Press Release

SGI Increases the Velocity of Research in Universities Throughout Australia and New Zealand

SGI Altix Systems with Intel© Itanium© 2 and Xeon© processors and SGI InfiniteStorage Solutions to Increase Capacity of Research

SYDNEY, Australia (September 17, 2007)—ac3 Research Consortium, Central Queensland University (CQU), Massey University, Otago University and University of Sydney have all recently installed new SGI® Altix® servers and SGI® InfiniteStorage systems from SGI (NASDAQ: SGIC) to increase the capabilities of their research facilities and their internationally recognised research projects.

As well as these five wins, Altix servers and storage are being installed at the Australian Partnership for Advanced Computing (APAC) national facility, South Australian Partnership for Advanced Computing (SAPAC) facility and at Queensland University of Technology this year. SGI Altix is fast becoming the industry standard solution for scientific research and development in this region. SGI has developed a range of standard and specialized solutions to assist with a variety of customer challenges. Each of the five universities has different requirements and in each case SGI was able to offer the best value system consistent with existing systems and future needs.

"SGI has worked closely with universities for a number of years, and has a great wealth of specialized knowledge on the research and development being undertaken at these institutions," said Bill Trestrail, Vice President, Asia Pacific Region, SGI.

All of the SGI Altix servers included either Intel® Itanium® 2 or Intel® Xeon® processors. Matthew Jones, Group Manager, Digital Healthcare — Oceania, Intel, said, "The research community in the past has had to develop and utilize proprietary systems to meet their requirements. Access to the specialists required to develop, manage, implement and service these proprietary systems places unnecessary constraints on their scarce resources. Now they can purchase from an industry standard product set, whilst having the reassurance that these products are well known, tried and trusted. This allows them to maintain their focus on the process of innovation."

ac3 Research Consortium
Since the SGI® Origin® 2400 system was decommissioned in 2004, the NSW based ac3 Research consortium was without a genuine capability system to support the advanced computational requirements of researchers in its partner universities. To remedy this, the University of Technology Sydney (UTS), on behalf of a consortium of five universities; University of Sydney, University of New South Wales, Macquarie University and University of Wollongong, led a successful application for Infrastructure funding from the ARC in 2005. With ARC funds, together with matching contributions from partner universities, the ac3 consortium sought a genuine capability facility with a very large system memory in order to accommodate the advanced modeling requirements of a number of prominent research groups in NSW—which include two ARC Centres of Excellence, two ARC Networks, a number of Australian Federation and Professorial Fellows, and research which is supported by a large number of competitive grants.

Amongst these are the areas of:

• The modeling of advanced photonic devices through the ARC Centre of Excellence, CUDOS, which is developing key enabling technologies for the next generation of ultra-high bandwidth optical communication systems
• Climate change modeling and earth system science, involving the development of a new climate model ACCESS for use by Australian researchers
• A wide range of computational chemistry, including research into the chemistry of free radicals (through an ARC Centre of Excellence) that have a role in various human disorders including Alzheimer's disease
• A gamut of simulations in nanotechnology and material science
• The solution of heat and fluid transfer problems in computational engineering
• Computational physics including quantum many-body problems, space physics and brain dynamics

All of these applications require access to high-performance computation and very large shared and distributed memory.

"The SGI Altix 4700 was a natural choice, since it was extremely fast and offered the largest single image shared memory in the Asia-Pacific, as well as being able to handle the broad range of applications encountered in a shared facility such as ac3", said Lindsay Botten, Professor of Applied Mathematics at UTS, who led the grant and the acquisition of the new facility.

SGI delivered in August, an SGI® Altix® 4700 with 256 cores and one terabyte of shared memory, a 15TB SGI® InfiniteStorage NAS 4050 system and Spectralogic Tape Library. The system comes with 128 Dual-Core Intel Itanium 2 9000 series processors.

Central Queensland University (CQU)
CQU required a new system because their last system had reached capacity. In June, SGI provided a new system for mathematical modeling to support engineering and bioinformatics research, which centers on computational processing rather than data processing.

Three vendors' code samples in cluster configurations were tested, and the architecture proposed by SGI offered a comprehensive solution that included best-of-breed products that could be incorporated into CQU's environment with installation, support and training.

"CQU required a significant increase in computational capacity. We have a significant amount of in-house developed software that supports a variety of research both industrial- and theoretical-based which comes under the heading of ‘Non-Linear Intelligent Systems'. The primary goal for CQU was to find a system that performed well with CQU's unique computational code," said CQU's Prof. Jennelle Kyd, "The architecture proposed by SGI provided the best performance with our in-house code. The computational tasks were clearly understood by the Altix system and it responded well."

To continue its Non-Linear Intelligent Systems research, CQU purchased a configuration consisting of SGI® Altix® XE310 computational nodes, an Altix® XE240 file server and Altix® XE210 administration and head nodes powered by 43 Quad-Core Intel Xeon 5300 series processors and an SGI® InfiniteStorage 220 system with SGI® Data Migration Facility (DMF) Hierarchical Storage Management System software environment.

Key areas of research of the ‘Non-Linear Intelligent Systems' include:

• Non-linear modeling
• Control
• Optimization
• Neural networks
• Fuzzy logic
• Evolutionary computation
• Robotics
• Image processing
• Train and vehicle dynamics

Massey University
To support the BeSTGRID project*, Massey University chose SGI Altix XE310 compute nodes and Infiniband Switches, capable of 2.2 teraflops using 52 Quad-Core Intel Xeon 5300 series processors, which provide the high-performance computing capacity it required. SGI delivered the cluster to Massey in August.

"More than 5 times the computation can take place on the new SGI Altix high- performance computer," said Dr. Chris Messom, Massey University. "We are excited about the Infiniband connection between the nodes, being able to share memory across nodes and access high speed I/O. The higher speed Altix system allows communication between processing elements with high bandwidth, and lower latency, resulting in increased productivity."

Research projects which will use the new SGI system include:

• Theoretical and Computational Chemistry group studying relativistic effects in chemistry, quantum electrodynamic effects in atoms, electroweak interactions, superheavy elements
• Theoretical and Computational Physics group studying quantum lattice chromodynamics and solitons in quantum gases
• Computer science researching parallel machine algorithms and Grid computing tools
• Molecular Ecology and Evolution group studying molecular rates of evolution, biodiversity and molecular anthropology

Otago University
Otago University required a new computational system able to cater to forty users, all working on large scale numerical applications in several areas of research including molecular modeling, quantum field theory and atom optics.

In July, SGI installed an SGI® Altix® XE1200 cluster solution comprising the ScaliManage, PBSPro, SGI® ProPack™ software, and Intel software development environment. The hardware comprised SGI Altix XE240 and XE210 nodes using 27 Quad-Core Intel Xeon 5300 series processors and an InfiniteStorage 220 disk subsystem for the shared file system. The system was installed within three days, and the staff has found the transition to the new system very easy.

"SGI has demonstrated an excellent understanding of our needs and the ability to architect a solution to effectively address those needs. They offered us a comprehensive solution that was also excellent value," said Otago University's Professor Rob Ballagh.

University of Sydney
Sydney University's original motivation to upgrade was that the prior multiprocessor machine in Physics was aging and SGI was able to offer the best all-round package of hardware, system architecture, and software.

This package consisted of the Altix XE1300 with 600 Xeon processor cores (150 Quad-Core Intel Xeon 5300 series processors), 1.2TB distributed Memory, SGI® InfiniteStorage NAS Gateway 1050 Storage node with NAS Manager software, 8TB SGI® InfiniteStorageTP9300 Storage Array, and Spectralogic Tape Library. SGI is expected to deliver this solution in October.

The main projects in the physics department to use the new Altix system will include research into complex systems, neurodynamics, self-organization, plasma theory, space physics, biophysics, condensed matter physics, molecular dynamics, and astrophysics. Professor Peter Robinson, University of Sydney said, "The system will have the processor and memory capacity to handle the next generation of mid-range problems that yield scientific end results themselves and/or are stepping stones to more intensive supercomputer-level computations."

"With the range of complex research being conducted in Australia and New Zealand it is important to provide solutions that can compute increasing amounts of data," added SGI's Bill Trestrail. "By providing industry standard solutions for universities, SGI is demonstrating its commitment to increasing the velocity of research in the region."

* BeSTGRID is a Tertiary Education Commission Innovation and Development Fund Project 2006-2008, focused on how to make eResearch work, to create a fully-functional eResearch ecosystem for New Zealand. BeSTGRID will deliver mechanisms, methods and tools that facilitate collaboration on shared information, sharing of computational resources and online visualization of instruments and experiments.

SGI - Innovation for Results™
SGI (NASDAQ: SGIC) is a leader in high-performance computing. SGI delivers a complete range of high-performance server and storage solutions along with industry-leading professional services and support that enable its customers to overcome the challenges of complex data-intensive workflows and accelerate breakthrough discoveries, innovation and information transformation. SGI helps customers solve significant challenges whether it's enhancing the quality of life through drug research, designing and manufacturing safer and more efficient cars and airplanes, studying global climate, providing technologies for homeland security and defense, or helping enterprises manage large data. With offices worldwide, the company is headquartered in Sunnyvale, Calif., and can be found on the Web at sgi.com.

© 2007 SGI. All rights reserved. SGI, Altix, Origin, the SGI cube and the SGI logo are registered trademarks, and ProPack is a trademark, of SGI in the United States and/or other countries worldwide. Intel, Itanium and Xeon are trademarks or registered trademarks of Intel Corporation or its subsidiaries in the United States and other countries. All other trademarks mentioned herein are the property of their respective owners.

This news release contains forward-looking statements regarding SGI technologies and third-party technologies that are subject to risks and uncertainties. These risks and uncertainties could cause actual results to differ materially from those described in such statements. The reader is cautioned not to rely unduly on these forward-looking statements, which are not a guarantee of future or current performance. Such risks and uncertainties include long-term program commitments, the performance of third parties, the sustained performance of current and future products, financing risks, the ability to integrate and support a complex technology solution involving multiple providers and users, and other risks detailed from time to time in the company's most recent SEC reports, including its reports on Form 10-K and Form 10-Q.