Patrick O'Leary, Kitware Inc.
Burlen Loring, LBNL
- Enable in situ visualization for global hybrid (electron fluid, kinetic ions) simulations used to study the interaction of the solar wind with planetary magnetospheres such as the Earth and Mercury
- Directly embed ParaView Catalyst in a tightly coupled manner with UH3D code to perform scalable in-situ analysis at run time
Impact (Already made and/or Expected)
- Reduce I/O bottleneck by drastically decreasing storage demands
- Lessen time-to-insight in longer running, more complex global hybrid simulations
- Eliminate the need to migrate data to a visualization cluster
- Exclude the I/O bottleneck in traditional post-processing visualization software
- Do away with the need to adapt the output format from global hybrid simulations to input readable for traditional post-processing visualization software
Line integral convolution (LIC) visualization of magnetic field.
Progress (and/or Accomplishments)
Delivered UH3D/ParaView Catalyst integration with:
H. Karimabadi, P. O'Leary, B. Loring, A. Majumdar, M. Tatineni, and B. Geveci, "In-situ visualization for global hybrid simulations," In Proceedings of the Conference on Extreme Science and Engineering Discovery Environment: Gateway to Discovery (XSEDE '13). ACM, New York, NY, USA, Article 57, 8 pages, 2013
- In-situ visualization for global hybrid (electron fluid, kinetic ions) simulations used to study the interaction of the solar wind with planetary magnetospheres such as the Earth and Mercury.
- The term space weather has been coined to describe the conditions in space that affect the Earth and its technological systems. Particularly vulnerable are (i) Global Positioning System satellites, (ii) geosynchronous communication and weather satellites, (iii) large-scale power grids on the ground, and (iv) navigation and communications systems through the magnetosphere and ionosphere.
- ParaView Catalyst is designed to be tightly coupled with simulation codes, and was directly embedded into UH3D to perform scalable in-situ analysis at run time.
- The adaptor code translates the internal UH3D data structures to the VTK- based data structures. This translation was performed using zero-copy mechanisms, something that the existing I/O capabilities of UH3D fail to achieve.
- The line integral convolution (LIC) visualization technique convolves noise with a vector field along vector field tangents to visualize vector streamlines. Typically a serial GPU dependent.
- A parallel algorithm was developed in ParaView for multi-block datasets that are common in high performance computing that can also be used on systems that do not have GPU's.
- This work leveraged Mesa 9.2.0 , where the OS Mesa driver is replaced by Mesa’s Gallium llvmpipe software rasterizer. The Gallium llvmpipe driver is threaded and uses LLVM for just-in- time (JIT) OpenGL GLSL shader compilation.
- One major issue with the current paradigm of running the simulations and saving the data to disk for post- processing is that it is only feasible to save the data at a small number of time slices. This low temporal resolution of the saved data is a serious handicap in many studies where the time evolution of the system is of principle interest.
- One way to address this I/O issue is through in-situ visualization strategies. The idea is to minimize data storage by extracting important features of the data and saving them, rather than raw data, at high temporal resolution.
- Parallel file systems of current petascale and future exascale systems are expensive shared resources and need to be utilized effectively, and similarly archival storage can be limited and both of these will benefit from in-situ visualization as it will lead to intelligent way of utilizing storage.
- Preliminary results include both high-temporal resolution images and offscreen GPU independent LIC images using our in-situ visualization capabilities for global hybrid (electron fluid, kinetic ions) simulations which are used to study the interaction of the solar wind with planetary magnetospheres such as the Earth and Mercury.
- In addition, we demonstrated low the overhead and effect on code performance associated with the inline computations of in-situ visualization.