The university is using the Altix system's open-source environment to make it easily available to high-performance computing (HPC) applications in multiple disciplines, including mathematics, statistics, virtual photonics, atomic physics, chemistry, biology and even architecture. Strathclyde selected Altix 3000 over a PC-based cluster solution because of its flexibility in accepting different numerical codes, easy administration, low latency and peak power performance.

"The entire science faculty is excited about using the Altix 3000 system because its Linux OS-based, open-source architecture allows researchers to develop individual HPC projects much faster and more efficiently", stated

Professor Gian-Luca Oppo of the Department of Physics at the University of Strathclyde. "Altix allows a wide variety of scientific projects to be optimized on the same architecture and paves the way for collaboration and cross-disciplinary computation across many scientific subjects."

The SGI Altix 3000 supercluster offers global shared-memory across multiple nodes, scaling to hundreds of Intel Itanium 2 microprocessors. Each node in an SGI Altix 3000 supercluster can combine up to 256 processors in a single Linux operating system image. With its unique global shared-memory capabilities and industry-leading SGI NUMAlink interconnect fabric, the SGI Altix 3000 supercluster is ideal for managing complex data sets and complete work flows, enabling the highest levels of innovation for technical users.

The Parallel Computing Research Infrastructure (PCRI) project at Strathclyde University covers a multitude of scientific projects. For instance, researchers will leverage the SGI Altix 3000 supercluster to study virtual photonics, in which numerical models simulate new laser systems.

"The development of new photonic devices is at the heart of the future expansion of communication and information processing systems", added Professor Oppo. "High-performance computing and distributed visualization are crucial elements in virtual photonics, where new devices are invented and tested for performance ahead of expensive prototypes. Key collaborations between academic institutions and small- and medium-sized enterprises are based on the accuracy and speed of the numerical simulations and require a good balance of HPC and visualization to properly assess the performance of the virtual devices. It is paramount to have a good HPC-visualization balance, which was one of the most fundamental reasons we choose SGI servers when

developing new applications in virtual photonics."

"We are delighted to be assisting the University of Strathclyde in its scientific endeavours with a true end-to-end, fully supported Linux server that delivers record-setting performance levels. Our relationship with the

university spans a decade, and today's achievement with SGI Altix 3000 proves that this remarkable collaboration continues to show the world what is possible with computing technology in scientific research fields", stated Tim Butchart, Managing Director UK for SGI.

Launched at the beginning of 2003, the SGI Altix 3000 family of products has already been adopted by more than 200 customers, including more than 25 major research organisations around the world, such as the University of