Applications
Misc.
MPI
Papers
Misc.
- Papers Related to the MATCH Project
http://www.ece.northwestern.edu/cpdc/Match/papers.html
- Matlab MPI
http://www.ll.mit.edu/MatlabMPI/
-
Matpar:
Parallel Extensions to MATLAB
http://hpc.jpl.nasa.gov/PS/MATPAR/
-
300x Faster Matlab using MatlabMPI
Jeremy Kepner (kepner@ll.mit.edu)
http://ernie.ecs.soton.ac.uk/opcit/cgi-bin/pdf?id=oai%3AarXiv.org%3Aastro-ph%2F0207389
-
FileExchange
Distributed Processing
http://www.mathworks.com/matlabcentral/fileexchange/load..20Processing
-
Survey
http://supertech.lcs.mit.edu/~cly/survey.html
-
Matlab parallel processing
http://www.people.virginia.edu/~md9c/DistributePP/#limitations
-
PARMATLAB
http://www.mathworks.com/matlabcentral/fileexchange/loadFile.do?objectType=file&objectId=217
-
My Tools from Peter Rydesater
http://petrydpc.itm.mh.se/tools/
-
Parallel MATLAB ... Doing it right
http://www-math.mit.edu/%7Eedelman/homepage/papers/pmatlab.pdf
-
MPI Toolbox papers
http://atc.ugr.es/~javier/investigacion/papers/mpitb_papers.html
-
Parallelization of the
Nanoscale Device Simulator nanoMOS2.0 Using a 100 Nodes Linux Cluster
Sébastien Goasguen, Ali. R. Butt, Kevin D. Colby and Mark S. Lundstrom
http://atc.ugr.es/~javier/investigacion/papers/SebNano02.pdf
-
II-Linux cluster description and performance Our Linux cluster is composed of 100 nodes dual-Athlon 1.2 GHz with 1 GB of RAM and 10 GB of
disk. The head node is a quad Xeon 700 MHz with 2 GB of RAM and 160 GB of RAID storage. The
operating system is Red-Hat 7.2 with the 2.4.17 kernel. PBS is installed to schedule batch jobs and
monitor the usage of the cluster. Parallelization is done through the Message Passing Interface
(MPI) routines. The Automatically Tuned Linear Algebra Sub-programs (ATLAS) has been used to create the
BLAS and LAPACK library optimized for the slave nodes. The High Performance Linpack (HPL) is used
to benchmark the performance of the cluster. After careful tuning of the HPL benchmark we obtain
around 130 GFLOPS for the whole cluster. These performances unofficially put the cluster within the
TOP500 supercomputers. Fig. 1 shows the cluster in its final configuration.
-
Supporting Runtime Tool Interaction for Parallel
Simulations (1998)
Christopher W. Harrop, Steven T. Hackstadt, Janice E. Cuny, Allen D. Malony, and Laura S.
Magde
Proceedings of Supercomputing '98 (SC98), Orlando, FL, November 7-13, 1998 (Best Student Paper
Finalist)
Compressed
Postscript
|