From the beginnings of GPU processing to HPC for all
Chris has had a passion for computers as long as he can remember. He cites an early example: “For my Year 3 presentation in primary school, I took apart a Compucorp 625 Mark II computer to show to the class, complete with Zilog Z80 8-bit processor, 64 KB of RAM, dual 5.25” disk drives, 80×16 character display, and built-in dot matrix printer. Soon after that I started learning to program in BASIC on an Amiga 1000.”
That interest persisted to a PhD focusing on the use of GPUs for signal and image processing, and his research played a role in pioneering their use for radio astronomy signal correlation. From Year 3, his developing skill set was tailor-made for Pawsey, and he joined the Supercomputing Technology and Application Programming Team in 2012.
Since then, Chris has worked across many facets of Pawsey. “I started out providing front-line helpdesk support for Pawsey researchers – helping them install software, set up workflows, and use the systems at scale. But over time I’ve also coordinated the merit allocation processes for our supercomputers, led our uptake projects, and been involved in our outreach, training and user feedback activities.”
“I really enjoy the fact that there is no ‘normal working day’ at Pawsey. There’s such a variety of work, depending on the projects currently under way. I could be working on procurement documents, technical reviews of merit applications, or calls for uptake projects. There’s always technical discussions with colleagues, and travel to provide face-to-face training and hold user forum discussions.”
Chris is always looking ahead as HPC develops, exemplified by his work on technology demonstrator projects. “I loved working on our Advanced Technology Cluster ‘Athena’ – collecting the user requirements, coordinating the technical review before procuring the system, running the early adopter process and reporting on the appropriateness of the technology.”
Projects like Athena have been critical in allowing Pawsey users to try out emerging technologies in preparation for the current capital refresh. Equally, the user forums that Chris coordinates are opportunities for both Pawsey users and staff to discuss Pawsey services and infrastructure, allowing staff to better understand user needs, and users to better understand Pawsey services and infrastructure. The resulting regular feedback is used to improve many aspects of Pawsey operations, and inform new developments like the capital refresh.
Chris points out that it’s all fundamentally about enabling research. “I really enjoy working with researchers to help them get the most out of Pawsey’s infrastructure. Coordinating the merit allocation calls, I’ve developed a real appreciation for the breadth and depth of computational research that we support across hundreds of projects and users from so many different scientific disciplines.”
“It’s so rewarding to see the transformative effect our services can have on our users’ research. We’ve taken workflows that take months to process on workstations and moved them to supercomputing systems where they complete in hours. This increases the researchers’ productivity by drastically reducing the time to result, and often also improves research quality, by allowing researchers to iterate and improve on those results in a reasonable timeframe.”
“I think HPC is becoming a mainstream necessity as the scale of data and associated computation grows across all areas of scientific research. It will be an ongoing process for us to become more accessible for a wider range of end users.”