Quake 3 Benchmark Setup
Resolutions:
800x600, 1024x768, 1600x1200
Graphic mode: HIGH QUALITY
GL Driver: DEFAULT
GL extensions: ON
Color : 16 Bit
Full screen: ON
Lighting: LIGHTMAP
Geometric detail: HIGH
Texture detail: 4
Texture quality: 32 BIT
Texture Filter: BILINEAR |
Game Options
Marks on walls: NO
Ejecting brass: NO
Dynamic lighting: NO
Identify target: YES
High quality sky: YES |
|
16 Bit Color |
|
Benchmark Type |
800x600 |
1024x768 |
1600x1200 |
| Q3 Demo0001 No
FSAA |
109.5 |
99.0 |
42.5 |
| Q3 Demo0002 No
FSAA |
108.0 |
101.5 |
43.0 |
| Q3 Demo0001 2X
FSAA |
83.3 |
57.5 |
Not run |
| Q3 Demo0002 2X
FSAA |
84.5 |
58.6 |
Not run |
| Q3 Demo0001 4X FSAA |
43.1 |
26.9 |
Not run |
| Q3 Demo0002 4X FSAA |
43.5 |
27.2 |
Not run |
|
32 Bit Color |
|
Benchmark Type |
800x600 |
1024x768 |
1600x1200 |
| Q3 Demo0001 No
FSAA |
99.9 |
70.3 |
25.6 |
| Q3 Demo0002 No
FSAA |
101.8 |
72.5 |
26.3 |
| Q3 Demo0001 2X
FSAA |
47.1 |
33.7 |
Not run |
| Q3 Demo0002 2X
FSAA |
48.7 |
34.7 |
Not run |
| Q3 Demo0001 4X FSAA |
Not run |
Not run |
Not run |
| Q3 Demo0002 4X FSAA |
Not run |
Not run |
Not run |
Without FSAA running, the Voodoo 5 5500
proved to be no match for a GeForce2 GTS. Its 1600x1200 score completes more
with the older GeForce 256 than the new GTS. I knew the Voodoo 5 5500 was not
going to be faster than a GeForce2 but I didn't expect a 20fps performance gap
at 1600x1200 Quake 3.
With FSAA enable, performance takes quite a
hit. At 1600x1200 it was so slow I didn't bother to let the demo finish. Then
again, if you're running your games at 1600x1200, you don't need FSAA. It would
seem the idea resolution to run would be 800x600 with 4X FSAA or 1024x768 with
2X FSAA. Both at 16 bit color.
Like the GeForece2 video cards, there is a
big drop in frame rates when going to 32 bit color at high resolutions.
800x600x32 with 4X FSAA was so slow I didn't complete a benchmark. because,
well, I got tired of waiting for it to finish. Let's just say that it's not a
playable option even with an 825Mhz P3.
Next page: FSAA in action
[Page
1]
[Page 2]
[Page
3]
[Page 4]
[Page
5]
[Page 6] [Home]
