Z-fill?

[Lord]

Hey guys,
while reading and trying to understand all those reviews and posts I found in here, i stumbled over a question i couldn't find an appropriate answer to: What exactly does "Z-Fillrate" mean, what does it refer to? Z stands for the height describing coordinate in three dimensional space, I guess - but I don't exactly understand how you could make a fillrate out of that? Someone enlighten me, please :)

Greetings,
Lord

[arjan de lumens]

Under normal 3d graphics rendering, a Z value is maintained for each pixel (or sample, in case of multisampling), in order to represent a depth coordinate for the pixel. This is part of a technique called "Z-buffering" (http://en.wikipedia.org/wiki/Z-buffering).

Some GPUs can process pixels much faster if you only need the Z value for the pixels you are rendering and not also the color values; the "Z-fillrate" is a measure of how fast the GPU can render pixels when only rendering Z values.

[Humus]

Z stands for the height describing coordinate in three dimensional space, I guess

In 3D graphics you're dealing with a number of different vector spaces, or coordinate systems if you will. The Z you're talking about is in World Space, which is basically just the scene put into one global coordinate system. The Z referred to in this context however is the z value after the perspective divide and viewport transformation, which is what's stored in the depth buffer. In this space X is to the right on the screen, Y is up and Z is the depth into the screen. Z-buffering is often referred to as depth-buffering, which is a better name for it.

[Lord]

Thank you, mates :)