Getting started

If you're just getting started with HPC, you may also want to check out a video recording of our Intro to HPC workshop in addition to this quick start.

This page is designed to give users an overview of how to run their work on our systems. By the end of this, you should know:

What HPC is
How to log in
What the bastion host, login nodes, and compute nodes are
What a job scheduler is
How to access software
How to run interactive and batch jobs on HPC

If you have not already, you will need to Register for an HPC Account to follow along.

Contents

What is HPC?

If you're looking to tackle complex problems or speed up your data analyses, HPC might be just what you need!

HPC is an acronym for High Performance Computing and is often used interchangeably with supercomputing. As a UArizona affiliate, you can be sponsored by a faculty member (faculty members can sponsor themselves) to receive free access to our three supercomputers Puma, Ocelote, and ElGato. These are clusters of computers that are housed in the lower level of the UITS building and are available for your analyses.

If you're interested in a more in-depth overview of what a supercomputer is, see our page Supercomputing In Plain English.

Logging in

HPC Structure

Logging into the HPC is not quite the same as logging into your personal computer. Normally, when you log into your personal computer, you have access to all of your files and can run programs on all available CPUs right away. In contrast, the HPC is made up of a number of "nodes," each of which serves a different purpose. Below is a diagram and brief description of each node type.

Bastion Host: This is the server you will encounter when first logging into HPC via SSH. You will not have access to any storage or files from here.
Login Node: This node can be accessed from the bastion host, and has access to the main HPC filesystem. This node is essentially a single computer shared by all active users, so it is not ideal for running your code as it will impact other users. This node can be used for general organizational purposes and simple software configuration, but more complex installations or computations should be run from compute nodes. Running intensive code on the login nodes will slow down the system for other users and such processes will be shut down by administrators.
Compute Nodes: These are the powerful CPUs and GPUs that are intended to run intensive computations, and have access to the main HPC filesystem. Processors are allocated to specific jobs to avoid impacting other users' performance. There are multiple ways to access them, and multiple can be accessed simultaneously, which allows for access to large amounts of RAM and numbers of CPUs, enabling highly parallelized and memory intensive jobs.
File Transfer Node: This node connects the main HPC system to tertiary file systems such as rental storage. As its name suggests, its primary purpose is to move data around and is not intended for running computations.

HPC Credentials

HPC access does not require its own set of credentials. To log in, all you'll need is your UArizona NetID and password with two-factor authentication enabled.

How to Connect

Users can connect to HPC remotely either through our browser-based graphical interface Open On Demand, or through a local terminal. In this quick start, we'll use the terminal.

The steps for connecting depend on your operating system. Choose the relevant panel below for instructions:

Mac/Linux Instructions

Connecting with Mac/Linux

Use the Terminal (on a Mac, this can be found under Applications → Utilities → Terminal). On the command line, enter the following, replacing netid with your own NetID:

Establishing a connection

$ ssh netid@hpc.arizona.edu

You will then be prompted for your password. Note: when you enter it, you will not see characters appear. This is a security feature and completely normal.

After a successful login, you will be connected to the bastion host. This is a single computer that provides a gateway to our three clusters. This is the only function the bastion host serves. It is not for storing files, running programs, or accessing software. The hostname of this machine is gatekeeper.

Accessing the bastion host

Success. Logging you in...
Last login: 
This is a bastion host used to access the rest of the RT/HPC environment.

Type "shell" to access the job submission hosts for all environments
----------------------------------

[netid@gatekeeper 13:53:23 ~]$ hostname
gatekeeper.hpc.arizona.edu
[netid@gatekeeper 13:53:26 ~]$

Windows Instructions

Connecting with Windows

You will need an ssh client such as PuTTY. Open a connection and enter hpc.arizona.edu under Host Name and press Open. This will open a terminal. At the prompt, enter the following, replacing netid with your own NetID: