Matthias Hopf a écrit :
On Oct 25, 06 02:07:50 +0200, Joachim Schrod wrote:
Fred A. Miller wrote:
I've also noticed that the performace of the nVidia commercial driver stinks! I have a PCI Express 16X card with the Geforce FX 5300 chipset. In the past, it's proved to be very quick.......NOT now!! Below are my results with glxgears.......this is a P-4 3.2Ghz. box.
Yes, I read that. I also double checked to make sure my user ID is listed as "video" in the user setup, and it is. I get the same results you do.....not worth noting the difference.
Here too, with a GeForce FX 5200, and a P4 3 GHz, I get roughly 1000 FPS in the small window.
In full screen mode, 1600x1200 at 24bpp, glxgears gets reduced to 55 FPS! This seems to be awfully slow, or is this normal for that card?
Well - it doesn't look *too* bad. Though if you had higher performance in fullscreen windows before, this is a regression. glxgears is actually a valid indicator for fillrate issues, if used fullscreen. Just the higher fps numbers in windowed mode don't make any sense at all.
I find that very bad, with a GeForce 7800 GTX, I get with glxgears 24bbp: -default resolution of glxgears which is about 300x300 pixels. - Full screen is 1280x1024. SuSE 10.1 64 bits Digital entry : 300x300 -> 16606 fps 1280x1024 -> 3410fps SuSE 10.1 64 bits Analog entry : 300x300 -> 16791 fps 1280x1024 -> 3383 fps SuSE 10.1 32 bits Digital entry : 300x300 -> 15710 fps 1280x1024 -> 3410fps My monitor is an DFP EIZO Flexcan L778 with digital and analog entries. Michel.