Boing.
I realise it is difficult for them since there isn't anything else out yet to try out PCI-x, but assuming the AGP slot could keep up with the data transfer then PCI-x isn't going to show any real difference.
All they are really testing here is GPU performance, and the differences are so small they can easily be put down to other things going on on the pc, such as moving the mouse or the joystick. That's why some "benchmarks" show AGP slightly faster, others show PCI-x slightly ahead.
You can only test the speed of bus interfaces if you are going to saturate the bus, clearly this is not happening.
.Political correctness is based on the principle that it's possible to pick up a turd by the clean end.
If theres not much difference why the hell are they gona make new chipsets and stuff for it..and make new cards for it..?
[SIZE=1]AMD 4200 X2 @ 2.65Ghz, ASRock 939-VSTA
1.75GB PC3200, 2 X 160GB Seagate w/ 8MB Buffer
HIS Radeon X800 Pro, Antec Super Lanboy Aluminum
I'm saying that the stuff they are using to test with isn't saturating the AGP bus, but they may be pretty close.
Newer software/games may well reach saturation.
Also, don't forget that other devices will be using the PCI-x busses so there is much more potential with PCI-x than with AGP (which is really just high speed PCI).
.Political correctness is based on the principle that it's possible to pick up a turd by the clean end.
Bookmarks