I realize that that I may be beating a dead horse here, but why in god’s name does the GeForce2 take as much as a 50% performance hit when rendering in wireframe instead of polygon’s. I have an app that calls glutWireTorus a few times inside a display list. That display list is only initialized once. When I run the program, it goes about 50% faster if the tori are rendered versus when they are not. This is unacceptable. Can anyone from NVIDIA explain to me why this is so? There may be a good reason for it, but right now I’m not seeing it.
Tone
[This message has been edited by Tone (edited 03-13-2001).]
nVidia has said their consumer hardware is slower at drawing lines than at drawing textured polygons. That’s just a design decision they made. We have to live with it, or buy another card, I guess.
If you think of a line as a very thin polygon, it makes sense, because there’s two rendered triangles per line, and three lines per triangle you give to OpenGL. I don’t know if that’s how they implemented it, however.
I just found this out too … In FILL mode I get 1400 FPS, in LINE mode I get 100 FPS! In POINT mode I get 97 . These tests were on a GeForce2. On a GeForce DDR I get 1050 FPS in FILL mode, 500 FPS in LINE mode and 500 in POINT model. Go figure. My Voodoo3 runs LINE mode at 250 FPS.
This is where profesional gfx boards come in, they have a far faster geometry engine but a slower fillrate than consumer boards (except quadro’s which are basicly overpriced consumer boards, sorry NV ).