CSIRO has partnered with Dell to develop an advanced high-performance computing (HPC) system named Virga. 

The new computing cluster, based on Dell PowerEdge XE9640 servers, is designed to enhance scientific research while promoting energy efficiency through direct liquid cooling.

Virga represents a significant step forward in Australia’s scientific infrastructure. 

The system is equipped with NVIDIA H100 Tensor Core GPU accelerators to support deep learning, machine learning, and AI. 

Each GPU features 94GB of high-bandwidth memory. The cluster also includes a Transformer Engine, 4th Gen Intel Xeon Scalable processors, and hybrid direct liquid cooling to minimise energy use compared to traditional air cooling systems.

Professor Elanor Huntington, CSIRO’s Digital, National Facilities and Collections Executive Director, says Virga will provide the critical computing infrastructure necessary for machine learning and AI, thereby supporting the growth of Australia’s industry and economy.

AI is increasingly integral to CSIRO’s research across diverse fields, including flexible printed solar panels, fire prediction, wheat crop measurement, and vaccine development. “High-performance computing systems like Virga also play an important role in CSIRO’s robotics and sensing work and are crucial to the recently launched National Robotics Strategy to drive competitiveness, and productivity of Australian industry,” Prof Huntington says. 

Housed at CDC’s Hume Data Centre in Canberra, Virga is named after the meteorological phenomenon where rain evaporates before reaching the ground. The name is a nod to CSIRO’s extensive work on modelling cloud and rain physics.

Dr Jason Dowling from CSIRO’s Australian e-Health Research Centre says there is also an  urgent need for advanced computational power in medical image analysis due to the increasing volume and complexity of medical imaging data. 

“The new HPC facilities will allow researchers in our Australian e-Health Research Centre to train and validate new computational models, which will help us develop translational software in medical image analysis for image classification, segmentation, reconstruction, registration, synthesis, and automated radiology reporting,” Dr Dowling said. 

“One collaborative project with the Queensland Children’s Hospital that will benefit from the new cluster is the training of artificial intelligence (AI) models to diagnose pathology from MRI scans of the lungs in children with cystic fibrosis.”