RS/6000 Fact Sheet

The machine behind the machine IBM RS/6000 SP Offers Real-World Solutions to Customers Somers, NY . . . The world's of business and science rely on the IBM* RS/6000* SP* to help solve the most complex and challenging problems. In technical computing, the SP aids scientists and engineers in the simulation and calculation behind much of today's progress in fields like nuclear research, medicine and oil exploration. The SP has also become the premier commercial computing platform for banks, retailers, insurers, manufacturers, healthcare and communications providers engaging in data warehousing, transaction processing, server consolidation, or on-line analytical processing (OLAP). Below are few examples of real-world solutions that IBM has delivered to its customers. Lawrence Livermore National Laboratory awarded IBM a $93 million contract to build the world's fastest supercomputer. By December 1998, the computer, an RS/6000 SP, will push computing speed to one trillion calculations per second. Known as tera-scale computing, this capability has many commercial and scientific applications, including medical simulations, global climate modeling, and aerospace and automotive design. The SP and tera-scale computing are part of the Department of Energy's (DOE) Accelerated Strategic Computing Initiative (ASCI). The first application of the SP will be in the Stockpile Stewardship and Management Program, a DOE effort designed to safely and reliably maintain the nation's remaining store of nuclear weapons. Advanced computations, specifically three-dimensional (3D) modeling and simulation capability, will be used to construct 3D simulations that will allow scientists to analyze the effects of aging and its implications on nuclear warheads or their components. Lloyd's of London, one of the world's premier insurance and re-insurance markets, transacts business worth billions of pounds in premiums every year. Between 1993 and 1995, they estimated the growth in the quantity of data processed to exceed 20 percent per year. In order to effectively manage their information resources, Lloyd's has chosen the SP to serve as the platform to facilitate direct access to their centrally held accounting and claims data. With the processing power and scalability of the SP, Lloyd's will be able to better manage their growing information management requirements. Lloyd's estimates that this system will provide significant cost savings over the next five years. Petroleum Geo-Services (PGS) is a geophysical service company that provides 3D seismic surveys to its customers to help them explore for oil and natural gas. PGS employs seismic survey vessels to record data using towed receiver cables, known as streamers. The streamers collect an enormous volume of data, 1 to 10 terabytes per survey, that must be rigorously processed. Previously, PGS would wait until the end of data acquisition and then transfer the data to a land based facility for processing, which would cause delays up to several months for the final product. By using the SP, PGS Tensor, the seismic data processing division of PGS, can now begin processing the data on board the survey vessels concurrent with data acquisition, allowing their customers quicker access to the final processed data sets. Additionally, the scalability of the SP allows PGS Tensor to easily expand or contract their on board computer capacity on short notice to meet their rapidly changing processing needs. Stichting Academisch Rekencentrum Amsterdam (SARA) is the foundation for academic computing in Amsterdam, Netherlands. SARA is using an SP to serve as the most powerful mathematical processing system in the Netherlands. Researchers from the Vrije University and the University of Amsterdam will use the system to carry out projects in various areas, including molecular design, research into certain medicines, advanced simulation and financial modeling. In the health care arena, researchers will use the SP to carry out analysis of the walls of the heart, blood cells, blood circulation and dental structures. Researchers from other scientific organizations will also be able to gain access to the SP via the Netherlands Organisation for Scientific Research (NWO). Eastman Kodak is converting the system used to manage its largest warehouses around the world from a multiple mainframe to a single client/server based warehouse management system. The move supports global supply-chain initiatives by Kodak and helps the company capitalize on advances in warehouse automation technology. Their goal is to configure each warehouse to take on additional value-added responsibilities such as flexible labeling, lot tracing, dating, variable picking strategies, and to utilize an array of logistics technologies. After looking at many options, Kodak decided to use two SPs to support their US and Canadian rollout. The cluster will support six warehouses ranging in size from 120,000 to 2 million square feet, with over 700 users, 800 RF terminals, printers, monitors, automated guided vehicles, and several high volume sorter/conveyor systems. The SP will also manage a detailed multi-year data warehouse supporting a variety of performance metrics, like activity-based costing, network configuration and flows, transportation planning, and carrier alliance compliance. The IBM RS/6000 SP is a general-purpose scalable parallel system based on a massively parallel architecture, that scales to accommodate as many as 512 processors to perform numeric-intensive and data-intensive tasks. The SP is used in such complex and large data processing environments as financial modeling, computation fluid dynamics, numerical analysis, data mining and decision support, LAN server consolidation and on-line transaction processing. Additional information on the IBM RS/6000 SP and related products is available through the World Wide Web on the Internet (http://www.rs6000.ibm.com).