CURRENT RESEARCH AT SCL

Research at the Scalable Scientific Computing Laboratory broadly deals with the area of parallel scientific computing.

Dr. Raghavan's group main area of research concerns sparsity as a unifying abstraction from computational science to computer architecture, toward increasing computational performance by constant factors to orders of magnitude.  Sparse representations of data arise from many different contexts, for example, from partial differential equations models where an associated matrix is populated mostly with zeroes and the number of non zeroes in the matrix is bounded by a small constant times its dimension, or from largely local connections in a system represented as a sparse graph or network. Her publications concern: parallel scientific computing; energy-aware supercomputing, i.e., performance and power scalability of advanced computer systems; and scalable algorithms for knowledge extraction.

The group focuses on the following research areas:

• Energy Aware Performance Scaling

• Computational Modeling, Simulation and Knowledge Extraction

• Parallel Scientific Computing

Dr. Shontz's group performs research in the design and analysis of meshing and numerical optimization algorithms and their applications.  They are also design efficient algorithms for linear algebra and model order reduction.  Meshing projects include the design of efficient algorithms for mesh optimization, mesh untangling, and mesh warping for cardiology, elasticity, and other applications. Optimization projects involve the development of an effective shape-matching algorithm, the design of efficient geometry optimization methods for electronic structure calculations, and the use of approximation models in parallel nonlinear optimization.  Additional projects include:  the design of efficient parallel numerical linear algebra algorithms for acoustics and model order reduction for electronic circuits and microarray data.

The group focuses on the following research areas: