3DMark 2001SE

Because of its synthetic tests, 3DMark 2001 ends up being a good candidate for measuring the low-level benefits of increasing the core and memory frequency of a graphics card.

 
3DMark Score @ 1024x768
Multitextured Fill rate (MTexels/s)
High Polygon Count - 1 Lights (MTriangles/s)
High Polygon Count - 8 Lights (MTriangles/s)
EMBM (fps)
DOT3 (fps)
Vertex Shader (fps)
Pixel Shader (fps)
Advanced Shader (fps)
Point Sprites (MSprites/s)
Gigabyte GV-R9700Pro
13916
2556.8
69.3
14.9
183
202.8
184.7
185.1
190.2
35.3
Gigabyte GV-R9700Pro (400/674)
14518
3125.6
79.8
18.1
189.4
225.3
187.1
190.3
220.2
39.8
ATI Radeon 9700 Pro
13963
2531.7
70.0
14.9
182.7
193.0
185.4
183.0
189.4
37.0
ATI Radeon 9700 Pro (378/674)
14459
2929.3
77.5
17.4
188.5
212.8
187.0
188.3
212.6
40.5
NVIDIA GeForce4 Ti 4600
11618
2321.8
52.4
12.6
136.5
152.2
102.1
123.2
88.8
30.3

You can already see from the fill rate numbers that there's a huge improvement in fill rate caused by the R300 GPU running at 400MHz. The unfortunate problem with this is that without a similar increase in memory bandwidth, the additional fill rate will go to waste in real world situations. Improvements in vertex shader performance are not all that great, signifying limitations elsewhere.

You'll also notice that the ATI Radeon 9700 Pro that didn't overclock as high as the Gigabyte card, still managed to outperform the overclocked Gigabyte card. This goes back to the prerelease BIOS issues we mentioned earlier, and it's a trend that you'll see continue as we explore our gaming benchmarks under Unreal Tournament 2003.

The Test Unreal Tournament 2003 Performance
Comments Locked

0 Comments

View All Comments

Log in

Don't have an account? Sign up now