Download PowerPoint-Präsentation

Graphics Hardware Trends
performance
• Faster development than Moore’s law
– Double transistor functions every 6-12 months
– Driven by game industry
• Improvement of performance and functionality
– Multi-textures
– Pixel operations (transparency, blending, pixel shaders)
– Geometry and lighting modifications (vertex shaders)
graphics
CPU
network
time
Interactive Visualization of Volumetric Data on Consumer PC Hardware:
Introduction
Daniel Weiskopf
Transistor Functions
transistors (millions)
NVIDIA GeForce FX 5800 (125M)
120
110
100
ATI
Radeon 9700 Pro (110M)
90
80
70
NVIDIA GeForce4 (63M)
60
50
40
NVIDIA GeForce3 (57M)
ATI Radeon 8500 (60M)
30
20
Riva 128 (3M)
10
0
9/97 3/98 9/98 3/99 9/99 3/00 9/00 3/01 9/01 3/02 9/02 3/03
time (month/year)
Interactive Visualization of Volumetric Data on Consumer PC Hardware:
Introduction
Daniel Weiskopf
Typical GPU Characteristics
Brand
ATI Radeon 9800 P
NVIDIA GeForceFX 5900 U
Transistors
107 M
130 M
Technology
Clock rate
Mem bandwidth
Fill rate (peak)
Pixel pipelines
Textures per unit
FSAA
Bits per channel
Tri transform (peak)
Tris (3Dmark)
Vertex shaders
0.15 micron
380 MHz
22 GB/s
3 GPixel/s
8
8
6 x 18 Gsample/s
10
380 M
19 M
4
0.13 micron
450 MHz
27 GB/s
1.8/3.6 GPixel/s
4/8
16
4 x 27 Gsample/s
10
315 M
28 M
4+
Source: www.tomshardware.com
Interactive Visualization of Volumetric Data on Consumer PC Hardware:
Introduction
Daniel Weiskopf
Modern Scientific Visualization
• Traditional plotting techniques are not appropriate for
visualizing the huge datasets resulting from
• computer simulations (CFD, physics, chemistry, ...)
• sensor measurements (medical, seismic, satellite, …)
„The purpose of computing is insight not numbers“
• Map abstract data onto graphical representations
• Try to use colorful 3D raster graphics in
• expressive still images
• recorded animations
• interactive visualizations
„To see the unseen“
Interactive Visualization of Volumetric Data on Consumer PC Hardware:
Introduction
Daniel Weiskopf
Visualization Pipeline and Classification
visualization pipeline
sensors
simulation
raw data
• lines
• surfaces
• voxels
attributes:
data bases
volume rend.
3D isosurfaces
2D
filter
geometry:
classification
height fields
color coding
stream
ribbons
topology
arrows
LIC
glyphs
icons
attribute
symbols
1D
vis data
scalar
map
renderable
representations
vector
tensor/MV
different grid types  different algorithms
• color
• texture
• transparency
render
visualizations
images videos
interaction
3D scalar fields
Cartesian
(eg. medical datasets)
3D vector fields
un/structured
(eg. CFD)
Interactive Visualization of Volumetric Data on Consumer PC Hardware:
Introduction
trees, graphs, tables,
data bases
InfoVis
Daniel Weiskopf
GPU and Visualization Pipeline
• Renderer
– Texture-based techniques (e.g., for volume rendering)
– Large textured terrain height fields
• Mapper
–
–
–
–
Classification in volume rendering
Integrate ray segments (in unstructured volumes)
Integrate particle traces (in flow fields)
Assign color and transparency for NPR
• Filtering
– Data filtering in graphics memory
– Compression/decompression (of textures)
Interactive Visualization of Volumetric Data on Consumer PC Hardware:
Introduction
Daniel Weiskopf
Visualization of Volumetric Data
• Direct volume rendering of
scalar fields
• Flow visualization in 3D
• Focus on regular grids
Interactive Visualization of Volumetric Data on Consumer PC Hardware:
Introduction
Daniel Weiskopf
Visualization of Volumetric Data
Interactive Visualization of Volumetric Data on Consumer PC Hardware:
Introduction
Daniel Weiskopf