This pillar stimulates and leverages advances in physical theory, applied math, data science, algorithms, and diverse approaches to high-performance computing to accelerate predictive capability. The other pillars, in turn, leverage this pillar to achieve their own long-term goals. This pillar supports an integrated set of pillars in a discovery-to-prediction Laboratory.
Computing Platforms: Architecture, Technologies, and Infrastructure
The hardware, supporting physical infrastructure, and system management software appropriate to multiple modes of scientific computing.
Computational Science: Methods, Algorithms, and Implementations to Advance Prediction
Existing, new, and emerging theory, methods, and algorithms for advancing prediction and implementation of these approaches in scientific areas of interest at relevant scale and complexity.
Data science: Analytics, Management, Sharing
A partial set of capabilities which manage, analyze, and interact with data to advance predictability for mission.