MOSR: Next Powerbook has G4 - Page 3

Well, as one who has been out of the portable Mac market since the debut of the PB 3400, I'd hope that my next PB would be a step up in technology from what's currently available.

Be it a unit powered by a new Motorola process, or a brand new IBM driven G5....I just want a great laptop that I can build web pages with, play games and run Office.


That said...if I had to choose, my decision would be based solely on who can provide the faster computing experience. Battery life is important, but most of my portable computing will be done near an outlet.

Since I have the SNDF presentation in hand, here is the relevant chart on the Moto dual core processor:



Looks like an integrated host processor (like the 8560) to me. Not sure if it would be useful for Apple.

Whether you offer a dual core CPU depends on what sort of cores you have. Intel won't, because their whole design philosophy is to make one core as powerful as possible. IBM already has, but only at the high end.

Motorola's in a different position than many of the other manufacturers, so what other manufacturers do isn't really relevant. Mot/Freescale has a cool-running, proven and powerful embedded core in the G4+. The die shrink to 90nm will officially make it small, so putting two on one die would be a simple way to boost the line's capabilities inexpensively. Plus, as mentioned above, it's something that Motorola has explicitly talked about doing.


Crolles 2 is online and operating. New fabs usually ramp up making simpler things than CPUs, so it's not making G4s yet. Mot/Freescale expect to have 90nm CPUs rolling off the lines this summer.

Amorph, Thanks for the response.

I suspect that designing a dual core processor will present unique challanges that Motorola will only discover during the development, testing and evaluation phases. Somewhat off topic, but Motorola hasn't even begun production of the MPC8540/MPC8560 processors which incorporate Rapid I/O, and these aren't desktop processors. They are so far behind in their roadmap, I don't hold much hope for any significant advances in any cpu that can be used in an Apple desktop/laptop computer for a while(2005/2006??).

IBM has a tremendous lead over Motorola in that they are and have been producing a dual core processors, even if it is a high end cpu. It seems to me that all the lessons IBM learned in advancing the technology of dual cores in the Power4 will be of immediate use to IBM in designing a desktop version. My only question is whether they will be able to do this @ 0.09µm or must wait until 0.065µm. I'd be willing to bet that this decision has already been made(possibly a long time ago - over a year maybe longer), now we await the results of this decision.

I wish Motorola good luck, but I still suspect IBM would be first to manufacture a dual core cpu for Apple.

I will be happily surprised if Motorola beats IBM producing a dual core cpu for Apple, as it would be beneficial to Apple.


Now that's good news. The G4 @ 0.09µm would be a definite winner in a laptop, especially if they can bump the FSB.

More wild speculation today from MOSR clone.

I'm rooting for a moto rebound. The G4 is still a solid portable contender. Wintelon has really only caught it and passed it with the better Centrino machines. Even there, Apple tends to spec a slightly anemic 45-55w/hour battery. If they'd just spec something in the 60-65 range, battery life would equalize nicely. Even some of the new features of this 7447A should help.

If Apple can sell me a machine with the same power as my 12" PB867, but it can run all day, and I mean really run, as in 8 hours plus of constant use, and maybe 5+ hours of intensive use, full disk access, writing CD's and DVD's, watching a film, etc etc... yet still keep it the same size and weight of my current machine, then I'm buying, and I don't care what it runs.

G5 or G4 ?

For laptops it all depends on two things:

Price, how could it not?

and

Battery life/heat.

If sometime in the summer, Apple wants to sell me a 1.5Ghz G4 iBook, I'm game for that, particularly if they increase the battery life and add a Superdrive and DVI out monitor spanning.

Yes, that's dangerously close to a PB feature set, but one would hope that the PB has become "the first 64 bit UNIX notebook" by then. And with FW800, PC cards, major graphics, more standard RAM, bigger HDDs, widescreens etc etc, the pro/wanabee pro market can still buy those.

Off topic yes, but hoping Apple does something aboutthe awful white of the iBook line. A nice silver, perhaps a gunmetal color, or even iPod mini like tints.

Whatever.

We need Moto.

12 months will pass sooner than later. The G5 is still mired at 2Ghz, Apple has added a second dual to compensate... sound familiar?

I don't understand what you mean. The G4 is trounced by Pentium-M and Pentium 4-M in raw processing power. With Altivec aware applications the situation may be different.


It is true there is a wide margin to improve battery life in the Powerbooks.


How about processor performance too?


Apparently you missed the other thread.


That's true. But I think they can hardly compete in today's notebook processor market. I would like though to be soon surprised by Moto.

The G4 was really a very good mobile chip, as was the G3 for a long time.

It's been passed by the better centrino solutions, but IMHO, P4-M was never a competitor. They were faster at the higher clock rates, yes. But apart from high end models in at least PB price territory and beyond (like Think Pads) they were also heavy, very hot, and had generally very poor battery life.

As for the 1st 64 bit thing, Apple will just keep adding adjectives untill it's true.

Ie, the first widescreen aluminium 64bit Unix supercomputer for your lap!

I'd like to see Apple sell fully featured iBooks and pBooks.

iBooks for price and superior run time, throw in superdrives and DVI-out spanning.

pBooks for power and fully loaded specification. Toss in 128MB VRAM. Add 2 independent FW800 buses for the video editors/musicians, all Pb's currently have all FW connetors sharing one bus. 512MB BASE RAM! And if they were really keen, they'd build that 512 into the Mobo like they do on the iBook/PB12 (albeit 128-256 in those cases) and have TWO EMPTY RAM SLOTS. In short, have the connectivity features that pros can really appreciate if they're looking for a portable workstation.

For something like the PB17, I'd say why not go with 3-4 RAM slots? It has the room. Basically, lets see the PB get features that would really stand out for the pro markets. BIG storage, BIG standard and max RAM capacities, BIG I/O. I/O is industry leading already, but why not turn a home run into a grand slam? Especially the FW buses. Lets have two independent FW800 buses (one of which carries a FW400 port for compatibility reasons) For A/V this would be a boon. Connect to cheap drives and a video camera/capture system simultaneously. Full speed access all the way, no bottlenecks. Excellent.

It is true that the P4-M based laptops were not a competitor with their weight, heat and poor battery life. But it is also true that their processors are today (and from some time now), MUCH more powerful than the speediest G4. At least so it seems from benchmarks available now and then.

The Centrino based notebooks addressed the weight, heat and battery life issues, while rising the performance even more in some cases. Furthermore, they often have equipment and components (optical drives, wireless, etc.) equivalent to those found in the Powerbooks and for a more attractive price, generally. So, what remains in the Powerbooks? Well, apart the exceptional hardware design, MacOS X and the bundled applications.


While this may be true, I don't think they will try to market a 64-bit Powerbook as a first to something, when in the market there are/will be at least 3-4 manufacturers of 64-bit notebook computers. Perhaps Apple will wait the 65 nm IBM chips to make the move and market the new Powerbooks as something really unique. Not just to market them as such, but since these chips will certainly allow a 64-bit processor to go into a slim or slimmer than today Powerbook.


G4 or G5, I believe this is a certain change for the next Powerbook revision. Especially now that we have this ATI Radeon Mobility 9700.


Adding more RAM slots to the 17" Powerbook seems quite a logical step, but I am no sure if it is indeed possible technically. This would be the last and ultimate step to fully exploit the 32-bit G4 processor with 4 GB of RAM in a notebook. That would very nice, and I wonder why Apple did not make already the modification in the second generation of the 17" Powerbooks.

There is something more missing: option to include a 7200 rpm hard drive.

Oh, yes, one more: improve the display quality. And I don't mean resolution, though many people would like so. I mean display quality: brightness and viewing angles. At least the viewing angles, are very poor in the Powerbooks, even in the last ones. Look for example here, where a last generation Titanium Powerbook is being compared with a Sager 5660, and you will see what I mean. And people keep saying that the new Aluminum Powerbooks have displays no better than the ones in the Titaniums (actually, the display in the 17" Powerbook is considered a little dimmer). The 12" ones are out of game of course, in the iBook territory...

Couldn't agree more.

I recently had the opportunity to compare an Apple 15" PB with several PC laptops at Best Buy.

The clarity of the PC laptop displays really stood out compared to the Apple screen.

That's my experience to, but those I've compared to was about twice as thick as the screens in the Powerbooks. That must count for something. I rather have a slim lid than a very bright screen, but that's just me.. I want my laptom to be as small as possible. If Apple could make a laptop that didn't have any optical drive, used the 1.8" drive in an iPod, a 10" display with XGA resolution, and not that bulky pastic case that the iBook have, i'd buy it in an instant!

IMHO there is a point where form doesn't get top billing over function.

The display on a Powerbook should be as clear as possible. At least as clear as similarly priced PC laptops.

The thought of not getting a PowerBook with a next-gen CPU for that long is a bit scary. But could it be true?

True. Just look at the 20" iMac display . But I think this would affect more brightness than viewing angles. Perhaps I am wrong. Anyone knows for sure?

Quite possible. Until some weeks ago, there was no roadmap fo the G4 from Motorola. This 7447A came from the nowhere. Or it looks like that, since the lack of roadmap made the G4 look like dead. Well, it appears that this was not the case. Not for now at least.

So, Apple almost certainly will put this one in the next Powerbook revision.

On the other had, Motorola publicly talked about dual core G4 for sometime this year. If Apple uses this new 7447A, and this would happen around this summer, they give Motorola the time to further enhance the G4 (for example with a move to 90 nm, which could bring significant improvements). So, Motorola would have practically one year from now to make a substantially better G4 for Apple, a successor to the 7447A. If they deliver indeed as scheduled , this gives IBM the time they need to finish a 65 nm G5 or whatever else of this size.

Three things to remember here:

1) Motorola management incompetence was killing Mot SPS (along with the rest of the company). Mot SPS is now becoming Freescale, its own company, and the guy in charge has a great deal of respect as an engineer and a manager.

2) Motorola's dismal yields are the fault of their fabs, which they couldn't afford to run clean. But Crolles 2 is the fab for 90nm and the next few process generations, and Motorola is only contributing to that. They have partners, both of whom are quite capable of running a clean, efficient fab. So one of SPS' biggest problems is overcome.

3) SPS promised a "high performance PowerPC" this year.

Now, I'm as mindful as anyone else of all the historical caveats, but at this point there are more unknowns than anything else. Freescale appears to have licked its two biggest problems, so now we'll just have to see how things pan out starting this summer.

Meanwhile, IBM quietly reduced the top clockspeed of the 750GX from 1.1GHz to 1.0GHz.

G4's future is very obscure right now. But if Motorola or Freescale manage to produce a G4 at 90 nm, with all the current G4 issues addressed, I don't see the need for a G5 in the Powerbook. Not now, nor the next one-two years.

The G4 @ 90 nm would be an excellent mobile chip. Until the Powerbooks get 3-4 memory slots and 2 GB RAM modules for notebooks become available, there is no point to put a 64-bit chip in a Powerbook, if the G4 @ 90 nm is proved to be on par with the G5 in performance at equal clock frequency (actually it is mainly the floating point performance that suffers).


Do you read anything on that?

Specifically, what problems with the current G4 could a 90nm version rectify? Heat, yes. Higher clock speeds, yes. But some work needs to be done to enhance the framework around the CPU, right? Or is the G4 of its own divination incapable of being assisted in any way beyond a certain limit? (i.e., its crippled--compared to 970--pipeline will always remain relatively so?)

An integrated memory controller would remove the FSB bottleneck, for instance.

For me this is totally unthinkable. As a recent switcher I can say that there is no way that I would switch back to Windows as long as long as it looks and works the way it does. Apple's computers are nice but if they were running Windows or Linux I would much rather have paid less for a (less classy) but cheaper IBM-clone. What made me buy a Mac was Mac OS X. I love this system and G4 Powerbook or not, I am not going back to Windows any time soon.

[Edit] Also I must admit, half the fun with the Mac is you guys. There is no place like this in the Wintel world. I'm loving it here. And you.

Just did a quick-n-dirty analysis of the life span of a PowerBook G4. The
shortest time between intro and end-of-life is just over 6 months (the 550/667 TiBook--the one I have, of course). The longest was for its successor, the 867/1GHz TiBook at just over 10 months. I didn't look at the previous generations or the iBook for more datapoints...

The average is just over 8 months -- given that, we might expect speed bumps in a couple of months or so.

Major architectural changes seem to happen every 1 1/2 to 2 1/2 years.

Given that, I'd not count on a G5 any sooner than this summer.

I am frequently wrong, however.

Haha, and the world keeps on spinning, we keep on wishing. Will Apple be delivering...

Hopefully in my lifetime we see a G5 PowerBook...especially my lifetime, this summer, right before college begins for me.

Ooooops, double post.

Anyway, I am posting here the rest. I found an interesting and, I dare say, informative SNDF presentation from Motorola. It must certainly be well known to the people here, but I would like to cite something from the end:


It is interesting that they have a roadmap showing integrated host G4 processors for this year.

Finally:


25 Watts power dissipation at 1.5 GHz AND dual core? With DDR AND RapidIO? Looks very promising. But where is this chip? Their roadmap indicates it would happen this year.

So "dual core," this means two cores running at 1.5 GHz dissipating 25W? We need some techno-meisters in here to clarify this. It could be a great CPU, but the 970FX @2 GHz uses 24.5W according to early specs. Given that's a single-core chip, though...

That's what I get from the context. Two cores @ 1.5 GHz each, for 25 W total.


The numbers are rather misleading. IBM and Motorola quote the power dissipation data in a different way. If I am not mistaken, 24.5W for the 970FX @2 GHz is typical power, susceptible to go up to 50 W (was that the exact value?). And this is only the processor without the monstrous system bus. That's why the idea of a SoC 970FX has been recently discussed in these boards.

Dual Core Moto G4s sound interesting, but do you think this makes sense for the PowerBooks? How can Apple release a dual-core G4 and then go back to a single-core G5 in only some month's time?

The Dual Core G4 (if it will ever be available) might still be a nice option for eMacs and iBooks.

I hope they would just sit tight until they can put a 970FX in it. Unless the 7447A appears in significant quantities within the next six months, I don't see (or rather, hope that it doesn't happen) another Motorola product going into the professional notebook line.

I think Smircle sums up very well the situation separating facts from speculation.


Yet, given the dark in the suitability of the 970FX for Powerbooks (well, IBM publicly said it can be used in notebooks, but the thin form factor of a Powerbook could be the source of issues), this (i.e. 7447A based Powerbook) is the most realistic scenario at this moment.

As for the adoption of a dual core G4 (or whatever else), if and when comes into the light of the day: I think that if ever Apple takes this path, that will mean that the future PPC generations would easily have low power dual core versions for notebooks and that at least one of the Powerbooks (say the 17"), will always be dual core. Just like the Power Macs (dual processor here, but dual core would be the mobile version of the dual processor concept).

wow!! except for the "A" part of the 7447, i thought we came to this conclusion in september 2003 already:
next rev. powerbooks will sport the 90nm ppc 970 OR a faster ppc7447

I guess it would be unwise to purchase a PB anytime soon then.

However, how's this scenario for the many who are sitting on the PB G5 fence?

Keep the next Rev. PB G4 based but offset the disappointment by cutting the price.
Knock off $500 across the board, then introduce a top tiered and priced G5 PB late in 2004 or early 2005.

Silly question if you ask me. I'd take a dual core G4 @ 1.5GHz over any of the current G5s. This woud especially be the case if the processor has its own memmory interface and is hitting the power points described.

Can you imagine just how responsive such a chip would make the PowerBooks or IMac replacements? If they implemented the G4's extended addressing all the better. Addressable memory would be a major concern, this is the only real advantage that the G5 has at the present.

Given the above and very low power usage I'd have to say go for it Apple. The reality of Motorola does indicate that the processor is going to be late by a year or two, so that does blunt the enthusasme for such a machine.

Dave

Personnally I don't care to much one way or the other. What I want is the best power disapation at around 1.8 GHz. I'm thinking a 1.8 GHz would give us a nice competitive machine, go with the lowest power processor in that performance range.

The fact that the 7447A has not been announced at that frequency range is a bit of a negative. I don't see a Powerbook revision running at 1.4 GHz, with either chip, as being worthwhile at all.

But that is me, and I see the current hardware as being a little long in the tooth.

Dave

Apple needs at least a 2.5GHz G4 to compete. Centrino is much faster than the G4 clock-for-clock and some say even faster than the G5 on a clock-normalized basis. Dothan, Centrino's sucessor will come this summer and have more L2 cache (2MB) and faster clockspeeds (2GHz+). The PowerBooks right now are simply pathetic.

Can those Centrino's or will that Donthan processor be able to support OS X? Nope, which means the PowerBook is not pathetic at all, it is much better. Plus, the PowerBook has incredible I/O. Just needs an update because its cycle has run its course.

I think Existence's point is that the Powerbooks suffer in the processor sector. Of course the Powerbook is not pathetic as a package (we have to not forget the bundled OS AND software), but in raw processor power it is trounced by the Pentium-M. Simply, the performance chasm between the Pentium 4 and the G4, is propagating to the mobile space. As long as the G4 stagnates, this is unavoidable.

The Powerbook needs a G4 with drastic architectural changes to keep up and stay competitive, even in the immediate future. Or a completely new processor.

Do you people ever think that sometimes updates are ready a very long time before the time that they get introduced? I think that there may be some truth to it, because people can not afford to keep buying new computers every 6 months, there are probably equations that Apple marketing people use to project when the upgrade cycle of the user is. Sometimes, I think we lose sight of the fact that just because we yearn for more speed, we may be in a minority(not that we want speed but that we would buy a new computer just after buying one within 2 years)...and we may only notice this speed issue because we are so closely involved with computing.

Apple knows when the best time to release products is, for their bottom line. There are times where they wish they could have released things more quickly, there are probably also times where they hold releases because the market is not ready. Maybe most of the market is simply not ready to handle a new influx of machines and inventory, especially since hardcore users upgraded less than a year ago to the G5.

Centrino microarchitecture


Motorola 745x microarchitecture


The G4's primary disadvantage is processor bus bandwidth and poor compilers. Even so, the G4 should be as good as the Centrino clock-for-clock. The Centrino is mostly a P6 microarchitecture (PPro, PII, PIII) with better processor bus bandwidth, better branch prediction, and excellent power usage features. Based on architecture alone, the G4 and PM should be able the same clock-for-clock.

Of course we all know what the problems are, G4s above 1.5 GHz won't be available for awhile, while PMs are available at 1.7 GHz now.


I like the form factor. But yes, in terms of performance, either Apple has to get the G4 processor bus fixed or use a G5 to play in the benchmark wars.