Attention: Confluence is not suitable for the storage of highly confidential data. Please ensure that any data classified as Highly Protected is stored using a more secure platform.
If you have any questions, please refer to the University's data classification guide or contact ict.askcyber@sydney.edu.au

Research Computing at Sydney

 

Welcome to the Research Computing wiki! This space is home to technical documentation for research computing infrastructure at the University of Sydney.

How to use this guide

Please see the following sections of our documentation for details on how to access and use services available to USyd researchers: 

Category

What does it include?

Computing platforms

Different research computing infrastructures available to USyd researchers. Includes: high performance computing (HPC), cloud computing, virtual machines, and virtual research desktops.

Data storage

Working, immediate, and long-term storage options.

Data movement

Transferring data across different platforms and storage locations.

Code management

Version control and collaborative code management with the git version control system.

List of platforms available to USyd researchers

Research Data Platforms

The University provides and supports a suite of different tools to aid with storing and managing your research data.

Please refer to the following pages for information on University-supported research data platforms:

Research Compute Platforms

Researchers from Sydney use many different computational platforms and services to undertake their research. The platforms listed below may help cover the various computing needs you face from day-to-day administrative tasks, running complex analyses, using specialised software, collaborating or presenting results, and more.

National Infrastructure

Sydney Managed

Sydney Supported (Paid by researcher)

Other Compute Options

These are not supported or endorsed by Sydney University, nevertheless they are used by researchers.

  • DUG - Commercial HPC provider.

  • Google Colab (See NCI’s ARE and Nirin, or Nectar for alternative).

  • NVIDIA GPU Cloud (NGC)

  • Google Cloud Platform

  • Other commercial cloud offerings

  • Personally hosted servers

 

Glossary of research computing terms

 

  • Allocation: cluster usage granted per project. Allocations define the number of service units you can consume

  • Compute cluster: a set of networked, often identical computers (nodes) that work together to act as a single, powerful computer system

  • Containerisation: packaging applications and their dependencies into isolated environments

  • High-performance computing (HPC): use of supercomputers or clusters to solve complex, large-scale computational problems

  • Job submission: a user-defined task (often in the form of a script or program) that is submitted to the cluster scheduler to run on specific nodes. Often includes resource requests like CPUs, GPUs, memory, walltime

  • Node: an individual server within a cluster, equiped with CPUs, memory, network interfaces, sometimes GPUs

    • Login node: the gateway system where users log in, compile code, or launch jobs. It is not for heavy computation

    • Compute node: nodes performing actual computation, they are scheduled to run jobs

  • Research data store: USyd’s secure storage platform for research datasets

  • Scheduler: software that queues job submissions, allocates resources, and schedules jobs

  • Supercomputer: an extremely powerful computer system, often a massively parallel cluster

  • Virtual machine: lightweight dedicated environments ideal for persistent services, small-scale analyses, hosting web apps and databases

  • Virtual research desktop: on-demand GPU-enabled environments designed for interactive computation and visualisation

  • Virtual private network (VPN): mechanism of secure access to the University’s network from off-campus