Mission
The Scientific Computing and Visualization (SCV) group, within Information Services & Technology at BU, provides resources and services for high-performance computing and visualization in support of research that has specialized or highly intensive computation, bandwidth, storage or graphics requirements. Typical applications include scientific and engineering simulation, data analysis, visualization, and graphic arts.
Resources are managed in collaboration with the BU Center for Computational Science and with close consultation of the Research Computing Governance Committee. Most of Boston University’s Scientific Computing Facilities are available free of charge to the Boston University research community, collaborators, and students. Consulting, training, and educational services are also offered without charge.
Computation Facilities
The University’s most powerful system is an IBM Blue Gene supercomputer. It has 1,024 compute nodes, each containing a dual-processor PowerPC chip and 512 MB of memory, and a peak performance of 5.7 TFLOPS (trillion floating point operations per second).
Complementing the Blue Gene is the Katana Linux Cluster, which is made up of machines of several configurations. The cluster currently has 99 blades with a combined 588 processors.
These systems offer a wide range of programming languages, parallelizing compilers, mathematical and scientific libraries, graphics and visualization software, and discipline-specific application packages.
Data storage for these systems is provided by an 84 terabyte (TB) disk array which incorporates a parallel file system. A 500TB robotic tape system is available for backup and archival storage.
Networking
Boston University has a long history of running high performance networks both within the campus and in connecting with the rest of the region and the world.
The campus core utilizes 10-Gigabit Ethernet. SCV uses InfiniBand, Myrinet , 10-Gigabit Ethernet, and Gigabit Ethernet to interconnect its computing systems and connect to the campus core.
Boston University is a member of Internet2, a consortium of more than 200 U.S. research institutions, which develops and deploys advanced network applications and technologies in support of research and education. The University is also a founding member of Northern Crossroads (NoX), two dozen institutions committed to advanced networking in New England. The NoX and Internet2 maintain regional and national networks that interconnect at 10 GB/s; the NoX serves as an even higher-performance regional exchange to New England participants and commodity Internet service providers.
The University is also connected to the Metro Ring—144 optical fibers over a 7.4 mile radius through Boston and Cambridge.
Boston University’s optical fiber connections provide multiple GB/s of bandwidth and afford scalable connectivity to regional collaborators, the Internet, Internet2, and international research networks.
Visualization
SCV provides graphics and virtual reality facilities, and staff expertise supporting a broad range of research application areas.
The Computer Graphics Lab houses general access high-end graphics workstations and virtual reality facilities. Workstations include Windows and Linux computers running scientific visualization, molecular design, graphics, modeling, and animation software.
The Deep Vision Display Wall is a 15 x 8 foot stereoscopic display wall that can display high-resolution (18 MPixel) images and animations. The passive stereo rear-projection system uses multiple projectors driven by Linux-based computers. Integrated with the Display Wall is an 8-channel audio system, which can be used for digital audio playback, sonification (the auditory equivalent of visualization), 3D audio localization for immersive environments, and more. Software used to power the virtual environments includes SGI OpenGL Performer, OpenGL, OpenSceneGraph and SCV’s in-house DAFFIE system. Application areas include Biomedical Engineering, Chemistry, Space Weather modeling, Computational Fluid Dynamics, Archaeology, and the Fine Arts.
Consulting
SCV’s broadly trained staff assist members of the research and education community to accelerate the advancement of research and education.
- Scientific programming consultants assist in program parallelization, performance measurement, and code tuning, as well as with numerical methods and computational techniques.
- System consultants help in the use of the supercomputing systems and their various programming tools and packages, as well as data management.
- Graphics consultants assist researchers in developing visualizations, virtual environments, and graphics problem solving. They cover the use of graphics and visualization packages, the development of custom computer graphics tools, the application of graphics algorithms, the conversion of data into effective graphical representations, and creating virtual environments.
Training and Education
SCV offers live and online tutorials customized for the specifics of BU’s installation. A live tutorial series is offered each semester covering parallel programming techniques and languages, code tuning and optimization, and scientific visualization. On request, SCV will customize presentations for courses and seminars and hosts workshops, user groups, symposia, and vendor training sessions.
Software
Software on the IBM Blue Gene and Linux systems supports the development of parallel programs using data-parallel, message-passing, and shared-memory paradigms. Parallelizing compilers for FORTRAN and C are also available, as are mathematical and scientific subroutine libraries. Other packages are available for distributed computing, modeling and rendering, animation, image manipulation, scientific visualization, and plotting.
Programming languages
- C
- C++
- FORTRAN (F77, F90, and F95)
- High-performance FORTRAN
Parallel libraries
- OpenMP
- Message Passing Interface (MPI)
Job management
- Load Leveler, load balancer, and batch manager (Blue Gene)
- Load Sharing Facility (LSF)—load balancer and batch manager (pSeries)
- Portable Batch System (PBS)—load balancer and batch manager (Linux)
- Sun Grid Engine—load balancer and batch manager (Linux)
Mathematics, data analysis, and plotting
- IBM ESSL—Engineering and Scientific Subroutine Library
- IBM PESSL—Parallel Engineering and Scientific Subroutine Library
- Mathematica—symbolic math package with graphic display
- MATLAB and MATLAB PCT—math and engineering package
- Maple—interactive computer algebra system
- GAUSS—mathematical and statistical system
- Interactive Data Language (IDL)—data analysis and visualization package
- R—statistical analysis and graphing package
- Gnuplot—plotting package
- Xmgrace—plotting package
Scientific and engineering applications
- CHARMm—molecular modeling
- Gaussian—quantum chemistry
- Accelrys Discovery Studio—molecular modeling and simulation
- ABAQUS—computer-aided engineering and analysis
Visualization
- VTK—visualization tool library
- ParaView—visualization application
Animation and graphics
- Autodesk Maya—modeling and animation software
- OpenSceneGraph—high-performance graphics toolkit
- SGI OpenGL Performer—high-performance graphics package
- OpenGL/GLUT—3D direct rendering library and window utilities
- DAFFIE—distributed virtual environment software
- ImageMagick—image manipulation programs
Contact Information
Scientific Computing and Visualization
Information Services & Technology
111 Cummington Street
Boston, MA 02215
617-353-8190
E-mail: scv@bu.edu
http://www.bu.edu/tech/research/scv/
An equal opportunity, affirmative action institution.