About T&L on GeForce

Started by
-1 comments, last by Shinkage 24 years, 4 months ago
I believe NVIDIA has some slideshows put up on their web site explaining some of the details about programming for hardware T&L. A couple of the points:

Always try to use indexed vertex buffers when rendering polygons. For that matter, always use indexed rendering whenever possible.

The GeForce has a vertex cache in which a certain number of the most recently used vertices are stored (16 I think). In order to best utilize this cache, always try to reuse vertices as soon as possible. Try to order your vertices so this is accomplished.

ProcessVertices commands will not make use of hardware transformation. Try to limit the use of this as much as possible and use as low detail models for it. For example, transform bounding boxes instead of the whole model.

Hmmm, that's all I can remember from it right now. A couple notes on using it though. First, it will be automatically used whenever possible with the TNL device in Direct3D. Just set your matrices and they will be passed on to the video card for use in geometry processing.

This topic is closed to new replies.

Advertisement