Skip to content

GitLab

Paralellization scheme of time dependent codes · Changes

Page history
Update Paralellization scheme of time dependent codes authored by Gabriel Wlazłowski's avatar Gabriel Wlazłowski
Show whitespace changes
Inline Side-by-side
Paralellization-scheme-of-time-dependent-codes.md
View page @ 737af119
...@@ -17,7 +17,7 @@ where: ...@@ -17,7 +17,7 @@ where:
* in input file: `gpuspernode 4`. * in input file: `gpuspernode 4`.
When executing the code, the following mapping `MPI Process <--> GPU` will be applied: When executing the code, the following mapping `MPI Process <--> GPU` will be applied:
![td-scheme-2](uploads/99212e2b948a20702ce6c6937c176b8e/td-scheme-2.png) ![td-scheme-2](uploads/392f47833c1d499edf7ce504cfbf277f/td-scheme-2.png)
In this case each MPI process is connected to one GPU. In this case each MPI process is connected to one GPU.
Alternatively, one can use: Alternatively, one can use:
......