HPC Training

Important

We request all the users to pass the following quiz before submitting jobs, Make sure to log in to your NYU Google Apps Account (e.g., NYU Google Mail) first! Otherwise, your access to the Quiz will be denied (Quiz)

There are essentially four important components a HPC user needs to be familiar with to be able to use it successfully for computational research.

Component	Description
System	More about the system hardware and how to access it.
Storage	Page about user quotas and different types of storage available ($SCRATCH, $HOME, $ARCHIVE).
Software	More about the software stack and modules, and what it offers you in terms of your application (MPI,Python,R,Singularity)
Jobs Management	Section containing details of Job submission
HPC Web Interface	An interactive interface to remote computing resources

We shall have a brief introduction to each of the components below. The users can also refer to the corresponding sections above for a more detailed explanation, or reach us out at jubail.admins@nyu.edu for any assistance.

System

• System
  • Overview
  • Hardware
  • Typical Workflow
  • Summary of Nodes
  • Access

Storage

• Storage
  • Summary
  • Know Your Quota
  • Best Practices

Software

• Software
  • Software modules
  • Miniconda (Important for Python Users)
    • Miniconda on the HPC
      • One Time Setup
      • Application Installation
      • Managing Environments
      • Migrating / Sharing Environment
      • Submitting Job Scripts

Job Submission

• Quick Intro to Job Submission
  • Things to Keep in mind
  • Partitions Summary
  • Sample Job Script
  • #SBATCH with -n , -c and -N
  • Basic SLURM Commands
  • Requesting a GPU node
  • Preempt Partition
• Job Array
  • Examples of the --array parameter
  • A simple example
  • File of commands

HPC Web Interface

• HPC Web Interface
  • Overview
  • Connection to OOD
  • OOD Features

Quiz

• Quiz