For my thesis, I developed and compared a sequential CPU and parallel GPU implementation of a ray tracer written in C++ and CUDA respectively. Here are the presentation slides from my thesis defense.

1. 1
GPU Ray Tracing
with CUDA
BY TOM PITKIN
Bill Clark, PhD
Stu Steiner, MS, PhC

2. Objectives

Develop a sequential CPU and parallel GPU ray tracer

Illustrate the difference in rendering speed and design of a CPU and
GPU ray tracer
2

3. Outline

Introduction to Ray Tracing

CUDA

Parallelization with CUDA / Results

Future Work

Questions
3

4. What is Ray Tracing?

Rendering technique used in computer graphics

Simulates the behavior of light

Can produce advanced optical effects
4

5. Light in the Physical World
5
Light Source
Film
Object with
Red Reflectivity
Pinhole

6. The Virtual Camera Model

Eye Position – camera location in 3D space

Reference Point – point in 3D space where the camera is pointing

Orientation Vectors (u, v, n) – camera orientation in 3D space

Image Plane – projected plane of the camera’s field of view
Reference Point
v (Up Vector)
n
u
Eye Position
6

7. Ray Generation

Map the physical screen to the image plane

Divide the image plane into a uniform grid of pixel locations

7
Send a ray through the center of each pixel location
𝐼𝑚𝑎𝑔𝑒 𝑃𝑙𝑎𝑛𝑒 𝐻𝑒𝑖𝑔ℎ𝑡
𝑆𝑐𝑟𝑒𝑒𝑛 𝐻𝑒𝑖𝑔ℎ𝑡
Pixel
Eye Position
𝐼𝑚𝑎𝑔𝑒 𝑃𝑙𝑎𝑛𝑒 𝑊𝑖𝑑𝑡ℎ
𝑆𝑐𝑟𝑒𝑒𝑛 𝑊𝑖𝑑𝑡ℎ

8. Ray Intersection Testing

Ray – Sphere Intersection

Ray – Triangle Intersection
8

9. Phong Reflection Model
Ambient
+
Diffuse
+
Specular
9
=
Phong Reflection

10. Specular Reflection

Recursive Ray Tracing
10

11. Outline

Introduction to Ray Tracing

CUDA

Parallelization with CUDA / Results

Future Work

Questions
11

12. What is CUDA?

Compute Unified Device Architecture (CUDA)

Parallel computing platform

Developed by Nvidia
12

13. Kernel Functions

Specifies the code to be executed in parallel

Single Program, Multiple Data (SPMD)
13

14. Kernel Execution

Grids

Blocks

Threads
14

15. Memory Model

Global Memory

Constant Memory

Texture Memory

Registers

Local Memory

Shared Memory
15

16. Outline

Introduction to Ray Tracing

CUDA

Parallelization with CUDA / Results

Future Work

Questions
16

17. Thread Organization

2D array of blocks

2D array of threads

17
Each thread represents
a ray
Block (0, 0)
Block (1, 0)
Block (2, 0)
Block (0, 1)
Block (1, 1)
Block (2, 1)
Image Plane

18. Testing Environment

OS – Ubuntu Gnome Remix 13.04

CPU – Core i7-920


Core Clock – 2.66 GHz
GPU – Nvidia GTX 570

Core Clock - 742 MHz

CUDA Core - 480

Memory Clock - 3800 MHz

Video Memory - GDDR5 1280MB
18

19. Test Objects

Teapot



Surfaces: 1
Triangles: 992
Al



Surfaces: 174
Triangles: 7,124
Crocodile

Surfaces: 6

Triangles: 34,404
19

20. Single Kernel
20
Single Kernel
Single Thread
160 (0.16 sec)
Teapot
(Surfaces: 1)
(Triangles: 992)
23,003 (23 sec)
411 (0.41 sec)
Al
(Surfaces: 174)
(Triangles: 7,124)
55,260 (55.26 sec)
5,867 (5.87 sec)
Crocodile
(Surfaces: 6)
(Triangles: 34,404)
1,617,160 (26.95 min)
1
10
100
1,000
10,000
Milliseconds
100,000
1,000,000
10,000,000

21. Kernel Complexity and Size

Driver timeout

Register Spilling
21

22. Replacing Recursion

Iterative Loop

Layer based stack


Layers store color values returned from rays
Final image from convex combination of layers
22

23. Multi-Kernel
23
Multi-Kernel
Single Kernel (Previous Kernel)
381 (0.38 sec)
Teapot
(Surfaces: 1)
(Triangles: 992)
160 (0.16 sec)
967 (0.97 sec)
Al
(Surfaces: 174)
(Triangles: 7,124)
411 (0.41 sec)
13,217 (13.22 sec)
Crocodile
(Surfaces: 6)
(Triangles: 34,404)
5,867 (5.87 sec)
1
10
100
1,000
Milliseconds
10,000
100,000

24. Multi-Kernel with Single-Precision Floating Points
24
Multi-Kernel with Single-Precision
Floating Points
Multi-Kernel (Previous Kernel)
46 (0.05 sec)
Teapot
(Surfaces: 1)
(Triangles: 992)
381 (0.38 sec)
118 (0.12 sec)
Al
(Surfaces: 174)
(Triangles: 7,124)
967 (0.97 sec)
1,556 (1.56 sec)
Crocodile
(Surfaces: 6)
(Triangles: 34,404)
13,217 (13.22 sec)
1
10
100
1,000
Milliseconds
10,000
100,000

25. Caching Surface Data

Object’s surface data stored on shared memory

All threads in same block have access to cached surface data

Removes duplicate memory requests

Data reuse
25

26. Multi-Kernel with Surface Caching
26
Multi-Kernel with Surface Caching
Multi-Kernel with Single-Precision
Floating Points (Previous Kernel)
30 (0.03 sec)
Teapot
(Surfaces: 1)
(Triangles: 992)
46 (0.05 sec)
133 (0.13 sec)
Al
(Surfaces: 174)
(Triangles: 7,124)
118 (0.12 sec)
1,007 (1.01 sec)
Crocodile
(Surfaces: 6)
(Triangles: 34,404)
1,556 (1.56 sec)
1
10
100
Milliseconds
1,000
10,000

27. Simplifying Mesh Data
27

Triangle data originally stored as three points (vertices)

Optimize data by storing triangles as one point (vertex) and two edges

Calculate edges on host before kernel call
0.5, 1
0, 0
0.5, 1
1, 0

28. Multi-Kernel with Mesh Optimization
28
Multi-Kernel with Mesh Optimization
Multi-Kernel with Surface Caching
(Previous Kernel)
27 (0.03 sec)
Teapot
(Surfaces: 1)
(Triangles: 992)
30 (0.03 sec)
127 (0.13 sec)
Al
(Surfaces: 174)
(Triangles: 7,124)
133 (0.13 sec)
873 (0.87 sec)
Crocodile
(Surfaces: 6)
(Triangles: 34,404)
1,007 (1.01 sec)
1
10
100
Milliseconds
1,000
10,000

29. Final Results
29
Multi-Kernel with Intersection
Optimization
Single Thread
27 (0.03 sec)
Teapot
(Surfaces: 1)
(Triangles: 992)
23,003 (23 sec)
127 (0.13 sec)
Al
(Surfaces: 174)
(Triangles: 7,124)
55,260 (55.26 sec)
873 (0.87 sec)
Crocodile
(Surfaces: 6)
(Triangles: 34,404)
1,617,160 (26.95 min)
1
10
100
1,000
10,000
Milliseconds
100,000
1,000,000
10,000,000

30. Outline

Introduction to Ray Tracing

CUDA

Parallelization with CUDA / Results

Future Work

Questions
30

31. Future Work

Spatial partitioning

Multiple GPUs

Optimize code for different GPUs
31

32. Questions?
32