You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Copy file name to clipboardExpand all lines: docs/software/sciapps/lammps.md
+31-3Lines changed: 31 additions & 3 deletions
Display the source diff
Display the rich diff
Original file line number
Diff line number
Diff line change
@@ -19,7 +19,7 @@ The current version of LAMMPS is written in C++.
19
19
### Loading LAMMPS Interactively
20
20
21
21
On Alps, LAMMPS is precompiled and available in a user environment [uenv].
22
-
LAMMPS has been built with the Kokkos and GPU packages separately.
22
+
LAMMPS has been built with the [Kokkos](https://docs.lammps.org/Speed_kokkos.html) and GPU packages separately.
23
23
24
24
To find which LAMMPS uenv is provided, you can use the following command:
25
25
@@ -154,7 +154,7 @@ You may need to make the `wrapper.sh` script executable (`chmod +x wrapper.sh`).
154
154
run $t
155
155
```
156
156
157
-
### Running LAMMPS+GPU on the HPC Platform
157
+
### Running LAMMPS + GPU on the HPC Platform
158
158
159
159
To start a job, two bash scripts are required: a [Slurm][ref-slurm] submission script, and a wrapper for [CUDA MPS][ref-slurm-gh200-multi-rank-per-gpu].
160
160
@@ -226,7 +226,35 @@ To enable oversubscription of MPI ranks per GPU, you'll need to use the `mps-wra
0 commit comments