DEVELOPER TOOLS

BOOTH #525 CLUSTERWORLD EXPO, SAN JOSE, CA. -- The current TOP500 list shows a continued growth in clusters coupled with an increasing widespread use of Quadrics QsNet technology within Linux, especially within the capability sector of supercomputing. Now 6 of the top10 Supercomputers in the world use QsNet as the underlying network. The list of QsNet users (http://www.top500.org/list/2003/06/) includes 22 HP AlphaServer SC systems and 5 Linux based clusters. These include the second most powerul system in the world at LANL called ASCI-Q and the number 3 system, the world’s most powerful Linux cluster, at LLNL called MCR. "Quadrics is continuing to extend its market share within capability computing and production level supercomputing. With the advent of the next generation of QsNet later this year we anticipate further gains in this sector as performance, reliability and scalability become overriding factors in the large cluster configurations" commented John Taylor, Technical Marketing of Quadrics. The TOP500 list (www.top500.org), published twice a year at the ISC conference in June and at the SC conference in November, ranks supercomputers worldwide according to their performance on the LINPACK benchmark. The figure RMAX, or the efficiency of LINPACK, is a strong indicator of network performance for a given architecture. Systems in the TOP500 fall into a number of classifications. The most widely used architecture is based on the paradigm of distributed memory. This is broken down into MPP or cluster, depending ostensibly on the proprietary nature of the interconnect. For the classification of clusters, over 28% of the TOP500, based on RPEAK, deploy QsNet and these contribute to over 43% of achieved LINPACK performance for the group. The following list provides a snapshot of QsNet systems in the top 10. #2 The ASCI-Q system at Los Alamos National Laboratory is the most powerful supercomputing system in the US. It is based on the HP™ AlphaServer™ SC. It deploys 2 rails of Quadrics QsNet and spans over 2000 nodes. #3 The MCR system at Lawrence Livermore National Laboratory is the most powerful Linux supercomputer in the world at 7634 GigaFlops. The system was built by Linux Networx and is comprised of a customized QsNet network that is designed to provide an optimal balance of capacity and capability computing potential. #6 The ASCI Linux Cluster provides an identical (but slightly smaller) configuration to MCR and uses an IBM implementation of the Intel™ Xeon™ Processor. Its RMAX figure is 6586 GigaFlops. #8 The system at Pacific Northwest National Laboratory is the most powerful Itanium system in the top500 at 4881 GigaFlops. It uses Hewlett Packard™ 2 processor Itanium™ 2 nodes and will be the first system to deploy the Quadrics QsNet II native PCI-X based network. #9 The system at Pittsburgh Supercomputing Center entered the top500 some 2 years ago as the most powerful Supercomputer in academia with an RMAX of 4463 GigaFlops. It was the first system to deploy the federated QsNet network and comprises 2 rails of Quadrics QsNet. The system is based on the HP™ AlphaServer™ SC product and combines Intel processing nodes using Linux to create a customized heterogeneous QsNet configuration for high performance graphics and storage. #10 The system at CEA is the most powerful Supercomputer in Europe with an RMAX of 3980 GigaFlops. It is based on the HP™ AlphaServer™ SC product.