Differences
This shows you the differences between two versions of the page.
Both sides previous revision Previous revision Next revision | Previous revision Next revisionBoth sides next revision | ||
doku:memory [2014/07/30 11:04] – [ad 2. parallel environment with fewer processes per node] ir | doku:memory [2015/09/17 12:32] – [ad 3. increased virtual memory] jz | ||
---|---|---|---|
Line 28: | Line 28: | ||
==== ad 2. parallel environment with fewer processes | ==== ad 2. parallel environment with fewer processes | ||
- | Jobs using more than 2 GB per process can be executed in one of the parallel environments | + | *[[doku:ompmpi|Hybrid OpenMP/MPI jobs ]] |
- | | + | |
- | * mpich4: 8 GB per core | + | |
- | * mpich2: 16 GB per core | + | |
- | * mpich1: 32 GB per core | + | |
- | ---- | ||
- | * mpich: 2 GB or 1 cores per process | ||
- | * mpich8: 4 GB or 2 cores per process | ||
- | * mpich4: 8 GB or 4 cores per process | ||
- | * mpich2: 16 GB or 8 cores per process | ||
- | * mpich1: 32 GB or 16 cores per process | ||
- | |||
- | ---- | ||
- | |||
- | Please keep in mind that on VSC-2 jobs are node-exclusive and therefore your contingent of CPU-hours will be computed by full nodes and are therefore significantly more expensive. | ||
- | For example, < | ||
- | |||
- | |||
- | ---- | ||
- | |||
- | |||
- | The variable NSLOTS_REDUCED is set to the number of cores requested, whereas the variable NSLOTS is set to the number of cores allocated in the queueing system, which corresponds to the cost calculation of the previous paragraph. | ||
- | Replace your calls to mpirun accordingly to< | ||
- | |||
- | |||
- | ---- | ||
- | == what to include in your jobscript == | ||
- | < | ||
- | mpirun -machinefile $TMPDIR/ | ||
- | The parallel environment option (pe) includes the type of the requested environment (mpich2) and the number of processes (4). -ppn is the number of processes per node (2). | ||
- | Within the machine file the variable NSLOTS_REDUCED is automatically set to the number of processes/ | ||
- | |||
- | === Hybrid MPI/OpenMP jobs === | ||
- | < | ||
- | export OMP_NUM_THREADS=8 | ||
- | mpirun -machinefile $TMPDIR/ | ||
- | In this example, each process allocates 8 cores and starts 8 threads. | ||
==== ad 3. increased virtual memory ==== | ==== ad 3. increased virtual memory ==== | ||
- | Some programs allocate more memory than they use. This was especially true in old FORTRAN 77 programs, which had to decide at compile time how much memory will be used. These programs are allowed to allocate 50% more memory than available by '''# | + | Some programs allocate more memory than they use. This was especially true in old FORTRAN 77 programs, which had to decide at compile time how much memory will be used. These programs are allowed to allocate 50% more memory than available by ''# |
==== ad 4. swap space (still experimental) ==== | ==== ad 4. swap space (still experimental) ==== | ||
A novel feature of the VSC-2 is remote swap space (implemented using 'SCSI RDMA Protocol', | A novel feature of the VSC-2 is remote swap space (implemented using 'SCSI RDMA Protocol', |