When NCI conducted stakeholder consultations and a survey of our users in early 2021, the demand for training was clear. Since then, our training program has grown into a comprehensive set of courses for new and existing researchers – those getting started on their journey into high-performance computing (HPC), and those looking for the advanced skills that can take their science to the next level. Over the past 18 months, NCI has conducted more than 8,000 hours of workshops, sessions and courses with over 4,000 participants. As Steve McMahon, CSIRO’s HPC National Partnerships Advisor said, “The investigation of existing training needs and the planning of future training is excellent.”
To offer introductory courses quickly and effectively, we have worked with Intersect Australia, a national eResearch support agency. These fundamental programming courses, offering some of the core skills of supercomputing and data analysis, have been a great success. Dr Tracy Chew, Bioinformatics Group Lead from The University of Sydney said, “The introductory courses NCI offers in partnership with Intersect provide an inroad for different cohorts of researchers to accelerate their transition to computational platforms and frameworks.”
To date, this Skill Sharpening program has trained 675 attendees from more than 40 institutions across Australia. Professor Attila Mozer from the University of Wollongong, a chemical engineer, physical chemist and laser spectroscopist who also benefited from the program said, “I kept coming back to this training program because the courses were nicely organised. I've done all the Python machine learning courses because I'm very interested in that. There was also a logical progression from one course to the other. The training is very accessible, and I can learn some practical relevance to what I'm doing.”
Since the start of our training program, we have also been developing our own, NCI-designed training courses in partnership with academics and computational professionals from our community. To address the skill and knowledge gaps that off-the-shelf online training does not fill – and given the growing researcher interest in topics beyond the traditional HPC topics – we decided to design and develop a series of Artificial Intelligence (AI) and Machine Learning (ML) courses to showcase the research possibilities of this technology. Our courses, known as the AI/ML Applications on Gadi series, advise researchers on how to implement these techniques in their respective science domains. In addition, we partner with industry collaborators and subject matter expert organisations, such as NVIDIA and the HPC-AI Council, to facilitate other learning activities. This creates hands-on and easy-to-adapt educational pathways for science students and researchers to innovate using AI and ML.
“By way of the AI and ML course series, we are contributing to the Australian HPC-AI community. We hope to contribute new opportunities through these courses to foster collaboration among researchers, enable equitable access, and empower an inclusive open-source community,” said NCI Director Professor Sean Smith. By addressing participants’ immediate needs and showing them what can be done to make their codes run more efficiently, the courses have been hugely successful at raising the skill levels of researchers. The popularity of the courses thus far is testament to their effectiveness.
The wide array of training courses and workshops, run on a regular basis, cover an ever-growing range of topics including introductions to the Gadi supercomputer and the NCI cloud-based Australian Research Environment (ARE), parallel programming, AI/ML Applications on Gadi, code profiling, data-intensive analytics, and technical and scientific webinars. A major benefit of NCI’s training is that users can easily take what they have learned and apply it to their existing workflows running on our computing platforms. Improved performance and higher accuracy with unchanged resource needs are some of the obvious benefits for the researchers. Abdullah Shaikh and Doris Harrison from UNSW Research Technology Services said that “NCI’s training program fills the gap of upskilling the researchers around using HPC and computationally intensive code on Gadi. The program was very timely and was much needed.”
Supported by the Advisory Board and Executive Team, NCI’s training program has gone from strength to strength. Operating under a collaborative framework, expert members of NCI staff are involved throughout course development and delivery to help produce the best course for our users. Dr Jingbo Wang, Training and Research Engagement Manager at NCI said, “It makes the NCI Training Team feel lucky to work within such a supportive environment.” Training statistics, post-workshop reports and demand analysis help us capture the landscape of skills and needs out in the community. This data is critical for our stakeholders, Advisory Board and Executive Team for road mapping and planning future courses. In fact, this process is already bearing fruit. Abdullah Shaikh and Doris Harrison from UNSW Research Technology Services said, “The NCI course catalogue is a great mixture of the suite of services a researcher would need. Their post-training follow-up and reporting to the stakeholders is also very valuable.”
According to Dr Christina Hall from the Australian BioCommons, “NCI uplifts different scientific domains through a range of tailored activities. Biologists and bioinformaticians are supported by training outreach events like the Australian BioCommons webinar High performance bioinformatics: submitting your best NCMAS application, and this research community also benefits from NCI's participation in the National Bioinformatics Training Cooperative.” The successes of NCI’s training program after 18 months of development and growth are clear. The benefits to the researchers are becoming apparent through every session we run. All of it is only possible due to the valued support from our Collaborators, stakeholders and research community. As Dr Tracy Chew said, “The Sydney Informatics Hub is excited to partner with NCI in the future, to complement their training program with our internal approaches, balancing the requirements of our diverse users, from clinicians to scientists, and linguists to policy advocates - and everyone in between. We look forward to additional courses, showcasing the diversity of the research possible using High-Performance Computing, inspiring new users to join the platform and become power compute magicians.”
NCI’s training team will continue to consult with our community to develop new courses which address ever growing demands, and which bring research awareness of new technologies to advance science and research. We look forward to offering more exciting learning opportunities to national and international HPC-AI-Data-Science communities.
All of NCI's training offerings are outlined on our Training and Educational Events page.