The adaptability of this framework for use with large-data simulations will drive new levels of innovation in the computational sciences.
(PRWEB) April 24, 2012
Kitware, a leader in large-data scientific computing, today announces the award of $150,000 in SBIR funding from the United States Department of Energy (DOE) for the development of a cosmology analysis framework for in-situ analysis and data-reduction of cosmological simulations. Such simulations play an important role in the DOE’s High Energy Physics Cosmic Frontier program and are critical for understanding dark energy.
Cosmological observations currently produce terabytes of data per simulation, and data size is predicted to grow to exascale within a decade. The DOE needs an infrastructure in place to realize, encapsulate, and interpret the enormous wealth of information contained in such massive datasets. Traditional workflows run simulation codes for days and analysis is performed as a post-processing step.
This project requires an evolutionary shift in the way cosmological predictions are obtained, and therefore addresses critical challenges such as workflow I/O and the lack of domain-specific data analysis algorithms. An infrastructure will be developed that provides data reduction for minimizing input and output, robust and efficient halo-extraction methods, online tracking of halos to capture halo formation dynamics, and in-situ and co-visualization capabilities.
In addition to its use in cosmology simulations, the developed framework will have wide applicability in industries that use particle-based simulation techniques, including astrophysics, ballistics and defense, volcanology, oceanology, solid mechanics modeling, and a range of maritime applications.
“The adaptability of this framework for use with large-data simulations will drive new levels of innovation in the computational sciences,” said Berk Geveci, Technical Leader at Kitware and Principal Investigator on this project. “It will facilitate the very real and necessary transition from terascale work to peta- and exascale computing.”
Kitware is an open-source solutions provider for research facilities, government institutions, and corporations worldwide. Founded in 1998, Kitware specializes in research and development in the areas of visualization, medical imaging, computer vision, quality software process, data management, and informatics. Kitware is headquartered in Clifton Park, NY, with offices in Carrboro, NC and Villeurbanne, France. More information can be found at http://www.kitware.com.