SIGGRAPH 2013 Shaping the Future of Visual Computing Building Ray - - PowerPoint PPT Presentation

▶
siggraph 2013
SMART_READER_LITE
LIVE PREVIEW

SIGGRAPH 2013 Shaping the Future of Visual Computing Building Ray - - PowerPoint PPT Presentation

Dec 29, 2022 977 likes •1.45k views

SIGGRAPH 2013 Shaping the Future of Visual Computing Building Ray Tracing Applications with OptiX David McAllister, Ph.D., OptiX R&D Manager Brandon Lloyd, Ph.D., OptiX Software Engineer Why ray tracing? Ray tracing unifies rendering of

slide-1
SLIDE 1

SIGGRAPH 2013

Shaping the Future

  • f Visual Computing

Building Ray Tracing Applications with OptiX™

David McAllister, Ph.D., OptiX R&D Manager Brandon Lloyd, Ph.D., OptiX Software Engineer

slide-2
SLIDE 2

Why ray tracing?

Ray tracing unifies rendering of visual phenomena

fewer algorithms with fewer interactions between algorithms

Easier to combine advanced visual effects robustly

soft shadows indirect illumination transparency reflective & glossy surfaces subsurface scattering depth of field

slide-3
SLIDE 3

Recursive Ray Tracing

Whitted-style recursive Reflection and refraction per hit Beer’s Law attenuation Depth cut-off Importance cut-off

slide-4
SLIDE 4

Ray Tracing Algorithms

Computational Power Interactive Real-time Batch

today

slide-5
SLIDE 5

Real Time Path Tracing

What would it take?

4 rays / sample 50 samples / pixel 2M pixels / frame 30 frames / second = 12B rays / second

GeForce GTX 680: 350M rays / second

Need 35X speedup

1 shading sample 1 AA sample 9 shading samples 1 AA sample 18 shading samples 2 AA samples 72 shading samples 8 AA samples 144 shading samples 16 AA samples 36 shading samples 4 AA samples

Good enough for games

slide-6
SLIDE 6

How to optimize ray tracing (or anything)

Better hardware (GPUs) Better software (Algorithmic improvement) Better middleware (Tune for the architecture)

slide-7
SLIDE 7

GPUs – the processor for ray tracing

Abundant parallelism, massive computational power GPUs excel at shading Opportunity for hybrid algorithms

slide-8
SLIDE 8

Acceleration Structures

slide-9
SLIDE 9
slide-10
SLIDE 10

Callable Program

OptiX Execution Model

rtContextLaunch Ray Generation Program Exception Program Selector Visit Program Miss Program Node Graph Traversal Acceleration Traversal

Launch Traverse Shade

rtTrace Closest Hit Program Any Hit Program Intersection Program

slide-11
SLIDE 11

C Host API Sample

RTresult rtContextCreate (RTcontext* context); RTresult rtContextDestroy (RTcontext context); RTresult rtContextDeclareVariable (RTcontext context, const char* name, RTvariable* v); RTresult rtContextSetRayGenerationProgram (RTcontext context, unsigned int entry_point_index, RTprogra RTresult rtBufferCreate (RTcontext context, unsigned int bufferdesc, RTbuffer* buffer); RTresult rtBufferSetFormat (RTbuffer buffer, RTformat format); RTresult rtBufferMap (RTbuffer buffer, void** user_pointer); RTresult rtBufferUnmap (RTbuffer buffer); RTresult rtProgramCreateFromPTXString (RTcontext context, const char* ptx, const char* program_name, RTresult rtProgramCreateFromPTXFile (RTcontext context, const char* filename, const char* program_name RTresult rtContextLaunch2D (RTcontext context, unsigned int entry_point_index, RTsize image_width, RTsize

slide-12
SLIDE 12

C++ Host API Sample

Context* context = Context::create(); context["max_depth"]->setInt( 5 ); context["scene_epsilon"]->setFloat( 1.e-4f ); // Ray gen program Program ray_gen_program = context->createProgramFromPTXFile( “myprogram.ptx”,"pinhole_camera" ); context->setRayGenerationProgram( 0, ray_gen_program ); BasicLight lights[] = { ..... }; Buffer light_buffer = context->createBuffer(RT_BUFFER_INPUT); light_buffer->setFormat(RT_FORMAT_USER); light_buffer->setElementSize(sizeof(BasicLight)); light_buffer->setSize( sizeof(lights)/sizeof(lights[0]) ); memcpy(light_buffer->map(), lights, sizeof(lights)); light_buffer->unmap(); context["lights"]->set(light_buffer);

slide-13
SLIDE 13

OptiX 3.0.1

Bug fix release Available this week

slide-14
SLIDE 14

OptiX Feature Sneak Peek

Next release coming in a few months

slide-15
SLIDE 15

Compilation Optimization

rtContextCompile()

3-7X faster

Still, you should avoid recompiles if possible.

slide-16
SLIDE 16

Updated Support Visual Studio 2012 support CUDA 5.5 support Quadro K6000 support

slide-17
SLIDE 17

Visual Studio OptiX Wizard

slide-18
SLIDE 18

Acceleration Builder Options

Sbvh has world class ray tracing performance Lbvh is extremely fast and works on very large datasets

Slow Build Fast Render Fast Build Slow Render

Sbvh Bvh MedianBvh Lbvh

slide-19
SLIDE 19

Acceleration Builder Options

Slow Render Slow Build

Bvh

Fast Build Fast Render

Sbvh MedianBvh Lbvh

slide-20
SLIDE 20

Fast BVH Builds

LBVH + refinement

A new approach introduced in OptiX 3.0  Very fast to build  Good for animation  Quality does not approach optimal

slide-21
SLIDE 21

TRBVH

LBVH + refinementnew + splits

New work by NVIDIA Research  VERY fast to build

40M tris / sec on a GeForce GTX Titan World’s fastest high quality BVH builder

 Quality averages 91% of SBVH HPG 2013 paper: https://research.nvidia.com/publication/fast- parallel-construction-high-quality-bounding-volume-hierarchies

slide-22
SLIDE 22

Acceleration Builder Options

Slow Render Slow Build

Bvh

Fast Build Fast Render

Sbvh MedianBvh Lbvh

Trbvh

slide-23
SLIDE 23

OptiX Low-Level Library

Specialized for ray tracing (no shading) Replaces rtuTraversal Improved performance

Uses latest algorithms from NVIDIA Research

ray tracing kernels [Aila and Laine 2009; Aila et al. 2012] Treelet Reordering BVH (TRBVH) [Karras 2013]

Can use CUDA buffers as input/output Support for asynchronous computation

Distributed as DLL and static library Designed with an eye towards future features

slide-24
SLIDE 24

API Overview

C API with C++ wrappers API Objects

Context Buffer Descriptor Model Query

slide-25
SLIDE 25

Context

Context tracks other API objects and encapsulates the ray tracing backend Creating a context

OLLresult

  • llContextCreate(OLLcontexttype type, OLLcontext* context)

Context types

OLL_CONTEXT_TYPE_CPU OLL_CONTEXT_TYPE_CUDA

Default for CUDA backend uses all available GPUs

Selects “Master GPU” and makes it the current device Master GPU builds acceleration structure

slide-26
SLIDE 26

Context

Selecting devices:

  • llContextSetCudaDeviceNumbers(

OLLcontext context, int deviceCount, const int* deviceNumbers ) First device is used as the master GPU

Destroying the context

destroys objects created by the context synchronizes the CPU and GPU

slide-27
SLIDE 27

Buffer Descriptor

Buffers are allocated by the application Buffer descriptors encapsulate information about the buffers

  • llBufferDescCreate(

OLLcontext context, OLLbufferformat format, OLLbuffertype type, void* buffer, OLLbufferdesc* desc )

Specify region of buffer to use (in elements)

  • llBufferDescSetRange( OLLbufferdesc desc, int begin, int end )

Context BufferDesc

slide-28
SLIDE 28

Buffer Descriptor

Variable stride supported for vertex format

  • llBufferDescSetStride

Allows for vertex attributes

slide-29
SLIDE 29

Buffer Descriptor

Formats

OLL_BUFFER_FORMAT_INDICES_INT3 OLL_BUFFER_FORMAT_VERTEX_FLOAT3, OLL_BUFFER_FORMAT_RAY_ORIGIN_DIRECTION, OLL_BUFFER_FORMAT_RAY_ORIGIN_TMIN_DIRECTION_TMAX, OLL_BUFFER_FORMAT_HIT_T_TRIID_U_V OLL_BUFFER_FORMAT_HIT_T_TRIID …

Types

OLL_BUFFER_TYPE_HOST OLL_BUFFER_TYPE_CUDA_LINEAR

slide-30
SLIDE 30

Model

A model is a set of triangles combined with an acceleration data structure

  • llModelCreate
  • llModelSetTriangles
  • llModelFinalize

Asynchronous finalize

  • llModelFinalizeWait
  • llModelFinalizePoll

Context Model BufferDesc BufferDesc

indices vertices

slide-31
SLIDE 31

Query

Queries perform the ray tracing on a model

  • llQueryCreate
  • llQuerySetRays
  • llQuerySetHits
  • llQueryExecute

Query types

OLL_QUERY_TYPE_ANY OLL_QUERY_TYPE_CLOSEST

Asynchronous query

  • llQueryWait
  • llQueryPoll

Context Model BufferDesc BufferDesc

indices vertices

Query BufferDesc BufferDesc

rays hits

slide-32
SLIDE 32

20 40 60 80 100 120 140 160 180 200

Speedup vs. SBVH in rtuTraversal

Build Performance

slide-33
SLIDE 33

0.0 0.5 1.0 1.5 2.0 2.5 3.0

Speedup vs. SBVH in rtuTraversal

Ray Tracing Performance

slide-34
SLIDE 34

0.0 50.0 100.0 150.0 200.0 250.0 300.0 350.0 400.0

Mrays/s

Ray Tracing Performance

slide-35
SLIDE 35

Future

Features we want to implement

Animation support (refit/refine) Instancing Large-model optimizations

slide-36
SLIDE 36

Art And Animation Studio

slide-37
SLIDE 37

Vertex Light Baking

Working with Bungie

But will be made available. Contact us if interested. Kavan, Bargteil, Sloan, “Least Squares Vertex Baking”, EGSR 2011

“NVIDIA's Optix has been instrumental when baking Ambient Obscurance (AO) over the extremely complex geometry in the worlds of Destiny. The high performance and ability to quickly explore various formulations of AO were invaluable.”

slide-38
SLIDE 38

Vertex Light Baking

Working with Bungie

But will be made available. Contact us if interested. Kavan, Bargteil, Sloan, “Least Squares Vertex Baking”, EGSR 2011

Compared to textures…

Less memory & bandwidth No u,v parameterization Good for low-frequency effects

slide-39
SLIDE 39

Typical Scene

Linear interpolation Static mesh Coarse mesh

slide-40
SLIDE 40

Piecewise Linear Approximation

Sample illumination on surface Each sample is a hemisphere of rays Reconstruct values at vertices

1

x x x x x x x x x x x x x x x x

slide-41
SLIDE 41

Weighted Averaging

slide-42
SLIDE 42

Least Squares Vertex Baking

slide-43
SLIDE 43

Pixar Research & NVIDIA OptiX Interactive Lighting Collaboration

slide-44
SLIDE 44

NVIDIA Visual Computing Theater in NVIDIA Booth

  • Tue. 4:00 – 4:30
  • Wed. 2:00 – 2:30
  • Thur. 10:40 – 11:10
slide-45
SLIDE 45

See OptiX at Siggraph

Pixar Research Interactive Lighting

NVIDIA Visual Computing Theater

  • Tue. 4:00, Wed. 2:00, Thur. 10:40

Mental ray with OptiX-powered Final Gather

NVIDIA booth

Bunkspeed Shot / Move / Drive with IRay

Bunkspeed booth, NVIDIA booth NVIDIA Visual Computing Theater: Thur. 1:20

Course: “Physically Based Shading in Theory and Practice”

  • Thur. 9:00

10:00: “Crafting a Next-Gen Material Pipeline for The Order: 1886”

slide-46
SLIDE 46

OptiX Resources

Slides & Video

GTC OptiX Introduction GTC OptiX Optimization http://www.gputechconf.com/gtcnew/on- demand-gtc.php?topic=39 This talk: http://www.nvidia.com/object/siggraph2013- tech-talks.html

Download OptiX

Available for free: Windows, Linux, Mac http://developer.nvidia.com

OptiX forum

https://devtalk.nvidia.com/default/board/90