LASCA Summer 2007 Activities

We offer information, training, and consulting opportunities throughout the year. We also welcome discussions about possible research collaborations. Computational science and engineering researchers are encouraged to take advantage of the opportunities listed below.

• Parallel programming short courses.

  These short courses are designed to help you get started with parallel programming. After participating you will be able to parallelize simple (FORTRAN, C or C++) codes, and you will have a good idea of the steps needed to parallelize more complex codes. Possible next steps might include taking the parallel programming summer FDI course, taking a parallel computation course from the CS department, or collaborating with a parallel computing researcher.

  The courses are "hands-on"---participants will run simple examples on parallel platforms. Participants will also have the opportunity to "bring-their-own-code" and discuss its potential for parallelization with the instructors.

  No previous experience with parallel programming is assumed. Some facility with Unix is strongly recommended, however.

  Short courses are open to all VT faculty, and to post-docs and graduate students with the endorsement of their advisors. Dates for Summer 2007 offerings are pending.

  • Introduction to Parallel Programming with MPI, An introduction to the Message Passing Interface (MPI), the standard API for developing parallel codes on distributed-memory machines such as System X.

  • Introduction to Parallel Programming with OpenMP, An introduction to OpenMP, a standard API for developing parallel codes on shared-memory machines such as Inferno.

• Consulting. Feel free to email lasca@cs.vt.edu with your questions about using VT's high-end resources. Routine questions can be handled by email. Appointments can be set up for more challenging problems.

• HPC and computational science consulting. By appointment.

  LASCA faculty and students are available to meet with researchers to discuss the application of high performance parallel computing resources to particular computational science problems. Contact for information.