Got a tip for us?
Let us know
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
New PowerBooks on Tuesday.
- Thread starter MacRumors
- Start date
- Sort by reaction score
Re: Re: Re: Ok here is the scoop
I am actually hoping it will be a G4 update, then we can continue this discussion for another half year (I told it was too hot! Look at the heatsinks man."
Originally posted by displaced
*/me hits forehead on desk..... repeatedly*
I am actually hoping it will be a G4 update, then we can continue this discussion for another half year (I told it was too hot! Look at the heatsinks man."
That is because you can run a current G5 at that clock rate. Look at the links that have been all over MacRumors (e.g. on the power savings) plus the links in the posts. Look at the processor energy saving features (well documented in posts I've made and others have made plus on the front page of rumors too). At *worst* Apple could just run it in the "low power mode" all the time.
*Current* PB G4s and G5s are comparable. Favorably in favor of the G5s even because they don't have the cache that the G4 has.
I *hope* it will happen, but I *don't* think it will. But the reason won't be heat issues.
Re: Re: Re: Re: Ok here is the scoop
Hmm on the one hand, that'd be funny. On the other hand I think I'd lose what's left of my sanity. I don't remember having seen so much bickering on these forums for some time...
(Just the thought of it is starting to make me sweat profusely... I'm the one who need heatsinks!)
LOL
Hmm on the one hand, that'd be funny. On the other hand I think I'd lose what's left of my sanity. I don't remember having seen so much bickering on these forums for some time...
(Just the thought of it is starting to make me sweat profusely... I'm the one who need heatsinks!)
How about performace hits on the G5 of lower frequency, lower bus, etc... Doesn't that defeat the purpose?
Some, but the bus would still be faster than the G4 bus since it runs at 1/2 the speed of the G5.
A good point NicoMan... the current 1.6 G5s aren't that much faster than the 1.25 G4s. Part of this is that the bevy of code to take advantage of the G5 to it's fullist isn't written yet (like when we went from G3 to G4). Eventually the G5 will be a monster lion, but while it's stuck with so much legacy code it's a bit of a lamb!
Go to any site, from barefeats.com to macaddict and you'll see the G4 1.25 DP even in SP mode can hold it's own vs. the G5. In some cases the G5 gets beaten by the current G4, even with the faster MHz. It raises an excellent point, how much faster would the G5 be short term in the Powerbook clocked down? Fact is, for some things, it wouldn't.
And this would lead more credence to why even it do it until Apple has faster chips in good supply. A 1.2 G5 in the current offering would be more of a show pony than a trouncing megacomputer, a 1.25 G4 by many accounts based on the current test at the sites above and elsewhere would actually beat the G5... making the G5 pointless. The figures on these sites don't lie. So finally I've got even more reason to doubt the G5, you'd be risking actually speed bumping down in some regards... that's aside from the fact it's NOT going to happen anyways for a while.
You have to love the people who say
'G5s aren't possible cus they're too hot. Dual G4s are more likely.'
Last I heard the 17 barely sold enough to justify its existance. It is, afterall, an expensive replacement for a desktop, rather that apealing to the traditional laptop market. Thats why i think you'll find the 15 and 17 specs will be more or less the same; certainly for the top model in the 15 range.
Concivably they might put a G5 in the 17 and leave the rest G4 to boost 17 sales.
DP laptops would requre probably more of an internal redesign than a G5, in terms of Mobo etc.
Further on the G4/G5 debate:
Is the G5 not cheaper than the G4? And I beleve IBM can produce them faster and more reliably than Moto can the G4.
They aren't too hot. Look at the posts re power dissapation at the start.
It is posible that they could intorduce a G5 laptop, either in just the 17 or both the 17 and 15. The 12 seems unlikey, only because its small screen pitches it at the pure laptop market rather than the 'portable computer' market (ie you'll run Word and Excel and use it on the train but probably not serously use FCP4 etc.). This could explain the update to the 15 and 17 but not the 12. (I'd sort of expect the 12 and 17 to be updated together.)
AS to the need for devleopment work for the G5 lets not forget that the G5 has more than a passing resmblance to the Power4 chip so it's not like they would need to design a computer completly from scratch.
I'm not saying that they will be G5s or G4s, just saying that there's enough doubt either way to make it very difficult to guess. Safe (-er) money says 1.25-1.33 GHz G4s.
Not too long to wait now.
The next question is: If they're G4s then how long 'till the G5s and is it woorth the wait?
'G5s aren't possible cus they're too hot. Dual G4s are more likely.'
Last I heard the 17 barely sold enough to justify its existance. It is, afterall, an expensive replacement for a desktop, rather that apealing to the traditional laptop market. Thats why i think you'll find the 15 and 17 specs will be more or less the same; certainly for the top model in the 15 range.
Concivably they might put a G5 in the 17 and leave the rest G4 to boost 17 sales.
DP laptops would requre probably more of an internal redesign than a G5, in terms of Mobo etc.
Further on the G4/G5 debate:
Is the G5 not cheaper than the G4? And I beleve IBM can produce them faster and more reliably than Moto can the G4.
They aren't too hot. Look at the posts re power dissapation at the start.
It is posible that they could intorduce a G5 laptop, either in just the 17 or both the 17 and 15. The 12 seems unlikey, only because its small screen pitches it at the pure laptop market rather than the 'portable computer' market (ie you'll run Word and Excel and use it on the train but probably not serously use FCP4 etc.). This could explain the update to the 15 and 17 but not the 12. (I'd sort of expect the 12 and 17 to be updated together.)
AS to the need for devleopment work for the G5 lets not forget that the G5 has more than a passing resmblance to the Power4 chip so it's not like they would need to design a computer completly from scratch.
I'm not saying that they will be G5s or G4s, just saying that there's enough doubt either way to make it very difficult to guess. Safe (-er) money says 1.25-1.33 GHz G4s.
Not too long to wait now.
The next question is: If they're G4s then how long 'till the G5s and is it woorth the wait?
G5 heat/PB FAQ & Summary
This is the summary/FAQ on the G5 in the PB discussion. It deals with all the questions and information that several of us have been providing for several months. I think it summarizes all the questions, has all the links to the relevant documents etc. Nothing is speculation. It is solely based on currently (Sept 14, 2003) available information. If you have more, *please* share it!
I am going to save it and post it ASAP whenever the "New PBs" coming arrises. Perhaps it will be moot with a PB G5 in Paris tomorrow, but if not...
1. Compared to the PBs shipped through September 14,2003, the G5 is comparable in heat usage at similar clock speeds..
According to IBM's PowerPC 970 documentation , a 1.2 GHz PowerPC 970 comes in at 19W power dissipation (typical), while a 1.8 GHz PowerPC 970 comes in at 42W.
Meanwhile, Motorola's PowerPC 74xx documentation indicates that a 1 GHz PowerPC 7455, the latest processor about which data is available, dissipates 35.5W (typical) (and 50W max).[a]
Power dissipation (typical) per MHz:
PowerPC 7455: 0.0355W
PowerPC 970 (1.1v): 0.0158W
PowerPC 970 (1.3v): 0.0233W
It's easy to see that, clock for clock, the PowerPC 970 consumes much less power - and therefore dissipates less heat - than the PowerPC 74xx. The current 1.8 GHz G5 numbers aren't significantly higher than a 1.0 GHz G4...it's therefore not hard to imagine the G5 making its debut in a portable machine in the future.
Yes, the prevailing rumors indicate that the PowerPC 7457 will be used in any new (none out as of Sept 14 yet) PowerBook, and will presumably consume less power, and therefore dissipate less heat, per clock. However, since the clock speed is likely being increased, total power dissipation may be the same (http://e-www.motorola.com/files/32bit/doc/fact_sheet/PPCSALESFACT.pdf). This PDF shows them to be released sometime in 2003 (and as of Sept 14, 2003 they are not out).
2. People keep saying it: "it needs 9 fans to keep it cool." From what I read on Apple's site, it does NOT *need* 9 fans.
They *use* 9 fans at really low speeds to keep the noise way down (think of the so-called "wind-tunnel Macs" last year) whereas 1 fan at a higher speed would work. A lot of this has to do with the clocking of the chip (the high speeds generate lots of heat), whereas at, say, 1.2Ghz it would be much cooler. In short, the point of the 8/9 fans was to use more at a lower speeds to decrease the sound, not because 1 fan couldn't handle it.
3. Regarding the power saving mode:
Summary: the 970 has got doze/nap/sleep mode, which were on previous PowerPCs. It can turn off some parts of the processor when things are quiet--under OS control. For thermal management, it has a diode on chip for monitoring temperature with leads that go off chip allowing external control.
4. The total CPU/cache combo uses *less* powersince the G5 doesn't need a L3 cache which uses a fair amount of power (it too has to be fast memory, and fast=more power). Thus the power usage *could* be *less* than for a PB G4 at, say, 1.2Ghz.
Although I'd *love* to see it happen soon, but am skeptical that it will. I think the things like the "needs 8 fans" become conventional wisdom without facts to back them up.
(G5:
http://www-3.ibm.com/chips/techlib/...7256C5200611780
7457:
http://e-www.motorola.com/files/32b...t/MPC7457EC.pdf
G5 power management:
http://www.arstechnica.com/cpu/03q2...nterview-2.html
http://developer.apple.com/document...rMacG5/1Introduction/chapter_2_section_4.html
)
4. " PowerBooks are not possible this soon. Apple has officially said so. "
No, what they said was:
The G5 was "not going in PowerBook anytime soon". That was back in June. They never defined "anytime soon." That could mean, the summer or whatever they want. It is vague enough to mean anything. And I *think* that was a good move. Unlike them saying "ITMS for Windows by the end of the year." That was silly because it encouraged everyone to try to beat them to the bandwagon.
5. Have the figures on the G5 heat used changed since October 2002 when IBM published their specs? If the figures from Oct 2002 were significantly different then IBM would update them. They are trying to sell chips and need accurate information for their customers. Their customers don't want surprises.
6. The dual G5 dissipates 97 watts at 2 Ghz. So around 48.5 for each G5 which is reasonably close to the October 2002 figures of 42 watts at 1.8 GHz. Thus, the lower speed figures are probably pretty accurate too.
7. The IBM Power4 consumes about 35 watts at 1.2GHz. Given the differences in the chip and manufacturing, the October 2002 figures for the 970 are probably pretty accurate. (from various reading regarding the Power4).
8. There are also considerations of watts/sq. mm, e.g. density.
March 2003:
"Unlike the Itanium2, IBM's PowerPC 970 isn't that different from the chips Apple is already using. Its die size, power dissipation and core voltage are all in line with what Apple would need for 64-bit PowerBooks. " (http://www.technewsworld.com/perl/story/21144.html )
And for completeness:
9. Here is a pic of the 970:
http://www.theinquirer.net/images/articles/ppc970.jpg
10. Can the G5 be used in a current (Sept 14, 2003) design. No, but Apple has had 18 months or so to design one.
11. When Intel, IBM or Motorola or whomever makes chips, they don't make specific speeds. They make chips and then test the chips to see what speeds they work reliably at. This is why they speak of "yields" for certain speeds. All the chips might work at 100Mhz, but none at 100 Ghz. 10% might at 2Ghz, 90% at 1Ghz.
[a] As of Sept 14, 2003
This is the summary/FAQ on the G5 in the PB discussion. It deals with all the questions and information that several of us have been providing for several months. I think it summarizes all the questions, has all the links to the relevant documents etc. Nothing is speculation. It is solely based on currently (Sept 14, 2003) available information. If you have more, *please* share it!
I am going to save it and post it ASAP whenever the "New PBs" coming arrises. Perhaps it will be moot with a PB G5 in Paris tomorrow, but if not...
1. Compared to the PBs shipped through September 14,2003, the G5 is comparable in heat usage at similar clock speeds..
According to IBM's PowerPC 970 documentation , a 1.2 GHz PowerPC 970 comes in at 19W power dissipation (typical), while a 1.8 GHz PowerPC 970 comes in at 42W.
Meanwhile, Motorola's PowerPC 74xx documentation indicates that a 1 GHz PowerPC 7455, the latest processor about which data is available, dissipates 35.5W (typical) (and 50W max).[a]
Power dissipation (typical) per MHz:
PowerPC 7455: 0.0355W
PowerPC 970 (1.1v): 0.0158W
PowerPC 970 (1.3v): 0.0233W
It's easy to see that, clock for clock, the PowerPC 970 consumes much less power - and therefore dissipates less heat - than the PowerPC 74xx. The current 1.8 GHz G5 numbers aren't significantly higher than a 1.0 GHz G4...it's therefore not hard to imagine the G5 making its debut in a portable machine in the future.
Yes, the prevailing rumors indicate that the PowerPC 7457 will be used in any new (none out as of Sept 14 yet) PowerBook, and will presumably consume less power, and therefore dissipate less heat, per clock. However, since the clock speed is likely being increased, total power dissipation may be the same (http://e-www.motorola.com/files/32bit/doc/fact_sheet/PPCSALESFACT.pdf). This PDF shows them to be released sometime in 2003 (and as of Sept 14, 2003 they are not out).
2. People keep saying it: "it needs 9 fans to keep it cool." From what I read on Apple's site, it does NOT *need* 9 fans.
They *use* 9 fans at really low speeds to keep the noise way down (think of the so-called "wind-tunnel Macs" last year) whereas 1 fan at a higher speed would work. A lot of this has to do with the clocking of the chip (the high speeds generate lots of heat), whereas at, say, 1.2Ghz it would be much cooler. In short, the point of the 8/9 fans was to use more at a lower speeds to decrease the sound, not because 1 fan couldn't handle it.
3. Regarding the power saving mode:
Summary: the 970 has got doze/nap/sleep mode, which were on previous PowerPCs. It can turn off some parts of the processor when things are quiet--under OS control. For thermal management, it has a diode on chip for monitoring temperature with leads that go off chip allowing external control.
4. The total CPU/cache combo uses *less* powersince the G5 doesn't need a L3 cache which uses a fair amount of power (it too has to be fast memory, and fast=more power). Thus the power usage *could* be *less* than for a PB G4 at, say, 1.2Ghz.
Although I'd *love* to see it happen soon, but am skeptical that it will. I think the things like the "needs 8 fans" become conventional wisdom without facts to back them up.
(G5:
http://www-3.ibm.com/chips/techlib/...7256C5200611780
7457:
http://e-www.motorola.com/files/32b...t/MPC7457EC.pdf
G5 power management:
http://www.arstechnica.com/cpu/03q2...nterview-2.html
http://developer.apple.com/document...rMacG5/1Introduction/chapter_2_section_4.html
)
4. " PowerBooks are not possible this soon. Apple has officially said so. "
No, what they said was:
The G5 was "not going in PowerBook anytime soon". That was back in June. They never defined "anytime soon." That could mean, the summer or whatever they want. It is vague enough to mean anything. And I *think* that was a good move. Unlike them saying "ITMS for Windows by the end of the year." That was silly because it encouraged everyone to try to beat them to the bandwagon.
5. Have the figures on the G5 heat used changed since October 2002 when IBM published their specs? If the figures from Oct 2002 were significantly different then IBM would update them. They are trying to sell chips and need accurate information for their customers. Their customers don't want surprises.
6. The dual G5 dissipates 97 watts at 2 Ghz. So around 48.5 for each G5 which is reasonably close to the October 2002 figures of 42 watts at 1.8 GHz. Thus, the lower speed figures are probably pretty accurate too.
7. The IBM Power4 consumes about 35 watts at 1.2GHz. Given the differences in the chip and manufacturing, the October 2002 figures for the 970 are probably pretty accurate. (from various reading regarding the Power4).
8. There are also considerations of watts/sq. mm, e.g. density.
March 2003:
"Unlike the Itanium2, IBM's PowerPC 970 isn't that different from the chips Apple is already using. Its die size, power dissipation and core voltage are all in line with what Apple would need for 64-bit PowerBooks. " (http://www.technewsworld.com/perl/story/21144.html )
And for completeness:
9. Here is a pic of the 970:
http://www.theinquirer.net/images/articles/ppc970.jpg
10. Can the G5 be used in a current (Sept 14, 2003) design. No, but Apple has had 18 months or so to design one.
11. When Intel, IBM or Motorola or whomever makes chips, they don't make specific speeds. They make chips and then test the chips to see what speeds they work reliably at. This is why they speak of "yields" for certain speeds. All the chips might work at 100Mhz, but none at 100 Ghz. 10% might at 2Ghz, 90% at 1Ghz.
[a] As of Sept 14, 2003
Than you will wake up in heaven and God will say to you that the G5 powerbook was avaible few day after your death! but he can have you a:
Quadra G6 Powerbook
0.5 inch thick
a projector integrated
16 Mo cache/processor
Ati Radeon 10 500 Pro with 512 Mo DDR IV
250 Go of RAM (no more need of HD cause it's now a no power still conserve the data type)
Holographic cube burner (1 To of Data)
Then god will look at you and say: "Hoo sorry, I missidentify you, you have to go to hell and use a DELL P1, 16 RAM, Floppy drive, onboard graphic and onboard sound, PS2 keyboard if unplug and replug no more working, etc..."
Sorry, my divaguation!
You made some interesting points, but why do you even start about heat dissipation? Even if you let alone the fact that the posts you mentioned don't mean zilch, because no one has actual figures for the 7457s, the 7455 revB and so on. So stop citing nonexistant fact. I'm sure you mean well but with all due respect you haven't brought anyhing new to the debate. The bickering about heat dissipation has been utterly grotesque. Let's put it to rest once and for all.
I'm not sure I understand... Could you explain? (It's not a flame, it's a genuine question)
I'm glad you like it. When I typed it I wasn't even sure it existed in English. I've just checked now. It does exist.
I mean that the relationship with the Power4 means that there has already been quite a lot of development work done on it -- IBM has a wealth of experience about what does and doesn't work, where the problems in Mobo design are etc.
As to the bickering about heat dissipation I intended to agree with your sentiment; my argument is perhaps better put as:
'What little data is available seems to suggest that there is little difference in heat dissipation. Thus other factors are the important ones: Mobo development, Marketing etc.'
I intended the post's main emphasis to be on the other areas. Was in a bit of a rush when I wrote it, so sorry for not making myself clearer.
In conclusion: G5s are not 'certain' nor are they 'impossible', all I'm trying to say is that there are reasons why the new PBs could be either G4 or G5, or a mix of both.
Spelling Corrected at 15.05GMT
OK - Isn't there anybody out there who has even the sligtest bit of REAL news about the PB Range - The rumours and knowing there's only 18 hours until big Steve gets on the Podium is driving me insane - Someone shed some light -
Re: G5 heat/PB FAQ & Summary
that's about as far as I went. Anyone patient enough to read through all of this and see if there is something new in there?
And I agree, spelling out the actual facts about the G5 has been like explaining the colour blue to an endless stream of blind people.
I have now officially died and gone to a better place where mac rumours don't exist and we all use dos.
c:\ya\later >
Agreed. 100%. (even though I am a bit sceptical about a mixed G4/G5 PB range)
edit: corrected the spelling of sceptical... oops
But when will they be available?
What's the over/under in weeks for the availability of the alleged new PBooks?
Immediately?
2 weeks?
4 weeks?
6 weeks?
I'm going to be furious if its more than that.
Any thoughts?
.a
What's the over/under in weeks for the availability of the alleged new PBooks?
Immediately?
2 weeks?
4 weeks?
6 weeks?
I'm going to be furious if its more than that.
Any thoughts?
.a
What about the iMacs? If the G5s where cheaper/faster/colder(?) than the G4 they would for sure be used in the iMac.
They would never have comparable speeds, so consumer/pro would not really be a big thing. This means that there is some reason that apple know of that meant it made sense to upgrade the G4 there.
They would never have comparable speeds, so consumer/pro would not really be a big thing. This means that there is some reason that apple know of that meant it made sense to upgrade the G4 there.