"Mother of all challenges" for G5 Powerbook

ibook911 · January 12, 2005 4:57PM

On the conference call today, they said a G5 Powerbook would be the mother of all challenges to cool. I really got the impression that a G5 Powerbook is not looking good anytime soon (i.e. the next year).

What do you guys think? You should listen to the last ten minutes of the conference call, and you might see what I mean.

mr. h · January 12, 2005 5:02PM

You beat me to it!

I was about to start a thread. It did not sound good. I think anyone expecting a G5 powerbook this year is going to be very disappointed

Or he was bluffing.

Let the conspiracy theories commence!

ibook911 · January 12, 2005 5:19PM

Quote:

Originally posted by Mr. H

You beat me to it!

I was about to start a thread. It did not sound good. I think anyone expecting a G5 powerbook this year is going to be very disappointed

Or he was bluffing.

Let the conspiracy theories commence!

Yep, maybe he was bluffing. I don't know, but I think I'm going ahead and buying a Powerbook when they release a G4 update, hopefully in the coming weeks. It isn't worth waiting anymore. I can always find a way to buy a G5 Powerbook, if and when, one exists.

pb · January 12, 2005 5:32PM

Quote:

Originally posted by ibook911

Yep, maybe he was bluffing.

I don't think so. Sorry guys, I just don't see the Powerbook G5 coming out this year. Freescale offer reasonable options for at least one year from now, they even (will) offer a dual-core e600 chip capable to go in a slim Powerbook without power or heat problems. It should be ready for production in a year from now.

ompus · January 12, 2005 9:15PM

So Apple has thrown in the g5 towel for the time being. I'll run with that...

Development of the Powerbook g4 hinged on the availability of the 970. If the g5 was around the corner, Apple would cease development of the current Powerbook and milk it for profits. We now know the g5 is out of the the short to mid-term future.

Realizing that it's stuck w/ the g4, Apple will probably spend more than it wants trying to stay competitive. Sure, clockrates will creep up, but the easiest speed bump is already on the table... a move to the 200 mhz MPX bus.

Now, I can't find anything firm evidence to support the conclusion the 7447b will have a 200 Mhz bus. However, a faster bus is the only compelling reason I see for Freescale moving to the stop gap 7447b.

This is fairly wild conjecture, but I don't think it's beyond the pale. Accordingly, in the next few weeks, I see Apple bumping the Powerbooks to 1.667 Ghz w/ a 200 Mhz bus. If the chip had some powersaving features, all the better. And if Apple could cut prices, they might even be able to keep the laptop wolves at bay.

pb · January 13, 2005 4:06AM

Quote:

Originally posted by Ompus

Now, I can't find anything firm evidence to support the conclusion the 7447b will have a 200 Mhz bus. However, a faster bus is the only compelling reason I see for Freescale moving to the stop gap 7447b.

Indeed, there is no word right now about the bus speed of the 7447B. However, since Freescale is advertising the 7448 as a chip running on a 200 MHz, it seems safe to assume that the 7447B runs on a 167 MHz bus, like the 7447A. But this is nothing more than an asumption from my part.

a j stev · January 13, 2005 5:51AM

Quote:

Originally posted by PB

Indeed, there is no word right now about the bus speed of the 7447B. However, since Freescale is advertising the 7448 as a chip running on a 200 MHz, it seems safe to assume that the 7447B runs on a 167 MHz bus, like the 7447A. But this is nothing more than an asumption from my part.

I reckon your right...167Mhz bus is the most likely bus figure. The 7447B is a process improvement, not a new chip. It might get a new system controller but Freescale isn't going to know that kind of thing. That's up Apple's alley.

You'll have to wait for the 7448 for the 200Mhz stuff.

A 1.667 G4 is not too bad if coupled with faster hard-drive (SATA?) and a Mobile Radeon 9800.

ompus · January 13, 2005 7:02AM

1. Does a process improvement merit a change in name (even as slight as 7447a -> 7447b)? I would think that no matter how much the process changed, a 7447a would be a 7447a.

2. How much of an engineering challenge would it be to bump the 7447 to a 200 Mhz bus? That is, if Apple committed to the chip, could Freescale deliver it?

electric monk · January 13, 2005 9:23AM

Considering the 7447a probably didn't deserve an "a" added then yeah, the "b" moniker seems reasonable. The bus. Oh god the bus. The MaxBus was never really designed to scale. So it has been a huge pain just to climb up to 167 MHz. The 200 MHz bus alone means the 7448 deserves a number bump and as such is really not the easiest speed bump as someone suggested.

At least for Freescale this has been a huge problem (think moving beyond 3/3.2 Ghz P4's at Intel. They sat there for over a year. Imagine Intel facing the same problem plus having horrible fabs, bad management, spinning off part of the company, and a way smaller R&D budget), they're fixing it with the e600's system on a chip with a 667 MHz for the bits left over.

If it ships. Cough.

Would Apple buy a million 7448's? Yes. They can stick them in PB's, iBooks, eMacs, and the Mac Mini. They sold 600 000 (700 000?) odd of those models last quarter and the Mac Mini could easily add 300 000 more by the time they update to the 7448, counting growth for current product lines and good numbers on the Mac Mini's continuing sales Apple would probably buy a couple of million 7448's as soon as Freescale can deliver them.

The problem is that Motorola has a horrible track record. You can blame it on many things; infamously bad management (Hopefully not a problem since Freescale is now separate), dirty small fabs (solved by Crolles), and probably more stuff that I'm forgetting.

They also make some incredible designs, but producing and scaling them has always been the problem. The e600 is amazing. Flat out fantastic. An e600 is a better chip for laptops/Mac Minis by far then a G5. Now a G5 would kill current G4's at similar clock speeds due to bus and memory speed (although there is some debate). A dual-core e600 would eat a G5 for lunch at similar clock speeds. And use a heck of a lot less heat.

Before anyone mentions 64 bits, it doesn't matter. It applies only to the 4 GB limit single program addressing and a handful of other math things. And yes, all this has been said a million times. Search is your friend.

Now the 7448 samples in Jan. Assuming they don't subcontract out to IBM or someone else they'll build them at Crolles which is a state of the art fab. Also we haven't heard any 90nm horror stories from them like IBM and Intel ran into. This would seem to put them in good shape for fairly fast ramp up to production. I don't know how fast they can get them out the doors though. That is the question.

What it boils down is that Apple laptops, and incidentally the Mac Mini, are stuck with Freescale for the foreseeable future. Assuming 7448's and the e600's come out fast that's cool (The e600 also gives Apple a P-M killer). Delays and we have situations like where we are now with the PB's having small speed boosts for a year or more.

emig647 · January 13, 2005 11:05AM

I've been saying the g5 970fx is still too much to put into a book form that is less than 1.5" thick. They can clock it down, clock the fsb down, add whatever cooling they can, and it still would be too hot and hog battery like a LA crackwhore.

I personally think we're not going to see a tru 970 derivitive in a powerbook. I don't think I'd want one anyways. I know battery is going to be hit and if you've guys noticed the goal in the laptops is to get them cooler and use way less battery than normal by clocking down the VCORE.

Here's another fact, Males that use laptops on their laps have been found with lower sperm counts (that could be good or bad I suppose). But the heat generated from laptops is killing off sperm. (This article was in Wired Magazine and on the site a few months ago).

So if you want to be sterile get a g5 powerbook baby!!! hehe. Then you can get all the condom-less tail you want and the girls will never know. Just play dumb

.

Anyways, I can wait a year for dual core e600... I'd much rather have that than 1 g5 proc.

Also the argument that 64bit doesn't have a place except for over 4gb memory and a few math calcs is insane. I do a lot of inline assembly in my programs. I have been waiting for 64bit for about 5 years now. You can do more code executions if you're wise with the way you place info in the registers. Less clock cycles. It's plain and simple. I wish apple's whole lineup was 64bit so I could stick to just 64bit assembly instead of inserting checks which are a BITCH.

stoo · January 13, 2005 2:13PM

Quote:

Now a G5 would kill current G4's at similar clock speeds due to bus and memory speed (although there is some debate)

Roughly,

G4 wins on: integer, non-linear code, cache bound Altivec

G5 wins on: floating point, bandwith limited code.

Clock for clock, the G5 certainly doesn't always kill the G4.

hasapi · January 13, 2005 6:27PM

Id say Apple had every intention of getting the Powerbook G5 out by MWSF.

They just could not get the Powerbook G5 they have in pre-production to be reliable enough and within the thermal paramaters they had hoped for.

It looks as though the low power version is still just too hot for a laptop - but was cool enough for an iMac.

rhumgod · January 13, 2005 6:42PM

Too hot for a PowerBook but cool enough for an iMac for one reason - physics (well and battery life). The iMac sits vertically, the PowerBook sits horizontally.

I still find it hard to believe that the iMac can have a boatload of "desktop" components which run hotter, but the PowerBook with it's low-power, low-heat generating parts is still too hot.

I think we will see a variation of the G5 (sort of like the Pentium M in that realm) in the PowerBook this year. Call it a hunch, but Apple knows damn well that their PowerBook sales suck, and it has been 9 full months since they last updated them (with a mere speed bump and graphics refresh). Ugh.

We will see something soon, April-ish I am guessing. Whether that will be the G5 PowerBook or not is anyone's guess at this stage. A full year from the last bump is not a good thing, and Apple knows it.

hasapi · January 13, 2005 7:08PM

You would think with falling sales and the unlikelyhood of a PB G5 at MWSF they could have bumped them to 1.7G in November?.

Id say the 7447 could not get to 1.7 and the G5 wont work (yet) - so the PB is exactly where its at - stuck dead until an option becomes available.

rhumgod · January 13, 2005 7:57PM

Quote:

Originally posted by hasapi

so the PB is exactly where its at - stuck dead until an option becomes available.

Oh I think they realized this quite a while ago and worked on the G5, but when that didn't pan out (and it's clear it didn't), then they had to go with an alternate plan. What that might be remains to be seen.

nd32k3 · January 13, 2005 9:12PM

Im happy about this! lol

I just bought a iBook and that means the Powerbook G5 will be coming out just as I'm about to go into college. So that willl be an escuse to get myself it.

sunilraman · January 14, 2005 4:09AM

"Mother of all challenges" is a line that is just a ruse to deflect attention... Taiwanese contractors appear to have inadvertently leaked

iBook G5 and PowerBook G5 in 2nd quarter 2005

http://www.digitimes.com/news/a20050114A7040.html

sunilraman · January 15, 2005 8:04AM

Quote:

Originally posted by Mr. H

You beat me to it!

I was about to start a thread. It did not sound good. I think anyone expecting a G5 powerbook this year is going to be very disappointed

Or he was bluffing.

Let the conspiracy theories commence!

Oppenheimer was not bluffing.

But, there will be a G5-class powerbook, just not the current G5 somehow jammed into a portable....

The key is a Power5 variant, not Power4... The PowerBook G5mobile

MacNewsWorld, MacObserver reports just such a thing - the Power5 chip from IBM... all refer to an InfoWorld article on Power5

http://www.macnewsworld.com/story/39327.html

http://www.macobserver.com/article/2004/12/15.7.shtml

This Tom Yager guy seems to be the main cheerleader for Power5 and possible Power5 in Apple...

http://www.macworld.co.uk/news/inde...mp;NewsID=10441

programmer · January 15, 2005 1:22PM

Oddly enough, the POWER5 is a POWER4 variant. They took the basic POWER4 architecture and added some things -- power management, multi-threading hardware, etc. Is it any surprise that these technologies will be applied to the whole range of Power products? That's what power.org is all about... building an ecosystem of related products built off the same technologies. We will see those technologies in other Power products, and Apple will be able to use some (perhaps many) of them.

charless · January 15, 2005 1:24PM

Quote:

Originally posted by sunilraman

"Mother of all challenges" is a line that is just a ruse to deflect attention... Taiwanese contractors appear to have inadvertently leaked

iBook G5 and PowerBook G5 in 2nd quarter 2005

http://www.digitimes.com/news/a20050114A7040.html

Digitimes.com is the least reliable of all the sources of Mac rumors on the web. I'm not aware of a single prediction they made that actually came true, but there have been many rumors by them that have turned out to be BS.

http://www.macrumors.com/pages/2003/...17033706.shtml

What digitimes.com says is a more reliable indicator of what's not going to happen than what is.

mawsdz · January 15, 2005 1:43PM

My thought is that they will deliver with a new PB design, initially using the dual core G4 processors. After they can engineer new variants or figure out how to cool the current G5, they will come out with a rev that has the G5 in it.

what do you think?

"Mother of all challenges" for G5 Powerbook

Comments