Parallel Computing Research Groups, Projects and Services

• Alewife Project at MIT. This is a large-scale multiprocessor which integrates both cache-coherent, distributed shared memory and user-level message-passing in a single integrated hardware framework. Each Alewife node consists of a 33 MHz Sparcle integer unit, an off-the-shelf FPU, 64 kbytes of direct-mapped cache, and 4 Mbytes of globally-shared main memory. The nodes communicate via messages on a two-dimensional mesh network. The current implementation scales directly to 512 nodes. A single-chip Communication and Memory Management Unit (CMMU) on each node holds the cache tags and implements the memory coherence protocol by synthesizing messages to other nodes.
• Computer Architecture page at University of Wisconsin.
• BECAUSE - Benchmarking concurrent architectures at Informatics Department, Rutherford Appleton Laboratory, UK.
• BRIM (Brunel, RAL, Intercept, Meiko) - Parallelising compiler for Fortran at Informatics Department, Rutherford Appleton Laboratory, UK.
• CacheMire research group at Department of Computer Engineering, Lund University, Sweden. The main focus of the project is design and performance issues of memory systems for shared-memory multiprocessors including hardware- and compiler-based optimization techniques and performance prediction methods and tools.
• Supercomputing, Parallel Computing, and High-Performance Computer Architecture Research Groups (Jonathan Hardwick - jch@cs.cmu.edu) at Carnegie Mellon.
• The Charm portable parallel programming system at Programming System Laboratory, University of Illinois.
• Data Diffusion Machine (DDM) project at SRF-PACT - a virtual shared memory architecture where data is free to migrate through the machine.
• Dome - Distributed objects for parallel computing at Carnegie-Mellon.
• Stanford FLexible Architecture for SHared memory (FLASH) project: Multiprocessor research developed in partnership with the Supercomputer Systems Division at Intel and ARPA.
• Fortran 90 and High Performance Fortran
  • Fortran 90 Tutorials based on the Fortran 90 book by Michael Metcalf.
  • Fortran 90, Fortran 77 and C comparisons
  • NAG Ltd's Fortran 90 library.
  • Fortran 90 Software Repository at NAG Ltd.
  • Evaluation of High Performance Fortran at ICASE.
  • The Fortran M Programming Language. FM is a small set of extensions to Fortran that supports a modular approach to the design of task parallel, message-passing programs.
  • Fortran Tools Repository at Rice University (Ken Kennedy - crpc@cs.rice.edu)
  • PARTITA: an interactive FORTRAN parallelism analyzer, part of the FORTRAN reverse-engineering system FORESYS. (Bernard Dion - Bernard.Dion@sophia.inria.fr)
• Fortran M - Task-parallel Fortran at ANL.
• General-Purpose Parallel Computation at Queens University, Kingston, Ontario, Canada.
• Parallel Genetic Algorithms:
  • local FAQ list.
  • comp.ai.genetic FAQ at RTFM
  • Work done by the Genetic Algorithms Research and Applications Group (GARAGe) group at Michigan State University (MSU).
  • Summary of parallel GA responses at University of Salzburg, Austria.
• Hive OS group, part of the FLASH project at Stanford. It is being designed to cope with a machine that is able to operate effectively in a traditional supercomputer environment as well as in a general-purpose, multiprogrammed environment.
• IDENTIFY (Interactive Design using a Network of Transputers in Fluids) - Parallel mathematical routines for applications in fluid flow at Informatics Department, Rutherford Appleton Laboratory, UK.
• Local Area Multicomputer (LAM)
  A programming environment and development system for a message-passing parallel machine constituted with heterogeneous UNIX computers on a network. With LAM, a dedicated cluster or an existing network computing infrastructure can act as one parallel computer solving one compute-intensive problem.
  • LAM project home -- including a complete implementation of the Message Passing Interface, MPI, tutorials on MPI and LAM and the full MPI specification and the MPI FAQ.
  • Local copy of LAM distribution.
• Linda Research Group at Yale University. This is part of the Programming Languages and Software Systems group in the Department of Computer Science and explores a variety of research topics in parallel and distributed computing, adaptive computation, and parallel programming languages.
• Legion Research Group at University of Virginia.
• Message Passing Interface (MPI). MPI is the new standard message passing interface for parallel computing that was specified by the authors of various current message passing systems, including the authors of P4, PVM, Chameleon, NX, Zipcode, and many others. It is specifically designed to allow programmers to write portable,high performance parallel programs and libraries.
  • Local MPI area
  • MPI page at Mississippi State University.
  • MPI page at ANL. This site contains the MPI draft standard, News, FAQs and errata about MPI, and pointers to implementations.
  • MPI page at Parallel Distributed Computing Team/CIC-8, Los Alamos, USA.
  • MPI: It's Easy to Get Started tutorial.
  • MPI FAQ.
  • Info on the Implementation of MPI on the Fujitsu AP1000 multicomputer located at the Australian National University including users guide and some papers.
  • See also the LAM project above.
  • Proceedings of MPIDC95.
• Massively Parallel Processing (MPP) Project at LLNL.
• Maths Topics at Finnish national supercomputer center (CSC) including mathematical animations, and an editable list of best mathematical services.
• Mentat Research Group at University of Virginia.
• MCS Graphics at ANL.
• MQM: A Tool for Graphical Management of Application Message Passing (Don Breazeal at Intel, Jerry Yan at NASA Ames)
• High-Performance Networking
• The Opus Environment: Integrating Task and Data Parallelism at ICASE.
• Parallel C / C++
  • Europa Working Group On Parallel C++ project at the London Parallel Applications Centre, UK.
  • pC++/Sage++ Information Home Page. pC++ is a portable parallel C++ for high performance computers. It is a language extention to C++ that permits data-parallel style opertations using "collections of objects" from some base "element" class.
  • The CC++ Programming Language . This is a parallel programming language based on the C++ programming language.
• Parallel I/O bibliographies, papers and information (David Kotz - dfk@moosilauke.dartmouth.edu
• PARIX
  • Virtual topologies library of the PARIX run-time system at the Paderborn Center for Parallel Computing (PCPC), University of Paderborn, Germany.
  • A library with global functions (operations) for PARIX at PCPC.
• P4
  • P4 Parallel Programming System at ANL.
  • Local copy of P4 distribution.
• Paradyne Parallel Performance Tools Project at University of Wisconsin - exploring new approaches to building scalable tools for parallel program performance.
• Program for Computational Reactive Mechanics (PCRM).
• PDDP - Data Parallel Programming Language Tool at NERSC (Brent C. Gorda - brent@nersc.gov)
• Parallel Data Lab at Carnegie Mellon University, USA. Research focuses on parallel file systems, aggressive prefetching in file systems, and Redundant Arrays of Independent Disks (RAID).
• PRIPS - Parallel Reconfigurable Image Processing System at Informatics Department, Rutherford Appleton Laboratory, UK.
• Psi Compiler Project at the University Of Missouri-Rolla.
• PVM: See Local for distribution and many other links.
• Piranah and Adaptive Parallelism at Yale University.
• PSI Array Compiler at University of Missouri - an array compiler that performs algebraic reductions on the expressions which eliminate unnecessary computation and temporary storage.
• The Scalable Concurrent Programming Laboratory conducts research in compiler technology and programming methods, with emphasis on message-driven multicomputing and irregular applications, and is based at the California Institute of Technology.
• Scalable Unix Commands for Parallel Computers (Rusty Lusk at ANL, Dan Frye at IBM, Jim Cownie at Meiko)
• PETSc - Software for scientific and parallel computing at ANL.
• The System of Systolic Structures Synthesis (S4CAD) high level interactive software tool for designing, analysis and modeling of systolic VLSI-array processors. S4CAD runs on PCs under Windows 3 and uses the technologies like operating in windows and hypertext. (Stanislav Sedukhin - sedukhin@u-aizu.ac.jp)
• The SCANDAL group at Carnegie-Mellon University (CMU) - The nested data-parallel language (NESL) and supercomputer programming environments.
• The Stanford SUIF Compiler Group - includes research into parallelization of programs and parallel languages.
• The Sisal Project: A High-Performance Functional Parallel Programming Language at LLNL.
• The Synergy Project at the Center for Advanced Computing and Communications (CACC), Temple University. Designed to provide an efficient and stable platform for distributed and parallel processing.
• TallShiPP - High-level Sharing for Parallel Programming at Informatics Department, Rutherford Appleton Laboratory, UK.
• INMOS Transputer processor archive (Jonathan Bowen - Jonathan.Bowen@comlab.ox.ac.uk) at Oxford University.
• Parallel Algorithms for Unstructured Grid Computation at ANL.