[slurm-users] ntasks and cpus-per-task

[Patrick Goetz]

On 02/22/2018 07:50 AM, Christopher Benjamin Coffey wrote:
> It’s a big deal if folks use -n when it’s not an mpi program. This is because the non mpi program is launched n times (instead of once with internal threads) and will stomp over logs and output files (uncoordinated) leading to poor performance and incorrect results.
> 

I have a LAMMPS user who has been bugging me for weeks about a problem 
where if she runs the code by hand she gets useable output, but when she 
submits it with sbatch she gets garbage and couldn't figure out why. So, 
I finally made her come to my office to show me what she's doing, and 
sure enough, I found this in her submit file:

    #SBATCH --ntasks=4

and wouldn't have thought about it if I hadn't read your email this morning.

So, thanks!  <:)