High Performance Computing

There are several resources for High Performance Computing, i.e., running computational jobs that require multiple cores and larger amounts of memory than what is available on a Desktop Computer.

  1. MAE Linux cluster
  2. MAE Windows Compute Servers
  3. Carleton’s Research Computing Group
  4. Digital Alliance of Canada

1 – MAE Linux Cluster

The Department of Mechanical and Aerospace Engineering has an updated Linux cluster available to graduate students for research and undergrad students for Capstone project work.  The cluster provides computing resources for computational simulations that are not practical to run on the lab desktops due to compute time or memory limits.  On average, simulations will run 4 times faster on the cluster (using 2 nodes, 40 cores) than a lab desktop (i7 with 8 cores) and can run in the background without the need to stay logged onto the system.

The cluster is running CentOS 7 with SLURM as the cluster management and queuing system.  It is currently capable of running:

  • Abaqus 2022 with user subroutines using Intel oneAPI compilers
  • Ansys Fluent 2023R2
  • Ansys CFX 2023R2
  • Ansys LS-Dyna 2023R2

More programs may be added depending on demand.

To obtain access to the cluster submit a request to neilb.mcfadyen@carleton.ca with the following information:

First Name:

Last Name:

Email address:

Student Number:

Supervisor if you are a graduate student:

Capstone Project Name if you are an undergrad:

Name of the software you need to run:

The amount of memory needed to run your jobs in GB:

(Hint – check the log file)

The estimated length of time to run one of your jobs:

The estimated amount of disk space needed in GB:

Include the output of a log file from a run on a lab desktop computer:

Have you taken a FEM course (yes/no):

if yes, course number:

Are you familiar with Linux:

 

If your request is approved you will get an account for logging into the cluster. The cluster is accessed by SSH/SFTP to the Head Node.  A VPN connection is required to connect from off campus so you Carleton account will need to be enabled for Azure MFA VPN.

Student Requiring VPN Access

You will need to learn some Linux commands so you can edit files in a command terminal, create directories and transfer files.  Here is a resource you can use:

Workshops

2- Windows Compute Servers

There are 3 Windows servers allocated for running Ansys CFX/Fluent, Matlab for undergraduate students doing Capstone projects.

server1 – 64 cores, 256gb memory

server2 – 32 cores, 128gb memory

server3 – 10 cores, 32gb memory

Capstone Students can request access to these servers.