PARADOXical Training, 14 October 2011, Belgrade, Serbia

Contribution List

1. Opening

14/10/2011, 10:00

0. Introduction to High-Performance Computing

Dr Antun Balaz (Institute of Physics Belgrade)

14/10/2011, 10:05

- Clusters - Supercomputers - HPC vs Grid Computing - Overview of HPC projects

2. Hands on session: PARADOX Cluster job management

Nikola Grkic (Institute of Physics Belgrade)

14/10/2011, 11:05

- Access to login node - Preparing job submitting scripts - Submitting jobs: - Serial jobs - OpenMP jobs - MPI jobs - Hybrid jobs - Application specific jobs (NAMD, CPMD) - Batch system job control and monitoring

3. Hands-on session: Overview and usage of PARADOX software stack

Vladimir Slavnic (Institute of Physics Belgrade)

14/10/2011, 12:50

- Overview of software installed - Compilers - Debuggers - Libraries - Application software - Using compilers (ICC, GCC, PGI) - Useful flags - Using numerical libraries - Intel MKL libraries (linking guide) - Debugging simple codes using GDB - Performing simple profiling using gprof - Finding memory leaks with Valgrind

Back to the event