Dear Discovery Cluster Users,

Northeastern University, ITS, Research Computing will be holding a seminar on “Parallel Computing Using MPI, OpenMP, CUDA with examples and debugging, tracing, profiling of parallel programs on the Discovery Cluster” on Tuesday, December 16th, 2014 that is open to all Faculty, Staff and Students. There will be one session from 10.00am to 12.00pm. Seating is limited to 40 for the session. 

For signup and venue details use the link http://discovery3.neu.edu/cgi-bin/signup/index.cgi. The registration is on a first come first served basis.

More details about this seminar is available here - http://nuweb12.neu.edu/rc/wp-content/uploads/2014/12/NU-RC_par_computing_12-16-2014.pdf.

This will enable users to get advanced details on using, tracing, debugging, and profiling MPI, OpenMP and CUDA/OpenCL programs on the Discovery cluster. It will also help users with serial code get some introduction on parallelization paradigms to start parallelization projects that can exploit the 8000+ compute cores on the Discovery Cluster. Users familiar with parallel computing would benefit from examples shown using all three parallel computing methodologies – shared-memory intra-node parallelization (OpenMP), TCP-IP and RDMA based inter-node parallelization (MPI) and GPGPU based multi-node acceleration (CUDA/OpenCL).

Hope to see you there.

Best
Nilay

########################################################################

To unsubscribe from the DISCOVERY list, click the following link:
http://listserv.neu.edu/cgi-bin/wa?SUBED1=DISCOVERY&A=1