Running GRMHD module of Athena++ in Pleiades after upgradation to TOSS4 #541
Replies: 3 comments
-
Can you say (1) what exact environment you are loading, (2) what configure commands you are using, and (3) how you are calling the executable in the job script? (I've never actually used Pleiades, so I can only be of indirect assistance here.) |
Beta Was this translation helpful? Give feedback.
-
Also, this sounds like an MPI issue. I assume you are following the recommendation here. |
Beta Was this translation helpful? Give feedback.
-
Hi Chris, Thanks! Please note that the same set up was giving 'good' results before the upgrade. Here are the details: Version used: Athena++ downloaded in 2019 Here I modified configure.py to use avx accelerator and mpi-hpe. in line 637: Modules used (recommended by the NAS support in private communication):
|
Beta Was this translation helpful? Give feedback.
-
Hi, I am running a GRMHD simulation of an accretion disk using Athena++ in Pleiades Hasswll nodes in the NASA HECC facility. After the upgrade to TOSS4, I found weird artefacts in my GRMHD simulations along the mesh blocks. The same set-up is running fine in other clusters. Has anyone also encountered this kind of issue in Pleiades? Any suggestions will be really helpful.
Beta Was this translation helpful? Give feedback.
All reactions