2
# Batch options (SLURM)
3
# =====================
7
#SBATCH --partition=para
8
#SBATCH --output nameandcaseo.%j
9
#SBATCH --error nameandcasee.%j
10
#SBATCH --job-name=nameandcase
12
# -t<time>, --time=<time> : walltime in minutes, minutes:seconds,
13
# hours:minutes:seconds, or
14
# days-hours:minutes:seconds
15
# -N, --nodes=<minnodes[-maxnodes]> : number of allocated nodes
16
# --ntasks=24, -n24 : number of total tasks
17
# --ntasks-per-node=<ntasks> : number of tasks per node
18
# --ntasks-per-socket=<ntasks> : number of tasks per socket
19
# --ntasks-per-core=<ntasks> : number of tasks per core
20
# --cpu-bind=cores, sockets : bind CPUs to cores or sockets
22
# --contiguous : force use of contiguous nodes for
23
# better latency and minimal latency
24
# --partition=<name>, -p<name> : partition (queue) (run "sinfo -s"
25
# for list of partitions)
26
# --reservation=<name> : allocate resources from reservation
27
# -o<pattern>, --output=<pattern> : output file name pattern
28
# -e<pattern>, --error=<pattern> : error file name pattern
29
# -J<jobname>, --job-name=<jobname> : job name
31
# Warning: when coupling with SYRTHES, 1 processor will be reserved for each
32
# instance of SYRTHES. The Kernel will be executed on the remaining
33
# processors, so make sure to reserve a sufficiently high number