PowerPC 970's at 90nm?

how many watts do the current g4s in the powerbooks put out?

During the reign of the 1Ghz Ti-Book, people citing publically available Motorola documents reported numbers of slightly below 20W @1Ghz. For the 7447/7457 @ 1.25/1.33Ghz, again using Motorola documents, slightly lower numbers where cited. But it remains disputed whether Apple wasn't or isn't using lower voltage versions, which would draw less power.
With a battery capacity of roughly 60 Wh (1Ghz Ti-Book), and screen, harddrive and GPU power consumption of 5 to 10W, and a real-life 'uptime' under heavy usage of 2h+, 20W typical processor consumption would make sense.
For the current Powerbooks with slightly reduced battery uptime (officially 4.5h compared to 5 hours for the 1Ghz Ti-Book) and a battery capacity which is roughly 70% of that of the old model, this would indicate a processor consumption of 15W or slightly less.

Sounds like they're on target, perhaps even a little ahead of schedule.

I don't know about you guys/gals, but I think this is phenominal. The G5 is already speed king for many tasks. I, for one, like hearing about IBM's constant success rather than Motorola's constant failure.

A dual 3GHz G5 would be ridicuously powerful. I played around with the 1.8 Ghz G5 at the Apple SoHo store and it was damn zippy. I've been spoiled by using Panther, however, so the speed increase was not as visible in the GUI. Panther makes the GUI on even a slow machine seem very snappy... as snapy as OS 9. No, I'm not kidding.

Imagine Panther on a dual G5 and you've got yourself an amazing machine. These are exciting times and there has never been a better time to buy IMHO.

I'm hoping Apple surprises us with a January 3GHz G5 and G5 Powerbooks. They need to keep the momentum growing and a good surprise of that magnitude along with Panther would decimate the competition for some time to come.

"with a 2GHz at only 34W, down from 57W. The 3.2GHz chip reportedly at 71W."

That's a nice drop on 2Ghz chip. However for the 3.2 Ghz chip, that's sounds a bit much.

I'm skeptical of these numbers--they sound a little "too good to be true" for my taste.

But hey, it'd be nice if it is true, and you never know. I'd love to see a DP3.0 come next year as much as the next Mac fan (though the part of me that just ordered the DP2.0 secretly wishes that it'll take longer, so I won't feel like I should've waited another four months).

I'm skeptable, just because.

It's not an unreasonable story. Chips comming off a test run will vary a lot in quality. Some will be junk, some will be slow and a very few will go very very fast. There are, afterall, a very few P4s that will OC extremely high (like close to 4GHz) without a compressor chilling system.

I wouldn't be surprised if the .09micron 970s would operate in this frequency range. I'd be surprised if Apple let the information leak so easily (unless they wanted it to leak).

As for 71W... that's less than athlons and Athlon64s. The current high end P4s put out 85watts... Prescott is rumored to be 106 watts at launch (at .09 micron).
71 watts is hot, but the cooling system that Apple has in the G5 should easily handle that. It might have to spin the fans up a bit more, but they won't have to run at 5000rpm or anything.

i really think the g5 @ 90nm can be the chip to make apple's future so so bright. look at where they are going and they just started. moto wasnt that interested, ill refrain from name calling but they got rid of another ceo and there stock looks like crap. The more we hear the brighter the 970's future looks.

oh... my... god.... IBM I LOVE YOU

Originally posted by Makosuke
I'm skeptical of these numbers--they sound a little "too good to be true" for my taste.

But hey, it'd be nice if it is true, and you never know. I'd love to see a DP3.0 come next year as much as the next Mac fan (though the part of me that just ordered the DP2.0 secretly wishes that it'll take longer, so I won't feel like I should've waited another four months).

If you think about it, Apple said 3.0 GHz within a year. That's next June. It says that volume production begins in Q2, 2004, which _ends_ on June 1st of 2004.

Sounds to me as if it's right on time. If the 3.0 chip is somewhere around 60 Watts, then a dual is not out of the question.

Then again, as you point out in your post, the facts above match up so nicely it could be a fabrication. That doesn't make a 3.0 .90nm chip in June of next year incorrect, though. It just means that this rumor could be a fabrication of known facts.