Topic 12: Numerical Methods and Applications

The solution of large-scale problems in Computational Science and Engineering requires the availability of accurate, robust, and efficient numerical algorithms and software that exploit the potential of modern computer architectures. Such algorithms provide the means to further the development of existing applications, and the building blocks to prototype new methodologies. Ultimately, the objective is to relieve users from issues related to numerical methods and from implementation aspects strongly influenced by the computing environment.

This topic will provide a forum for discussing recent developments in the design and implementation of parallel and distributed numerical algorithms. The focus is on fundamental algorithmic concepts, efficient implementations on modern parallel architectures (e.g., multicore and hybrid platforms, multi-GPU systems), design and prototyping of scientific simulation software, performance analysis of numerical methods, and application studies.

• Dense and sparse linear algebra
• Discrete algorithms in scientific computing
• Combinatorial scientific computing
• Solvers (PDE, ODE, DAE)
• Large-scale data analysis and data mining
• Tensor decompositions and contractions; low-rank approximations
• Methods for uncertainty quantification
• Differential, integral, differential algebraic equations
• Optimization
• Transforms (wavelets, FFTs, …)
• Large-scale parallel applications and workflows
• Nonlinear systems