Différences
Ci-dessous, les différences entre deux révisions de la page.
| Les deux révisions précédentes Révision précédente Prochaine révision | Révision précédente | ||
| ansys [2019/12/13 14:42] – [Parallel ansys using multicore (1 node)] admin | ansys [2021/09/07 12:47] (Version actuelle) – admin | ||
|---|---|---|---|
| Ligne 3: | Ligne 3: | ||
| ===== Installed version ===== | ===== Installed version ===== | ||
| - | * 19.2 | + | * 19.2b |
| Use the command <code bash>$ module avail ansys_inc</ | Use the command <code bash>$ module avail ansys_inc</ | ||
| + | |||
| + | <note important> | ||
| ===== How to use it? ===== | ===== How to use it? ===== | ||
| Ligne 15: | Ligne 17: | ||
| Open a terminal and load the ansys_inc module: | Open a terminal and load the ansys_inc module: | ||
| <code bash> | <code bash> | ||
| - | $ module load ansys_inc/ | + | $ module load ansys_inc/ |
| </ | </ | ||
| Ligne 39: | Ligne 41: | ||
| ==== Parallel ansys using multi-core (1 node)==== | ==== Parallel ansys using multi-core (1 node)==== | ||
| - | <code bash> | + | In this example we will execute ansys in parallel (shared memory) using 16 cores. |
| + | * We request 16 cores (one can request from 1 to 16 cores. harware limitation) | ||
| + | * We use default | ||
| + | * We request 4G of memory by core i.e total 64G (the whole memory in the node). | ||
| + | |||
| + | <note warning> | ||
| + | |||
| + | < | ||
| #!/bin/bash | #!/bin/bash | ||
| #$ -N test_ansy | #$ -N test_ansy | ||
| - | #$ -pe openmp | + | #$ -q all.q |
| + | |||
| + | #$ -pe openmp | ||
| #$ -o $JOB_NAME.$JOB_ID.out | #$ -o $JOB_NAME.$JOB_ID.out | ||
| Ligne 52: | Ligne 64: | ||
| ## load ansys module | ## load ansys module | ||
| - | | + | |
| ## start application | ## start application | ||
| Ligne 58: | Ligne 70: | ||
| ansys192 -np $NSLOTS -j inputansys23 < inputansys23.txt | ansys192 -np $NSLOTS -j inputansys23 < inputansys23.txt | ||
| + | # delete rst files | ||
| + | rm *.rst | ||
| + | |||
| + | </ | ||
| + | |||
| + | Submit the script to SGE : | ||
| + | |||
| + | <code bash> | ||
| + | $ qsub ansy.sge | ||
| </ | </ | ||
| + | |||
| + | |||
| ==== Parallel ansys using MPI (distributred nodes)==== | ==== Parallel ansys using MPI (distributred nodes)==== | ||
| - | FIXME TODO | + | FIXME TODO \\ |
| + | NO MPI VERSION AVAILABLE | ||