Parallel Computing

Many students now own multicore laptops on which they can potentially write and execute parallel programs. Also, Rutgers has recently acquired an IBM supercomputer on which high performance computing applications can be run. However, our undergraduates who know how to program are not exposed to parallel computing.

Vardit, Brian and Kyle researched the current state of practice and education in the field: what kinds of parallel architectures and systems are currently used and for what kinds of applications, what programming languages, tools, and techniques are used to write parallel programs, and what undergraduate courses nationwide and abroad teach parallel computing.

Topics

• Platforms
• OpenMP on Shared Memory Archiectures
• MPI on Supercomputers
• CUDA on GPUs
• Other Languages
• Bitonic Sort
• Sorting Overview
• OpenMP
• MPI
• CUDA
• Linear Algebra
• BLAS, LINPACK, and LAPACK
• Gaussian Elimination
• Algorithm
• OpenMP
• MPI
• CUDA
• Education
• Glossary