3DMark06 Under the Magnifying Glass

Other Tests

Like its predecessors, 3DMark06 has feature tests to break out certain segments of the graphics subsystem. Five of these tests were in 3DMark05 and I will go over them briefly. The first two tests are the fill rate tests; fill rate was first introduced in 3DMark2000. The fill rate is measured in millions of texels per second. A texel is defined as a "single element in a texture" per Microsoft's Direct X specification. The theoretical calculation for the texture fill is the number of pixel pipes times the clock speed, per Nvidia. ATI, with its fragmented design, classifies this as the number of texture units times the clock speed.

The third test is called the pixel shader test. This test involves the view of a canyon wall from the game test Canyon Flight in 3DMark05, which is pivoted while the light source moves across the sky. This test score is measured in frames per second.

The fourth and fifth tests are the vertex shader tests. Test four runs by using four of the creatures from the 3DMark05 Canyon Flight game test and rotating them in space. The fifth test uses a rotating square section of ground covered in blades of grass.

Both vertex scores are measured in millions of vertices per second. Nvidia and ATI calculate the theoretical limit the same way, as the number of vertex units times the clock frequency, divided by four (as only one complete calculation is performed over four clock cycles).

Create a new thread in the US Reviews comments forum about this subject
This thread is closed for comments
No comments yet
Comment from the forums
    Your comment