Transformation. All models are made out of vertices (which are just 3-dimensional points) which are ordered into polygons (such as triangles or quadrangles). These models are then placed in a virtual world on the computer. For example, if I were creating a first-person shooter, I might call the center of my world a reactor core at the center of a level, and then all the character positions would be in reference to that - for example, you could imagine a situation where the player is located 10 meters to the left of the reactor, and 5 meters down from it.
The problem with this system, though, is that you are not always drawing the world as seen by the reactor. In fact, in a first-person shooter, everything is drawn from the player's perspective. This means that we need to change our coordinates - rather than referring to everything with the reactor as the center of the world, we want to refer to everything with the player (or, in 3rd person games, the player's camera) being the center of the world! This process is known as transformation, and I cover the mathematics behind it in Part 1 of my 3D pipeline series on Gaming Future (you don't need to read that, but if you are interested in the theory behind graphics, I recommend it). However, despite transformation being a relatively straightforward process, not all systems handle it equally.
Voor normale gamers is het genoeg om te weten dat een GeForce 2 GTS een NV15 chip heeft en de XBox een NV25 .