The Computational Science and Mathematics Research (CSMR) department does research and development (R&D) in support of large-scale modeling, simulation, and data analysis by developing advanced algorithms, establishing relevant theory, creating high-performance software, and engaging scientists, engineers, and analysts to solve real-world problems.

The department's research includes optimization, uncertainty quantification, linear and nonlinear solvers, preconditioning, partial differential equations, verification and validation, and machine learning algorithms. Results of this research regularly appear in leading applied math, computer science, and computational science journals. The methods are implemented in freely-available software libraries and frameworks such as Trilinos and Dakota.

Department members work closely with scientists and engineers at Sandia to impact a wide variety of applications in engineering design, computational biology, systems studies, national security, etc.

Software

The following are CSMR-related software packages.

• APPSPACK - A parallel pattern search method for optimization.
• DAKOTA - A multilevel parallel, object-oriented framework for design optimization, parameter estimation, uncertainty quantification, and sensitivity analysis.
• OPT++ - An object-oriented nonlinear optimization package.
• SUNDANCE - Rapid development of high-performance parallel PDE simulators.
• Tensor Toolbox - Classes for manipulating dense, sparse, and structured multidimensional arrays in MATLAB.
• Trilinos - A collection of solver components.