I don't buy this one. I think Apple is a long way off a G5 Powerbook. Heat = wasted energy = short battery life. Even if they could cool the G5 it still seems to be using too much energy to have a useful life. I doubt Apple has any choice but to continue using a G4 processor for some time to come, the G5 problems were not planned, we all know what has publicly happened to the 3ghz G5.
It would be a public relations disaster to bring on a G5 PB with a slower clock rate than the G4 PB. I wish there were some Rabbits and the duel cored G4 would be a good solution, possibly better than than a single G5, except that Apple needs it right now.
I'm with Addison on this one. I'm skeptical. It would not be surprising if Apple has prototypes on campus running G4s at clock rates up to 2.0 GHz. It would not be surprising if they had G5 powerbook prototypes. Certainly they have to build these to learn how well the models fit reality. However, engineering prototypes don't mean that a manufacturable solution is near at hand.
Based on the report from IBM that was floating around a while back it sounds like there is a lot of variability in the power consumption of the G5s from one unit to the next. Throw in the added power for the faster bus and you have a machine that drains batteries fast. It stands to reason that a laptop G5 is a ways off.
On the other hand, with a nice padded carrying case the new iMac G5 could serve as a transportable computer for those who need to travel to a customer's site to work. That might relieve some pressure for the need for a G5 laptop.
Okay, correct me if I'm wrong here, but isn't it impossible for the G4 to go above 1.5Ghz?
Secondly, if this is true, and if the G5s are way off, is there any reason that the powerboks can't have dual G4 chips (say, 2 individual 1.3 or 1.0 Ghz G4 chips) and use the watercooling setup from the G5s (or a knock-off smaller modified one?)
From what I understand, dual-core is along way off, G5 is a long way off, and higher clockspeeds are a long way off...so we're either fdsked in terms of innovation or advancement for a year or so, or there's something I'm missing.
Dual G4 processors + Watercooling unit = Battery life < 1hr.
2 x G4 1.0 Ghz < 1 x G4 1.5 GHz (for most uses).
So why bother? Apple is not in the business of desktop replacements, which is the only rationale I can think of for the above.
There is precious little point in bringing out 'innovations' if they have either no practical use nor add any particular value. The problem with Apple's processor replacements lie in the fact that its main supplier, IBM, is in the business of heavy duty servers running heavy duty OSes. I'm not sure, if Apple is an important enough client for IBM to want to bother with throwing too much in the way of resources in trying to design a chip to compete with the x86 mobile versions, for which Apple would conceivably be the only client. Not to say that developments in IBM's primary range won't lead to exciting prospects for Apple, but I do not see it as IBM's main focus. That mound of cash Apple is sitting on may prove persuasive, if the shareholders do not object, however.
Probably time for Apple to concentrate on other matters, such as battery life, screen quality, improving the logic board bus throughput, notebook housing which does not fry an egg nor dents if you so much as look at it etc... If nothing else, there may be some price cuts to come, which is always welcome. The PBs are no longer competitive with the x86 competition at the prices they sell for here in Europe.
If there is any consolation (and it is a small one, indeed) then owners of the current range of PBs, such as myself, will have a year or so in which we will not rue waiting just a little bit longer to be able to enjoy the latest and greatest.
.... Now, it appears they are only waiting for low-k 970fx chips.
The wait may not be as long as people think. I can't locate the links, but a testing firm that analyzes the structure and process used by chip manufacturers published results on the latest G5 about a month ago. At the end of the article a representative of that company was quoted as saying that IBM was qualifying the low-k process and his company was expecting samples for testing in about a month(re: it's been about a month since the report.) If true, Apple should be receiving production run samples of a low-k G5 soon, if not already, for testing in laptops. What would the power consumption of a 1.0 - 1.1v G5 chip be??
The wait may not be as long as people think. I can't locate the links, but a testing firm that analyzes the structure and process used by chip manufacturers published results on the latest G5 about a month ago. ... At the end of the article a representative of that company was quoted as saying that IBM was qualifying the low-k process and his company was expecting samples for testing in about a month(re: it's been about a month since the report.)
..."IBM has been conservative with the 970FX, and used almost the same dielectric stack as in their 130nm back-end", said James. "The dielectric matrix is a bi-layer, with an FSG inter-metal layer, with oxide at the via levels, and oxynitride cap layers on the upper metal levels. There are no etch-stop layers for the metal trenches, but historically IBM have not used them. They have changed the cap layers on the lowest four copper levels to SiOCN to reduce the effective dielectric constant a little."
"They have been reported to be qualifying low-k during the last few months, and we are expecting to see low-k product soon. We want to compare it with the other low-k processes that we have analyzed."
Quote:
If true, Apple should be receiving production run samples of a low-k G5 soon, if not already, for testing in laptops. What would the power consumption of a 1.0 - 1.1v G5 chip be??
An approximation could be made by muiltiplying the current power numbers with (0.75 for low-k) * (Vnew/Vold)^2. Last Spring, IBM abandoned the original Applied Materials based low-k process they were using for a "proprietary one". So hopefully, 6 months is enough to implement this new process.
The lower-right portion of the curve shows the change in frequency and power from 0.8 to 1.3 V in 25-mV increments. The very top line of the envelope is the change in power at a fixed voltage of 1.3 V as the frequency decreases from 2.5 to 0 GHz, while the lowest line is at a fixed voltage of 0.8 V. These lines show the linear change in power following the ½CV²f power relation.
Using the FSG dielectric 970fx, and counting down the 25 mV steps, it appears a 970fx running at 1 V and 1.8 GHz will burn 35 Watts or so. Low-k should lower that another 20 to 30%.
Thanks for clarifying. So based on your post, do you think that low-k dialectric alone would allow Apple to use a G5 in their laptops, knowing how picky Apple is about heat/weight etc.?
Also, has anyone heard any rumors concerning progress IBM is making on Strained Silcon Directly on Insulator?
So based on your post, do you think that low-k dialectric alone would allow Apple to use a G5 in their laptops, knowing how picky Apple is about heat/weight etc.?
Yes, I think so.
The IBM chart shows that the 1.0V 1.8 GHz FSG 970fx should run ~30 Watts max. The 1.3V 1.5 GHz 7447A in the current Powerbooks run ~30 Watts max, so they'll be within the Powerbook G4 processor power box. The low-k should guarantee better yields at lower voltages, hopefully less leakage, and they should "comfortably" be within the power budget compared to the FSG ones.
The german tech mag c't has a 20" iMac G5 test in its latest edition. They measured temperatures of 34 degrees celsius in the bottom right corner of the screen, and up to 48 degrees celsius in the upper left corner. According to c't, TFT screens are specified for up to 40 degrees celsius, and normally run at around 30 degrees in normal use. They say that this might possibly reduce the life span of the screen.
Certainly Apple are aware of this -- and still they were not able (or willing: maybe it doesn't matter at all) to cool the iMac better.
But anyway: this thing does get hot, and I find it difficult to imagine that it would work in a PB enclosure at this time.
Thank you for your reply. Always enjoy your posts. I'm not in the maket for a laptop, but am nearing a decision on buying a new or used desktop and a small boost in the G5 might occur with the introduction of a G5 using a low-K dialectric.
The german tech mag c't has a 20" iMac G5 test in its latest edition. They measured temperatures of 34 degrees celsius in the bottom right corner of the screen, and up to 48 degrees celsius in the upper left corner. According to c't, TFT screens are specified for up to 40 degrees celsius, and normally run at around 30 degrees in normal use. They say that this might possibly reduce the life span of the screen.
Certainly Apple are aware of this -- and still they were not able (or willing: maybe it doesn't matter at all) to cool the iMac better.
But anyway: this thing does get hot, and I find it difficult to imagine that it would work in a PB enclosure at this time.
The current G5 does not use a low-k dialectric. Different process = cooler chip = suitable for laptops, maybe.
The current G5 does not use a low-k dialectric. Different process = cooler chip = suitable for laptops, maybe.
That would be ideal, wouldn't it. Even better would be intelligent power management as well, like Intel does with the Pentium-M and XScale. I'm not sure where IBM stands on such issues, although they have used it in the now defunct PPC405LP.
The performace improvement in the G5/1.8 is minimal versus G4/1.5- at least with the slower bus as in the iMac. I think we will need a newer(cooler and faster) single G5 to make the transition worth making. More likely we will get another patheticly slow(developing-not just performance) G4.
The german tech mag c't has a 20" iMac G5 test in its latest edition. They measured temperatures of 34 degrees celsius in the bottom right corner of the screen, and up to 48 degrees celsius in the upper left corner. According to c't, TFT screens are specified for up to 40 degrees celsius, and normally run at around 30 degrees in normal use. They say that this might possibly reduce the life span of the screen.
Did the guys in c't say how they measured the temperature?
The performace improvement in the G5/1.8 is minimal versus G4/1.5- at least with the slower bus as in the iMac. I think we will need a newer(cooler and faster) single G5 to make the transition worth making. More likely we will get another patheticly slow (developing-not just performance) G4.
You do realize that a Powerbook G5 would have at least a Radeon 9700 Mobility as its GPU? Given the same GPU, a 1.8 Powerbook G5 will be faster than a 1.5 GHz Powerbook G4 by ~20% in integer heavy code, ~80% in floating point heavy code, and ~20% in AltiVec code. Those are all much better improvements than the ~10% performance increase from the previous 1.33 GHz PB G4 highend to the current 1.5 GHz PB G4 high end.
The transition to a 1.8 GHz Powerbook G5 is worth making in the next 3 or 4 months. Longer than that, Apple will either need a 2+ GHz 970fx or the dual-core G4 to be competitive. And we all know a dual-core G4 isn't coming until 1H 06, so they really don't have many choices.
They are in a bit of a conundrum for the next Powerbook update. If they want to stay G4, they'll have to let the current 1.5 GHz G5 be the top end until at least May 05 for pilot production 7448 CPUs or shoehorn a 970fx for release in the next 3 to 4 months. They should just sell a 1.5" thick Powerbook G5 in addition to the thinner Powerbook G4 and be done with it.
And we all know a dual-core G4 isn't coming until 1H 06, so they really don't have many choices.
They are in a bit of a conundrum for the next Powerbook update. If they want to stay G4, they'll have to let the current 1.5 GHz G5 be the top end until at least May 05 for pilot production 7448 CPUs or shoehorn a 970fx for release in the next 3 to 4 months.
Indeed, there are not choices for Apple right now. If IBM cannot deliver a "mobile" G5 in the next 3-4 months, I don't see nothing more than a possible speed bump with some faster variant of the 7447A. Or even worse, stagnation of the Powerbook line (CPU-wise) until the production of the faster 7448. With a feature bump in the meantime.
And I have doubts that even a low voltage G5 could go in a such a slim notebook like the Powerbook.
And I have doubts that even a low voltage G5 could go in a such a slim notebook like the Powerbook.
Agreed... "IF" we see a powerbook g5 with in the next year... bet on a redesigned case... Personally I still don't see the advantage of moving to g5 in a powerbook... it is going to EAT battery life, EAT component life (heat), EAT performance. Give me a 1.8ghz g4 with a little FSB bump and I'd be happy.
Comments
It would be a public relations disaster to bring on a G5 PB with a slower clock rate than the G4 PB. I wish there were some Rabbits and the duel cored G4 would be a good solution, possibly better than than a single G5, except that Apple needs it right now.
Based on the report from IBM that was floating around a while back it sounds like there is a lot of variability in the power consumption of the G5s from one unit to the next. Throw in the added power for the faster bus and you have a machine that drains batteries fast. It stands to reason that a laptop G5 is a ways off.
On the other hand, with a nice padded carrying case the new iMac G5 could serve as a transportable computer for those who need to travel to a customer's site to work. That might relieve some pressure for the need for a G5 laptop.
Secondly, if this is true, and if the G5s are way off, is there any reason that the powerboks can't have dual G4 chips (say, 2 individual 1.3 or 1.0 Ghz G4 chips) and use the watercooling setup from the G5s (or a knock-off smaller modified one?)
From what I understand, dual-core is along way off, G5 is a long way off, and higher clockspeeds are a long way off...so we're either fdsked in terms of innovation or advancement for a year or so, or there's something I'm missing.
Please enlighten me.
Dual G4 processors + Watercooling unit = Battery life < 1hr.
2 x G4 1.0 Ghz < 1 x G4 1.5 GHz (for most uses).
So why bother? Apple is not in the business of desktop replacements, which is the only rationale I can think of for the above.
There is precious little point in bringing out 'innovations' if they have either no practical use nor add any particular value. The problem with Apple's processor replacements lie in the fact that its main supplier, IBM, is in the business of heavy duty servers running heavy duty OSes. I'm not sure, if Apple is an important enough client for IBM to want to bother with throwing too much in the way of resources in trying to design a chip to compete with the x86 mobile versions, for which Apple would conceivably be the only client. Not to say that developments in IBM's primary range won't lead to exciting prospects for Apple, but I do not see it as IBM's main focus. That mound of cash Apple is sitting on may prove persuasive, if the shareholders do not object, however.
Probably time for Apple to concentrate on other matters, such as battery life, screen quality, improving the logic board bus throughput, notebook housing which does not fry an egg nor dents if you so much as look at it etc... If nothing else, there may be some price cuts to come, which is always welcome. The PBs are no longer competitive with the x86 competition at the prices they sell for here in Europe.
If there is any consolation (and it is a small one, indeed) then owners of the current range of PBs, such as myself, will have a year or so in which we will not rue waiting just a little bit longer to be able to enjoy the latest and greatest.
Originally posted by THT
.... Now, it appears they are only waiting for low-k 970fx chips.
The wait may not be as long as people think. I can't locate the links, but a testing firm that analyzes the structure and process used by chip manufacturers published results on the latest G5 about a month ago. At the end of the article a representative of that company was quoted as saying that IBM was qualifying the low-k process and his company was expecting samples for testing in about a month(re: it's been about a month since the report.) If true, Apple should be receiving production run samples of a low-k G5 soon, if not already, for testing in laptops. What would the power consumption of a 1.0 - 1.1v G5 chip be??
Originally posted by rickag
The wait may not be as long as people think. I can't locate the links, but a testing firm that analyzes the structure and process used by chip manufacturers published results on the latest G5 about a month ago. ... At the end of the article a representative of that company was quoted as saying that IBM was qualifying the low-k process and his company was expecting samples for testing in about a month(re: it's been about a month since the report.)
CHIPWORKS DISCOVERS IBM STRAIN AT 90NM, DELAY IN LOW-K
..."IBM has been conservative with the 970FX, and used almost the same dielectric stack as in their 130nm back-end", said James. "The dielectric matrix is a bi-layer, with an FSG inter-metal layer, with oxide at the via levels, and oxynitride cap layers on the upper metal levels. There are no etch-stop layers for the metal trenches, but historically IBM have not used them. They have changed the cap layers on the lowest four copper levels to SiOCN to reduce the effective dielectric constant a little."
"They have been reported to be qualifying low-k during the last few months, and we are expecting to see low-k product soon. We want to compare it with the other low-k processes that we have analyzed."
If true, Apple should be receiving production run samples of a low-k G5 soon, if not already, for testing in laptops. What would the power consumption of a 1.0 - 1.1v G5 chip be??
An approximation could be made by muiltiplying the current power numbers with (0.75 for low-k) * (Vnew/Vold)^2. Last Spring, IBM abandoned the original Applied Materials based low-k process they were using for a "proprietary one". So hopefully, 6 months is enough to implement this new process.
The IBM PowerPC 970FX power envelope and power management
Maximum power envelope
The lower-right portion of the curve shows the change in frequency and power from 0.8 to 1.3 V in 25-mV increments. The very top line of the envelope is the change in power at a fixed voltage of 1.3 V as the frequency decreases from 2.5 to 0 GHz, while the lowest line is at a fixed voltage of 0.8 V. These lines show the linear change in power following the ½CV²f power relation.
Using the FSG dielectric 970fx, and counting down the 25 mV steps, it appears a 970fx running at 1 V and 1.8 GHz will burn 35 Watts or so. Low-k should lower that another 20 to 30%.
Thanks for clarifying. So based on your post, do you think that low-k dialectric alone would allow Apple to use a G5 in their laptops, knowing how picky Apple is about heat/weight etc.?
Also, has anyone heard any rumors concerning progress IBM is making on Strained Silcon Directly on Insulator?
Originally posted by rickag
So based on your post, do you think that low-k dialectric alone would allow Apple to use a G5 in their laptops, knowing how picky Apple is about heat/weight etc.?
Yes, I think so.
The IBM chart shows that the 1.0V 1.8 GHz FSG 970fx should run ~30 Watts max. The 1.3V 1.5 GHz 7447A in the current Powerbooks run ~30 Watts max, so they'll be within the Powerbook G4 processor power box. The low-k should guarantee better yields at lower voltages, hopefully less leakage, and they should "comfortably" be within the power budget compared to the FSG ones.
Certainly Apple are aware of this -- and still they were not able (or willing: maybe it doesn't matter at all) to cool the iMac better.
But anyway: this thing does get hot, and I find it difficult to imagine that it would work in a PB enclosure at this time.
Originally posted by THT
Yes, I think so.
...
Thank you for your reply. Always enjoy your posts. I'm not in the maket for a laptop, but am nearing a decision on buying a new or used desktop and a small boost in the G5 might occur with the introduction of a G5 using a low-K dialectric.
Originally posted by philby
The german tech mag c't has a 20" iMac G5 test in its latest edition. They measured temperatures of 34 degrees celsius in the bottom right corner of the screen, and up to 48 degrees celsius in the upper left corner. According to c't, TFT screens are specified for up to 40 degrees celsius, and normally run at around 30 degrees in normal use. They say that this might possibly reduce the life span of the screen.
Certainly Apple are aware of this -- and still they were not able (or willing: maybe it doesn't matter at all) to cool the iMac better.
But anyway: this thing does get hot, and I find it difficult to imagine that it would work in a PB enclosure at this time.
The current G5 does not use a low-k dialectric. Different process = cooler chip = suitable for laptops, maybe.
Originally posted by rickag
The current G5 does not use a low-k dialectric. Different process = cooler chip = suitable for laptops, maybe.
That would be ideal, wouldn't it. Even better would be intelligent power management as well, like Intel does with the Pentium-M and XScale. I'm not sure where IBM stands on such issues, although they have used it in the now defunct PPC405LP.
David
http://www.barefeats.com/imacg5.html
Originally posted by philby
The german tech mag c't has a 20" iMac G5 test in its latest edition. They measured temperatures of 34 degrees celsius in the bottom right corner of the screen, and up to 48 degrees celsius in the upper left corner. According to c't, TFT screens are specified for up to 40 degrees celsius, and normally run at around 30 degrees in normal use. They say that this might possibly reduce the life span of the screen.
Did the guys in c't say how they measured the temperature?
Originally posted by vdoc
The performace improvement in the G5/1.8 is minimal versus G4/1.5- at least with the slower bus as in the iMac. I think we will need a newer(cooler and faster) single G5 to make the transition worth making. More likely we will get another patheticly slow (developing-not just performance) G4.
http://www.barefeats.com/imacg5.html
You do realize that a Powerbook G5 would have at least a Radeon 9700 Mobility as its GPU? Given the same GPU, a 1.8 Powerbook G5 will be faster than a 1.5 GHz Powerbook G4 by ~20% in integer heavy code, ~80% in floating point heavy code, and ~20% in AltiVec code. Those are all much better improvements than the ~10% performance increase from the previous 1.33 GHz PB G4 highend to the current 1.5 GHz PB G4 high end.
The transition to a 1.8 GHz Powerbook G5 is worth making in the next 3 or 4 months. Longer than that, Apple will either need a 2+ GHz 970fx or the dual-core G4 to be competitive. And we all know a dual-core G4 isn't coming until 1H 06, so they really don't have many choices.
They are in a bit of a conundrum for the next Powerbook update. If they want to stay G4, they'll have to let the current 1.5 GHz G5 be the top end until at least May 05 for pilot production 7448 CPUs or shoehorn a 970fx for release in the next 3 to 4 months. They should just sell a 1.5" thick Powerbook G5 in addition to the thinner Powerbook G4 and be done with it.
Originally posted by THT
And we all know a dual-core G4 isn't coming until 1H 06, so they really don't have many choices.
They are in a bit of a conundrum for the next Powerbook update. If they want to stay G4, they'll have to let the current 1.5 GHz G5 be the top end until at least May 05 for pilot production 7448 CPUs or shoehorn a 970fx for release in the next 3 to 4 months.
Indeed, there are not choices for Apple right now. If IBM cannot deliver a "mobile" G5 in the next 3-4 months, I don't see nothing more than a possible speed bump with some faster variant of the 7447A. Or even worse, stagnation of the Powerbook line (CPU-wise) until the production of the faster 7448. With a feature bump in the meantime.
And I have doubts that even a low voltage G5 could go in a such a slim notebook like the Powerbook.
Originally posted by DHagan4755
Oh.
I thought the iMac G5 was using the same chips as the Power Mac G5. 90nm or not, I thought the issue of voltage was separate.
The Rev B Powermacs are 90nm 970fx chips.
Originally posted by PB
And I have doubts that even a low voltage G5 could go in a such a slim notebook like the Powerbook.
Agreed... "IF" we see a powerbook g5 with in the next year... bet on a redesigned case... Personally I still don't see the advantage of moving to g5 in a powerbook... it is going to EAT battery life, EAT component life (heat), EAT performance. Give me a 1.8ghz g4 with a little FSB bump and I'd be happy.