I Suspect that if Motorola could get its manufacturing act together there would be a great deal of upsid epotential in the G5. Probally scallable to 2 GHz. To be really usefull though such a revison would need its own memory port on board. Motorola has experience with such memory controllers but for some reason never pursued it with the G4.
Apple has very much hit a design wall with the G4 and the powerbooks mostly due to manufacturing issues with Motorola. If motorolas new devision is to survive they need to one of two things one would be to get out of the System processor market completely or two actually deliver a chip that is so compelling that Apple would have little resistance to its use.
As to the G5 yes heat would be an issue, much more so than many want to accept. So far the information indicates that the power figures IBM has shown so far are typical. Apple indicates that the max is quite a bit higher. I'd be surprised if Apple could deliver a G5 PowerBook with acceptable power usage and performance characteristics. Of course what is acceptable is high dependant on the user of the machine.
The big issue with the G5 is that it really isn't a stellar performer relative to the G4 at similar clock rates. If one can manage similar performance at say 1.4 GHz on both processors why not go with the lower power solution. What that solution is no one publicly has the answer to. So you really have a guestion here that asks: is there a low power solution to the G5 that would allow Apple to deliver a competitive Laptop?
The biggest issue is that Apple can't scale down any processor to far. The competition is rather intense in the battery driven market. What they really need is a 2GHz processor drawing 5 watts. Wether or not the rumored 750VX hits that point is an open question, hell we are not even sure that it exists.
AS to feasability of a 32 bit processor over a 970 based solution it really comes down to marketing and performance. There is a noteable advantage to being able to advertise a 64 bit laptop. Being able to do that does not imply however that the solution is viable. This in a nutshell is Apples problem can they put a 970 into a laptop and get results that are acceptable to their customer base. Sure they will have no problem selling to the small segment of their customer base that jumps at every new release blindly. The bigger question is would the G5 laptop perform well enough to attrack the intelligent portion of its customer base. That is an open question as their simply nothing out there to support the 970 in a laptop. Obviously Apple is working on such support now and probaly has been for some time now.
If that work of Apple's produces a Laptop with good on battery time, with a reasonable performance increase then we are all set. I loved to be proved wrong but I think Apple may be forced to go 32 bit for awhile longer. Either that or the PowerBooks become a tethered portable and the iBooks take up the slack as a battery driven laptop. Apple has everything in place to put the 970 everywhere but the laptops based on information in the public domain, laptops are a differrent issue - we really don't know.
Originally posted by shetline
Credible story or not, I find myself wondering how far past the current 1.33 GHz can you really push the G4. Apple pushed the MDD Power Macs up to 1.42 GHz, but in a big box with big heat sinks and very loud fans.
Has Motorola (our whatever their spun-off chip division is now called) done anything to advance the G4 at all, or would this just be a case of pushing the same old design harder while picking only the cream of crop from the productions yields?
Wouldn't heat and battery life issue be even worse than using G5s? I'd think a 90nm G5 running at something like 1.4 GHz would be a lot more appealing than a G4 pushed that high or higher, and probably use less power.
Unless I'm all wrong about the power issues (I've never seen specs for a 90nm G5 run at such low speeds, so I can only guess how much the power would scale down), Apple just can't get new G5 PowerBook mobo's out the door any faster than this, or 90nm G5s are simply in short supply, I can't see much reason for a stop-gap release of badly overclocked G4 PowerBooks.
Does it make sense to use some new IBM-made G4 for only one short generation of PowerBooks, or would the introduction of new IBM G4s put off the day of the G5 PowerBook even longer?