Abe Stephens Solomon Boulos, James Bigler, Ingo Wald, Steven Parker - - PDF document

abe stephens
SMART_READER_LITE
LIVE PREVIEW

Abe Stephens Solomon Boulos, James Bigler, Ingo Wald, Steven Parker - - PDF document

An Application of Scalable Massive Model Interaction using Shared-Memory Systems Abe Stephens Solomon Boulos, James Bigler, Ingo Wald, Steven Parker EGPGV 2006. Braga Portugal. Thursday May 11, 2006 10:45 - 11:10 Introduce yourself, explain


slide-1
SLIDE 1

1

An Application of Scalable Massive Model Interaction using Shared-Memory Systems

Abe Stephens

Solomon Boulos, James Bigler, Ingo Wald, Steven Parker EGPGV 2006. Braga Portugal.

Thursday May 11, 2006 10:45 - 11:10

Introduce yourself, explain that you’re a new member of the group, specializing in mesh generation. The authors were unable to be here, so you will be presenting. Read title – explain BioPSE = Biomedical Problem Solving Environment

slide-2
SLIDE 2

2

Scientific Computing and Imaging Institute, University of Utah

Talk Overview

Feasibility Demonstration:

  • Task adopted from real process.
  • Rendering one component.
  • System demonstrated at SC2005.

Software Architecture:

  • Shared Memory Interactive Ray Tracer.
  • Modular, extensible design.
  • Transparency and ambient occlusion.
slide-3
SLIDE 3

3

Scientific Computing and Imaging Institute, University of Utah

Feasibility Demonstration

Engineering Task

1. Quality Analyst identifies problem with aircraft (in the field). 2. Contacts Liaison Engineer, begins remote collaborative session. 3. Use 3D model to visualize problem. 4. Discuss best course of action using video and desktop sharing tools.

Remote Visualization

  • Necessary to share large servers.
  • Increases data security.
  • Collaborative visualization a

byproduct.

  • Performance and resolution factor
  • f distance.
slide-4
SLIDE 4

4

Scientific Computing and Imaging Institute, University of Utah

Massive Model Rendering

Boeing 777-200 model provided by The Boeing Company

Boeing 777-200 Aircraft

  • ~350 M Triangles.

Why Ray Trace?

  • sub-linear scaling in # triangles.
  • Easy transparency, global effects.

Acceleration Structure

  • kd-tree acceleration structure.
  • Size and quality depend on build time.
  • Memory use not a concern.
  • 29 GB kd-tree.
slide-5
SLIDE 5

5

Scientific Computing and Imaging Institute, University of Utah

Manta Software Architecture

Manta Pipeline

  • Display followed by rendering.
  • Modular and extensible components.
  • Transaction state changes applied each stage.
  • Barrier synchronization between stages.

Rendering Stack

  • Modular call stack invoked by rendering stage.
  • Scene intersection and shading performed by

Renderer component.

Modular Design

  • Allows Manta to be embedded in other

programs.

  • Supports multiple primitives:
  • Massive triangle models.
  • Massive volumes.
  • Sphere glyph (MPM) rendering.

MIT Licensed, Open Source Software.

slide-6
SLIDE 6

6

Scientific Computing and Imaging Institute, University of Utah

Transparency

Transparent rendering reveals intricate details while preserving the context of the model.

kd-tree traversal modified

  • Continues until ray attenuates.
  • Intersections are stored in an array.
  • Array is sorted and each intersection

color blended.

Simplifications:

  • No solid objects, blend surfaces only.
  • Fixed number of intersections for

constant alpha.

slide-7
SLIDE 7

7

Scientific Computing and Imaging Institute, University of Utah

Ambient Occlusion

Adapted Instant Global Illumination

  • Interleaved sampling of secondary rays based on

NxN pixel neighborhood.

  • Filter truncates based on distance threshold.
  • Sensitive to normal between neighbors.

Filter Kernel based on dot product between normals.

Compute Weights Filter

slide-8
SLIDE 8

8

Scientific Computing and Imaging Institute, University of Utah

Exploring the Model

Cutting Planes Hiding Objects

Users employed cutting planes and object hiding to locate a certain region of the model, then adjusted opacity to examine fine details and occluded structures.

Object Hiding & Cutting Planes

slide-9
SLIDE 9

9

Scientific Computing and Imaging Institute, University of Utah

Rendering Performance

128 p 1.6 Ghz Itanium2

  • 92% linear at 64p 82% at 126p
  • Resolution 1024x768
  • Demonstration system

62 p 1.4 Ghz Itanium2

  • Resolution 800x600
slide-10
SLIDE 10

10

Scientific Computing and Imaging Institute, University of Utah

Rendering Performance Cont’d

Ambient Occlusion 62 p 1.5 Ghz

  • Resolution 512x512
  • 3-5 fps depending on quality.
  • Not fast enough to be useful

during feasibility demonstration.

slide-11
SLIDE 11

11

Scientific Computing and Imaging Institute, University of Utah

Acknowledgments

David Kasik and Jim Troy of Boeing Hansong Zhang of Silicon Graphics Inc. Rocky Rhodes, Kenny Hoff, Dan McLachlan, Jimmy Wang and other members of the now defunct Visual Systems Group at SGI. All 3D model data was provided by the Boeing Company. Manta is open source software and contains contributions from many individuals. SGI has been, and continues to be, a principle contributor. Funding Agencies

  • DOE ASC CSAFE
  • Utah Center of Excellence for Interactive Ray-Tracing and Photo Realistic Visualization

Feasibility Demonstration was a joint project between:

  • Silicon Graphics Inc.
  • Intel Corporation
  • The Boeing Company
slide-12
SLIDE 12

12

Scientific Computing and Imaging Institute, University of Utah

Questions? Questions?