High Memory Nodes

We currently have 10 nodes (h01-h10) - Intel Xeon Silver 4116 2.1GHz, each with  24 CPUs, 768GB DDR4 RAM.

When to use the high memory queue

The hmem queue should be used when your job requires more than ~120Gb RAM,  Jobs of that size can run on the standard compute node, but they may affect the number of other jobs that can run on the node at the same time, because it is using more than the standard 4Gb per cpu allocation.
Using the hmem queue means you have access to much more memory, and are less likely to have an impact on other users' jobs.

To use this queue you need to be given access (hmem qos) - please email us at hpc.admin@uea.ac.uk

Submitting hmem jobs

  • To connect to the hmem queue for an interactive session use :

interactive-hmem

  • To connect to the hmem queue for an batch job, you need to select the hmem queue :

An example of an slurm submission script for a hmem job:

#!/bin/bash
#SBATCH --mail-type=ALL           #Mail events (NONE, BEGIN, END, FAIL, ALL)
#SBATCH --mail-user=<username>@uea.ac.uk    # Where to send mail
#SBATCH -p hmem-754                 #Which queue to use

#SBATCH --qos=hmem                # access to the queue
#SBATCH --mem=200G               # memory required
#SBATCH --time=0-03:00            # time (DD-HH:MM)
#SBATCH --job-name=testjob #Job name
#SBATCH -o test-%j.out       #Standard output log
#SBATCH -e test-%j.err       #Standard error log
#set up environment
module load gcc
#run the application
/gpfs/home/s154/gcc_code

  • Note there are two queues providing hmem resources

hmem-512 and hmem-754 - please see the queues page for details of resources available.