Parallel Volume Rendering for Ocean Visualization in a Cluster of PCs

1
Parallel Volume Rendering for Ocean Visualization
in a Cluster of PCs

• Alexandre Coelho
• Marcio Nascimento
• Cristiana Bentes
• Maria Clicia S. de Castro
• Ricardo Farias

Geomática/UERJ Geomática/UERJ
Geomática/UERJ IME/UERJ COPPE/UFRJ
2
Outline

• Volume Visualization Overview
• Parallel Rendering System
• Experimental Results
• Conclusions

3
What is Volume Visualization?

• Volumetric data ? images
• 3D data ? 2D plane
• Gains
• Understanding
• Visual analysis and interpretation
• Meaningful information

4
Volume Visualization Applications

• Medicine, Geology, Chemistry, Industry
• Geographical Information Systems
• Ocean Modeling
• Monitoring the atmospheric pollution
• Terrain Modeling
• Analyzing natural phenomena (cyclones)

5
Volumetric Data

• Generated
• Sensors (CT scanner)
• Simulation (Fluid Dynamic)
• Measured Data (Ocean Buoys)

• Representation
• 3D grid of voxels (Regular or Irregular)

6
Volume Visualization Methods

• Surface Rendering
• Generates image of the surface
• Throws away data between surfaces
• Direct Volume Rendering

• Treats object as semi-transparent
• Can see entire volume

7
Volume Visualization Methods
Surface Rendering
Volume Rendering
x
8
Volume Rendering Challenge

• Large scale 3D data
• Computational intensive
• Unacceptably long time on uniprocessors

9
Efficient Volume Rendering

• Parallel Processing
• Multiple processors
• Parallel Machines
• Cluster of PCs

10
Parallel Volume Rendering
Parallel Machines

• Good speedups
• Expensive

• Clusters of PCs

• Low cost
• High availability
• Easy to update

11
Our goal

• Parallel Volume Rendering System for Ocean
  Visualization
• Efficient and scalable
• Low-cost
• All software implementation
• Portable and Free software
• Out-of-core execution

12
Our goal

• Allows Visualization of Ocean Inner Structure
• Climate research
• Offshore industries
• Fishing and Mammal Management

13
The Parallel Rendering System

• DPZSweep
• Based on PZSweep
• Sweeping plane paradigm
• Projection of the faces in depth order
• Two modules
• Pre-processing
• Parallel Rendering

14
The Parallel Rendering System
Pre-processing
Parallel Rendering
Grid Generation
Irregular grid
Octree Creation
Parallel Algorithm
Ocean data
Octree
15
Pre-Processing

• Grid Generation
• Latitude/Longitude data ? irregular grid

16
Pre-Processing

• Octree Creation
• Out-of-core execution

17
Parallel Rendering Algorithm

• Parallelization
• Breaking the screen into rectangles - tiles

Image portion that can be computed independently
18
Parallel Rendering Algorithm

• Parallelization
• Breaking the screen into rectangles - tiles

19
Parallel Rendering Algorithm

• Tile distribution
• Random assignment
• Dynamic distribution

20
Parallel Rendering Algorithm

• Dynamic Load Balancing
• Rebalance the work
• Distributed information diffusion algorithms
• Work stealing

21
Load Balancing Algorithms

• Nearest Neighbor (NN)
• Steal work from the nearest neighbor
• Longest Queue (LQ)
• Steal work from overloaded node
• Token ring to distribute load information
• Circular Distribution (CD)
• Dynamic distribution with token ring

22
Experimental Results

• Cluster
• 16 processors
• 512M bytes
• Fast Ethernet 100Mbits/sec
• Linux 2.4.20
• MPI

23
Ocean Dataset

• Gulf of Mexico Data NRL/ERC-MSU
• Thanks to Dr. Robert Moorhead
• Resolution 1 degree latitude and longitude
• 6 depth levels
• 1 time step Velocity
• 3 tetrahedralized versions
• Ocean (44K cells)
• Ocean1 (356K cells)
• Ocean2 (2854K cells)

24
Performance Analysis
25
Performance Analysis
26
Performance Analysis
27
Ocean Results
28
Conclusions

• Distributed parallel volume rendering for ocean
  datasets on cluster of PCs
• Dynamic load balancing low overhead
• Out-of-core execution
• Portable and free software infrastructure
• Great reductions in execution time
• Allows ocean researchers to interactively
  visualize large volumes of 3D data

29
Future Work

• Fault-tolerance
• Grid execution
• Handheld interface
• Handling Time-varying data

30
Thank you
For your attention.
31
(No Transcript)
32
(No Transcript)
33
Load Balancing Algorithms

• Nearest Neighbor

34
Load Balancing Algorithms

• Nearest Neighbor

Proc 1
Proc 0
Proc 2
35
Load Balancing Algorithms

• Nearest Neighbor

Proc 1
Proc 0
Proc 2
36
Load Balancing Algorithms

• Nearest Neighbor

Proc 1
Proc 0
Proc 2
37
Load Balancing Algorithms

• Nearest Neighbor

Proc 1
Proc 0
Proc 2
38
Load Balancing Algorithms

• Longest Queue

39
Load Balancing Algorithms

• Longest Queue

Proc 1
Proc 2
Proc 0
Proc 3
40
Load Balancing Algorithms

• Longest Queue

Proc 1
Proc 2
Proc 0
Proc 3
41
Load Balancing Algorithms

• Longest Queue

Proc 1
Proc 2
Proc 0
Proc 3
42
Load Balancing Algorithms

• Circular Distribution

43
Load Balancing Algorithms

• Circular Distribution

Proc 1
Proc 2
3
Proc 0
5
Proc 3
2
44
Load Balancing Algorithms

• Circular Distribution

Proc 1
Proc 2
3
8
Proc 0
5
Proc 3
2
45
Load Balancing Algorithms

• Circular Distribution

Proc 1
Proc 2
3
8
Proc 0
5
Proc 3
2
46
Load Balancing Algorithms

• Circular Distribution

Proc 1
Proc 2
3
8
Proc 0
5
Proc 3
2