CLOUD

By Steve Fisher, Editor In Chief -- Sun Microsystems of Canada Inc. recently celebrated the official opening of the Sun Center of Excellence (COE) for Visual Genomics at the University of Calgary (U of C). This new Sun COE joins a worldwide network of academic institutions developing advanced technology in computational biology. To learn more, Supercomputing Online interviewed Sun Microsystems’ Stephan Unger, Global Education & Research, Jonny Akerberg, Sr. Product manager, Visualization Products, and Subra Mohan, Product Line Manager, Visualization Products. SCO: Please provide some background on Sun's academic Center of Excellence program. UNGER: Sun Global Education and Research establishes close collaborations with carefully selected schools and non-profit organizations throughout the globe in various areas, including science and engineering, HPC, eLearning, libraries, administration, etc. The schools share our vision of building communities of users, and are willing to make a joint investment. Within computational biology, we have established COEs with the University of Wisconsin, Virginia Bioinformatics Insititute, Delaware Biotechnology Institute, Beijing Genomics Institute, University of Chicago, as well as the University of Calgary. More are planned, and we have been able to foster several collaborative projects among our COEs. SCO: Tell us about the high-end systems from Sun that were delivered to the University of Calgary. UNGER: University of Calgary is building a Sun Campus Grid for a 3-D representation biological system with a Java-3D enabled CAVE solution, with Sun Fire 6800, 4 Sun Expert3D Graphics Accelerator, 5 TB of SunStorage T3s, SunRays, and Sun Grid Engine. Sun's High Performance and Technical Computing (HPTC) strategy is to provide end-to-end solutions that incorporate technical computing requirements for efficient, highly scalable computation, efficient access and utilization of large sets of computational resources. Our goal is to offer a balanced approach that addresses data management and movement as well as scalable and parallel computation. Today, it is clear that the future of supercomputing is network-centric, and lies in both the shared power of smaller, parallel web-serving systems and large, cooperating clusters of SMP servers. Rather than focusing solely on traditional islands of computing power in a single location, Sun is bringing disparate commercial and scientific communities the hardware they need to share data and resources across the globe. The SunFire 6800 Server and Sun StorEdge T3 systems are is a key parts of the installation at the Center of Excellence at the University of Calgary, and a key part of any Visualization and HPC solution. Another interesting feature within the installation at the University of Calgary is the Java 3D-enabled CAVE solution. Java 3D is an Application Programming Interface (API) that allows developers to quickly create 3D applications using the Java programming language. Java 3D allows developers to concentrate on their content by programming to a scene-graph model. By using Java 3D, developers get to program in Java, have access to a scene-graph, and get the many people-years of experience of taking advantage of OpenGL in the most efficient manner. SCO: Why do you think Sun was chosen over the competition? AKERBERG: Working technology, shared vision and partnership, and Sun's long-term commitment in the Technical Computing segment. SCO: Being as specific as you can, please tell us how UC will be utilizing Sun technology. UNGER: The University of Calgary is creating a virtual CyberCell which will be displayed in a Java 3D technology-enabled CAVE(r) from FakeSpace Systems Inc. The relevant biochemistry will be displayed, and analyzed, within the cell. Also, a visible human will be displayed and biochemistry represented within it. In addition, work on enhancing their MAGPIE and BLUEJAY bioinformatics portals will be undertaken, as well as building a local portal between schools within Alberta (University of Calgary, University of Alberta and University of Lethbridge), as well as tapping into the NRC CBR bioinformatics grid for collaborative projects. SCO: How is Sun doing in the market for collaborative visualization solutions? MOHAN: We view the collaborative visualization solutions market as ranging from individual desktop solutions to Small Work-group (5-10) solutions to Large Group Visualization (25-50). Until recently, most of the deployment activity has been around Large Group Visualization solutions that cost $1M-2M which typically include custom display solutions and an expensive SGI system, which has limited the growth in this segment. Lately, we are seeing a lot of growth in Small Work-Group visualization solutions, as customers are trying to provide such technology to a broader group of users spread across their organization. Sun is benefiting from this trend, since the Sun Blade 1000 workstations equipped with Expert3D graphics accelerators, have hit the mark with respect to price-performance. We have had several successes in the Petroleum industry with our Sun Blade 1000 systems configured with dual Sun Expert3D graphics accelerators, used for Collaborative Visualization solutions. SCO: Sun recently announced Sun Expert3D graphics accelerator support on the Sun Fire 6800 server, what sort of impact do you see this having on customers and potential customers. Why should people take note of this announcement? MOHAN: Support for Sun Expert3D graphics on the Sun Fire 6800, is the first incarnation of product configurations which Sun is targeting at the technical markets which have both demanding computational and visualization needs. There are several markets in the HPC sector such as Life Sciences, Medical Research, CFD, etc.; where scientists and researchers run large simulations while operating on extremely large data-sets, and rely on visualization technology to enable them to drive the simulations, and verify their results. An integrated solution incorporating support for multiple Expert3Ds on a Sun Fire 6800, will enable increased interactivity with their datasets in a Collaborative Visualization environment, without having to move their large datasets from their compute system to a separate visualization system. SCO: Is there anything you'd like to add? MOHAN: This is just the beginning, and a sign of great things to come from Sun in this area. We are extremely excited about our role in this market, and look forward to introducing more powerful products that truly address the performance and quality needs of the visualization customers at price points that will enable them to truly leverage such technologies across their organizations, reducing their development costs in the process.