Users schedule jobs to run on NCAR's high-performance computing, data analysis, and visualization clusters by submitting them through the PBS Pro workload management system. Procedures for submitting jobs are similar, but there are some differences between the systems' hardware and queue structures.

Other documentation listed below is relevant to all of the HPC clusters that CISL manages.

Page contents

Submitting jobs

Cheyenne job-submission procedures and example batch scripts are provided here:

• Starting Cheyenne jobs
• Cheyenne job script examples

Casper job-submission procedures and example batch scripts are provided here:

• Starting Casper jobs
• Casper job script examples

Gust (pre-Derecho test system) documentation is in development.

Managing jobs

Information on these pages is relevant to users of all systems.

• Determining computational resource needs
• Managing and monitoring jobs
• Peer scheduling between Cheyenne and Casper
• Personalizing start files
• Propagating environment settings to a PBS job
• Running Singularity containers on NCAR systems
• Systems Accounting Manager