Introduction to Programmable Hardware

1
Introduction to Programmable Hardware
2
Traditional Graphics Pipeline
transform lighting
(per vertex operations)
setup rasterizer
(per primitive operation)
texture blending
(per fragment operation)
frame-buffer anti-aliasing
3
Programmable features

• Vertex Programming
• Pixel Shader
• Texture shader
• Register combiner
• Based on nVIDIA architecture

4
Vertex Program (contd)

• Vertex Programming offers programmable TL unit

User-defined Vertex Processing
transform lighting
setup rasterizer
texture blending
Gives the programmer total control of vertex
processing.
frame-buffer anti-aliasing
5
Vertex Program (contd)
Vertex Program
transform lighting
setup rasterizer
texture blending
frame-buffer anti-aliasing
6
Vertex Program (contd)

• Vertex Program
• Assembly language interface to TL unit
• GPU instruction set to perform all vertex math
• Reads an untransformed, unlit vertex
• Creates a transformed vertex
• Optionally creates
• Lights a vertex
• Creates texture coordinates
• Creates fog coordinates
• Creates point sizes

7
Create Vertex Program

• Programs (assembly) are defined inline as
• character strings

static const GLubyte vpgm \!!VP1. 0\ DP4
oHPOS.x, c0, v0 \ DP4
oHPOS.y, c1, v0 \ DP4
oHPOS.z, c2, v0 \ DP4
oHPOS.w, c3, v0 \ MOV
oCOL0,v3
\ END"
8
Programming Model
V0 V15
Vertex Source
Program Constants
c0 c96
16x4 registers
OHPOS OCOL0 OCOL1 OFOGP OPSIZ OTEX0
OTEX7
Vertex Program
96x4 registers
R0 R11
Temporary Registers
128 instructions
12x4 registers
Vertex Output
15x4 registers
All quad floats
9
Instruction Set The ops

• 17 instructions total
• MOV, MUL, ADD, MAD, DST
• DP3, DP4
• MIN, MAX, SLT, SGE
• RCP, RSQ, LOG, EXP, LIT
• ARL

10
Pixel Shader
User-defined per pixel shading
11
Texture Mapping/Blending

• Traditional OpenGL texture mapping/blending

Vertex colors
Gouraud Shading
Fragment color
Texture Coordinate
Texture Unit
Blend colors
Fragment color output
12
Multitexturing

• An optional extension of OpenGL 1.2

fragment color input
texture unit 0
blend colors
texture unit 0
blend colors
texture unit 0
blend colors
texture unit 0
blend colors
fragment color output
13
Texture Compositing

• OpenGL 1.2

Fragment Color
Texture Environment 0
Texture Fetching
Tex0
Tex1
Specular Color Sum
Fog Application
Specular Color
Fog Color/Factor
14
Compositing Operator
Choice of 5 set functions for RGB and Alpha
Ct texture color At texture alpha
Cf incoming fragment color Af incoming
fragment alpha Cc color assigned to
GL_TEXTURE_ENV_COLOR Post-environme
nt specular color addition and fog application
Function RGB Alpha
Replace Ct At
Modulate Cf Ct Af At
Decal Cf (1 At) Ct At Af
Blend Cf (1 Ct) Cc Ct Af At
Add Cf Ct Af At
15
Pixel Shader (contd)

• Based on nVIDIAs GF3/4 architecture

• Texture shader
• 4 texture units
• 23 different texture shader operations
• Conventional (1D, 2D, 3D, texture rectangle, cube
  map)
• Special case (none, pass through, cull fragment)
• Dependent texture fetches (result of one texture
  lookup affects texture coords for subsequent
  unit)
• Dependent textures fetches with dot product (and
  optional reflection) calculations
• Register combiners
• 8 stages (general combiners) on GeForce3/4
• Per-stage constants

16
Pixel Shader

• Based on nVIDIAs GF3/4 architecture
• Texture shader register combiner

texture shader
fragment color input
texture unit 0
texture program
texture unit 1
texture program
texture unit 2
texture program
texture unit 3
texture program
register combiner
fragment color output
17
Texture Shader

• Texture program example conventional 2D texture

Tex
Texture Coords (S,T,R,Q)
Shader Operations
Texture Fetch
Bound Texture Target/Format
Output Color
2D Any Format
Texture 2D
Si
Ti
(R,G,B,A)
( , )
(Si,Ti,Ri,Qi)
i
Qi
Qi
18
Texture Shader (contd)

• Texture program example pass through

19
Texture Shader (contd)

• Texture program example dependent texture

20
Register Combiner

• GeForce 2 (only 2 general combiner stages)

4 RGB Inputs
Fragment Color
4 Alpha Inputs
General Combiner 0
3 RGB Outputs
Specular Color
3 Alpha Outputs
Fog Color/Factor
4 RGB Inputs
4 Alpha Inputs
General Combiner 1
Register Set
Texture 0
Texture Fetching
3 RGB Outputs
3 Alpha Outputs
Texture 1
Spare 0
Specular Color
Final Combiner
6 RGB Inputs
1 Alpha Input
21
Register Combiner (contd)

• Register-based programming
• All textures and colors available for each and
  every texture blending stage
• 8 Stages of blending in hardware, plus specular
  and fog
• Note that GeForce3 has 8 combiners, and 4
  textures.
• Signed color arithmetic

22
Diagram of a General Combiner
Input RGB, Alpha Registers
Input Mappings
RGB Function
RGB Scale/Bias
Next Combiners RGB Registers
A
A op1 B
B
RGB Portion
C op2 D
C
AB op3 CD
D
Input Alpha, Blue Registers
Input Mappings
Alpha Function
Alpha Scale/Bias
Next Combiners Alpha Registers
A
AB
B
Alpha Portion
CD
C
AB op4 CD
D
23
General Combiner Input Registers
Input RGB, Alpha Registers
Input Mappings
RGB Function
RGB Scale/Bias
Next Combiners RGB Registers
A
A op1 B
B
RGB Portion
C op2 D
C
AB op3 CD
D
Input Alpha, Blue Registers
Input Mappings
Alpha Function
Alpha Scale/Bias
Next Combiners Alpha Registers
A
AB
B
Alpha Portion
CD
C
AB op4 CD
D
24
The Register Set

• Primary (diffuse) color
• initialized to RGBA of fragments primary color
• Secondary (specular) color
• initialized to RGB of fragments
  secondary/specular color
• alpha not initialized
• Texture 0 and Texture 1 colors
• initialized to fragments filtered RGBA texel
  from numbered texture unit
• not initialized if numbered texture unit is
  disabled or non-existent
• Spare 0 and Spare 1
• Alpha of Spare 0 is initialized to alpha of
  Texture 0 color (if enabled)
• RGB of Spare 0 and all of Spare 1 is not
  initialized
• Fog
• RGB is current fog color
• alpha is fragments fog factor (only available
  in final combiner)
• read-only
• Constant color 0 and Constant color 1
• initialized to user-defined RGBA value
• read-only
• Zero

25
General Combiner Input Mappings
Input RGB, Alpha Registers
Input Mappings
RGB Function
RGB Scale/Bias
Next Combiners RGB Registers
A
A op1 B
B
RGB Portion
C op2 D
C
AB op3 CD
D
Input Alpha, Blue Registers
Input Mappings
Alpha Function
Alpha Scale/Bias
Next Combiners Alpha Registers
A
AB
B
Alpha Portion
CD
C
AB op4 CD
D
26
General Combiner Input Mappings
Signed Identity f(x) x -1, 1 ? -1, 1 Unsigned Identity f(x) max(0, x) 0, 1 ? 0, 1 Expand Normal f(x) 2 max(0, x) - 1 0, 1 ? -1, 1 Half Bias Normal f(x) max(0, x) ½ 0, 1 ? -½, ½
Signed Negate f(x) -x -1, 1 ? 1, -1 Unsigned Invert f(x) 1-min(max(0,x),1) 0, 1 ? 1, 0 Expand Negate f(x) -2 max(0, x) 1 0, 1 ? 1, -1 Half Bias Negate f(x) -max(0, x) ½ 0, 1 ? ½, -½
27
General Combiner RGB Function
Input RGB, Alpha Registers
Input Mappings
RGB Function
RGB Scale/Bias
Next Combiners RGB Registers
A
A op1 B
B
RGB Portion
C op2 D
C
AB op3 CD
D
Input Alpha, Blue Registers
Input Mappings
Alpha Function
Alpha Scale/Bias
Next Combiners Alpha Registers
A
AB
B
Alpha Portion
CD
C
AB op4 CD
D
28
General Combiner RGB Functions
Dot / Dot / Discard
Dot / Mult / Discard
Mult / Dot / Discard
A
A
A
A B
A B
AB
B
B
B
C D
CD
C D
C
C
C
D
D
D
Mult / Mult / Sum
Mult / Mult / Mux
A
A
AB
AB
B
B
CD
CD
C
C
AB CD
mux(AB, CD)
D
D
mux(AB, CD) (Spare0Alpha ? ½) ? AB CD
Dot products on RGB registers A B (Ared
Bred Agreen Bgreen Ablue
Bblue, Ared Bred Agreen Bgreen
Ablue Bblue, Ared Bred Agreen
Bgreen Ablue Bblue) Multiplication
on RGB registers AB (Ared Bred, Agreen
Bgreen, Ablue Bblue)
29
Diagram of the Final Combiner (OpenGL only)
Input RGB, Alpha Registers
Input Mappings
RGB Function
Available RGB Inputs
Input Mappings
A
Multiplier
E
EF
B
F
RGB Portion
RGB Out
AB (1-A)C D
Clamp to 0, 1
Color Sum Unit
C
Spare0
Sum
2nd-ary Color
D
Input Alpha, Blue Registers
Input Mapping
Alpha Portion
Alpha Out