Note NERSC queue limits:

Haswell nodes 1 & 2
queue: regular
QOS: normal
Max running: 50 [* 5/node = 250 nodes, but max is 100, => 20 jobs]
Max nodes in use: 100
Max queued jobs: 200

Haswell nodes 3-512
queue: regular
QOS: normal
Max running: 10
Max nodes in use: 512
Max queued jobs: 50

Example of declaring memory use on the shared queue
#!/bin/bash -l
#SBATCH -p shared
#SBATCH --mem=13GB
#SBATCH -t 02:00:00
#SBATCH -J my_job
#SBATCH -L SCRATCH
./something.exe

Can also specify -n 1 to use just one CPU slot

-N or --nodes
-n or --ntasks
--mem-per-cpu=8000      # default units are MB
