How to use ANSYS on the HPC

This page provides explains how to run ANSYS on UCI's High Performance Computing Cluster

For questions regarding the HPC, please visit their website

Nabil Khalifa HPC Presentation

A crash course in CFD analysis introducing the finite volume approach and Fluent software as an application. This is the final workshop out of a series of four. A brief introduction for using Fluent on High Performance Computing units is presented, while highlighting how to deal with Fluent through the Text User Interface (TUI).

Nabil Khalifa's Crash Course in CFD

How to run ANSYS on UCI's HPC

Get Started

1) Request an account

Scroll down to "Get an account"

2) Review HPC linux tutorials, beginner's guide, and tutorials.

Go through all of their user guides
Give the tutorials a serious try if you have no experience with Linux. Go for it!
PLEASE read the HowTo's specifically anything about queues

3) Download and install:

X2GO (access cluster) - this is the preferred client
WinSCP (file drop)

4) hpc3.rcic.uci.edu (host name)

5) Read the Beginners Guide to find out how to connect with your SSH client such as X2GO.

7) Please read through the HPC guide on how to connect via SSH.

The procedure has changed due to campus DUO authentication

8) Read how to use software modules

9) Check out Nabil's course (above).

Using queues with SSH, X2GO, ANSYS

NOTE: Some of this is out of date since the HPC switched to SLURM

1) First type ‘q’ (Private queues specify here ‘queue name’)

2) qrsh –q ‘queue name’ –pe openmp ‘# of cores’ (allows you to reserve cores on a queue)

a. qrsh – q ‘queue name’ –pe mpi ‘# of cores’ (WARNING: always tries to take all cores no matter #of cores you type. If max cores not typed it will error out)

3) qstat | grep ‘queue name’ (Find out who is using a queue via username)

4) module load ansys

For more information see software tutorials and using modules

5) fluent

This assumes you want to use Fluent

6) qstat –u ‘username’ (see job number and etc.)

7) qdel ‘job number’ (make sure your job is deleted when done so others can access)

8) exit

Need HPC HELP?

https://rcic.uci.edu/hpc3/getting-help.html

Please make sure to:

1. Check the HPC FAQ and Help Guides

2. If the issue is not documented, then check out this overview how to report bugs.

3. Finally email support at hpc-support@uci.edu

FYI: Be specific in emails!

CAUTION

· VPN may cause the system to “error” out

Updated 5/3/2023

Lisa Preble