Navigation


Welcome to the Utah State University High Performance Computing (HPC@USU) wiki page. Below you will find information on the basics of HPC and more specifics on how to begin working with our local USU team. If you wish to submit or view your previous tickets click the LOGIN link.

VERY IMPORTANT You must update your Preferences with your name and E-mail address. Without this information, we are unable to contact or email you.

HPC Clusters

HPC@USU offers two main clusters UINTA and WASATCH. (UINTA is the default cluster)

To access UINTA (the default cluster), run the batch script like this: (Note: A must be uppercase)

[A#@login-XX ~]$ qsub mybatch

To access WASATCH, run the batch script like this: (Must specify WASATCH)

[A#@login-XX ~]$ qsub -q wasatch mybatch

The -q option requires a valid queue name, either UINTA or WASATCH. If you do not list the queue, the request will default to UINTA.

Note: for UINTA ppn=4, for WASATCH ppn=8 and nodes=X. X will be determined by the total number of processors desired divided by the processors per node. For example, to use 16 processors on UINTA, nodes=4 ppn=4. For WASATCH, nodes=2 ppn=8.

Look here for more information on WASATCH cluster

Documentation Wiki for HPC@USU

Getting Started - Request an account, how to connect, running jobs on the cluster

Videos - All you need to know to get started in a video format.

Systems - List of HPC@USU systems

User Info - Basic instructions on using the cluster

Software - Software provided by HPC@USU

EL237 Visualization Lab - HPC@USU Visualization resources and lab

Software Tutorials - Help for using HPC@USU provided software

FAQs - Frequently Asked Questions

Benchmarking - Performance benchmarking analysis of HPC@USU software & systems

UACSI 2008 - Presentations, files, and tutorials

Report a problem - Found something that needs to be fixed or improved? Let us know

 Contact Us - How to reach us


* HPC@USU Load (see  http://ganglia.hpc.usu.edu/ganglia/ for more machine stats.)


Trac Powered

Visit the Trac open source project at
http://trac.edgewall.org/