As Pawsey’s Chief Technology Officer, Ugo Varetto not only manages Pawsey’s technical operations and scientific computing support, he also leads Pawsey’s overall technical strategy, putting him at the centre of Pawsey’s capital refresh and the integration of new technologies like quantum computing.
Ugo brings a wealth of experience to the role, with a background in software engineering and a more than 25-year history of developing software, designing IT solutions and leading research and development (R&D) projects. Working in Silicon Valley start-ups was only the beginning. “I have always had an interest in code optimisation, real-time computer graphics, low-latency networks and machine learning,” says Ugo. “Then all of the technologies I’d been working with for many years in commercial companies started to be applied to supercomputing. I got pulled into HPC applications in 2006 at the Swiss National Supercomputing Centre and have been contributing to R&D projects in HPC ever since.”
Pawsey has a number of groups that support researchers through both short- and long-term engagements using the HPC facilities. Ugo’s involvement with the senior scientists and university research leaders is pivotal to understanding the resources they need, allowing the various groups at Pawsey to best meet their requirements. Ugo notes that filling both an executive and a technical lead role can be challenging: “There is such diversity within the research community and the projects they undertake. I have an active role in planning and co-designing solutions with researchers in computing, visualisation and data management for their various research projects, and then also implement processes at Pawsey to make sure their resource needs can be met”.
Ugo is well-placed to advise on the best software or hardware technology to apply to a given problem, as he puts significant time into assessing and testing the latest technology. “At the Swiss Supercomputing Centre I contributed to evaluating and designing the first ever HPC solutions based on graphics accelerators. They’re now mainstream, and a significant part of our capital refresh. But that type of work is ongoing – I’m really proud that at Pawsey, we’re deploying and testing new technologies together with supercomputing centres around the world. From exploring quantum computing, to bringing our users into the exascale through activities like our PaCER program, it’s great to be able to offer the best-in-class support to our scientists.”
“Working at Pawsey (and HPC centres in general) is special because through collaborations with both researchers and computing vendors we can enable cutting-edge science through bleeding-edge technology. It’s really rewarding to make a tangible contribution to Australia’s research outcomes, and actively contribute to the design of future national supercomputing infrastructure at the same time. Pawsey in particular is unique because of the focus on large-scale real-time data processing and management brought in by the various radio astronomy facilities we support – the Murchison Widefield Array (MWA) and ASKAP – and in the near future the SKA.”
Radio astronomy is showing the potential for our next industrial revolution – Industry 4.0 – the digital transformation of production industries and industrial control systems. Its requirement for high-fidelity simulations and data analytics to manage the increasing size of data sets will require computational capabilities orders of magnitude above what is currently available in Australia, even with Pawsey’s capital refresh. Ugo expects to stay in the middle of it: “Exabytes of data are going to need exaflops of computing power….”