The Trusted Leader in High Performance Computing


Press Release

Swinburne University of Technology Selects SGI to Accelerate Astrophysics Research

Hybrid CPU- and GPU-Based System Will Serve as National Facility for Astronomers

Fremont, Calif. — July 13, 2011 — SGI (NASDAQ: SGI), a trusted leader in technical computing, today announced that the Swinburne University of Technology in Melbourne, Australia, has selected an SGI® Rackable™ and Altix® UV 10 high performance computing (HPC) solution to accelerate the study of astrophysics for astronomers, students and other researchers.

The chosen solution consists of 69 Rackable C3108 computer servers and four Altix UV10 systems as large-memory nodes with over 975 total processor cores built with both Intel® Xeon® processor 5600 series and E7 product family and NVIDIA® Tesla™ C2070 and M2090 GPUs. The networking technology connecting the systems is a Qlogic QDR Infiniband non-blocking network. For storage, the solution has 13 SGI IS5000 storage arrays, providing 1.8PB of Lustre storage. This phase of the system, known as gSTAR for GPU Supercomputer for Theoretical Astrophysics Research, will have a direct link to the Australia Telescope National Facility (ATNF) in Parkes, NSW, to process telescope data outputs.

"This new SGI system will deliver more than 130 teraflops of computing power, making it over 10 times more powerful than the Green Machine, our existing supercomputer," said Dr. Jarrod Hurley, manager of Swinburne's supercomputer. "This addition opens many new avenues for cutting edge simulations and rapid processing of telescope data."

"The interest in combining CPU and GPU processing in HPC solutions is increasing," said Bill Mannel, vice president of product marketing at SGI. "It's critical that research institutions such as Swinburne have the ability to quickly and accurately capture, process and share this kind of data."

The first phase of the system is expected to be operational in September 2011, and will be housed in Swinburne's new Data Center. Phase two of the project, which will focus primarily on CPU-based compute nodes, is expected to be completed by early 2012.

About SGI
SGI, a trusted leader in technical computing, is focused on helping customers solve their most demanding business and technology challenges. Visit for more information.

Connect with SGI on Twitter (@sgi_corp), YouTube (, and LinkedIn.

Contact Information:

Ogilvy Public Relations
Analisa Schelle

© 2011 SGI. SGI, Altix and Rackable are registered trademarks or trademarks of Silicon Graphics International Corp. or its subsidiaries in the United States and/or other countries. All other trademarks are property of their respective holders.