Using ORCA Systems

Speaking the (Linux) language

The ORCA supercomputer, Titan, runs CentOS and requires Linux commands to interact with the system (example: listing files and folders, editing files, submitting jobs, monitoring pending & running jobs, etc.). We strongly recommend users become familiar with Linux commands in order to effectively work with the supercomputer.

• http://linuxcommand.org/lc3_learning_the_shell.php
• http://www.ee.surrey.ac.uk/Teaching/Unix/

Logging into the supercomputer

SSH is the preferred method of logging into ORCA computers. Click here for detailed instructions on how to log into ORCA systems.

Transferring files to the supercomputer

SCP and SFTP are the two methods by which you can transfer files between your computer and the supercomputer. Click here for detailed instructions.

Get familiar with our batch scheduler

Learn how to properly submit jobs on SLURM by referring to the following resources:

• Running Jobs on Titan
• SLURM User Tutorial
  • Quick Start User Guide
  • Command/option Summary (two pages)
  • Man Pages
  • Rosetta Stone of Workload Managers
  • Job Array Support
  • Slurm Checkpoint/Restart with BLCR
  • Job Exit Codes

Data and Processing Restrictions:

Please note that the following types of data sets and processes are not allowed on Titan:

• HIPPA, FERPA, export controlled, and classified data sets shall not be stored or processed on Titan. The same applies to any data set that is controlled or has policies set by any US Governement entity.
• The following activities related to Aritificial Intelligence are alos not allowed on Titan:
  1. creating generative AI models using data scraped from the internet or other publicly available data;
  2. training generative AI models using licensed third-party data;
  3. providing generative AI services to customers (in a B2B or B2G setting);
  4. enabling consumers to interact directly with generative AI (in a B2C setting)

Still have questions?

Contact us at supportorca@oru.edu.