Training Basics

NOTE! MANY OF THESE DOCUMENTS NOT UPDATED to HPC3 specifics, yet!

HPC3 101 Documents

Software Carpentry Lessons are the best. If you don’t have basic Linux command line skills, we highly recommend that you go through their The Unix Shell lesson. Look for an instructor-led version of this, and their Version Control with Git in the near future.

For more in depth training, you can see our HPC3 Basics PDF.

UNIX Command Line Reference Card (PDF)

HPC3 101 Videos

HPC3: What is it?

A brief overview of Wharton’s High Performance Compute Cluster (HPC3).

Wharton VPN – Coming Soon!

mobaXterm Installation

Installation of mobaXterm, Wharton Research Computing’s preferred SSH and SFTP client for Windows. OSX (Mac) and Linux/UNIX users already have an SSH client: Terminal, so you can skip this video … see our Access Page for details on connecting for other OSes.

Batch Jobs with qsub

Use Univa Grid Engine’s ‘qsub’ command to run research code on Wharton’s HPC3.

The next four training videos were part of March 2, 2021 HPC3 training.

Best Practices

This section will be going over best practices on HPC3 like using project folders, group permissions, symlinks and git repo.

Parallel vs Array Jobs

This section will be going over parallel vs array jobs on HPC3. It will discuss when to use parallel vs array jobs and why using array jobs might be a better route.

Python on HPC3

This section will be going over how to use Python in the HPC3 environment like creating a virtual environment and installing modules.

Advanced Tricks

This section will be going over advanced tricks to use in the HPC3 environment like using rsync, rclone, holding jobs and email/text alerts for job completion.