MSI FX5950 Ultra - Page5

3DMark03

3D Mark is a widely used and accepted benchmark that stress tests the DirectX performance of a VGA card. For testing the performance of each card, we use the 4 game benchmark. The first is a DirectX 7 game, the second and the third use DirectX 8 and the last one stress tests graphics under DirectX 9. A very strong point of 3DMark is that its VGA card measurements does not require any CPU power. So the resulting fps are a good reference of a VGA card's rendering performance.

3Dmark03 also includes sound and CPU tests as well as some other feature tests. In this page we offer the average FPS for the game benchmarks, having the MSI 5950 Ultra compared with ATI's X800XT.

- Game Test 1 - Wings of Fury (DX7)

This test is a combat flight simulator written for older hardware (DirectX 7). Particles are used a lot in this test - smoke and vapor trails, flak and gunfire, and explosions are produced using point sprites and quads.

There are plenty of planes but their polygon frame rate is low, and single textured background objects are also used. Hence, the average fps (frames per second) are high for all cards.

Not bad at all. The FX5950 card can perform reasonably well with DirectX 7.0 at around 70% the performance of the ASUS AX X800.

- Game Test 2 - Battle of Proxycon (DX8)

This test is a simulation of first person shooter game types. 1.1 and 1.4 Vertex shaders are widely used since all character models are skinned using vertex shaders. This makes this test a good vertex shader comparison for VGA cards.

Vertex shaders brought the worst out of the FX5950U, which couldn't even output half the frames of the ASUS card. Let's see how it runs the DirectX 8.1 game test, Troll's Lair.

- Game Test 3 - Trolls' Lair (DX8)

This test should be the favorite of all RPG lovers. It is a cut scene of a female warrior facing two malicious trolls. Again, the same vertex and pixel processing is used as in game test 2.

This test also uses post-processing effects, such as Depth of Field and Bloom effects which are widely used in today's game cut scene sequences.

Another devastating result for the FX family. This test is also based on DirectX 8.1 code like Battle of Proxycon but a bit more GPU intensive as you can see. The X800 outperforms the FX5950 without really trying. However, keep in mind that 3DMark03 is all about GPU power. Should it have been a real life game as with the ones we'll see later on, I think MSI's gladiator wouldn't have had any trouble delivering decent framerates close to the limit your CPU can handle.

- Game Test 4 - Mother Nature (DX9)

This game test represents the level of effects and realism that are possible using 2.0 vertex and pixel shaders, plus some other features that DirectX 9 offers.

We've said before that the Radeon series is known to be slightly more capable, processing vertex and pixel shaders 2.0, according to this test.

No surprises here. The Geforce FX chipsets are well known for their dramatically low DirectX 9 performance. Let's see what the final 3DMark03 scores are.

- 3DMark Official score

If you test your machine with 3DMark, you can post the results at the 3DMark' online result browser. For more information visit futuremark.com.

To tell you the truth, I'd be ashamed to post that on the Futuremark website.
The FX5950 can hardly breathe after this struggle with the X800XT.
Is Fx5950 that bad or is the X800XT such a performance monster?
I think a bit of both would be the correct answer. Keep in mind this was an unfair comparison to start with.