The MPG supercomputer Hydra
Basic information
==> The Hydra cluster was taken out of operation on Nov. 14, 2018 <==
- About the system
- Node architecture, communication network, peak performance
- Usage statistics
- Availability and daily, weekly, monthly usage information
Using the System
- Access
- How to login. About ssh and scp.
- Configuration
- How much memory is available? How many CPUs?
- File systems
- Where to place your files. About AFS, the /tmp, and GPFS filesystems. About data handling, backup and migration.
- Software
- Available software packages, modules.
- Batch System
- How to submit a batch job. With sample scripts.
- Getting started on Hydra
- Hints for the migration and the porting of codes.
Documentation
- Compilers
- Available compilers (Fortran, C, C++, CUDA).
- Parallel Programming
- Getting started with MPI and OpenMP. Running programs under the "Parallel Operating Environment" (POE). NVidia Kepler K20X and Intel Xeon Phi accelerators.
- Libraries
- Which libraries (e.g. numerical, I/O) are available?
- Scientific applications
- A list of popular scientific application packages (e.g. GROMACS) installed on the system
- Programming Tools
- Information on debugging and profiling.
- IBM documentation:
- Parallel Environment, LoadLeveler (see menu 'Cluster Software')
- IBM Performance Guide for HPC Applications on the IBM System x iDataPlex.
- IBM Redbooks website.