The SGI^® NUMAflexTM global shared-memory architecture maintains balanced performance regardless of the configuration. The SGI Altix 3000 family has already established a world-record result for memory bandwidth scalability for microprocessor-based systems, as measured by the industry-standard STREAM benchmark. The shared-memory capabilities of NUMAflex allow users to create and analyze extremely large data models without having to artificially decompose them into smaller bits. As a result, SGI Altix 3000 is helping scientists and engineers push the boundaries of what can be done with open-source solutions.

Standard System, Serious Innovations

SGI Altix 3000 is binary compatible with industry-standard 64-bit Linux^®,bringing thousands of developers together in a common operating environment. Users with codes that already run on 64-bit Linux will hit the ground running; without any additional effort, applications will take advantage of the built-in, lightning-fast SGI^® NUMAlinkTM clustering fabric. Furthermore, SGI Altix 3000 offers layer upon layer of differentiated tools for superior data management, performance monitoring, load balancing, and resource management, so it's easy to tune your code to harness the full power of the SGI Altix 3000 supercluster.

SGI Altix 3000 was designed for technical users in such fields as research and education, manufacturing, and energy, who are striving to reach new heights in their high-productivity, open-source applications. Some of the innovation areas include:

Research/Education: Computational chemistry, physics, and biology involve the manipulation of extremely complex calculations or simulations, the outcome of which could never be guessed a priori even when the underlying theory is well known and the equations simple. For example, to study molecular dynamics simulations of the "melting" of argon clusters, one must evaluate billions of interatomic forces, updating atom positions and velocities and extracting meaningful quantities by time-averaging over thousands of time steps, a job that can only be accomplished with extremely powerful computers. The enormous processing power and speed of SGI Altix 3000 gives researchers the opportunity to load an entire data set into memory simultaneously. Every processor can access all the data as a single model, rather than analyzing separate components in a piecemeal fashion.

Manufacturing: Automobile crash simulation research is conducted using finite element analysis, a computer-based numerical technique for calculating the strength and behavior of engineering structures by breaking down the structure into many small simple elements. The behavior of each can be described with a relatively simple set of equations, which are then merged an extremely large set of equations that describe the behavior of the whole. By using a high-performance computer such as the SGI Altix 3000, researchers can efficiently and quickly analyze the astronomical number of calculations involved in testing any large structure.

Energy: The key to a profitable oil business is not just finding energy, but extracting it. An increase of just a few percent in the recovery rate can result in additional billions of dollars in revenue. To find those energy deposits and determine the exact spot to drill requires the analysis of literally terabytes of seismic information data. SGI Altix 3000 offers rapid, accurate analysis of the extraordinary amounts of seismic data necessary to discover deposits, reduce exploration costs, and maximize the productive yield of each well.

(continued...)	page 1 | 2
SGI Linux Environment 7.2 is based on Red Hat Linux 7.2 but is not sponsored by or endorsed by Red Hat, Inc. in any way. Linux penguin logo created by Larry Ewing.