Virtually all modern computational and experimental science projects rely on software technology for gaining insight into data of increasing size and complexity. Such software technology and tools for analyzing and understanding data are a critical part of the scientist's working environment. Our project focuses on R&D algorithmic and software technology that targets DOE science problems that are challenged by the combination of the rapid growth in data size and complexity from experiments and computations as well as the quickly changing computational landscape.
We pursue R&D activities in three interrelated areas that target three primary challenges on the critical path for DOE scientific data understanding methods and software:
- We aim to better understand how to effectively take advantage of an evolving architectural landscape where there are an increasing number of cores per processor, deepening memory and storage hierarchies, and increasingly complex computational platforms.
- We target increasing scalability of key algorithms and methods in response to growing data sizes, changing architectures, and emergent design and execution patterns.
- We strive to increase the value and usefulness of data through the design, development, and application of new methods and approaches in the areas of scientific and information visualization, data analysis, and machine learning/artificial intelligence. Together, advances in these three focus areas will help increase scientific knowledge discovery in a way that leverages the convergence of computing and data.