxbus FSB. Therefore, the G5 will have the faster FSB.
However, that comparison is misleading. Others have a better memory than me, but I was under the impression that, because of the length and width of the G5 processing pipeline, the G5 was more bandwidth hungry than the G4. Hence, the G5 needs a comparatively larger FSB to keep all the execution units pumping (and minimising wasted CPU resources and heat expenditure). The speed battle between the G4 and G5 is a line ball when clockspeed between the two is close and the FSB is throttled back from 1/2 to the lower ratios.
As to your other question, buggered if I know
Except, again, we don't know if thise figures are correct. This is a low power chip, it may not have the ability to use all of the bandwidth the full power chips use.
Apple probably knew at least a year or more ago the situation with the chips IBM recently announced.
They knew in mid-2003, and were designing hardware around them for 2004 releases. IBM is just 1 year, and counting, late with the "low power" 970fx. The 970mp isn't so much the issue. All of Apple's 2004 Mac plans depended on the 970fx and IBM's 90 nm fab performance.
IBM was expecting to ship a 2 GHz, 39 Watts max, 970fx CPU in Q1 04. But their fab never even came close to that, about 55 Watts max at 2 GHz, and not only that, they couldn't deliver on quantity either. They couldn't do that at any stage of that year. This was the chip that was planned to be at 3 GHz.
Remember, the 1st Xserve G5, a 1U server, was 3 to 4 months late (Jan announcement, Apr "shipment"). If IBM was able to produce this 2 GHz, 39 Watts max 970fx CPU rather than 55 Watts max, it meant a quieter Xserve. If they could ship it in quantity, Xserve G5s wouldn't have been that late.
If they could ship it in quantity, the iMac G5 would have been released in June 04, instead of September, and be quieter, possibly less buggy. The 2.5 GHz PowerMac G5 could have been released in February 04 instead of June 04. Lastly, a 1.6 to 1.8 GHz Powerbook G5 would have been possible in Fall 04.
They were counting on a "low power" 970fx for 2004. The current announcement of 1.2 to 1.6 GHz, <16 Watts typical, processors is really very little improvement from the current 970fx processors. 16 Watts typical translates to something like 25 Watts max. A 1.8 GHz 970fx G5 might be possible in September, but it is quite late, when dual core Yonahs are coming a few months later.
It really should be extremely little, extremely late, not too little, too late.
Lastly, I bet the 970mp was slated for a mid-summer 2005 Power Mac release, but Apple probably can't ship them until Q1 06.
Comments
Originally posted by NOFEER
all this talk of missed opportunity with this ibm thing. hey jobs knew of the path and potential of the ppc and is not just looking down 1 year....but future development. don't you think that SJ decribed exactly what he wants and where he wants to go with chips --intel, amd etc??? this i think is more for internal propaganda and to tell the other users of the ppc that ibm has a future with this chip so other manufacturers don't go away like apple. it's a defensive catch up mode.
I think Apple knew about these chips when IBM first proposed the designs.
I also think that Apple was already in the throes of deciding whether they would stay or leave.
Either IBM had these specs on hand then, and Apple decided months ago, but waited until the Dev Conf to make a decision, or IBM told Apple that the chips specs would be better than they finally came up with, and THAT straw broke the camel's back.
I don't believe that IBM intro'ed these chips to show Apple up. They aren't good enough for that.
The fact that these chips have been announced by IBM prior to Apple producing a computer using them is probably a bad sign. It means they are available for the highest bidder?
As for G5 Powerbooks with the low power 970FX, there could be all sorts of issues that held them up besides the G5's heat. Bridge chips, low power mode programming, sleep issues, power management, battery life. (Key employess leaving?) Steve is pissed (which I think has already been mentioned.)
I know the process is continuous as I knew an Apple employee who worked on the first Apple Portable with the black and white screen and worked there through Duo's, Wallstreets and Pismo, leaving just before the Tibook was released.
The first Portable wasn't released for months due to a design problem with the black and white screen. Everybody else was releasing portables where the mouse tracks ghosted on the screen and Apple wouldn't do it, they had to have dynamic refresh.
Originally posted by Silverdog
The fact that these chips have been announced by IBM prior to Apple producing a computer using them is probably a bad sign. It means they are available for the highest bidder?
It probably signals that the PowerBook G5 is in fact a finished product. I bet this presentation and information was set for release on that date before WWDC, with IBM's knowledge of the PowerBook G5's timetable and that it would have been announced by then.
Which newly announced processor has the faster front side bus?
-The new IBM 16W 1.6GHz G5
or
-The new Freescale 15W 1.67GHz G4
Also:
With the introduction of the new G5 chips, what would be the possibility of Apple putting the processors in the eMac and/or Mac-Mini, whilst putting the new G4 processors in the iBooks and/or PowerBooks? Would it make sense for Apple to have all their desktop computers powered by the G5 line of processors (single core for the eMac, Mac-Mini, and iMac; dual core for the PowerMac), and have all their mobile laptop computers powered by the G4 line of processors?
Originally posted by 00100011
Quick question:
Which newly announced processor has the faster front side bus?
-The new IBM 16W 1.6GHz G5
or
-The new Freescale 15W 1.67GHz G4
Also:
With the introduction of the new G5 chips, what would be the possibility of Apple putting the processors in the eMac and/or Mac-Mini, whilst putting the new G4 processors in the iBooks and/or PowerBooks? Would it make sense for Apple to have all their desktop computers powered by the G5 line of processors (single core for the eMac, Mac-Mini, and iMac; dual core for the PowerMac), and have all their mobile laptop computers powered by the G4 line of processors?
The problem right now is that nobody knows. The specs aren't out.
Apple has been putting the mobile chips into the Mini, and into the EMac as well. It's a power, heat, price thing.
Originally posted by 00100011
Quick question:
Which newly announced processor has the faster front side bus?
-The new IBM 16W 1.6GHz G5
or
-The new Freescale 15W 1.67GHz G4
Also:
With the introduction of the new G5 chips, what would be the possibility of Apple putting the processors in the eMac and/or Mac-Mini, whilst putting the new G4 processors in the iBooks and/or PowerBooks? Would it make sense for Apple to have all their desktop computers powered by the G5 line of processors (single core for the eMac, Mac-Mini, and iMac; dual core for the PowerMac), and have all their mobile laptop computers powered by the G4 line of processors?
The 970 has the option of 1/2, 1/3 or 1/4 FSB vis-a-vis clock speed. So a 1.8 Ghz G5 has the option of an FSB of 900, 600 or 450Mhz. The 7448 is a G4 with a 200Mhz Maxbus FSB. Therefore, the G5 will have the faster FSB.
However, that comparison is misleading. Others have a better memory than me, but I was under the impression that, because of the length and width of the G5 processing pipeline, the G5 was more bandwidth hungry than the G4. Hence, the G5 needs a comparatively larger FSB to keep all the execution units pumping (and minimising wasted CPU resources and heat expenditure). The speed battle between the G4 and G5 is a line ball when clockspeed between the two is close and the FSB is throttled back from 1/2 to the lower ratios.
As to your other question, buggered if I know
Originally posted by a j stev
The 970 has the option of 1/2, 1/3 or 1/4 FSB vis-a-vis clock speed. So a 1.8 Ghz G5 has the option of an FSB of 900, 600 or 450Mhz. The 7448 is a G4 with a 200Mhz Maxbus FSB. Therefore, the G5 will have the faster FSB.
However, that comparison is misleading. Others have a better memory than me, but I was under the impression that, because of the length and width of the G5 processing pipeline, the G5 was more bandwidth hungry than the G4. Hence, the G5 needs a comparatively larger FSB to keep all the execution units pumping (and minimising wasted CPU resources and heat expenditure). The speed battle between the G4 and G5 is a line ball when clockspeed between the two is close and the FSB is throttled back from 1/2 to the lower ratios.
As to your other question, buggered if I know
Except, again, we don't know if thise figures are correct. This is a low power chip, it may not have the ability to use all of the bandwidth the full power chips use.
Originally posted by DHagan4755
Apple probably knew at least a year or more ago the situation with the chips IBM recently announced.
They knew in mid-2003, and were designing hardware around them for 2004 releases. IBM is just 1 year, and counting, late with the "low power" 970fx. The 970mp isn't so much the issue. All of Apple's 2004 Mac plans depended on the 970fx and IBM's 90 nm fab performance.
IBM was expecting to ship a 2 GHz, 39 Watts max, 970fx CPU in Q1 04. But their fab never even came close to that, about 55 Watts max at 2 GHz, and not only that, they couldn't deliver on quantity either. They couldn't do that at any stage of that year. This was the chip that was planned to be at 3 GHz.
Remember, the 1st Xserve G5, a 1U server, was 3 to 4 months late (Jan announcement, Apr "shipment"). If IBM was able to produce this 2 GHz, 39 Watts max 970fx CPU rather than 55 Watts max, it meant a quieter Xserve. If they could ship it in quantity, Xserve G5s wouldn't have been that late.
If they could ship it in quantity, the iMac G5 would have been released in June 04, instead of September, and be quieter, possibly less buggy. The 2.5 GHz PowerMac G5 could have been released in February 04 instead of June 04. Lastly, a 1.6 to 1.8 GHz Powerbook G5 would have been possible in Fall 04.
They were counting on a "low power" 970fx for 2004. The current announcement of 1.2 to 1.6 GHz, <16 Watts typical, processors is really very little improvement from the current 970fx processors. 16 Watts typical translates to something like 25 Watts max. A 1.8 GHz 970fx G5 might be possible in September, but it is quite late, when dual core Yonahs are coming a few months later.
It really should be extremely little, extremely late, not too little, too late.
Lastly, I bet the 970mp was slated for a mid-summer 2005 Power Mac release, but Apple probably can't ship them until Q1 06.