Support and resources

Attention researchers

Please see the University of Sydney Research Computing documentation for a comprehensive overview of support available to you as a University of Sydney researcher using NCI resources.

Technical documentation

Familiarise yourself with Gadi using NCI’s documentation and the University of Sydney’s Research Computing documentation.

• NCI Gadi user guide
• USyd Research Computing documentation

Live and self-directed training

If you are new to HPC and/or NCI, we suggest you attend NCI’s Intro to Gadi courses.

There are various trainings offered by SIH and NCI on HPC and data analysis topics:

• SIH training calendar
• NCI training calendar

Demonstrations

HPC experts at SIH have created a series of videos to help you get jobs running on Gadi:

• Summarising the Sydney Scheme and how to access NCI
• Adapting Artemis job scripts to Gadi
• Data transfer between RDS and Gadi
• Working within the Gadi walltime limit
• KSU estimation and compute resource benchmarking on Gadi
• Running embarassingly parallel jobs on Gadi (replaces Artemis job arrays)

I need to talk to someone

Please contact the following people depending on your needs:

Type of issue	Who	How	Details
Service unit allocation for running jobs	Sydney Scheme managers	Make a request	Sydney Documentation
Request for increased g/data storage	Sydney Scheme managers	Send an email	Sydney Documentation
Request for increased scratch space	NCI Helpdesk	Log an NCI ticket	Provide brief justification for the need including number of files, size of files, data lifecycle
Request for increased iNode quota in gdata or scratch	NCI Helpdesk	Log an NCI ticket	Provide brief justification for the need including number of files, size of files, data lifecycle, efforts to reduce iNode footprint
A technical issue with NCI systems or compute jobs	NCI Helpdesk	Log an NCI ticket	Provide all relevant information such as your error, log file, jobid
A bug in an SIH pipeline	SIH	Submit an issue on Github	Provide details e.g. this issue
General research computing or data analysis advice	SIH	Log an SIH ticket	Provide relevant context, errors, tool names, scripts