Silicon Graphics Inc. used last week's LinuxWorld conference in New York City to show off two of its latest Linux entries: the Altix 350 mid-range server and the Visualization Developer Toolkit for Linux.
The Altix 350 server delivers the processing power required for life science research and other technical computing efforts. "The (Altix 350) is for the mid-range market encompassing technical database and high-throughput cluster applications," says Jacqueline Powers, marketing manager at SGI. "It's intended use is as a workgroup or departmental server."
The Altix 350 is roughly half the size of a common data center rack. It uses 64-bit Intel Itanium 2 processors, runs Linux, and makes use of SGI's NUMAflex shared memory architecture. The system can be configured with 1 to 16 Itanium processors and supports up to 192 GB of shared memory for an entire system.
SGI claims one of the advantages of the Altix 350 is that it allows a research group to scale up beyond the common two- or four-processor systems that are common in 64-bit Linux environments. At the same time, it can be used in processor- and database-intensive applications that today are typically run on large clusters with 32-bit nodes.
The 350 model rounds out SGI's line of Altix systems that includes the higher-performance Altix 3000 server and Altix 3700 Supercluster.
Starting price for the Altix 350 is US$12,199. A typically configured four-processor system has a starting list price of US$21,599.
Boosting Linux-based Visualization
Data visualization systems have typically relied on proprietary technology. At the LinuxWorld conference, SGI's demonstrated its Visualization Developer Toolkit for Linux, a software system that enables development of visualization applications for Linux systems.
The toolkit is part of a larger SGI Linux visualization initiative that includes the toolkit and collaboration with a number of open source graphics projects.
SGI says that this new effort is not intended to replace its existing Irix visualization products. Rather, the goal of the new initiative is to bring high-performance visualization to Linux-centric markets such as education and bioinformatics.