Comp 150-06 class: Introduction to High Performance Computing - Tools and Algorithms

Tufts Elist service

List informations

This course covers fundamental topics in High Performance Computing, with
emphasis on Parallel Computing. We will start from basics in programming with
C++, creating and compiling sequential codes and employing high performance
scientific numerical libraries such as LAPACK and BLAS. Next, we will learn MPI
programming and employ parallel linear algebra libraries such as ScaLapack for
dense and SuperLU for sparse matrix operations. The focus will be on load
balancing, efficient design of communication and multilevel parallelism required
for heterogeneous computing. We will continue with advanced topics such as:
parallel I/O and GPU-CPU computing. We will use computer laboratory for hands-on
tutorials. Grade will be based on 5-6 home assignments/mini-projects and a final
project with presentation. Students are encouraged to suggest projects related
to the area of their expertise.