Originally Posted by G-Omaha
Using that same "basis" would make all aspects of notebook computers bottlenecks. I guess if one is comparing apples, watermelons, and cherries, one should expect that the outcomes will be different.
Oh god, those two 7,200rpm RAID-0 HDD drives that I have on my 8790 are bottlenecks - now it's back to the desktop.
You obviously have a strong opinion.
Fact is, with all internal components staying the same
on the current high-end P4, A64 and Dothan systems, the platforms and CPUs are perfectly capable of better graphics performance with the single change being a more powerful graphics card.
As Robert602 prefers to put it...
Originally Posted by Robert602
we're really only saying the GPU is by far the most significant factor
This is independent of any comparison to desktops. It's not apples and oranges. Your sarcastic statement about the 7,200rpm RAID-0 drives seems to indicate that you're misinterpeting what we're talking about.
It's really a moot argument anyway, since the major mobile graphics companies clearly acknowledge the discrepancy between current mobile GPU performance and the rest of the system
, even if you don't. They've been working towards chips that greatly narrow that gap for some time. There were articles pointing to this trend last year, so they must have been working on the issue even longer.
The chip makers aren't thinking in terms of the incremental improvements of the 7500 > 9000/9200 > 9600 > 9700. They're working on possible 12-16 pipeline designs based on the current high-end desktop GPUs with 2-3 times the power requirements of current mobile "performance" GPUs. Will they match desktop GPU performance? No, but they will bring an unprecedented level of performance to high-end notebooks and narrow the gap.
The size and power requirements of these chips mean they won't work for smaller, battery conscious P-M notebook designs, so there will be a second line of low-power, mid-range performance chips which continue the 9600 > 9700 progression (likely pared down versions of the more powerful mobile chips, possibly with 8 pipelines).
If the mobile GPU isn't a bottleneck, Nvidia and ATI wouldn't be wasting their time and money developing two-tiers of performance chips. If the GPU isn't a bottleneck, why increase the performance at all if the CPU or some other area of the notebook would be holding it back?