Phase I effort focused on resolving bottlenecks in ultrascale visualization to support interactive exploration of massive datasets
NASA has awarded Kitware SBIR funding to develop an advanced visualization software solution to address the challenge of analyzing and obtaining insight from massive datasets. While the scientists producing these large datasets are accustomed to running batch jobs to perform visualization, scientific discovery is greatly enhanced by interactive exploration of the data. The software will address critical issues in ultrascale visualization in order to enable real-time investigation of these extremely large datasets using massive distributed memory architectures with up to 100,000 cores.
Kitware will be collaborating with California-based SciberQuest, leaders in kinetic modeling of space plasmas, to complete the project. This partnership provides Kitware with access to real world data from petascale simulations directly relevant to NASA’s missions and scientific goals, and allows the development to be guided by the ultimate users of the software.
In this investigative phase, Kitware will identify scaling bottlenecks in ParaView, an open-source visualization application currently used by NASA to explore the results of trillion element particle simulations on the Pleiades supercomputer. As the number of processors scales up past ten thousand, Kitware anticipates that the most critical issues will be data IO, architectural overhead, and the compositing of the partial results. Although the Phase 1 effort of this project is limited to developing prototypes and selecting a solution, if the Phase 2 effort is funded these improvements will be incorporated into ParaView and the underlying Visualization Toolkit (VTK) which will benefit tens of thousands of researchers world-wide.
“By customizing ParaView for use in different environments, we are able to accommodate a broad range of users, including NASA,” said Utkarsh Ayachit, a technical leader at Kitware who is leading efforts with NASA. “As datasets dramatically increase, it’s great to collaborate with an organization like NASA that requires facilitating large dataset computation and scientific collaboration in nearly every facet of its research.”