sunfluidh:domain_features_examples
Différences
Ci-dessous, les différences entre deux révisions de la page.
Les deux révisions précédentesRévision précédenteProchaine révision | Révision précédente | ||
sunfluidh:domain_features_examples [2016/11/18 18:37] – [2D domain configuration] yann | sunfluidh:domain_features_examples [2017/09/26 17:15] (Version actuelle) – yann | ||
---|---|---|---|
Ligne 1: | Ligne 1: | ||
===== Examples of data set ===== | ===== Examples of data set ===== | ||
+ | [[sunfluidh: | ||
The user finds here some examples illustrating different configurations related to the namelist [[sunfluidh: | The user finds here some examples illustrating different configurations related to the namelist [[sunfluidh: | ||
The data initialized by default, and not explicitly required, are generally not present for a sake of clarity.\\ | The data initialized by default, and not explicitly required, are generally not present for a sake of clarity.\\ | ||
Ligne 23: | Ligne 23: | ||
| | ||
| | ||
- | | + | |
| | ||
Ligne 43: | Ligne 43: | ||
| | ||
| | ||
- | | + | |
| | ||
| | ||
Ligne 67: | Ligne 67: | ||
| | ||
| | ||
- | | + | |
| | ||
| | ||
Ligne 77: | Ligne 77: | ||
<note important> | <note important> | ||
- | === Parallel setting : MPI Only (in MPI graphic | + | === Parallel setting : MPI Only (in MPI graph topology) === |
< | < | ||
No OpenMP parallelization is considered .\\ | No OpenMP parallelization is considered .\\ | ||
- | Domain decomposition approach (MPI parallelization) in MPI graphic | + | Domain decomposition approach (MPI parallelization) in MPI graph topology. The domain is divided on 4 subdomains : |
* 4 along the I-direction (maximum value) | * 4 along the I-direction (maximum value) | ||
* 2 along the J-direction (maximum value) | * 2 along the J-direction (maximum value) | ||
Ligne 95: | Ligne 95: | ||
| | ||
| | ||
- | | + | |
| | ||
- | | + | |
- | | + | |
| | ||
| | ||
| | ||
- | <WRAP important> | + | <WRAP important> |
- | The MPI-graphic | + | The MPI graph topology is used in cases where the domain configuration |
- | In a first step, the domain decomposition is carried out as if the MPI cartesian decomposition was used. The number of processes will be equal to the multiplication of the Max_Number_MPI_Proc_I_Direction by Max_Number_MPI_Proc_J_Direction by Max_Number_MPI_Proc_K_Direction. \\ | + | In a first step, the domain decomposition is carried out as if the MPI cartesian decomposition was used. The number of processes |
- | When some subdomains are totally occupied by solid parts, they are useless. They are therefore removed in order to reduce the MPI-process number. As the MPI topology is no longer cartesian due to the " | + | When some subdomains are totally occupied by solid parts, they are useless. They must therefore |
- | The software " | + | |
- | The user | + | The software " |
+ | * The user sets the MPI data in the data file as for a cartesian subdomain splitting. | ||
+ | * The user then runs the software that : | ||
+ | | ||
+ | | ||
+ | * reformulates the subdomain splitting by excluding the pointless subdomains (covering the solid parts of the domain). | ||
+ | * build the data file named " | ||
+ | * a map of the enabled MPI processes related to the new subdomain layout | ||
+ | * the maximum MPI process number implied in the MPI graph topology | ||
+ | |||
+ | The user must set the variable " | ||
</ | </ | ||
+ | |||
+ | [[sunfluidh: |
sunfluidh/domain_features_examples.1479490664.txt.gz · Dernière modification : 2016/11/18 18:37 de yann