Kitware announced its involvement in the Scalable Data Management, Analysis, and Visualization (SDAV) Institute, supported by the U.S. Department of Energy’s Scientific Discovery through Advanced Computing (SciDAC) program. The project will be led by Lawrence Berkeley National Lab and is a collaboration between prominent government laboratory, university, and industry partners for the purpose of creating a comprehensive set of tools and techniques that are essential for knowledge discovery on the DOE’s various computing platforms over the next five years.
The SDAV Institute’s primary goals are to actively work with application teams and assist them in achieving breakthrough science; and to provide technical solutions in data management, analysis, and visualization, all of which are broadly used by the computational science community. Such expertise is required to address the ever-increasing size, scale, complexity, and richness of scientific data.
Current users of DOE computing systems are faced with managing and analyzing their datasets for knowledge discovery, and must frequently rely on antiquated tools more appropriate for the teraflop era. While new techniques for interpreting large data have been developed, many in the field believe that these new algorithms and implementations will not work on their systems as we move into the many-core era. Without addressing these issues and beliefs, there will be a widening gap between computational and I/O capacity.
The SDAV Institute will work with DOE researchers to develop a comprehensive approach that will address these issues by encompassing all the stages of data analysis from initial data generation, orchestration of analysis tasks, and effective visualization of the results.
“We are extremely pleased to be part of the SDAV Institute,” said Berk Geveci, Kitware’s Principal Investigator on this project. “The aims of the Institute will greatly impact the computational science community while accelerating the pace of new discoveries.”
Kitware’s role is to support the DOE’s scientific teams with large-scale data analysis and visualization. This will involve making enhancements and extensions to current tools, such as ParaView and VisIt; introducing and supporting new technologies leveraging many-core and multi-core architectures; and coupling data analysis capability with simulation codes for in-situ analysis and co-processing.