NVIDIA GeForce2 User Manual - Page 18

D graphics card achieve 3D performance? - speed

Page 18 highlights

Q3 What does the set-up engine do in a graphics controller? Answer A set-up engine allows drivers to pass triangles in the form of raw vertex information; whereas, most common designs force triangles to be preprocessed for the rendering engine in terms of delta values for edges, color, and texture. Q4 Why does a 3D graphics chip need to have both a rendering engine and a setup engine? Answer Any "3D application", a game, VRML, or modeling package, can benefit from 3D rendering. This is especially true of an application that uses texturing extensively, because texturing and texture filtering are very intensive operations at the pixel level in terms of CPU operations and demands for memory bandwidth. Without a set-up engine in a graphics controller, the CPU has to calculate the delta values for edges, color, and textures; the drivers need to handle ten (10) times more extensive data. This results in slower 3D pipeline operations between the CPU and the graphics controller. Q5 If we use powerful CPUs, such as a Pentium™ 200, can a standard 2D graphics card achieve 3D performance? Answer Yes and no. Software rendering can take advantage of "tricks" learned by force of necessity through years of trial and error. With such stratagems, the speed of software rendering for simple scenes can approach that of low-level hardware 3D rendering. On the other hand, as scenes become more complex (or frame sizes become larger), there are conflicts between using the CPU for high-level game logic, geometry, lighting, and rendering, all of which increase their demands. No current CPU or system can perform advanced quality-enhancements (bilinear filtering and alpha blending) in real time. Even general case texture mapping with RGB lighting is too much for the current CPU generation. Q6 What does "software 3D" mean? Answer 18

  • 1
  • 2
  • 3
  • 4
  • 5
  • 6
  • 7
  • 8
  • 9
  • 10
  • 11
  • 12
  • 13
  • 14
  • 15
  • 16
  • 17
  • 18
  • 19
  • 20
  • 21
  • 22
  • 23
  • 24
  • 25
  • 26
  • 27

18
Q3
What does the set-up engine do in a graphics controller?
Answer
A set-up engine allows drivers to pass triangles in the form of raw vertex
information; whereas, most common designs force triangles to be pre-
processed for the rendering engine in terms of delta values for edges,
color, and texture.
Q4
Why does a 3D graphics chip need to have both a rendering engine
and a setup engine?
Answer
Any “3D application”, a game, VRML, or modeling package, can benefit
from 3D rendering. This is especially true of an application that uses
texturing extensively, because texturing and texture filtering are very
intensive operations at the pixel level in terms of CPU operations and
demands for memory bandwidth. Without a set-up engine in a graphics
controller, the CPU has to calculate the delta values for edges, color, and
textures; the drivers need to handle ten (10) times more extensive data.
This results in slower 3D pipeline operations between the CPU and the
graphics controller.
Q5
If we use powerful CPUs, such as a Pentium™ 200, can a standard
2D graphics card achieve 3D performance?
Answer
Yes and no. Software rendering can take advantage of "tricks" learned by
force of necessity through years of trial and error. With such stratagems,
the speed of software rendering for simple scenes can approach that of
low-level hardware 3D rendering. On the other hand, as scenes become
more complex (or frame sizes become larger), there are conflicts between
using the CPU for high-level game logic, geometry, lighting, and
rendering, all of which increase their demands. No current CPU or system
can perform advanced quality-enhancements (bilinear filtering and alpha
blending) in real time. Even general case texture mapping with RGB
lighting is too much for the current CPU generation.
Q6
What does "software 3D" mean?
Answer