Maybe some sort of higher clocked IBM "G4" chip? I think we've all realized how much IBM kicks Moto's ass by now, might there be some further love coming soon from an unexpected angle?
Unlikely. Apple just dropped IBM as their supplier for portable CPUs because they can't get their 750-designs to scale up. IBM is firmly stuck in the 1Ghz range and their roadmap indicates they won't come out of it soon. And, before you ask, it isn't because Apple had to keep a difference between the pro line and the 750-based consumer line.
Ex-Motos SPS will need some time to get their act together. The organizational shakeup will need to settle, they cannot just implement new ideas without research and development.
I see a nice drought coming for their mobile line - precisely at a time when intel is likely to push their Centrino designs to 2Ghz and beyond.
or the g5 powerbooks are right around the corner and they just want to keep g4 powerbook demand up. Does anyone find it fishy that in recent months apple execs have gone out of their way to say there will be no g5 powerbooks when apple doesn't typically comment on future hardware?????
or the g5 powerbooks are right around the corner and they just want to keep g4 powerbook demand up. Does anyone find it fishy that in recent months apple execs have gone out of their way to say there will be no g5 powerbooks when apple doesn't typically comment on future hardware?????
I am not acquainted with the details of the US law system, but publicly declaring a G5 design is not feasable right now and then selling the exact same thing in 3 month time sounds like an invitation to get sued.
Apple was indeed grasping the straw giving a damn about what would happen with their mobile line.
A friend of mine had an interesting idea. He said that the G5 is just too hot and power-hungry to really work as a mobile chip, but the G6 (PPC 980) is supposed to be out by late 2004 according to IBM, and it might be designed with mobility in mind. He speculated that it would go much like the launch of the G3 - until the G3, Apple laptops used the lower-end 603e processor instead of the 604e. The 604e (like the G5) was intended to be the next high-end CPU, but it required too much power to be used in a laptop. One company made a luggable Mac-clone notebook with a 604e, but it was almost ten pounds and required AC power. So, they came out with the G3 in PowerBooks and PowerMacs simultaneously, intending for the G3 to be the replacement for the 603e. Consumer desktops and the laptops (all laptops were pro laptops back then) would use G3s while the pro desktops, like the 9600, would use the 604e. Turns out the G3 was a whole lot faster than the 604e so it was used in all their machines.
Just replace "603e" with "G4," "604e" with "G5," and "G3" with "G6" in those sentences and you have the current situation (maybe). Laptops are using G4s because the G5 is too hot to work in a laptop. A new processor, the G6, is in development and it'll be out in late 2004 (and IBM is on time with their road maps unlike Moto). It'll be released for PowerBooks while the PowerMacs keep the G5. Actually I don't know really anything about the G6/980, but if it's intended for mobile use, they could give it a new naming scheme like "M6" so there aren't G6 PowerBooks at the same time as G5 PowerMacs. If it exceeds the performance of the G5 while still consuming little enough power to work in a laptop, it could be used in both PowerMacs and PowerBooks. I mean, if it takes the same amount of time to release a G6 processor as it does to release a mobile G5, why not go with the G6 and get a product line just as awesome as the late-1997 PowerMac G3/PowerBook G3 release?
Hopefully by that time, IBM will have come up with a newer, better low end processor incorporating AltiVec to be used in iBooks and iMacs. Apple could go all-IBM in late 2004, with new lower-end IBM processors (maybe like G4s, but scaling higher and with a better bus architecture) and G6s for PowerMacs and PowerBooks.
It seems kind of crazy, but it just might work. It seems plausible to me at least. I have no idea how the PowerBooks will continue for a whole extra year using the G4, and I'm afraid of what it'll do to the iBook and iMac. It seems like Moto has hit their second roadblock... first 500 MHz, now 1.4 GHz. It might be okay to bring the 12" PowerBook to 1.25 GHz and the iBooks to 1.0-1.25 GHz, but what about the 15" and 17" PowerBooks? A jump from 1.25 to 1.42 GHz would be okay, but 1.33 to 1.42 would be pretty pathetic. Can the G4s go any higher and still work in laptops? It would be cool to have dual processor laptops...
Anyway, maybe IBM will make some G4-like chip that Apple could call a G4, but it would be able to scale high enough to provide a few updates. I just don't know... we'll have to see what happens.
Unlikely. Apple just dropped IBM as their supplier for portable CPUs because they can't get their 750-designs to scale up. IBM is firmly stuck in the 1Ghz range and their roadmap indicates they won't come out of it soon.....
....Ex-Motos SPS will need some time to get their act together. The organizational shakeup will need to settle, they cannot just implement new ideas without research and development....
Let's see, IBM just introduced the 970 on 0.13µm., they have the Power 4 and soon the Power 5, the G3 has been on a 0.13µm process for over a year, they have been manufacturing a dual core processor with unbelievable amounts of L3 cache, extremely fast flexible bus design
yet, they just can't seem to get their act together to push the G3 chip to higher GHz and add SIMD.
Methinks, it is not IBM's capability to scale or improve the design of the G3, it has everything to do with markets, product demand and very possibly with contracts/agreements between Apple and Motorola over the G4. If Motorola has "some time to get their act together" it has nothing to do with IBM's capability and in all likelihood with agreements between Apple and Motorola.
If the 74x7 series scales a few hundred more megahertz, and Apple uses higher capacity batteries, fixes the display issues, and adds the fiber-optic keyboard across the line, it would make a worthy upgrade.
I don't think the G4 can be stretched for two updates though. After the winter/spring refresh, the next line of PowerBooks will need to use G5s.
The G4 may last a bit longer if IBM's rumored 750VX turns out not to be vaporware. ThinkSecret, MacEdition, and MacRumors have all posted intriguing information about this chip over the last year. It was supposed to be built on the 90 nm process (which IBM has just gotten up and running) with a faster FSB, begin sampling before the end of 2003, and debut at 1.5 GHz with a ceiling of 2.0 GHz.
It sounds like a nice portable chip, if it exists.
I can't see myself replacing the aging Ti400 with anything less than 1.8Ghz. Some 1.3Ghz 15" with crappy battery life, yesterdays screen definition and a plainly sucking RAM architecture surely will not entice me to shell out big bucks. Esp. since Apple has completely dropped the ball on quality control this time (even worse than with the original Tibooks).
It was a good idea to buy a Centrino to tide me over until Apple has chips and the determination to produce a killer laptop again. And you better beat that 5h battery life I get on the wintel.
Seems like a long wait...
You and me both, Smircle.
See you in the line at the cash register when they are eventually released.
And before any asks: no, I am not a power user. I surf, WP, e-mail, watch DVDs, do some light PS. I have absolutely no need for a G5 PowerBook.
It seems like IBM tends to under-promise and over-deliver so I wouldn't be surprised if some interesting things will happen with the Powerbooks. There is no way the current g4 powerbooks will survive until the end the 2004 (especially with upgrades of .09 clock speed) and still have strong sales. I am sticking to my theory that the g5 powerbooks will be here sooner than we expect.
PS manufacturers have no obligation to provide exact dates or timelines of future products. Even if apple execs spend the better part of this year assuring us no new g5 powerbooks will be out soon, they can easily say "Well or manufacturing is seriously ahead of schedule and we managed to develop cool-running chips for our powerbooks and they will be available in 2 weeks." That the nature of buying a computer: it is always hit or miss if you buy before the new/cooler/sleeker/faster one comes out. You just hedge your bets and getting the one that suits your needs now and for the foreseeable future. If nothing meets your requirements you wait it out or find someone else to develop it for you (if you have lots of money)
But it might be kind of interesting to sue computer-makers if you bought one right before the new one comes out, but that is just buyer's remorse.
... I can't see myself replacing the aging Ti400 with anything less than 1.8Ghz. ...
I have a Ti500, and I can't get excited over the current Powerbooks either. OK, the 17" maybe, if it was like $1000 cheaper. There just isn't enough improvement in */$ to justify purchasing it.
I am toying with the idea of the 933MHz iBook G4, however...
I'd say you can count on a G5 PowerBook by sometime in late '04. When's that jump to 3Ghz projected again? And for that matter, what about the the process shrink down to .09u ??? The two might go together, but not neccessarily.
I think Apple just might supply a 17" G5 powerbook a bit sooner than a 15" and certainly a 12" model. Getting "The world's first 64 bit supercomputer in your lap." will be a great coup for Apple and just the kind of stage show trick that Jobs lives for.
I also wouldn't count moto out, even now, and as tempting as that may be. Just a note, as bad as moto may have been for Apple in the past, now that IBM has restored workstation class cred to macland, it is Moto that supplies 4/5ths of all the product lines, and 75%+ of all the units Apple ships. The second tier isn't exactly being cruel to them nw, is it?
That brings up the question. If low power embedded is a core strength of Moto, you might still see something surprising from them, maybe even something 64 bit. Crolles is just starting out, and while SPS may on the block, it can quickly come off it as soon as the Moto-SMC partnership brings in any sort of marketable new parts. Moto might just come back sooner than you think. A FSB adjustment and some fabrication know how in a new clean facility suddenly makes the G4 viable all over again. Hell, when you think about it, moto's position at Apple has only improved since the intro of the G5. They've never before sold so many parts to Apple.
The G3, effectively a G4 with an inferior bus and no SIMD unit, probably has all the same problems a G4 has when it comes to scaling. It is extremely likely that IBM CAN NOT (and NEVER could) push it any better than moto could push the G4. The G5 is a clean sheet design that chose to scale down an industrial grade chip rather than scale up an embedded design -- read, they stripped down a POWER CPU rather than beef up a PPC. Do you think IBM knew something we didn't 'bout just how far the G3/4 class PPC could be pushed at reasonable cost and for reasonable results? I'm thinking they did, they studied, and decided it was better and cheaper to start again with a POWER based design. The G3 is dead, and I suspect, as far as IBM is concerned, so is any form of PPC-32. IBM realized that it was a dead end long before anyone here (up to 2 years ago)
In a nutshell that means that ALL the so called "info" that had been "widely reported" was just rumor-net grade bullshit. Either desert-tech (TM) G3 phantasy chips NEVER existed, and/or they were never quashed for anything other than purely technical reasons, and NOT AIM politics, as was so often erroneously assumed here.
An-u-way, for now, the G4 (especially at .13u) is not a bad laptop chip, it IS competitive with what Centrino offers on the overall balance of things. Laptops are one area where a tiny bit extra spent on quality goes a long way. I can't speak to the new 15" Al issues, but I've never had a problem with my original 867 PB12. The new PB's are even better. Perhaps they're not so drastically better than PC notebooks as Apple machines have sometimes been in the past, but neither are they worse.
The weight and sturdiness are class leading. The CPU power is competitive with true "mobile" machines. The screen resolutions actually makes sense (resolution crazed spec whores can go suck a fat one, Apple is right on this and you're acting like a bunch of retards) and the finishing touches are great -- great style, great I/O, keyboards, blinking light battery indicators, amazing sleep and dual head modes (that work just as you left them each and every time you plug in, without re-adjusting your settings each time) etc etc...
Great contribution to the thread Barto. Please explain to me how I was trolling.
I agree 110% that the G5 in it's current form can't be put into a powerbook. I did't say that I hope a 9 lb. G5 powerbook comes out tomorrow with 40 min battery life, just so we can have a G5 powerbook. I don't want the powerbook to lose it's current design (thin/light/battery life) anymore than anyone else. Of course it is going take some design wizadry and a lower power PPC9XX to get the job done. As I stated in my last post, it surely won't happen until at least the 970 hits its next die shrink.
Wanting one to happen soon doesn't make one a troll or a whiner, and telling people "if you want more power to buy a desktop" doesn't always work either. Yes there are people that want a more powerful powerbook just for the sake of wanting it, but there are also many people that make a living with a portable that need and can use more power. A G5 powerbook isn't just about a few more ticks on the clock cycle, it's the bandwidth baby. The G4 is a great little processor, it's just that is has a pathetic FSB.
Great contribution to the thread Barto. Please explain to me how I was trolling.
...
I seriously doubt Barto was calling you a troll. He was telling you not to feed the trolls... ie, don't even bother responding to geekmeet ( *cough* troll *cough*).
I seriously doubt Barto was calling you a troll. He was telling you not to feed the trolls... ie, don't even bother responding to geekmeet ( *cough* troll *cough*).
That's all.
In that case, my apologies to Barto. I didn't think of that. \
Mot SPS is widely considered to be in a better position for being spun off. Even the people remaining to it are talented, the head honcho is well regarded, and they'll be free of the stifling incompetence of Mot bureaucracy. The only cloud on the horizon is the amount of debt that bureaucracy decides to saddle an independent SPS with. Mot has a lot to get rid of.
Now, where's the G4 going? Well, as has been mentioned, Crolles is due to come on line in a couple of months. The G4's basic execution core is solid, and its AltiVec unit is almost perfect. A die shrink to 90nm - whether or not Mot gets its "black diamond" process going on that node - plus a graft of the on-die memory controller they're now shipping with the 85xx series, and you have a really nice mobile CPU. Take advantage of the die shrink to double up the FPU (the G4's FPU is very nice, it's just that there's only one of it) and you have something that will be able to defend its territory against 970 derivatives very effectively.
Too bad Apple can't claim to have the worlds first 64-bit notebook when the G5 PowerBook is finally announced. I saw a review of an Athlon64 notebook on AnandTech. Granted, the form factor sucks, but I really hope we see a G5 PowerBook before MWSF05. Competition is heating up on the PC side.
Amorph, I'm a little weak on the 85XX knowledge. How fast is its connecting memory bus speed. Faster than 167Mhz I hope. If Crolles does indeed come online in a few months, when could we expect to see the first fruits of their labor?
Will Mot SPS dedicate itself to the embedded market or would they try to put a bit more focus on the "desktop" arena?
It's going to take a while to get used to people speaking well of Mot again...
Amorph, I'm a little weak on the 85XX knowledge. How fast is its connecting memory bus speed. Faster than 167Mhz I hope.
It's RapidIO, whose bandwidth is comparable to HyperTransport. It's packet based and the channels can be variable in width, so clock speed isn't really a useful measure (any more than you'd measure your network bandwidth with a clock) but in terms of how much data it can move it smokes the current G4's bus. It wouldn't quite be the 970's bus, but it would also be cool and inexpensive enough to thrive in notebooks and iMacs.
Quote:
If Crolles does indeed come online in a few months, when could we expect to see the first fruits of their labor?
I honestly don't know. Almost nobody is crazy enough to break in a brand new process - let alone a brand new fab - by making something as complex as a CPU (I say almost nobody because Motorola was, in fact, crazy enough to do that once). So I wouldn't look for something like a G4 until the second quarter (optimistically) or the second half of '04. They'll break Crolles in with stuff like flash memory that's much easier to make.
Quote:
Will Mot SPS dedicate itself to the embedded market or would they try to put a bit more focus on the "desktop" arena?
Embedded is where their strength is, and the chip I'm speculating about would be aimed squarely at the embedded market. The design would be significantly different than a 970-class CPU, with different tradeoffs and different goals. But a notebook counts as an embedded application. So does, say, a blade server.
Quote:
It's going to take a while to get used to people speaking well of Mot again...
That's because we won't be speaking of Mot. Time will tell what SPS decides to call itself as an autonomous entity, but I don't think we'll have too much trouble getting used to speaking well of them, if everything goes well.
An-u-way, for now, the G4 (especially at .13u) is not a bad laptop chip, it IS competitive with what Centrino offers on the overall balance of things. Laptops are one area where a tiny bit extra spent on quality goes a long way. I can't speak to the new 15" Al issues, but I've never had a problem with my original 867 PB12. The new PB's are even better. Perhaps they're not so drastically better than PC notebooks as Apple machines have sometimes been in the past, but neither are they worse.
The weight and sturdiness are class leading. The CPU power is competitive with true "mobile" machines. The screen resolutions actually makes sense (resolution crazed spec whores can go suck a fat one, Apple is right on this and you're acting like a bunch of retards) and the finishing touches are great -- great style, great I/O, keyboards, blinking light battery indicators, amazing sleep and dual head modes (that work just as you left them each and every time you plug in, without re-adjusting your settings each time) etc etc...
While I certainly agree about the style and the finishing touches, I'm not so sure about the «G4 is competitive to Centrino» bit. I just read a review of the admittedly sensationally ugly Benq Joybook 5000, and compared it to the PowerBook 17" Rev B review in the same mag.
The Wintel thingy has a 1.3 GHz Pentium M, runs for 3hrs 30mins watching a DVD (PowerBook: 2hrs 30mins), 5hrs 30mins with very light CPU load, has a brighter screen, and makes absolutely no noise as the fan only comes on in 3D gaming. The Radeon 9000 achieves 145 frames/sec in Quake III (vs. the PBs 99); according to c't, it would go up to 200 fps if the Radeon 9000 had 64MB RAM and therefore full bus width, and I don't even want to know how fast it would be with the Radeon that is built into the PB. It's also got all the ports the PB has except for BlueTooth.
It costs about $150 more that the 12" iBook G4.
The Centrino competition is about to really heat up, I think, and Apple's current PowerBooks, while still classes better design-wise and detail-wise, will look more and more outdated when comparing CPU speed and battery life. I really hope the PB G5 is not a year away.
Comments
Originally posted by Chagi
Maybe some sort of higher clocked IBM "G4" chip? I think we've all realized how much IBM kicks Moto's ass by now, might there be some further love coming soon from an unexpected angle?
Unlikely. Apple just dropped IBM as their supplier for portable CPUs because they can't get their 750-designs to scale up. IBM is firmly stuck in the 1Ghz range and their roadmap indicates they won't come out of it soon. And, before you ask, it isn't because Apple had to keep a difference between the pro line and the 750-based consumer line.
Ex-Motos SPS will need some time to get their act together. The organizational shakeup will need to settle, they cannot just implement new ideas without research and development.
I see a nice drought coming for their mobile line - precisely at a time when intel is likely to push their Centrino designs to 2Ghz and beyond.
Originally posted by jade
or the g5 powerbooks are right around the corner and they just want to keep g4 powerbook demand up. Does anyone find it fishy that in recent months apple execs have gone out of their way to say there will be no g5 powerbooks when apple doesn't typically comment on future hardware?????
I am not acquainted with the details of the US law system, but publicly declaring a G5 design is not feasable right now and then selling the exact same thing in 3 month time sounds like an invitation to get sued.
Apple was indeed grasping the straw giving a damn about what would happen with their mobile line.
Just replace "603e" with "G4," "604e" with "G5," and "G3" with "G6" in those sentences and you have the current situation (maybe). Laptops are using G4s because the G5 is too hot to work in a laptop. A new processor, the G6, is in development and it'll be out in late 2004 (and IBM is on time with their road maps unlike Moto). It'll be released for PowerBooks while the PowerMacs keep the G5. Actually I don't know really anything about the G6/980, but if it's intended for mobile use, they could give it a new naming scheme like "M6" so there aren't G6 PowerBooks at the same time as G5 PowerMacs. If it exceeds the performance of the G5 while still consuming little enough power to work in a laptop, it could be used in both PowerMacs and PowerBooks. I mean, if it takes the same amount of time to release a G6 processor as it does to release a mobile G5, why not go with the G6 and get a product line just as awesome as the late-1997 PowerMac G3/PowerBook G3 release?
Hopefully by that time, IBM will have come up with a newer, better low end processor incorporating AltiVec to be used in iBooks and iMacs. Apple could go all-IBM in late 2004, with new lower-end IBM processors (maybe like G4s, but scaling higher and with a better bus architecture) and G6s for PowerMacs and PowerBooks.
It seems kind of crazy, but it just might work. It seems plausible to me at least. I have no idea how the PowerBooks will continue for a whole extra year using the G4, and I'm afraid of what it'll do to the iBook and iMac. It seems like Moto has hit their second roadblock... first 500 MHz, now 1.4 GHz. It might be okay to bring the 12" PowerBook to 1.25 GHz and the iBooks to 1.0-1.25 GHz, but what about the 15" and 17" PowerBooks? A jump from 1.25 to 1.42 GHz would be okay, but 1.33 to 1.42 would be pretty pathetic. Can the G4s go any higher and still work in laptops? It would be cool to have dual processor laptops...
Anyway, maybe IBM will make some G4-like chip that Apple could call a G4, but it would be able to scale high enough to provide a few updates. I just don't know... we'll have to see what happens.
Originally posted by Smircle
Unlikely. Apple just dropped IBM as their supplier for portable CPUs because they can't get their 750-designs to scale up. IBM is firmly stuck in the 1Ghz range and their roadmap indicates they won't come out of it soon.....
....Ex-Motos SPS will need some time to get their act together. The organizational shakeup will need to settle, they cannot just implement new ideas without research and development....
Let's see, IBM just introduced the 970 on 0.13µm., they have the Power 4 and soon the Power 5, the G3 has been on a 0.13µm process for over a year, they have been manufacturing a dual core processor with unbelievable amounts of L3 cache, extremely fast flexible bus design
yet, they just can't seem to get their act together to push the G3 chip to higher GHz and add SIMD.
Methinks, it is not IBM's capability to scale or improve the design of the G3, it has everything to do with markets, product demand and very possibly with contracts/agreements between Apple and Motorola over the G4. If Motorola has "some time to get their act together" it has nothing to do with IBM's capability and in all likelihood with agreements between Apple and Motorola.
Originally posted by Kecksy
If the 74x7 series scales a few hundred more megahertz, and Apple uses higher capacity batteries, fixes the display issues, and adds the fiber-optic keyboard across the line, it would make a worthy upgrade.
I don't think the G4 can be stretched for two updates though. After the winter/spring refresh, the next line of PowerBooks will need to use G5s.
The G4 may last a bit longer if IBM's rumored 750VX turns out not to be vaporware. ThinkSecret, MacEdition, and MacRumors have all posted intriguing information about this chip over the last year. It was supposed to be built on the 90 nm process (which IBM has just gotten up and running) with a faster FSB, begin sampling before the end of 2003, and debut at 1.5 GHz with a ceiling of 2.0 GHz.
It sounds like a nice portable chip, if it exists.
Originally posted by Smircle
Damn
I can't see myself replacing the aging Ti400 with anything less than 1.8Ghz. Some 1.3Ghz 15" with crappy battery life, yesterdays screen definition and a plainly sucking RAM architecture surely will not entice me to shell out big bucks. Esp. since Apple has completely dropped the ball on quality control this time (even worse than with the original Tibooks).
It was a good idea to buy a Centrino to tide me over until Apple has chips and the determination to produce a killer laptop again. And you better beat that 5h battery life I get on the wintel.
Seems like a long wait...
You and me both, Smircle.
See you in the line at the cash register when they are eventually released.
And before any asks: no, I am not a power user. I surf, WP, e-mail, watch DVDs, do some light PS. I have absolutely no need for a G5 PowerBook.
But I want one.
PS manufacturers have no obligation to provide exact dates or timelines of future products. Even if apple execs spend the better part of this year assuring us no new g5 powerbooks will be out soon, they can easily say "Well or manufacturing is seriously ahead of schedule and we managed to develop cool-running chips for our powerbooks and they will be available in 2 weeks." That the nature of buying a computer: it is always hit or miss if you buy before the new/cooler/sleeker/faster one comes out. You just hedge your bets and getting the one that suits your needs now and for the foreseeable future. If nothing meets your requirements you wait it out or find someone else to develop it for you (if you have lots of money)
But it might be kind of interesting to sue computer-makers if you bought one right before the new one comes out, but that is just buyer's remorse.
Originally posted by Smircle
... I can't see myself replacing the aging Ti400 with anything less than 1.8Ghz. ...
I have a Ti500, and I can't get excited over the current Powerbooks either. OK, the 17" maybe, if it was like $1000 cheaper. There just isn't enough improvement in */$ to justify purchasing it.
I am toying with the idea of the 933MHz iBook G4, however...
I think Apple just might supply a 17" G5 powerbook a bit sooner than a 15" and certainly a 12" model. Getting "The world's first 64 bit supercomputer in your lap." will be a great coup for Apple and just the kind of stage show trick that Jobs lives for.
I also wouldn't count moto out, even now, and as tempting as that may be. Just a note, as bad as moto may have been for Apple in the past, now that IBM has restored workstation class cred to macland, it is Moto that supplies 4/5ths of all the product lines, and 75%+ of all the units Apple ships. The second tier isn't exactly being cruel to them nw, is it?
That brings up the question. If low power embedded is a core strength of Moto, you might still see something surprising from them, maybe even something 64 bit. Crolles is just starting out, and while SPS may on the block, it can quickly come off it as soon as the Moto-SMC partnership brings in any sort of marketable new parts. Moto might just come back sooner than you think. A FSB adjustment and some fabrication know how in a new clean facility suddenly makes the G4 viable all over again. Hell, when you think about it, moto's position at Apple has only improved since the intro of the G5. They've never before sold so many parts to Apple.
The G3, effectively a G4 with an inferior bus and no SIMD unit, probably has all the same problems a G4 has when it comes to scaling. It is extremely likely that IBM CAN NOT (and NEVER could) push it any better than moto could push the G4. The G5 is a clean sheet design that chose to scale down an industrial grade chip rather than scale up an embedded design -- read, they stripped down a POWER CPU rather than beef up a PPC. Do you think IBM knew something we didn't 'bout just how far the G3/4 class PPC could be pushed at reasonable cost and for reasonable results? I'm thinking they did, they studied, and decided it was better and cheaper to start again with a POWER based design. The G3 is dead, and I suspect, as far as IBM is concerned, so is any form of PPC-32. IBM realized that it was a dead end long before anyone here (up to 2 years ago)
In a nutshell that means that ALL the so called "info" that had been "widely reported" was just rumor-net grade bullshit. Either desert-tech (TM) G3 phantasy chips NEVER existed, and/or they were never quashed for anything other than purely technical reasons, and NOT AIM politics, as was so often erroneously assumed here.
An-u-way, for now, the G4 (especially at .13u) is not a bad laptop chip, it IS competitive with what Centrino offers on the overall balance of things. Laptops are one area where a tiny bit extra spent on quality goes a long way. I can't speak to the new 15" Al issues, but I've never had a problem with my original 867 PB12. The new PB's are even better. Perhaps they're not so drastically better than PC notebooks as Apple machines have sometimes been in the past, but neither are they worse.
The weight and sturdiness are class leading. The CPU power is competitive with true "mobile" machines. The screen resolutions actually makes sense (resolution crazed spec whores can go suck a fat one, Apple is right on this and you're acting like a bunch of retards) and the finishing touches are great -- great style, great I/O, keyboards, blinking light battery indicators, amazing sleep and dual head modes (that work just as you left them each and every time you plug in, without re-adjusting your settings each time) etc etc...
I agree 110% that the G5 in it's current form can't be put into a powerbook. I did't say that I hope a 9 lb. G5 powerbook comes out tomorrow with 40 min battery life, just so we can have a G5 powerbook. I don't want the powerbook to lose it's current design (thin/light/battery life) anymore than anyone else. Of course it is going take some design wizadry and a lower power PPC9XX to get the job done. As I stated in my last post, it surely won't happen until at least the 970 hits its next die shrink.
Wanting one to happen soon doesn't make one a troll or a whiner, and telling people "if you want more power to buy a desktop" doesn't always work either. Yes there are people that want a more powerful powerbook just for the sake of wanting it, but there are also many people that make a living with a portable that need and can use more power. A G5 powerbook isn't just about a few more ticks on the clock cycle, it's the bandwidth baby. The G4 is a great little processor, it's just that is has a pathetic FSB.
Originally posted by opuscroakus
Great contribution to the thread Barto. Please explain to me how I was trolling.
...
I seriously doubt Barto was calling you a troll. He was telling you not to feed the trolls... ie, don't even bother responding to geekmeet ( *cough* troll *cough*).
That's all.
Originally posted by bauman
I seriously doubt Barto was calling you a troll. He was telling you not to feed the trolls... ie, don't even bother responding to geekmeet ( *cough* troll *cough*).
That's all.
In that case, my apologies to Barto. I didn't think of that.
Now, where's the G4 going? Well, as has been mentioned, Crolles is due to come on line in a couple of months. The G4's basic execution core is solid, and its AltiVec unit is almost perfect. A die shrink to 90nm - whether or not Mot gets its "black diamond" process going on that node - plus a graft of the on-die memory controller they're now shipping with the 85xx series, and you have a really nice mobile CPU. Take advantage of the die shrink to double up the FPU (the G4's FPU is very nice, it's just that there's only one of it) and you have something that will be able to defend its territory against 970 derivatives very effectively.
after all he has been wrong far more than me!
please.....just kill this whole g5 in a powerbook thing.
calling me names wont make those g5 powerbooks appear any faster.
Will Mot SPS dedicate itself to the embedded market or would they try to put a bit more focus on the "desktop" arena?
It's going to take a while to get used to people speaking well of Mot again...
Originally posted by opuscroakus
Amorph, I'm a little weak on the 85XX knowledge. How fast is its connecting memory bus speed. Faster than 167Mhz I hope.
It's RapidIO, whose bandwidth is comparable to HyperTransport. It's packet based and the channels can be variable in width, so clock speed isn't really a useful measure (any more than you'd measure your network bandwidth with a clock) but in terms of how much data it can move it smokes the current G4's bus. It wouldn't quite be the 970's bus, but it would also be cool and inexpensive enough to thrive in notebooks and iMacs.
If Crolles does indeed come online in a few months, when could we expect to see the first fruits of their labor?
I honestly don't know. Almost nobody is crazy enough to break in a brand new process - let alone a brand new fab - by making something as complex as a CPU (I say almost nobody because Motorola was, in fact, crazy enough to do that once). So I wouldn't look for something like a G4 until the second quarter (optimistically) or the second half of '04. They'll break Crolles in with stuff like flash memory that's much easier to make.
Will Mot SPS dedicate itself to the embedded market or would they try to put a bit more focus on the "desktop" arena?
Embedded is where their strength is, and the chip I'm speculating about would be aimed squarely at the embedded market. The design would be significantly different than a 970-class CPU, with different tradeoffs and different goals. But a notebook counts as an embedded application. So does, say, a blade server.
It's going to take a while to get used to people speaking well of Mot again...
That's because we won't be speaking of Mot.
Originally posted by opuscroakus
In that case, my apologies to Barto. I didn't think of that.
Hehe... no offense taken
(at least you got my point in the end
Originally posted by Matsu
An-u-way, for now, the G4 (especially at .13u) is not a bad laptop chip, it IS competitive with what Centrino offers on the overall balance of things. Laptops are one area where a tiny bit extra spent on quality goes a long way. I can't speak to the new 15" Al issues, but I've never had a problem with my original 867 PB12. The new PB's are even better. Perhaps they're not so drastically better than PC notebooks as Apple machines have sometimes been in the past, but neither are they worse.
The weight and sturdiness are class leading. The CPU power is competitive with true "mobile" machines. The screen resolutions actually makes sense (resolution crazed spec whores can go suck a fat one, Apple is right on this and you're acting like a bunch of retards) and the finishing touches are great -- great style, great I/O, keyboards, blinking light battery indicators, amazing sleep and dual head modes (that work just as you left them each and every time you plug in, without re-adjusting your settings each time) etc etc...
While I certainly agree about the style and the finishing touches, I'm not so sure about the «G4 is competitive to Centrino» bit. I just read a review of the admittedly sensationally ugly Benq Joybook 5000, and compared it to the PowerBook 17" Rev B review in the same mag.
The Wintel thingy has a 1.3 GHz Pentium M, runs for 3hrs 30mins watching a DVD (PowerBook: 2hrs 30mins), 5hrs 30mins with very light CPU load, has a brighter screen, and makes absolutely no noise as the fan only comes on in 3D gaming. The Radeon 9000 achieves 145 frames/sec in Quake III (vs. the PBs 99); according to c't, it would go up to 200 fps if the Radeon 9000 had 64MB RAM and therefore full bus width, and I don't even want to know how fast it would be with the Radeon that is built into the PB. It's also got all the ports the PB has except for BlueTooth.
It costs about $150 more that the 12" iBook G4.
The Centrino competition is about to really heat up, I think, and Apple's current PowerBooks, while still classes better design-wise and detail-wise, will look more and more outdated when comparing CPU speed and battery life. I really hope the PB G5 is not a year away.