VisIt: An End-User Tool for Visualizing and Analyzing Very Large Data 367
has approximately 100 regular users, and France’s Atomic Energy Com-
mission (CEA) at CESTA has approximately 50 regular users. Other
institutions, like Oak Ridge and Lawrence Berkeley, view VisIt as their
primary visualization and analysis tool, but do not keep user statistics.
• In terms of monetary support for developing VisIt, the U.S. Department
of Energy funds VisIt development through its Office of Science, National
Nuclear Security Agency, and Office of Nuclear Energy. Both of the US
National Science Foundation (NSF) XD centers on visualization actively
deploy and support VisIt as well.
• Another method for measuring usage is studying affiliations of users
who ask questions on the mailing list. The majority of these inquiries
come from none of the previously mentioned institutions, indicating that
usage goes beyond these sites.
FIGURE 16.3: Recent covers of the SciDAC Review Journal created using
VisIt.
Tracking individual user successes is difficult, although there is clear evi-
dence with certain types of usage. VisIt is used regularly to make images for
journal covers, a high-profile activity (see Fig. 16.3). Further, there have been
several notable instances of publications using VisIt to perform novel analysis:
• Analysis of laser wakefield simulations often amounts to finding key
particles [15], and query-driven visualization techniques were used to
search through terabytes of data to locate these key particles in as little
as two seconds.
• Simulations often deal with idealized meshes. VisIt’s comparative ca-
pabilities were used to quantify the importance of engineering defects
when differencing as-built and as-designed models [11].
• VisIt’s streamline code was used to find the toroidal magnetic fields
found in tokamaks by analyzing the fieldlines through a cross-sectional
slice and the topological “islands” they trace out [16].
368 High Performance Visualization
16.5 Future Challenges
Although VisIt is well suited for today’s supercomputing environment, the
project will face many challenges in the future. In the short term, I/O lim-
itations will force visualization and analysis activities to de-emphasize I/O.
The VisIt development team has invested in pertinent techniques, such as
multiresolution processing and in situ, but these techniques will need to be
further hardened to support production use. In the longer term, power limits
will constrain data movement, forcing much processing to occur in situ on
novel architectures, such as GPU accelerators. Unfortunately, VisIt’s existing
in situ implementation may be mismatched for this many-core future, for two
reasons. First, although VisIt can be easily multithreaded, using a pthreads or
OpenMP-type approach (see Chap. 12 to further understand the benefits of
hybrid parallelism), this approach may not be able to take advantage of these
architectures. The many-core future may require CUDA- or OpenCL-type lan-
guages; migrating the VisIt code base to this setting would be a substantial
undertaking. Second, although VisIt has been demonstrated to work well at
high levels of concurrency, some of its algorithms involve large data exchanges.
Although these algorithms perform well on current machines, they would vi-
olate the data movement constraints on future machines and would need to
be redesigned.
16.6 Conclusion
The VisIt project’s three focal points—understanding data, large data,
and delivering a product—together form a powerful environment for analyzing
data from HPC simulations. It is used in a variety of ways: it enables visual-
ization scientists, computational code developers, and the physicists that run
these codes to perform a broad range of data understanding activities, includ-
ing debugging, making movies, and exploring data. The user interface portion
of its design provides a powerful paradigm for analyzing data while the data
processing portion of its design is well suited for big data. This, in turn, has
led to many successes: in scaling up to high levels of concurrency and large
data sizes, in providing a “home” for large data algorithms, in understand-
ing how to best use supercomputers, and, most importantly, in helping users
understand their data. Further, despite significant upcoming changes in su-
percomputing architecture, VisIt’s future appears bright, as it enjoys vibrant
user and developer communities.
VisIt: An End-User Tool for Visualizing and Analyzing Very Large Data 369
References
[1] David Camp, Christoph Garth, Hank Childs, Dave Pugmire, and Ken-
neth I. Joy. Streamline Integration Using MPI-Hybrid Parallelism on a
Large Multicore Architecture. IEEE Transactions on Visualization and
Computer Graphics, 17:1702–1713, 2011.
[2] Hank Childs, Eric Brugger, Brad Whitlock, Jeremy Meredith, Sean Ah-
ern, Kathleen Bonnell, Mark Miller, Gunther H. Weber, Cyrus Harri-
son, David Pugmire, Thomas Fogal, Christoph Garth, Allen Sanderson,
E. Wes Bethel, Marc Durant, David Camp, Jean M. Favre, Oliver R¨ubel,
Paul Navr´atil, Matthew Wheeler, Paul Selby, and Fabien Vivodtzev.
VisIt: An End-User Tool For Visualizing and Analyzing Very Large Data.
In Proceedings of SciDAC 2011, July 2011. http://press.mcs.anl.gov/
scidac2011.
[3] Hank Childs, Eric S. Brugger, Kathleen S. Bonnell, Jeremy S Meredith,
Mark Miller, Brad J Whitlock, and Nelson Max. A Contract-Based Sys-
tem for Large Data Visualization. In Proceedings of IEEE Visualization,
pages 190–198, 2005.
[4] Hank Childs, Mark Duchaineau, and Kwan-Liu Ma. A Scalable, Hy-
brid Scheme for Volume Rendering Massive Data Sets. In Proceedings of
Eurographics Symposium on Parallel Graphics and Visualization, pages
153–162, May 2006.
[5] Hank Childs and Mark Miller. Beyond Meat Grinders: An Analysis
Framework Addressing the Scale and Complexity of Large Data Sets. In
SpringSim High Performance Computing Symposium (HPC 2006), pages
181–186, 2006.
[6] Hank Childs, David Pugmire, Sean Ahern, Brad Whitlock, Mark How-
ison, Prabhat, Gunther Weber, and E. Wes Bethel. Extreme Scaling of
Production Visualization Software on Diverse Architectures. IEEE Com-
puter Graphics and Applications, 30(3):22–31, May/June 2010.
[7] Thomas Fogal, Hank Childs, Siddharth Shankar, J. Kr¨uger,R.D.Berg-
eron, and P. Hatcher. Large Data Visualization on Distributed Memory
Multi-GPU Clusters. In Proceedings of High Performance Graphics 2010,
pages 57–66, June 2010.
[8] Christoph Garth and Ken Joy. Fast, Memory-Efficient Cell Location in
Unstructured Grids for Visualization. IEEE Transactions on Computer
Graphics and Visualization, 16(6):1541–1550, November 2010.
370 High Performance Visualization
[9] Cyrus Harrison, Hank Childs, and Kelly P. Gaither. Data-Parallel Mesh
Connected Components Labeling and Analysis. In Proceedings of Euro-
Graphics Symposium on Parallel Graphics and Visualization, pages 131–
140, April 2011.
[10] Martin Isenburg, Peter Lindstrom, and H. Childs. Parallel and Stream-
ing Generation of Ghost Data for Structured Grids. IEEE Computer
Graphics and Applications, 30(3):32–44, May/June 2010.
[11] Edwin J. Kokko, Harry E. Martz, Diane J. Chinn, Hank R. Childs,
Jessie A. Jackson, David H. Chambers, Daniel J. Schneberk, and Grace A.
Clark. As-Built Modeling of Objects for Performance Assessment. Jour-
nal of Computing and Information Science in Engineering, 6(4):405–417,
12 2006.
[12] Jeremy S. Meredith and Hank Childs. Visualization and Analysis-
Oriented Reconstruction of Material Interfaces. Computer Graphics Fo-
rum, 29(3):1241–1250, 2010.
[13] D. Pugmire, H. Childs, C. Garth, S. Ahern, and G. Weber. Scalable
Computation of Streamlines on Very Large Datasets. In Proceedings of
Supercomputing (SC09), 2009.
[14] David Pugmire, Hank Childs, and Sean Ahern. Parallel Analysis and
Visualization on Cray Compute Node Linux. In Proceedings of the Cray
Users Group Meeting, 2008.
[15] Oliver R¨ubel, Prabhat, Kesheng Wu, Hank Childs, Jeremy Meredith,
Cameron G. R. Geddes, Estelle Cormier-Michel, Sean Ahern, Gunther H.
Weber, Peter Messmer, Hans Hagen, Bernd Hamann, and E. Wes Bethel.
High Performance Multivariate Visual Data Exploration for Extemely
Large Data. In Proceedings of Supercomputing (SC08), 2008.
[16] A. R. Sanderson, G. Chen, X. Tricoche, D. Pugmire, S. Kruger, and
J. Breslau. Analysis of Recurrent Patterns in Toroidal Magnetic
Fields. IEEE Transactions on Visualization and Computer Graphics,
16(4):1431–1440, 2010.
[17] William J. Schroeder, Kenneth M. Martin, and William E. Lorensen.
The Design and Implementation of an Object-Oriented Toolkit for 3D
Graphics and Visualization. In Proceedings of the IEEE Conference on
Visualization (Vis96), pages 93–100, 1996.
[18] Gunther H. Weber, Vincent E. Beckner, Hank Childs, Terry J. Ligocki,
Mark Miller, Brian van Straalen, and E. Wes Bethel. Visualization Tools
for Adaptive Mesh Refinement Data. In Proceedings of the 4th High End
Visualization Workshop, pages 12–25, 2007.
VisIt: An End-User Tool for Visualizing and Analyzing Very Large Data 371
[19] Brad Whitlock, Jean M. Favre, and Jeremy S. Meredith. Parallel In Situ
Coupling of Simulation with a Fully Featured Visualization System. In
Proceedings of EuroGraphics Symposium on Parallel Graphics and Visu-
alization, pages 101–109, April 2011.
..................Content has been hidden....................
You can't read the all page of ebook, please click here login for view all page.