HSI worst case scenarios...

Mendel

Mr. Upgrade
Veteran
http://images.anandtech.com/graphs/initial pci express graphics_062004120650/2516.png

http://images.anandtech.com/graphs/initial pci express graphics_062004120650/2520.png

Pretty horrible.

edit: anand concludes "much ado about nothing" but personally I have to disagree. My conclusion: Ati does not immediately gain from PCIe, but Nvidia has a problem in some cases.

I mean: Why to upgrade to PCIe platform if there is no gain and possibly performance may even suffer


edit2: Well, since there is of course the possibility Anand might have made mistakes in measurement, I humbly request anybody with resources and time to make these benchmarking efforts again to see if we get similar results. Without more data my point seems moot. Of course there will be numerous cases where performance difference is close to none but it would be interesting to see if these particular cases can be verified.
 
Evildeus said:
Why is it funny? It's not the same map and the CPU seems to limit the FPS. The score is also ~70FPS @ 16*12.

okay then. Let's say they are completely cpu bound in that test. Then it wouldn't tell us much about the difference between agp and pcie variants of nv40, would it now?
 
Back
Top