Computing platforms

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

Computing platforms

Research computing comes in all shapes and sizes. In some cases, your compute needs are well met by your personal computer. In other cases, this is not sufficient, and this is where the platforms below can be critical to the timely analysis of your data.  


What platforms do I have access to?

We are in the midst of growing the Sydney Research Cloud ecosystem and this list will grow to include new systems as they are onboarded. Support for accessing and getting your workloads running on these systems is provided by Sydney Informatics Hub (SIH). Request our support here.

Type

System

Suitable for

Access for USyd researchers

Type

System

Suitable for

Access for USyd researchers

HPC

NCI Gadi

  • Parallel workloads for astronomy, climate modelling, genomics, fluid dynamics, engineering simulations, AI/ML

  • Virtual research environments

  • GPU-accelerated workflows

  • Not suitable for highly-protected data.

  • Fully subsidised access to capped allocations via Sydney Scheme

  • NCMAS (support provided by SIH)

  • ABLeS (for open source bioinformatics only)

 

Pawsey Setonix

  • Parallel workloads astronomy, climate modelling, genomics, structural biology, fluid dynamics, engineering simulations

  • GPU-accelerated workflows

  • Not suitable for highly-protected data.

  • Fully subsidised access to capped allocations via Sydney Scheme

  • NCMAS (support provided by SIH)

  • ABLeS (for open source bioinformatics only)

Interactive GPU

SIH GPU cluster

  • Interactive data exploration, visualisation, analysis

  • AI model inference workloads

  • GPU-accelerated workflows

Cloud

NCI Nirin

 

 

ARDC Nectar

  • Hosting research web services, databases, training environments, and lightweight analysis for collaborative projects

  • Free project allocations available via ARDC allocation schemes (support provided by SIH)

 

RONIN-AWS

  • Short-term, on-demand compute for population genomics, health informatics, AI training

  • Pay-as-you-go AWS billing access administered by ICT with SIH support

Virtual research desktops

Argus

  • Interactive accelerated desktops for sensitive health data, imaging, teaching, and other software requiring a graphical interface

  • Access via SIH, allocations and support provided for approved projects

High performance computing (HPC)

High performance computing refers to the use of parallel processing techniques to solve complex computational problems efficiently. HPC systems consists of clusters of interconnected computers, each equipped with multiple processors and large amounts of member. These systems are capable of handling massive datasets and perform computations at speeds far beyond those achievable by your personal computer. HPC provides a reliable and efficient means of analysing data of all shapes and sizes from research domains.  

To use HPC, it is not a requirement that your workflows make use of the multi-node architecture. There are many reasons why HPC would be justified:  

  • Large input data requiring vast physical storage for inputs and outputs  

  • High CPU or node requirement  

  • Batch processing GPU requirement  

  • High memory requirement  

  • Long walltime requirement  

  • Faster I/O operations than your local computer can handle  

  • Freeing up your local computer resources for other tasks or simply shut down for the day without stopping the analysis you’re running  

These systems are commonly used for astronomy, genomics, chemistry, earth and environmental sciences, and materials sciences research applications.

USyd researchers have access to nationally subsidised Tier-1 HPCs. Please refer the Sydney Scheme page for more information.

Gadi is a national HPC system that can be accessed by all Australian researchers. It is located in Canberra and is suited to high-throughput workloads that can run on Intel CPUs and NVIDIA V100 and DGX A100 GPUs. See our NCI Gadi guide for more information.

Setonix is a national HPC system that can be accessed by USyd researchers. It is located in Perth and is suited to high throughput workloads that can run on AMD EPYC CPUs and Instinct M1250X GPUs. See our Pawsey Setonix guide for more information.

Interactive GPU computing

Cloud computing

Cloud computing enables researchers to access scalable, on-demand computing resources over the internet from their laptop. Unlike HPC, which often involves scheduled batch jobs and shared queues, cloud platforms offer flexible environments well suited to interactive work, customised software stacks, and rapidly scaling workflows.  

Cloud services can be used for:  

  • Hosting virtual machines and notebooks for analyses  

  • Deploying web applications or APIs 

  • Prototyping and testing software in isolated environments  

  • Analyses requiring custom operating systems, tools, or workflows  

  • Training and deploying machine learning models  

Cloud computing can be used to complement HPC by offering more interactive, user-configurable approach to computation, particularly useful for tasks that don’t benefit from multi-node parallelisation or when self-managed environments are required.  

USyd researchers have access to:

Nirin is an on-demand environment suited to interactive analysis and service deployment. It offers researchers a dynamic alternative to traditional HPC, while remaining tightly integrated with NCI Gadi and data collections, enabling workflows that span both platforms.

Nectar is Australia's national research cloud platform, offering researchers self-service access to scalable virtual machines, GPUs, and large-memory instances for data analysis, software deployment, and collaborative research.

RONIN is an interface for creating virtual machines in the AWS cloud. The simplified user interface allows you to quickly turn-on new machines that are appropriate for your workload, use them, and destroy them when you are finished.

Virtual research desktops

Argus is the University of Sydney’s virtual research desktop (VRD) project providing remote, on-demand, interactive, graphically intensive compute environments. VRD supports common, convenient software and environments for processing data generated at the University's core research facilities. For more information, see the Service Portal.