2014 •
Linking performance data into scientific visualization tools
Authors:
Kevin Huck, Kristin Potter, Doug Jacobsen, Hank Childs, Allen D. Malony
Abstract:
Understanding the performance of program execution is essential when optimizing simulations run on high-performance supercomputers. Instrumenting and profiling codes is itself a difficult task and interpreting the resulting complex data is often facilitated through visualization of the gathered measures. However, these measures typically ignore spatial information specific to a simulation, which may contain useful knowledge on program behavior. Linking the instrumentation data to the visualization of performance within a spatial context is not (...)
Understanding the performance of program execution is essential when optimizing simulations run on high-performance supercomputers. Instrumenting and profiling codes is itself a difficult task and interpreting the resulting complex data is often facilitated through visualization of the gathered measures. However, these measures typically ignore spatial information specific to a simulation, which may contain useful knowledge on program behavior. Linking the instrumentation data to the visualization of performance within a spatial context is not straightforward as information needed to create the visualizations is not, by default, included in data collection, and the typical visualization approaches do not address spatial concerns. In this work, we present an approach that links the collection of spatially-aware performance data to a visualization paradigm through both analysis and visualization abstractions to facilitate better understanding of performance in the spatial context of the simulation. Because the potential costs for such a system are quite high, we leverage existing performance profiling and visualization systems and demonstrate their combined potential on climate simulation. (Read More)
Kevin A. Huck, Kristin Potter, Doug W. Jacobsen, Hank Childs, Allen D. Malony
2014 First Workshop on Visual Performance Analysis ·
2014
Data science |
Data mining |
We have placed cookies on your device to help make this website and the services we offer better. By using this site, you agree to the use of cookies. Learn more