Parallel Volume Rendering

Image

Problem Overview

Volume rendering is a visualization technique that uses a combination of color and opacity to allow an entire three-dimensional volume to be viewed at one time. In a parallel setting, the technique must employ special strategies, since each processor typically cannot generate an image for its own data in a way that can be combined with results from other processors. Instead, especially for unstructured data with complex overlaps between processors, partial results must be generated that can be merged with other processors' results to form a final image to form a final image.

Results

We have published multiple papers on this topic. Highlights include:

  • A three-dimensional rasterization technique that works on unstructured grids and avoids issues with load balancing and sorting, provided that a large buffer can be distributed over many processors (Childs/EGPGV06).
  • A study of how the three-dimensional rasterization technique can be mapped onto data-parallel primitives (Larsen/EGPGV15).
  • A study of how this algorithm scales at tens of thousands of cores and on data sets of trillions of cells (Childs/CGA10).
  • Two studies of the impacts of hybrid parallelism when using tens of thousands of cores and trillions of cores (Howison/TVCG12, Howison/EGPGV10).

People


Roba Binyahib
Ph.D. Student

Matt Larsen
Ph.D. Student (alum)

Hank Childs
CDUX Director

Publications


A Scalable, Hybrid Scheme for Volume Rendering Massive Data Sets
Hank Childs, Mark Duchaineau, and Kwan-Liu Ma
EGPGV, Braga, Portugal, May 2006

[PDF]     [BIB]

Volume Rendering Via Data-Parallel Primitives
Matthew Larsen, Stephanie Labasan, Paul Navratil, Jeremy Meredith, and Hank Childs
EuroGraphics Symposium on Parallel Graphics and Visualization (EGPGV), Cagliari, Italy, May 2015

[PDF]     [BIB]

Extreme Scaling of Production Visualization Software on Diverse Architectures
Hank Childs, David Pugmire, Sean Ahern, Brad Whitlock, Mark Howison, Prabhat, Gunther Weber, and Wes Bethel
IEEE Computer Graphics and Applications (CG&A), May 2010

[PDF]     [BIB]

Hybrid Parallelism for Volume Rendering on Large-, Multi-, and Many-Core Systems
Mark Howison, Wes Bethel, and Hank Childs
IEEE Transactions on Visualization and Computer Graphics (TVCG), January 2012

[PDF]     [BIB]

MPI-hybrid Parallelism for Volume Rendering on Large, Multi-core Systems
Mark Howison, Wes Bethel, and Hank Childs
EGPGV, Norrkoping, Sweden, May 2010
Best Paper

[PDF]     [BIB]