GPU PRO 360 Guide to GPGPU

Wolfgang Engel

• Publisher: A K Peters/CRC Press
• Year: 2018
• ISBN:
  9781351052085
• Edition: 1
• Topic: Informatica
• Subtopic: Grafica per computer

Wolfgang Engel's GPU Pro 360 Guide to GPGPU gathers all the cutting-edge information from his previous seven GPU Pro volumes into a convenient single source anthology that covers general purpose GPU. This volume is complete with 19 articles by leading programmers that focus on the techniques that go beyond the normal pixel and triangle scope of GPUs and take advantage of the parallelism of modern graphics processors to accomplish such tasks. GPU Pro 360 Guide to GPGPU is comprised of ready-to-use ideas and efficient procedures that can help solve many computer graphics programming challenges that may arise.

Key Features:

• Presents tips & tricks on real-time rendering of special effects and visualization data on common consumer software platforms such as PCs, video consoles, mobile devices

• 
  
  

• Covers specific challenges involved in creating games on various platforms

• 
  
  

• Explores the latest developments in rapidly evolving field of real-time rendering

• 
  
  

• Takes practical approach that helps graphics programmers solve their daily challenges

Contents

Introduction

Web Materials

1 2D Distance Field Generation with the GPU

Philip Rideout

1.1 Vocabulary

1.2 Manhattan Grassfire

1.3 Horizontal-Vertical Erosion

1.4 Saito-Toriwaki Scanning with OpenCL

1.5 Signed Distance with Two Color Channels

1.6 Distance Field Applications

Bibliography

2 Order-Independent Transparency Using Per-Pixel Linked Lists

Nicolas Thibieroz

2.1 Introduction

2.2 Algorithm Overview

2.3 DirectX 11 Features Requisites

2.4 Head Pointer and Nodes Buffers

2.5 Per-Pixel Linked List Creation

2.6 Per-Pixel Linked Lists Traversal

2.7 Multisampling Antialiasing Support

2.8 Optimizations

2.9 Tiling

2.10 Conclusion

2.11 Acknowledgments

Bibliography

3 Simple and Fast Fluids

Martin Guay, Fabrice Colin, and Richard Egli

3.1 Introduction

3.2 Fluid Modeling

3.3 Solver’s Algorithm

3.4 Code

3.5 Visualization

3.6 Conclusion

Bibliography

4 A Fast Poisson Solver for OpenCL Using Multigrid Methods

Sebastien Noury, Samuel Boivin, and Olivier Le Maître

4.1 Introduction

4.2 Poisson Equation and Finite Volume Method

4.3 Iterative Methods

4.4 Multigrid Methods (MG)

4.5 OpenCL Implementation

4.6 Benchmarks

4.7 Discussion

Bibliography

5 Volumetric Transparency with Per-Pixel Fragment Lists

László Szécsi, Pál Barta and, Balázs Kovács

5.1 Introduction

5.2 Light Transport Model

5.3 Ray Decomposition

5.4 Finding Intersections with Ray Casting

5.5 Application for Particle System Rendering

5.6 Finding Intersections with Rasterization

5.7 Adding Surface Reection

5.8 Shadow Volumes

5.9 Conclusion

Bibliography

6 Practical Binary Surface and Solid Voxelization with Direct3D 11

Michael Schwarz

6.1 Introduction

6.2 Rasterization-Based Surface Voxelization

6.3 Rasterization-Based Solid Voxelization

6.4 Conservative Surface Voxelization with DirectCompute

6.5 Solid Voxelization with DirectCompute

6.6 Conclusion

Bibliography

7 Interactive Ray Tracing Using the Compute Shader in DirectX 11

Arturo García, Francisco Ávila, Sergio Murguía, and Leo Reyes

7.1 Introduction

7.2 Ray Tracing

7.3 Our Implementation

7.4 Primary Rays Stage

7.5 Intersection Stage

7.6 Color Stage

7.7 Multipass Approach

7.8 Results and Discussion

7.9 Optimization

7.10 Conclusion

7.11 Further Work

7.12 Acknowledgments

Bibliography

8 Bit-Trail Traversal for Stackless LBVH on DirectCompute

Sergio Murguía, Francisco Ávila, Leo Reyes, and Arturo García

8.1 Introduction

8.2 Ray Tracing Rendering

8.3 Global Illumination

8.4 Stackless LBVH

8.5 The SLBVH in Action

8.6 Conclusion

8.7 Acknowledgments

Bibliography

9 Real-Time JPEG Compression Using DirectCompute

Stefan Petersson

9.1 Introduction

9.2 Implementation

9.3 Performance

9.4 Conclusion

9.5 Acknowledgments

Bibliography

10 Hair Simulation in TressFX

Dongsoo Han

10.1 Introduction

10.2 Simulation Overview

10.3 Definitions

10.4 Integration

10.5 Constraints

10.6 Wind and Collision

10.7 Authoring Hair Asset

10.8 GPU Implementation

10.9 Conclusion

Bibliography

11 Object-Order Ray Tracing for Fully Dynamic Scenes

Tobias Zirr, Hauke Rehfeld, and Carsten Dachsbacher

11.1 Introduction

11.2 Object-Order Ray Tracing Using the Ray Grid

11.3 Algorithm

11.4 Implementation

11.5 Results

11.6 Conclusion

Bibliography

12 Quadtrees on the GPU

Jonathan Dupuy, Jean-Claude Iehl, and Pierre Poulin

12.1 Introduction

12.2 Linear Quadtrees

12.3 Scalable Grids on the GPU

12.4 Discussion

12.5 Conclusion

Bibliography

13 Two-Level Constraint Solver and Pipelined Local Batching for Rigid Body Simulation on GPUs

Takahiro Harada

13.1 Introduction

13.2 Rigid Body Simulation

13.3 Two-Level Constraint Solver

13.4 GPU Implementation

13.5 Comparison of Batching Methods

13.6 Results and Discussion

Bibliography

14 Non-separable 2D, 3D, and 4D Filtering with CUDA

Anders Eklund and Paul Dufort

14.1 Introduction

14.2 Non-separable Filters

14.3 Convolution vs. FFT

14.4 Previous Work

14.5 Non-separable 2D Convolution

14.6 Non-separable 3D Convolution

14.7 Non-separable 4D Convolution

14.8 Non-separable 3D Convolution, Revisited

14.9 Performance

14.10 Conclusions

Bibliography

15 Compute-Based Tiled Culling

Jason Stewart

15.1 Introduction

15.2 Overview

15.3 Implementation

15.4 Optimization

15.5 Unreal Engine 4 Results

15.6 Conclusion

15.7 Acknowledgments

Bibliography

16 Rendering Vector Displacement-Mapped Surfaces in a GPU Ray Tracer

Takahiro Harada

16.1 Introduction

16.2 Displacement Mapping

16.3 Ray Tracing a Scene with Vector Displacement Maps

16.4 Ray Tracing a Vector Displacement Patch

16.5 Integration into an OpenCL Ray Tracer

16.6 Results and Discussion

16.7 Conclusion

Bibliography

17 Smooth Probabilistic Ambient Occlusion for Volume Rendering

Thomas Kroes, Dirk Schut, and Elmar Eisemann

17.1 Introduction

17.2 Smooth Probabilistic Ambient Occlusion

17.3 Approximating Ambient Occlusion

17.4 Results

17.5 Conclusion

Bibliography

18 Octree Mapping from a Depth Camera

Dave Kotfis and Patrick Cozzi

18.1 Overview

18.2 Previous Work and Limitations

18.3 Octree Scene Representation

18.4 Rendering Techniques

18.5 Results

18.6 Conclusion and Future Work

18.7 Acknowledgment

Bibliography

19 Interactive Sparse Eulerian Fluid

Alex Dunn

19.1 Overview

19.2 Introduction

19.3 GPU Eulerian Fluid Simulation

19.4 Simulation Stages

19.5 Problems

19.6 Latency Resistant Sparse Fluid Simulation

19.7 Performance

19.8 Sparse Volume Rendering

19.9 Results

19.10 Conclusion

Bibliography

About the Contributors

Introduction

Today, the use of GPGPU is prevalent in most modern architectures. With the parallel nature of the GPU, any algorithm that can process data in parallel can generally run orders of magnitudes faster than their CPU counterparts. Therefore it makes sense to take advantage of this power, considering that suitable hardware can now be found on most common hardware configurati...