Hmm, I've heard this before, but don't buy it.
Originally posted by MacBandit
Well for one the G3 in the iBook actually outperforms a G4 at the same MHz at CPU intensive tasks which UT was and I assume UT2003 is also because of the poor frame rates. Also I thin you're correct that the GeForce 440 is not on the same level as the ATI in the iBook.
It is incorrect though to blame all poor frame rates on NVidia cards as one of the fastest cards available for the Mac is the NVidia Ge4Ti. It's just that of all the graphics cards that Apple has used they have for the most part used slower NVidia cards as compared to the ATI cards they have used.
Well, I'm not blaming it all on the Nvidia cards, but browse through every page of this discussion, and you'll see a correlation. Note that I did not say
cause.
I take issue with this idea that the G3 800 is better at cpu-intensive stuff than the G4 867. The benchmarks at barefeats, macspeedzone, et al, do not show this. It shows the two (ibook and 12" pbook) as neck in neck on many tasks, with ibook slightly ahead on some, and 12" pbook on others. It is "comparable" on non-altivec cpu, single-task stuff.
After looking at many benchmarks and reader reports (here and at img and xlr8) I think it's fair to narrow the reasons for the 12"'s underperformance to: a) cache size and b) graphics card.
*I can't remember who it was, but someone with a Radeon 7500 tibook 667mhz was pulling really good frame rates for a system that's supposed to be under spec. That's a G4 that's slower than the ibook, with a boost in terms of having L3, and exactly the same graphics card as the ibook. Now, I wouldn't this say that this represents a g3 outperforming a g4, would you?
BTW, I encourage everyone to read up on how to run the UT benchmark, over at
www.xlr8yourmac.com and contribute your results to their database.