ENGINEERING

SAN FRANCISCO, CA -- Turbolinux(R), Inc., a leader in high-performance Linux for Internet infrastructure solutions, today announced the release of EnFuzion 7.0, a clustering technology that reportedly transforms a company's existing network of Linux, Unix, and Microsoft Windows servers or workstations into a supercomputer. No dedicated servers or workstations are necessary as EnFuzion integrates transparently into a business' current operating environment. EnFuzion is typically used for modeling complicated problems, from portfolio risk management assessment by investment banks to the calculation of frames for animated films by digital content creators to protein modeling by bioinformatics companies. EnFuzion allows users to take advantage of the processing power of their existing servers and workstations without needing to invest in specialized and dedicated systems or expensive supercomputer alternatives. In addition, EnFuzion runs existing customer applications without modification, and provides node failover and automated job scheduling that takes advantage of idle CPU cycles on cluster nodes. The concept behind EnFuzion is relatively simple and is sometimes referred to as grid computing or extreme clustering. Specifically, EnFuzion links computers within a company's network and emulates a supercomputer by drawing on the idle processing power of multiple desktop computers and servers connected to the company network. Companies can take advantage of the aggregate computing power without compromising their network security. "Businesses today face incredibly complex problems and need highly flexible computing solutions to help maintain their competitive edge," said Ly-Huong Pham, CEO of Turbolinux. "By creating pools of flexible processing power that can be accessed any time, anywhere, EnFuzion allows companies to free up supercomputing horsepower that already resides in their systems and leverage it for business results." Dynamic Workload Prioritization, Higher Throughput Means Faster Computing Times Through the power of parametric execution, EnFuzion's dynamic prioritization capabilities automatically match workload to existing computing resources. Specifically, EnFuzion splits user jobs into separate, discrete components for parallel execution. This allows multiple resources to simultaneously process the information for significantly reduced compute time and maximized resource utilization. Furthermore, no special coding is required so any application can be inserted into the EnFuzion network for parallel processing. AMP Asset Management, Australia's largest insurance company, is using EnFuzion to turn hundreds of its idle NT workstations at night into a clustered supercomputer. AMP uses artificial intelligence in financial modeling to uncover secrets hidden in vast amounts of data stored in its databases from markets worldwide. In one example, EnFuzion allowed AMP to run a financial model in a week that used to require 100 CPU days on a fast computer. According to AMR Research, a company that provides business application and technology research, grid computing technology is able to perform functions that are typically impossible in a standard desktop computing environment. The group predicts that grid computing will create huge cost-savings for business merely by effectively using existing computer equipment Latest Version Includes New Security Feature With SSL-based host authentication, EnFuzion allows only authorized hosts to use its capabilities. In addition, EnFuzion already supports the SSH protocol, user password encryption and user specified trusted hosts and executables. Specific EnFuzion features include: -- Support for clusters with up to 1000 nodes -- Support for Linux, Solaris, and Windows platforms -- Seamless integration with workstations and servers already deployed to take advantage of spare CPU cycles -- Speeds application calculations by up to 1000 times -- Integration with existing applications without need for modifications -- Automatic failed job rescheduling for node fault tolerance -- Automatic load balancing and resource sharing For further information visit www.Turbolinux.com