[djjava]

You know, with all this talk about Panther and the 64bit stuff, and beingoptimized for the new processor... what kind of differnece in performance can be expected between say, a 970 based mac, and say the PB17, or PB12 running the soon-to-be-old-school G4 Processor.

[djjava]

Originally posted by djjava:
You know, with all this talk about Panther and the 64bit stuff, and beingoptimized for the new processor... what kind of differnece in performance can be expected between say, a 970 based mac, and say the PB17, or PB12 running the soon-to-be-old-school G4 Processor.

come on people....

[parsec_kadets]

Nobody responded because nobody knows.

[djjava]

Originally posted by parsec_kadets:
Nobody responded because nobody knows.

i always thought that the people in the powerbook forum would be smartest in the entire message board... oh well.

[Pierre B.]

Originally posted by djjava:
... what kind of differnece in performance can be expected between say, a 970 based mac, and say the PB17, or PB12 running the soon-to-be-old-school G4 Processor.

We cannot presently make a comparison between the G4 and the 970 running Panther. However, even the "old" G4 powerbooks will see performance gains due to general optimisations and enhancement to already present technologies.

[cambro]

First of all, I don't know what I'm talking about. That said...

The fact that the new 970 chip is 64 bit will probably mean diddly squat in terms of performance. All it really means is that you'll be able to calculate pie with greater precision (an example of course...). That is to say, 64 bit means individual chunks of data can be bigger. But, if all you care about is 3.14159, 64 bit won't help speed at all. Maybe there are some fancy programming tricks that will allow you to store two pieces of data together in a single, larger variable that can be processed only once, but I don't see how this would be possible in most instances to improve speed.

What WILL make a huge performance difference is the bus and chip architecture of the 970. All indications suggest that this will be a major improvement over the current G4 set up on all fronts. The difference should be VERY IMPRESSIVE in desktops, but not so much in portables....here's why I think this:

In a portable, I would maintain that one of the principle bottle-necks isn't the processor but the hard disk. Those things are just too slow and are often the rate limiting step in most day-to-day tasks. This is why an 800 Mhz eMac with a 100 Mhz bus often seems faster than a Ghz Ti book when working. It's the much faster hard disk that compensates. I would wager that most of the time, 667+ Mhz Tibook processors are data starved, not data saturated.

My hunch is that Panther will be faster on all machines just because of the inevitable matruation of code, and that only relatively minor 970-specific performance gains will be realized (ignoring the effect of the different bus architectures). In short, if you own a Jan. 2003 Tibook (or any other for that matter), Panther should be a fun upgrade!

[Maflynn]

Seeing that neither Panther nor the 970 are out, people can only conjecture as to the speed increase in the OS and the 970, so that question is really premature.
If you were to use the M$ philosophy
then the next iteration will be slower but Apple seems to be on their game and make speed improvements with each release of the OS.
As for optimization on the 970 and running on the g4 who knows, the only real way to find out is to wait and see.

Mike