Large Ray Packets for Real-time
Whitted Ray Tracing
Ravi Ramamoorthi
Columbia University
Ryan Overbeck
Columbia University
William R. Mark
University of Texas at Austin
Intel Corporation
RT08, August ‘08
Current Real-time Ray Tracing
 Use large ray packets (16-256 rays)
 Focused on primary visibility and point-light shadows
 Images somewhat dull
 Easily generated using rasterization algorithms
RT08, August ‘08
This Work: Whitted Ray Tracing
 Whitted Ray tracing:




Primary Visibility
Point-light Shadows
Reflections
Refractions
RT08, August ‘08
Mission
 Study large ray packets for Whitted ray tracing
 Scene traversal (BVH)
 Partition Traversal: New!
 Frustum Culling
 General secondary ray packets: New!
RT08, August ‘08
Mission
 Evaluation of old and new algorithms
 Ray packet size
 Scene complexity
 Ray Recursion Complexity
RT08, August ‘08
Result: Real-time Whitted RT
RT08, August ‘08
Outline
 Motivation
 Frustum Culling for Secondary Rays
 Scene Traversal with Large Ray Packets
 Results
RT08, August ‘08
Frustum Culling
 Primary Rays
 Reshetov et al. 2005
 Point-Light Shadow Rays
 Boulos et al. 2006 , Wald et al. 2007
 Reflection + Refraction Rays
 New!
RT08, August ‘08
Frustum Culling: Point Light
1. Choose dominant axis:
2. Find
intersection points at
3. Get min/max
.
coords.
1
RT08, August ‘08
Frustum Culling: Point Light
1. Choose dominant axis:
2. Find
intersection points at
3. Get min/max
.
coords.
4. Create 4 corner rays.
5. Compute plane normals.
1
RT08, August ‘08
Frustum Culling: Reflection
RT08, August ‘08
Frustum Culling: Reflection
Scene
Bounding Box
Samples’ Bounding Box
RT08, August ‘08
Frustum Culling: Reflection
1. Choose dominant axis:
2. Find
1.
2.
intersections with 2 planes:
from scene AABB.
from samples’ AABB.
3. Get min/max
coords.
4. Create 4 corner rays.
5. Compute plane normals.
RT08, August ‘08
Outline
 Motivation
 Frustum Culling for Secondary Rays
 Scene Traversal with Large Ray Packets
 Results
RT08, August ‘08
3 Algorithms for Scene Traversal
 Masked Traversal – Control Method
 Based on Wald 2001, Reshetov 2005 (MLRT)
 Only good for small packets
 Ranged Traversal – State-of-the-art
 Wald et al. 2007
 Degrades performance for incoherent rays!
 Partition Traversal: New!
 Robust to incoherent rays and large packets
RT08, August ‘08
Masked Traversal
4
3
2
1
0
RT08, August ‘08
Masked Traversal
BVH Inner Cell
4
3
2
1
0
RT08, August ‘08
Masked Traversal
Near BVH Leaf Cell
4
3
2
1
0
RT08, August ‘08
Masked Traversal
Far BVH Leaf Cell
4
3
2
1
0
RT08, August ‘08
Masked Traversal
4
Alive
3
Alive
2
Alive
1
Alive
0
Alive
4
3
2
1
0
RT08, August ‘08
Masked Traversal
4
Alive
3
Alive
2
Alive
1
Alive
0
Alive
4
3
2
1
0
RT08, August ‘08
Masked Traversal
4
Alive
3
Alive
2
Alive
1
Alive
0
Alive
4
3
2
1
0
RT08, August ‘08
Masked Traversal
4
Alive
3
Alive
2
Alive
1
Alive
0
Alive
4
3
2
1
0
RT08, August ‘08
Masked Traversal
4
Alive
3
Alive
2
Alive
1
Alive
0
Alive
4
3
2
1
0
RT08, August ‘08
Masked Traversal
4
Alive
3
Alive
2
Alive
1
Alive
0
Dead
4
3
2
1
0
RT08, August ‘08
Masked Traversal
4
Alive
3
Alive
2
Alive
1
Dead
0
Dead
4
3
2
1
0
RT08, August ‘08
Masked Traversal
4
Alive
3
Alive
2
Alive
1
Dead
0
Dead
4
3
2
1
0
RT08, August ‘08
Masked Traversal
4
Alive
3
Alive
2
Alive
1
Dead
0
Dead
4
3
2
1
0
RT08, August ‘08
Masked Traversal
4
Dead
3
Alive
2
Alive
1
Dead
0
Dead
4
3
2
1
0
RT08, August ‘08
Masked Traversal
4
Dead
3
Alive
2
Alive
1
Dead
0
Dead
4
3
2
1
0
RT08, August ‘08
Masked Traversal
4
Dead
3
Alive
2
Alive
1
Dead
0
Dead
4
3
2
1
0
RT08, August ‘08
Masked Traversal
4
Dead
3
Alive
2
Alive
1
Dead
0
Dead
4
3
2
1
0
RT08, August ‘08
Masked Traversal
4
Dead
3
Alive
2
Alive
1
Dead
0
Dead
4
3
2
1
0
RT08, August ‘08
Masked Traversal
4
Dead
3
Alive
2
Alive
1
Dead
0
Dead
4
3
2
1
0
RT08, August ‘08
Masked Traversal
4
Dead
3
Alive
2
Alive
1
Dead
0
Dead
4
3
2
1
0
RT08, August ‘08
Masked Traversal
4
Dead
3
Alive
2
Alive
1
Dead
0
Dead
4
3
2
1
0
RT08, August ‘08
Masked Traversal
4
Dead
3
Alive
2
Alive
1
Dead
0
Dead
4
3
2
1
0
RT08, August ‘08
Masked Traversal
4
Dead
3
Alive
2
Alive
1
Dead
0
Dead
4
3
2
1
0
RT08, August ‘08
Masked Traversal
4
Dead
3
Alive
2
Alive
1
Dead
0
Dead
4
3
2
1
0
RT08, August ‘08
Masked Traversal
4
Alive
3
Alive
2
Alive
1
Alive
0
Alive
4
3
2
1
0
RT08, August ‘08
Masked Traversal
4
Alive
3
Alive
2
Alive
1
Alive
0
Alive
4
3
2
1
0
RT08, August ‘08
Masked Traversal
4
Alive
3
Alive
2
Alive
1
Alive
0
Dead
4
3
2
1
0
RT08, August ‘08
Masked Traversal
4
Alive
3
Alive
2
Alive
1
Dead
0
Dead
4
3
2
1
0
RT08, August ‘08
Masked Traversal
4
Alive
3
Alive
2
Alive
1
Dead
0
Dead
4
3
2
1
0
RT08, August ‘08
Masked Traversal
4
Alive
3
Dead
2
Alive
1
Dead
0
Dead
4
3
2
1
0
RT08, August ‘08
Masked Traversal
 Causes extra ray—cell tests
4
Dead
3
Dead
2
Alive
1
Dead
0
Dead
4
3
2
1
0
RT08, August ‘08
Ranged Traversal
4
3
2
1
0
RT08, August ‘08
Ranged Traversal
4
3
2
1
0
RT08, August ‘08
Ranged Traversal
4
3
2
1
0
RT08, August ‘08
Ranged Traversal
4
3
2
1
0
RT08, August ‘08
Ranged Traversal
4
3
2
1
0
RT08, August ‘08
Ranged Traversal
4
3
2
1
0
RT08, August ‘08
Ranged Traversal
4
3
2
1
0
RT08, August ‘08
Ranged Traversal
4
3
2
1
0
RT08, August ‘08
Ranged Traversal
4
3
2
1
0
RT08, August ‘08
Ranged Traversal
4
3
2
1
0
RT08, August ‘08
Ranged Traversal
4
3
2
1
0
RT08, August ‘08
Ranged Traversal
4
3
2
1
0
RT08, August ‘08
Ranged Traversal
4
3
2
1
0
RT08, August ‘08
Ranged Traversal
4
3
2
1
0
RT08, August ‘08
Ranged Traversal
4
3
2
1
0
RT08, August ‘08
Ranged Traversal
4
3
2
1
0
RT08, August ‘08
Ranged Traversal
4
3
2
1
0
RT08, August ‘08
Ranged Traversal
4
3
2
1
0
RT08, August ‘08
Ranged Traversal
4
3
2
1
0
RT08, August ‘08
Ranged Traversal
4
3
2
1
0
RT08, August ‘08
Ranged Traversal
3
4
2
1
0
RT08, August ‘08
Ranged Traversal
3
4
0
1
2
RT08, August ‘08
Ranged Traversal
3
4
0
1
2
RT08, August ‘08
Ranged Traversal
3
4
0
1
2
RT08, August ‘08
Ranged Traversal
3
4
0
1
2
RT08, August ‘08
Ranged Traversal
3
4
0
1
2
RT08, August ‘08
Ranged Traversal
3
4
0
1
2
RT08, August ‘08
Ranged Traversal
3
4
0
1
2
RT08, August ‘08
Ranged Traversal
3
4
0
1
2
RT08, August ‘08
Ranged Traversal
3
4
0
1
2
RT08, August ‘08
Ranged Traversal
3
4
0
1
2
RT08, August ‘08
Ranged Traversal
3
4
0
1
2
RT08, August ‘08
Ranged Traversal
3
4
0
1
2
RT08, August ‘08
Ranged Traversal
3
4
0
1
2
RT08, August ‘08
Ranged Traversal
3
4
0
1
2
RT08, August ‘08
Ranged Traversal
3
4
0
1
2
RT08, August ‘08
Ranged Traversal
3
4
0
1
2
RT08, August ‘08
Ranged Traversal
3
4
0
1
2
RT08, August ‘08
Ranged Traversal
 Reduces ray—cell tests
 Increases ray—triangle tests
3
4
0
1
2
RT08, August ‘08
Partition Traversal (New)
0
1
2
3
4
A
3
4
0
1
2
RT08, August ‘08
Partition Traversal (New)
0
1
2
3
4
A
3
4
0
1
2
RT08, August ‘08
Partition Traversal (New)
0
E B
1
2
3
4
A
3
4
0
1
2
RT08, August ‘08
Partition Traversal (New)
0
B E
1
2
3
4
A
3
4
0
1
2
RT08, August ‘08
Partition Traversal (New)
0
1
E B
2
3
4
A
3
4
0
1
2
RT08, August ‘08
Partition Traversal (New)
0
1
B E
2
3
4
A
3
4
0
1
2
RT08, August ‘08
Partition Traversal (New)
0
1
2
E B
3
4
A
3
4
0
1
2
RT08, August ‘08
Partition Traversal (New)
0
1
2
E
3
B
4
A
3
4
0
1
2
RT08, August ‘08
Partition Traversal (New)
0
1
4
E
3
2
B A
3
4
0
1
2
RT08, August ‘08
Partition Traversal (New)
0
1
4
3
E
2
B A
3
4
0
1
2
RT08, August ‘08
Partition Traversal (New)
0
1
4
3
2
A
3
4
0
1
2
RT08, August ‘08
Partition Traversal (New)
0
1
4
3
2
A
3
4
0
1
2
RT08, August ‘08
Partition Traversal (New)
0
E B
1
4
3
2
A
3
4
0
1
2
RT08, August ‘08
Partition Traversal (New)
0
B E
1
4
3
2
A
3
4
0
1
2
RT08, August ‘08
Partition Traversal (New)
0
1
E B
4
3
2
A
3
4
0
1
2
RT08, August ‘08
Partition Traversal (New)
0
1
E
4
3
2
B A
3
4
0
1
2
RT08, August ‘08
Partition Traversal (New)
0
4
E
1
3
2
B A
3
4
0
1
2
RT08, August ‘08
Partition Traversal (New)
0
4
1
3
2
E B A
3
4
0
1
2
RT08, August ‘08
Partition Traversal (New)
0
4
1
3
2
A
3
4
0
1
2
RT08, August ‘08
Partition Traversal (New)
0
4
1
3
2
A
3
4
0
1
2
RT08, August ‘08
Partition Traversal (New)
0
4
1
3
2
A
3
4
0
1
2
RT08, August ‘08
Partition Traversal (New)
0
4
1
3
2
A
3
4
0
1
2
RT08, August ‘08
Partition Traversal (New)
0
4
1
3
2
A
3
4
0
1
2
RT08, August ‘08
Partition Traversal (New)
0
4
1
3
2
A
3
4
0
1
2
RT08, August ‘08
Partition Traversal (New)
0
4
1
3
2
A
3
4
0
1
2
RT08, August ‘08
Partition Traversal (New)
0
E B
4
1
3
2
A
3
4
0
1
2
RT08, August ‘08
Partition Traversal (New)
0
B E
4
1
3
2
A
3
4
0
1
2
RT08, August ‘08
Partition Traversal (New)
0
4
E B
1
3
2
A
3
4
0
1
2
RT08, August ‘08
Partition Traversal (New)
0
4
E
1
3
2
B A
3
4
0
1
2
RT08, August ‘08
Partition Traversal (New)
0
4
1
3
2
A
3
4
0
1
2
RT08, August ‘08
Partition Traversal (New)
0
4
1
3
2
A
3
4
0
1
2
RT08, August ‘08
Partition Traversal (New)
 More ray—cell intersection
tests
0 4 1 3
2
A
3
 No unnecessary ray—triangle intersection tests
4
 Robust to incoherent rays
0
1
2
RT08, August ‘08
Outline
 Motivation
 Frustum Culling for Secondary Rays
 Scene Traversal with Large Ray Packets
 Results
RT08, August ‘08
3 Performance Variables
 Ray Packet Size
 Scene Complexity
 Ray Recursion Complexity
RT08, August ‘08
3 Performance Variables
 Ray Packet Size
4x4
8x8
16x16
32x32
 Scene Complexity
 Ray Recursion Complexity
RT08, August ‘08
3 Performance Variables
 Ray Packet Size
 Scene Complexity
ERW6
(804 Triangles)
Toasters
(11,141 Triangles)
Rings
Fairy
(172,669 Triangles) (217,812 Triangles)
 Ray Recursion Complexity
RT08, August ‘08
3 Performance Variables
 Ray Packet Size
 Scene Complexity
 Ray Recursion Complexity
Primary
Visibility
Primary
Visibility
+
Point-light
Shadows
Primary
Visibility
+
2-Deep
Refractions
Primary
Visibility
+
2-Bounce
Reflections
RT08, August ‘08
Traversal vs. Ray Packet Size
 Masked Traversal: performance degrades for > 4x4 packets
 Ranged Traversal: performance degrades for > 8x8 packets
Fairy Scene
2-Bounce Reflections
(Million CPU Cycles)
Time
 Paritition Traversal: performance increases with packet size
1500
1000
Ranged
Partition
Masked
500
0
Ray Packet 4x4
Size
8x8
16x16
32x32
RT08, August ‘08
Traversal vs. Recursion Complexity
 Masked Traversal: Degrades with recursion complexity
 Ranged Traversal: Degrades with recursion complexity
Fairy Scene
16x16 Packets
(Million CPU Cycles)
Time
 Partition Traversal: More robust to increasing recursion complexity
1200
Ranged
Partition
Masked
800
400
0
Recursion Primary
Complexity
Primary
+
Shadows
Primary
+
2-Deep
Refractions
Primary
+
2-Bounce
Reflections
RT08, August ‘08
Traversal vs. Recursion Complexity
Rings Scene
16x16 Packets
(Million CPU Cycles)
Time
 Difference between Partition and Ranged more pronounced on more
complex scenes
4000
3000
Ranged
Partition
Masked
2x
2000
1000
0
Recursion Primary
Complexity
Primary
+
Shadows
Primary
+
2-Deep
Refractions
Primary
+
2-Bounce
Reflections
RT08, August ‘08
Traversal vs. Scene Complexity
 Ranged Traversal: Degrades with scene complexity
(Million CPU Cycles)
Time
 Partition Traversal: More robust to increasing scene complexity
4000
3000
Ranged
Partition
Masked
2000
1000
0
Scene
Complexity
2-Bounce
Reflections
16x16 Packets
ERW6
Toasters
Fairy
Rings
RT08, August ‘08
Ranged vs. Partition Traversal
 Ranged Traversal
 Most ray—cell tests deep in BVH
 More extra ray—triangle tests for incoherent rays
 Partition Traversal
 Most ray—cell tests high in BVH
 No extra ray—triangle tests
RT08, August ‘08
Traversal: Summary
 Use Ranged Traversal
 Primary rays and coherent point-light shadow rays
 Small to medium packets: <= 8x8 rays
 Use Partition Traversal
 Incoherent rays: deep reflections + refractions
 Large ray packets: > =16x16 rays
RT08, August ‘08
Frustum Culling vs. Recursion Complexity
Toasters Scene
16x16 Packets
(Million CPU Cycles)
Time
 Frustum culling provides modest benefits
400
300
No Culling
With Culling
200
100
0
Recursion Primary
Complexity
Primary
+
Shadows
Primary
+
2-Deep
Refractions
Primary
+
2-Bounce
Reflections
RT08, August ‘08
Frustum Culling: Summary
 Primary + Point-light Shadows
 Up to 2x performance benefit possible
 But more often ~1.5x or less
 Reflections + Refractions
 Expect 1.2x – 1.3x performance benefit
RT08, August ‘08
Conclusion
 Use ranged and partition traversal in correct situation
 16x16 packets: 3x – 6x benefit over 2x2
 Whitted Ray Tracing is now interactive to real-time
11.8 FPS
6.7 FPS
8.5 FPS
4 FPS
RT08, August ‘08
Real-Time Whitted Ray Tracing
 Primary Visibility
 Ranged Traversal
 Point-Light
 Ranged Traversal
 1-Bounce Reflections
 Ranged Traversal
RT08, August ‘08
Conclusion
 Primary Visibility
 Ranged Traversal
 3-Bounce Reflections
 Partition Traversal
RT08, August ‘08
Conclusion
 Primary Visibility
 Ranged Traversal
 Point-Light Shadows
 Partition Traversal
 3-Deep Refractions
 Partition Traversal
 1-Bounce Reflections
 Partition Traversal
RT08, August ‘08
Acknowledgements
 Thanks to:
 Lab mates at Columbia University.
 Paid for by:
 NSF (grants CCF 03-05322, CCF 04-46916, CCF 0701775), a Sloan Research Fellowship, and an ONR
Young Investigator Award N00014-17-1-0900.
 Intel fellowship to Ryan Overbeck and related
equipment donations from Intel and NVIDIA.
RT08, August ‘08
Questions?
RT08, August ‘08