HPC Clusters

[one_fourth padding=”0 20px 0 0″]

Usage Guidelines

Request an Account

Request a Consultation

[/one_fourth]

[three_fourth_last padding=”0 0 0 0″]

Home and Work Directories

Ace and Turing each have separate home and work directories for each user.

Home directories are relatively small, and should only be used for local software installations, account settings, and temporary storage of final results from simulations.

Work directories are significantly larger, and should be used for writing simulation output during runs.

Currently, the home and work directories on Ace and Turing are protected against hardware failure. However, they are NOT backed up, and should therefore not be used for storage of important research data or job output.

General Cluster Information

Detailed hardware information for Ace (Turing in the future) can be viewed from the head node using the scluster command.

Both Ace and Turing use modules to manage scientific applications and libraries.

A variety of software and libraries are installed, and can be viewed using the following command:

module avail

To load the desired application or library from the available list, type:

module load SOFTWARE_NAME

To see the list of currently loaded software, type:

module list

The ARC managed HPC clusters use a batch manager called Slurm. Extensive documentation for Slurm can be found at http://slurm.schedmd.com/slurm.html

All jobs must be run through the batch manager. Any jobs found running on the compute nodes outside of the queueing system will be killed.

[/three_fourth_last]

Last Updated on May 1, 2017