$48m for Perth supercomputer

The Pawsey Supercomputing Centre has signed a $48 million contract with Hewlett Packard Enterprise to deliver the biggest upgrade to its computing infrastructure since the centre opened in 2009.

The new supercomputer will deliver up to 50 petaFLOPs, or 30 times more compute power than its predecessor systems Magnus and Galaxy.

Pawsey executive director Mark Stickells said the new investment would help the centre meet growing demand.

The centre is used by more than 1,600 researchers across 200 projects in fields such as medicine, artificial intelligence and radio astronomy and demand last year was double the centre’s capacity.

“More and more scientists are demanding computational resources to advance their research,” he told Business News.

“So there is the demand side, there is also cost and efficiency.

“We will be able to deliver more at a better return with new architecture.”

The Pawsey centre is home to one of two supercomputers in Perth, with the second owned by West Perth company DUG Technology.

Mr Stickells said it was great to have two high-performance computing facilities in Perth but noted they had different architecture, systems and business models.

While DUG is a commercial provider, Pawsey is a research facility that provides access via merit-based applications.

“We will have one of the most significant facilities in the southern hemisphere, connected into some of the largest computing facilities around the world,” he said.

The selection of HPE followed an open tender run by CSIRO, which is one of the owners of the Pawsey centre, alongside with WA’s four public universities.

Four compliant tenders were lodged.

HPE was selected after an assessment of three major criteria - energy efficiency, cost, and its integrated hardware and software solution.

The new supercomputer will be built using the HPE Cray EX architecture, following HPE’s $US1.4 billion acquisition of Cray last year.

The new system will be delivered in two stages.

Phase 1, available by Q3 2021, will provide a 45 percent increase in raw compute power in one-fifth of the size compared with the Magnus and Galaxy systems.

Full commissioning of the system will occur by the second quarter of 2022.

The new supercomputer will be at least ten times more power efficient than its predecessors Magnus and Galaxy, while providing a 30-fold increase in raw compute power.

The supercomputers are cooled by a groundwater cooling system specially developed by CSIRO for the supercomputing centre, which is offset by a 118kW solar photovoltaic system.

Pawsey said it will have more emphasis on accelerators with future-generation AMD EPYC™ CPUs and AMD Instinct™ GPUs and including expanded data storage capabilities with the Cray Clusterstor E1000 system.

The new supercomputer is part of a $70 million upgrade funded by a grant from the Australian Government in 2018.

Other upgrades include the commissioning of GPU clusters, procurement for new ingest nodes to support researchers using CSIRO’s Australian Square Kilometre Array Pathfinder (ASKAP) telescope and the upgrade of Pawsey’s Nimbus cloud service.

HPE general manager HPC and AI Nick Gorga said he was looking forward to collaborating with AMD to build Pawsey the most powerful system for the region and boost Australia’s research capabilities.

CSIRO postdoctoral fellow Dr Chenoa Tremblay illustrated the benefits of the Pawsey supercomputer by referencing his own research.

“We’re using Pawsey’s existing supercomputers to help scan more than ten million stars and analyse hundreds of terabytes of data to search for molecules that provide potential evidence of extraterrestrial life,” Dr Tremblay said.

“Doing this on my laptop would take 25 years, so having the power of the new supercomputer will help bring our research timelines from years down to days, giving us the power we need to analyse hundreds of thousands of images quickly.”