Faculty members involved in high performance computing and their areas
Computer architecture, parallel architecture and software, performance
Compilers, programming environments, performance modeling techniques
and tools, and computer architecture, with a particular focus on high-performance
computing and wide-area distributed computing.
General interests: Large-scale scientific computing, numerical
analysis, parallel computing.
Specific interests: Numerical linear algebra, sparse matrix
computation, numerical optimization, parallel performance visualization.
Parallel programming tools and techniques, parallel aplications in
science and engineering, parallel symbolic computations.
Compilers, machine organizations, performance prediction, programming
Reed, Department Head
Performance instrumentation and analysis techniques for large-scale
parallel systems, dynamic parallel resource management, parallel I/O access
Parallel algorithms and software for biomolecular modeling with an
emphasis on numerical methods.
Mesh generation, graph partitioning, computational geometry, parallel
scientific computing, combinatorial optimization, algorithms and cryptography.
Multithreading, processor-memory integration, scalability tools for
parallel applications, cache coherence, COMA vs NUMA, software based COMA,
speculative execution, database workloads, forwarding.
High performance I/O for parallel computing. The Panda project's goal
is to provide high-performance input and output by optimizing resource
utilization during movement of data between memory and secondary storage.