HP today introduced visualization technology designed to make high-performance computing (HPC) more affordable and productive for a broader range of customers. Launched at the SC/05 Supercomputing Conference, the HP Scalable Visualization Array (SVA) uses industry-standard components and leverages advances in clustering, graphics and networking technology to allow scientists and engineers to visualize and analyze complex sets of data. HP SVA is a key part of the HP Unified Cluster Portfolio, which is designed to integrate computation, data management and visualization in a single cluster environment, greatly enhancing productivity in HPC environments. Driven by this portfolio, HP continues to lead the HPC market in revenue with more than 30 percent market share, according to second quarter 2005 figures released by IDC. Customers in aerospace, oil and gas, automotive, manufacturing, defense, medical and scientific research organizations using the HP SVA can match their growing computational capability with powerful visualization tools. SVA technology can produce a vast, high- resolution display wall of 100 million pixels or more at approximately half the cost of competitive proprietary products. "Our ability to generate large data sets has outstripped our ability to visualize them, both for deriving science and verifying correctness," said Hugh Couchman, scientific director, SHARCNET, a prominent HPC institute in Ontario that conducts leading-edge research in science and engineering. "With HP SVA, integrated with the HP Scalable File Share and XC software, we will now be able to visualize and manipulate huge data sets in real time as an extension of computing and data management in the cluster." Along with the new array, HP announced additional enhancements to the HP Unified Cluster Portfolio, including: -- Version 3.0 of the HP XC System Software, including support for workstations operated as SVA nodes, large node options and new distributed management and provisioning features; -- Enhancements to HP StorageWorks Scalable File Share, including support for dual-gigabit Ethernet and Advanced Data Guard protection; -- Support for Microsoft Windows Compute Cluster Server 2003; -- Updates to HP ClusterPack for HP-UX, with new tools for easier custom configuration and console management; -- Expanded options for HP Cluster Platforms, including systems with dual-core Intel Xeon processors and HP Cluster Platform Express; -- Novell Validation Suite HPC and Grid Project updates. The HP SVA consists of a cluster of HP workstations running Linux, industry-standard graphics cards and network adaptors, and an integrated software system. System software includes HP XC System Software for cluster infrastructure and workload management, and HP Scalable Visualization Array software for configuration and job management. HP StorageWorks Scalable File Share and HP Remote Graphics software are optional. "The HP Scalable Visualization Array reduces the cost of visualization solutions through the use of industry standards and interoperable products," said Winston Prather, senior vice president and general manager, High Performance Computing Division, HP. "By increasing scalability and flexibility, and integrating computation, data management and visualization with our Unified Cluster Portfolio, HP is meeting the needs of more customers in more application areas and fostering faster scientific discovery around the globe." HP SVA with HP SFS and XC System Software In concert with the SVA launch, HP announced a new release of HP XC System Software and updates to HP SFS. XC version 3.0 supports HP workstations as SVA visualization nodes and large 16-processor node options, providing a single, managed cluster system that can support distributed and SMP applications. New features also include distributed management services, which allocate tasks to multiple nodes to improve system performance and availability, and expanded smart provisioning, which automatically determines cluster topology and the optimal configuration of system services. HP SFS updates include support for dual-gigabit Ethernet, providing a 50 percent increase in input/output bandwidth for Ethernet connected clusters; high-resiliency Advanced Data Guard protection, which protects against even double-simultaneous disk failures; support for double-density disks, providing up to 1,000 terabytes of storage per file system; and improved resiliency to better detect network failures. As part of a SC/05 StorCloud demonstration, HP plans to showcase a 64- terabyte, 40 gigabit-per-second HP SFS server connected over a Cisco Systems-powered wide-area-network to a cluster of HP Integrity servers located at Pacific National Labs 200 miles away. HP Unified Cluster Portfolio enhancements HP also announced that it will support Windows Compute Cluster Server 2003, which is scheduled for release in the first half of 2006. With the addition of Windows Compute Cluster Server 2003 to its Unfied Cluster Portfolio, HP will offer customers strong choices of HP-UX, Windows and Linux to meet their HPC operating environment requirements. Windows Compute Cluster Server 2003 is expected to be available on HP Cluster Platform 3000 and 4000 systems and Cluster Platform 4000 blade systems. The HP Message Passing Interface (HP-MPI), a leading implementation of the MPI standard and critical for creating and porting parallel applications, will be ported to Windows Compute Cluster Server 2003. In addition, HP and Microsoft are working with software vendors to port and optimize 64-bit applications to Windows Computer Cluster Server 2003. In support of HP-UX, HP ClusterPack 2.4 offers new tools for easier custom configuration and console management. It also includes many enhancements for easier deployment of HP-UX clusters in production environments where interruptions from adding or deleting compute nodes must be minimized. It is also now integrated with HP System Insight Manager software. The HP Cluster Platform 3000 now offers dual-core Intel Xeon processors as an option in the HP ProLiant DL380 control nodes. Dual- core AMD Opteron processors are available with both the HP Cluster Platform 4000 and the blade-based HP Cluster Platform 4000BL. HP Cluster Platforms are available in clusters of up to 1,024 nodes. HP Cluster Platform Express is a faster, easier way to configure and order single-rack cluster solutions compared to traditional customization methods. Customers can select from a menu of popular cluster components, which are then factory assembled to reduce integration costs while shortening the time to deployment. More information about HP SVA and the HP Unified Cluster Portfolio is available at its Web site. Novell Validation Suite - HPC and GRID Project updates In other HPC news, HP announced that the infrastructure software for the Novell Validation Suite - High Performance Computing and Grid Project, announced in March 2005 by HP and Novell, has been validated. To date, HP and Novell and participating software vendors have validated a range of infrastructure applications, including Altair, PolyServe Matrix Server and Cluster Volume Manager, Scali Connect and Scali Manage, as well as Grid resource managers from Axceleon, Data Synapse, TurboWorx and United Devices. CAE applications from ABAQUS, ANSYS and MSC NASTRAN have also been validated for the project. The Novell Validation Suite program focuses on defining and certifying integrated multi-application and platform stacks on Novell SUSE Linux to reduce the risk of deploying complex Linux solutions in HPC environments.