INTERCONNECTS

Quadrics announced its partnership with The Pennsylvania State University in building its latest high-performance computing cluster, Lion-XL, to further the research computing endeavors of scholars from a variety of departments and disciplines. The new Lion-XL cluster consists of a total 176 Dell PowerEdge 2650 servers, each configured with dual Intel Xeon processors, 4 GigaBytes of memory and a 36GigaBytes Ultra3 15K rpm SCSI drive. All of the Lion-XL nodes run RedHat's Linux operating system. The first subset of Lion-XL, 128 dual 2.4 GHz cpu nodes, are connected with Quadrics QsNet high performance network. The remaining 48 nodes, with dual 2.8 GHz cpus, are connected with gigabit ethernet. Penn State has become the latest customer to select Quadrics as a way to gain high-performance computing at an affordable price. Using Quadrics' QsNet high bandwidth, low latency interconnect, Penn State has built Lion-XL, a nearly 1.0 Teraflop machine, with standard components that offer the computation speeds comparable to specialized, and much more expensive, high-performance computers. Quadrics also recently announced its involvement in the fastest Linux cluster in the world (#5 in the Top500) at Lawrence Livermore National Laboratory (http://www.supercomputingonline.com/article.php?sid=2944). Quadrics is using these clusters to gain further endorsement of its QsNet products (which have, and still are, used in proprietary cluster SMP systems) as the leading high performance network supplier for commodity based clusters as well. Investment in commodity based Supercluster systems as opposed to traditional supercomputers can significantly reduce the overall cost of ownership for an organization. Penn State's Vijay Agarwala, director of graduate education and research services, said that "Quadrics is a leading vendor of high-speed interconnect technology, a crucial element in building scalable and well-balanced computational cluster." "Instead of researchers deploying small clusters in their offices...it's far more productive to build larger machines," Agarwala said. "Larger machines, when properly run, significantly lower the cost of ownership." "Working with leading organizations like Penn State will provide Quadrics valuable feedback on the optimization of Quadrics cluster operation for academia and provide opportunities for similar organizations to gain insight into the benefits of the Quadrics environment", said Drazen Stilinovic, General Manager of Quadrics Limited. "We look forward to continuing collaboration with Penn State in the future"