LAMMPS docs #80
LAMMPS docs #80
Conversation
rubber-duck-debug
requested review from
RMeli,
bcumming and
msimberg
as code owners
|
preview available: https://docs.tds.cscs.ch/80 |
|
|
||
| export MPICH_GPU_SUPPORT_ENABLED=1 | ||
|
|
||
| numactl --cpunodebind=$NUMA_NODE --membind=$NUMA_NODE "$@" |
There was a problem hiding this comment.
Can --cpunodebind be replaced by SLURM's --cpu-bind=core and --cpus-per-task=72/64 etc?
|
|
||
| ### Using LAMMPS uenv as an upstream Spack Instance | ||
|
|
||
| If you'd like to extend the existing uenv with additional packages (or your own), you can use the provide LAMMPS uenv to provide all dependencies needed to build your customization. See https://eth-cscs.github.io/alps-uenv/uenv-compilation-spack/ for more information. |
There was a problem hiding this comment.
Probably link to https://eth-cscs.github.io/alps-uenv/uenv-compilation-spack instead? I think some of the details could be left out from here and rely on the generic advice in https://eth-cscs.github.io/alps-uenv/uenv-compilation-spack instead, but I'm not completely against a more concrete step-by-step guide here either. We should just generally avoid 1. duplicating content and 2. having duplicated content give different advice.
|
Sorry @nickjbrowning, I had some un-submitted review comments 😬 Perhaps you can have a look in a follow-up PR? |
RMeli
left a comment
RMeli
left a comment
There was a problem hiding this comment.
Thanks @nickjbrowning! I just added some comments, most of which you can quickly accept/reject via the GitHub UI with minimal effort.
|
|
||
| ## Running LAMMPS | ||
|
|
||
| ### Loading LAMMPS Interactively |
There was a problem hiding this comment.
With the note above, I think this section can be streamlined and reduced. I would only mention the views that are available, since the instructions on how to use uenv are provided in the uenv documentation. But I can take of this after the PR is merged if you want.
|
Ouch, I was writing a review while the PR has been merged. I guess now it is no longer possible to automatically accept the changes. 😞 |
This reverts commit 0d2f0bb.
3 participants
still a WIP - need to define Eiger instructions