Here you will find documentation on the systems, hardware and software, tools and applications hosted on the CHPC super computers.

Update The new lustre3p filesystem is not yet reliable, therefore users are advised to use the old lustre filesystem until otherwise notified. The planned decommissioning of the old lustre filesystem has therefore been postponed.

With effect from 1 May 2021, a new 3 PB Lustre storage system has been implemented.

Please read Moving to the new 3 PB Lustre storage

The 3 PB Lustre is mounted on /mnt/lustre3p and, once you have copied over your files, you will need to change the paths in your job scripts.

Search and replace your script files and change /mnt/lustre/ to /mnt/lustre3p/

We will be updating all examples in this wiki to the new path, however, you may still come across an old example with the outmoded path. To make the example work on the new Lustre simply change the path from from /mnt/lustre to /mnt/lustre3p.

The CHPC has DELL cluster since mid-2016, named Lengau (cheetah), that can perform at petaFLOP level and replaces the old Tsessebe cluster.

Along with the new system are new accounts, new policies and new adventures.

• CHPC Accounts Policy Document Now updated to version 2.5.
• Procedure to follow for making use of CHPC resources

To register for the new system you need to apply online.

In order to obtain access to the cluster registration as either a Principal Investigator (PI) or Normal User on the new CHPC User Database is required. It is further required from approved PI's to register a single Research Programme to which registered users can be added as members. Job submissions will draw from cpu hour allocations associated with the respective Research Programmes. Full details of these new procedures and steps to follow to get registered may be found at the following link to the user database: https://users.chpc.ac.za/.

You should start here if you're in a hurry and familiar with HPC. The quick start guide briefly covers the main information a new users wants to know.

• Quick Start Guide

• How to ask for help

• Having Trouble Logging in?

• How To Guides for many specialised applications and tasks.

• The Users' Guide covers using the systems at the CHPC, from logging in to compiling and running your programs.

• Frequently Asked Questions with answers.

• How to ask for help — read this before asking us to help you.

• Workshops and Tutorials: future, current and previous training sessions.

• The Glossary describes common terminology used in high performance computing.

• Information on the Centre

• Helpful code that you might be able to use/adapt for your own problems
• Scaling – Check your code scales before running big jobs
• Getting an idea of Job Performance – Check your running job by logging into your job's node/s and using command htop
• Singularity in a Nutshell
• Need a python library that's not installed? Set up your own virtual environment
• More helpful Tips and Tricks on the HowTo page

• OpenNebula SR-IOV VMM Driver
• Application Benchmarks
• Change Management Control Cluster

• Overview
• Getting Started
• User Guide
• Best Practice and Policies
• Glossary of Terms
• FAQs

• Astronomy
• Bioinformatics
• CFD: Computational Fluid Dynamics
• Chemistry and Material Science
• Computational Space Physics
• Machine Learning

The CHPC is mandated to support high priority and time-critical projects for weather and pandemic forecasting which need to be prioritised for usage of CHPC resources from time to time. In order to achieve timeous results for national decision-making, the termination of some running jobs may be required without prior warning when cluster usage is high.

In compliance with the Protection of Personal Information Act (PoPIA) No. 4 of 2013, please read the CSIR's Privacy Notice.