Newsgroups: comp.sys.transputer,comp.parallel.pvm,comp.parallel.pvm
From: D.J.Beckett@ukc.ac.uk (Dave Beckett)
Subject: Internet Parallel Computing Archive (IPCA) Additions - October 1996
Keywords: transputer, occam, parallel computing, HPC
Organization: University of Kent at Canterbury, UK.
Date: Fri, 01 Nov 96 16:34:21 GMT
Message-ID: <144@mint.ukc.ac.uk>

                 INTERNET PARALLEL COMPUTING ARCHIVE

            Funded by JISC ISSC and hosted on HENSA Unix

                <URL:http://www.hensa.ac.uk/parallel/>
              <URL:ftp://unix.hensa.ac.uk/pub/parallel/>

The IPCA contains over 650M of freely distributable software and
documents about High Performance and Parallel Computing with special
interest in the Transputer processor concept and the occam language.

See URLs above for more details of mirror sites, indices, searching etc.


		     Additions for October 1996
		     ==========================

31st October 1996

/parallel/events/hpcapps-gwich
	"High Performance Computing And Applications"
	by LAI, Choi-Hong <C.H.Lai@greenwich.ac.uk>
	Call for papers for coruse being held from 13th-17th January 1997 at
	University of Greenwich, UK.
	
	This is an intensive 5-day course targeted at equipping participants
	with the experience and skills needed to build and exploit engineering
	/ commercial software on leading edge parallel high performance
	computing technology in the solution of practical problems.
	
	See also
	<URL:http://www.greenwich.ac.uk/research/cms/parallel/hpca.html>

/parallel/events/naecon97
	"IEEE National Aerospace and Electronics Conference" (NAECON '97)
	by Pan Yi <pan@cis.ksu.edu>
	Call for papers for conference being held from 14th-18th July 1997 at
	Dayton, Ohio, USA. Sponsored by IEEE.
	
	Topics: Parallel and Distributed Architectures; Parallel and
	Distributed Algorithms; High-Performance Scientific Computing;
	Software Systems/Environments/Tools; Fault Tolerance and Reliability
	Prediction; Load Sharing and Task Scheduling; Resource Allocation and
	Management; Network Routing and Communication Algorithms; Wireless
	Networks and Mobile Computing; Modeling and Simulation; Signal and
	Image Processing; Performance Monitoring and Visualization;
	Application Studies; Neural Networks and others.
	
	Deadlines: Abstracts: 16th December 1996; Complete Papers: 3rd
	February 1997; Notification: 15th March 1997; Camera-ready papers:
	15th April 1997.
	
	See also <URL:http://www.erinet.com/dayieee/NAECON/Welcome.html>

/parallel/occam/projects/occam-for-all/kroc/
	Release of Kent Retargetable occam Compiler (KROC) 0.8beta

/parallel/occam/projects/occam-for-all/kroc/README.html
	"KROC 0.8 Beta Release Announcement (from release)"

/parallel/occam/projects/occam-for-all/kroc/INSTALL.html
	"KROC 0.8 Beta Installation and Use (from release)"

/parallel/occam/projects/occam-for-all/kroc/NEWS.html
	"KROC 0.8 Beta Release History (from release)"

/parallel/occam/projects/occam-for-all/kroc/TARGETS.html
	"KROC Processor Targeting Information (from release)"

/parallel/occam/projects/occam-for-all/kroc/kroc-0.8beta-sparc-sun-sunos4.1.3_U1.tar.gz
/parallel/occam/projects/occam-for-all/kroc/kroc-0.8beta-sparc-sun-sunos4.1.3_U1.tar.Z
	"KROC 0.8 beta BINARY distribution for Sun Sparcs with SunOS 4.1.3U1"
	by Occam For All Group <ofa-bugs@ukc.ac.uk>
	Compiles the occam 2.1 language (RECORDS and DATA TYPEs) has an
	occam/C interface tool and has separate compilation support (#USE).
	Includes standard libraries (hostio.lib, streamio.lib etc.), language
	documentation, examples and exercises.

/parallel/occam/projects/occam-for-all/kroc/kroc-0.8beta-sparc-sun-solaris2.5.tar.gz
/parallel/occam/projects/occam-for-all/kroc/kroc-0.8beta-sparc-sun-solaris2.5.tar.Z
	"KROC 0.8 beta BINARY distribution for Sun Sparcs with Solaris 2.5
	(SunOS 5.5)"
	by Occam For All Group <ofa-bugs@ukc.ac.uk>
	Compiles the occam 2.1 language (RECORDS and DATA TYPEs) has an
	occam/C interface tool and has separate compilation support (#USE).
	Includes standard libraries (hostio.lib, streamio.lib etc.), language
	documentation, examples and exercises.

/parallel/occam/projects/occam-for-all/kroc/kroc-0.8beta-alpha-dec-osf3.0.tar.gz
/parallel/occam/projects/occam-for-all/kroc/kroc-0.8beta-alpha-dec-osf3.0.tar.Z
	"KROC 0.8 beta BINARY distribution for DEC Alphas with OSF 3.0"
	by Occam For All Group <ofa-bugs@ukc.ac.uk>
	Compiles the occam 2.1 language (RECORDS and DATA TYPEs) has an
	occam/C interface tool and the first release of Alpha separate
	compilation support. Includes standard libraries (hostio.lib,
	streamio.lib etc.), language documentation, examples and exercises.

/parallel/events/
	Added new event:

http://www.psc.edu/training/general/Course_Descriptions.html#message_passing/
	"Message Passing and Distributed Computing Workshop"
	by Stephanie Dobler <dobler@psc.edu>
	Details of workshop being held from 11th-14th November at Pittsburgh,
	PA, USA. Organised by: Pittsburgh Supercomputing Center.
	
	The purpose of this workshop is to familiarize researchers and
	programmers with the use of message passing to run parallel code on a
	cluster of homogeneous or heterogeneous machines. The goal is for a
	new user to successfully develop basic codes using message passing.
	There will be extensive talks and hands-on exercises on the Message
	Passing Interface (MPI).
	
	Topics: Message passing ; Parallel code and Message Passing Interface
	(MPI).
	
	Deadlines: Workshop application: Oct. 31, 1996.


28th October 1996

/parallel/architecture/communications/io/pario/STARFISH/starfish3.0.tar.gz
	"STARFISH Parallel file-system simulator V3.0"
	by David Kotz <dfk@cs.dartmouth.edu>,
	<URL:http://www.cs.dartmouth.edu/~dfk/>, Department of Computer
	Science, Dartmouth College, 6211 Sudikoff Laboratory, Hanover, NH
	03755-3510 USA; Tel: +1 603-646-1439
	Source code for parallel file-system simulator and associated stuff.

/parallel/architecture/communications/io/pario/papers/Kotz/rus:autonomous.ps.Z
	"Autonomous and Adaptive Agents that Gather Information"
	by Daniela Rus; Robert Gray and David Kotz <dfk@cs.dartmouth.edu>.
	In AAAI'96 International Workshop on Intelligent Adaptive Agents,
	August 1996.
	ABSTRACT:
	We have designed and implemented autonomous software agents.
	Autonomous software agents navigate independently through a
	heterogeneous network of computers. They can sense the state of the
	network, monitor software conditions, and interact with other agents.
	The network-sensing tools allow our agents to adapt to the network
	configuration and to navigate under the control of reactive plans. In
	this paper we illustrate the intelligent and adaptive behavior of
	autonomous agents in distributed information-gathering tasks.

/parallel/events/
	Added new event:

http://www.ce.unipr.it/ica3pp97/
	"The IEEE Third International Conference on Algorithms And
	Architectures for Parallel Processing - SPECIAL SESSION ON
	INSTRUCTION-LEVEL PARALLEL PROCESSING" (ICA3PP-97)
	by Alberto Broggi <broggi@ce.unipr.it>,
	<URL:http://www.ce.unipr.it/~broggi/>
	Details of event being held from 8th-12th December2 1997 at
	Melbourne, Australia. Organised by: IEEE Victoria Section; IEEE
	Computer Society, Deakin University; Faculty of Science and
	Technology, Deakin University.
	
	Topics: ILP-based architectures; Innovative ILP-based architectures;
	Design tools for ILP-based systems; Application-specific ILP-based
	systems for real-time image processing, numerical analysis, physical
	simulation; Languages and paradigms exploiting ILP; Compilation and
	code generation techniques for ILP; Code optimization for ILP;
	Performance analysis and prediction tools for ILP; ILP-based
	algorithms Compiler and architecture co-design; Applications of ILP
	techniques and Theoretical foundations of ILP.
	
	Deadlines: Papers: 17th March 1997; Notification of Acceptance: 18th
	July 1997; Camera-ready papers: 1st September 1997.


21st October 1996

/parallel/occam/compilers/ocpp/
	Updated OCPP: an occam pre-processor

/parallel/occam/compilers/ocpp/ocpp300.tar.gz
	"occam Pre-Processor V3.00 binaries" (OCPP)
	by Mark Ian Barlow <Mark@nlcc.demon.co.uk>; Tel: +44 (0)1207 562 154

/parallel/occam/compilers/ocpp/srci100.tar.gz
	"occam Pre-Processor V3.00 sources" (OCPP)
	by Mark Ian Barlow <Mark@nlcc.demon.co.uk>; Tel: +44 (0)1207 562 154



17th October 1996

/parallel/vendors/intel/paragon/software/nxlib/NXLibV1_1_4+LINUX.tar.gz
	"NXlib V1.1.4 distribution for Linux"
	Includes some bugfixes and a workaround for Linux


16th October 1996

/parallel/info/stats/
	Graphs of access statistics as GIF and PNG images for IPCA
	since start of archive and for just 1996.

/parallel/libraries/memory/global-array/
	Updated GA Toolkit to V2.2.
	Global Arrays is a portable shared Non-Uniform Memory Access (NUMA)
	programming environment for distributed and shared memory computers.

/parallel/libraries/memory/global-array/global2.2.tar.Z
	"Global Array (GA) Toolkit V2.2"
	by Jarek Nieplocha <j_nieplocha@pnl.gov>, Environmental Molecular
	Sciences Laboratory Pacific Northwest National Laboratory, MSIN:
	K1-87, Richland, WA 99352, USA; Tel: +1 (509) 372-4469
	Requires GNU make, IPC (semaphores).

/parallel/architecture/communications/io/pario/papers/Kotz/kotz:jdiskdir.Jan95.ps.Z
	"Disk-directed I/O for MIMD Multiprocessors"
	by David Kotz <dfk@cs.dartmouth.edu>, Department of Computer Science,
	Dartmouth College, Hanover, NH 03755, USA
	Submitted to TOCS January 15, 1995
	ABSTRACT:
	Many scientific applications that run on today's multiprocessors,
	such as weather forecasting and seismic analysis, are bottlenecked by
	their file-I/O needs. Even if the multiprocessor is configured with
	sufficient I/O hardware, the file-system software often fails to
	provide the available bandwidth to the application. Although libraries
	and enhanced file-system interfaces can make a significant
	improvement, we believe that fundamental changes are needed in the
	file-server software. We propose a new technique, disk-directed I/O,
	to allow the disk servers to determine the flow of data for maximum
	performance. Our simulations show that tremendous performance gains
	are possible both for simple reads and writes and for an out-of-core
	application. Indeed, disk-directed I/O provided consistent high
	performance that was largely independent of data distribution,
	obtained up to 93% of peak disk bandwidth, and was as much as 18 times
	faster than traditional parallel file systems.


14th October 1996

/parallel/simulation/communications/chaos/docs/
	Updated descriptions of:

/parallel/simulation/communications/chaos/docs/byost.ps.Z
	"Cost Effective Fault Tolerance for Network Routing"
	by William Yost, Dept. of Computer Science and Engineering,
	University of Washington, Seattle, Washington, USA
	Master of Science Thesis
	ABSTRACT:
	As the size of interconnection networks for multicomputers increases,
	it becomes clear that some degree of fault tolerance will be required
	in order to maintain system reliability at an acceptable level. The
	interconnection networks contain large numbers of relatively
	unreliable components. However, cost and performance concerns mandate
	that the addition of fault tolerance to the network have an
	unobtrusive impact on network design. The target design space requires
	reasonable reliability while avoiding exorbitant additional costs.
	Because routing messages around failed components may require
	non-minimal routes, it makes sense to examine routers which, by
	design, allow packets to take non-minimal routes. However, to provide
	for effective and efficient fault tolerance, a unified, coherent
	scheme, is augmented with a limited amount of hardware to support
	fault detection, identification, and reconfiguration so that the
	network can maintain reliable operation in the presence of faults. The
	Express Broadcast Network is introduced; a low overhead control of the
	network, and specifically, provides for global control and
	synchronization of fault management procedures for Chaos. A high-level
	design is presented in which fault tolerance is greatly enhanced by
	the addition of functionality to targeted blocks while maintaining the
	structure of Chaotic routing and avoiding large incremental costs.

/parallel/simulation/communications/chaos/docs/guide.ps.Z
	"A Guide to Literature on Chaotic Routing"
	by Chaotic Routing Project, Department of Computer Science and
	Engineering, University of Washington, Seattle, WA 98195, USA

/parallel/environments/pvm3/distribution/pvm_win32.tar.gz
/parallel/environments/pvm3/distribution/pvm_win32.zip
	Updated: "PVM 3.3.? for Win32"
	PVM 3.3.? for MS Windows 32-bit (Windows 95, Windows NT). Contains
	executables, documetnation, examples, source code, manual pages and
	support software.

/parallel/events/ipps97-hips97
	"2nd Workshop onHigh-Level Programming Models and Supportive
	Environments" (HIPS'97)
	by Rudi Hackenberg <hackenbe@informatik.tu-muenchen.de>
	Call for papers for IPPS97 workshop being held from 1st-5th April
	1997 at Geneva, Switzerland.
	
	Topics: Concepts and languages for high-level parallel programming;
	Concurrent object-oriented programming; Automatic parallelization and
	optimization; High-level programming environments; Performance
	analysis techniques and tools; Implementation techniques for
	high-level programming models; Distributed shared memory;
	Architectural support for high-level programming models and others.
	
	Deadlines: Manuscripts: 15th November 1996; Notification: 31st
	December 1996; Camera ready papers: 31st January 1997.
	
	See also <URL:http://cuiwww.unige.ch/~ipps97> (for IPPS97)

/parallel/events/ipps97-rtspp
	"Workshop on Runtime Systems for Parallel Programming at IPPS'97" (RTSPP)
	by Gregory D. Benson <benson@cs.ucdavis.edu>
	Call for papers for IPPS97 workshop being held on 5th April 1997 at
	Geneva, Switzerland.
	
	Topics: Techniques to reduce the tension between portability and
	efficiency in runtime systems; Relationship between runtime systems
	and programming models; Interfaces between compiler-generated code and
	runtime systems; Operating system support for runtime systems;
	Performance evaluation of runtime systems; The design and
	implementation of runtime systems for high-speed networks; The design
	and implementation of thread systems; Runtime systems for distributed
	shared memory; Extensibility and adaptability in runtime systems and
	others.
	
	Deadlines: Manuscripts: 6th December 1996; Notification: 10th January
	1997; Camera ready papers: 3rd February 1997.
	
	See also <URL:http://cuiwww.unige.ch/~ipps97> (for IPPS97)

/parallel/events/hpf-ug
	"1st High Performance Fortran (HPF) User Group Meeting"
	by Israel Gale <gale@wind.hpc.pko.dec.com>
	Call for speakers and participation in meeting being held from
	24th-26th February 1997 at Sante Fe, New Mexico, USA.
	
	Includes intro to HPF, HPF 2.0, invited talks by HPF users and
	developers, panel discussions, contributed talks, real HPF
	applications, software and development.
	
	Deadlines: Abstract: 2nd January 1997; Speaker Notification: 16th
	January 1997; (Early) Registration: 24th January 1997.

/parallel/job-digital-us
	"Digital Equipment parallel computing engineers."
	by Mark Gantly <mark@ilo.dec.com>
	Projects include distributed parallel runtime environments
	(supporting PVM, MPI and Fortran90/HPF), parallel tools (for debugging
	and performance profiling) and high performance I/O. Requirements: CS
	or equiv., UNIX & C experience.


7th October 1996

/parallel/events/wscg97
	"Fifth International Conference in Central Europe on Computer
	Graphics and Visualization 1997" (WSCG'97)
	by skala <skala@kiv.zcu.cz>
	Call for papers and participation for conference being held from
	10th-14th February 1997 at University of West Bohemia, nr Prague,
	Czech Republic.
	
	Topics: Fundamental algorithms, rendering and visualization, virtual
	reality, animation and multimedia, medical imaging, geometric
	modelling and fractals, graphical interaction, object-oriented
	graphics, standards, computer vision, parallel and distributed
	graphics, computational geometry, computer aided geometric design,
	CAD/CAM, DTP and GIS systems, educational aspects of related fields,
	usage of graphics within mathematical software in education.and
	others.
	
	Deadlines: Papers: 30th October 1996; Notification: 10th December
	1996.
	
	See also <URL:http://yoyo.zcu.cz/~skala/wscg97.html>,
	<URL:http://yoyo1.zcu.cz/~skala/wscg97.html> (maybe faster) or
	<URL:herakles.zcu.cz>

/parallel/events/mdt-mpp
	"Molecular Dynamics Techniques on MPP Platforms"
	by J-C Desplat <desplat@epcc.edinburgh.ac.uk>
	Call for participation for seminar being held on 25th December 1996
	at Royal Institution of Great Britain, London, UK.
	
	This multi-disciplinary seminar will deal with the application of
	Molecular Dynamics methods to a wide variety of scientific areas from
	materials structure to cosmology, but focused on implementations on
	massively parallel machines. With presentations by leading researchers
	from the US, Europe and the UK, this seminar will be of interest to
	computational scientists and engineers in any discipline using (or
	intending to use) MD techniques on MPP machines.
	
	See also <URL:http://www.epcc.ed.ac.uk/t3d/hpci/md_seminar.html>

/parallel/events/europar97
	"EURO-PAR'97"
	by Sergei Gorlatch <gorlatch@fmi.uni-passau.de>
	Call for papers for conference being held from 26th-29th August 1997
	at Passu, Germany.
	
	Topics: Parallel Computing (in Europe :-) ? ).
	
	Deadlines: Papers: 1st February 1997; Notification: 2nd May 1997;
	Final papers: 10th June 1997; Early Registration: 30th June 1997.
	
	See also <URL:http://www.uni-passau.de/europar97/> for updates and
	plans.

/parallel/events/ipps97-job-scheduling
	"3rd Workshop on Job Scheduling Strategies for Parallel Processing at
	IPPS'97"
	by Dror Feitelson <feit@humus.cs.huji.ac.il>
	Call for papers for IPPS97 workshop being held on 5th April 1997 at
	Geneva, Switzerland.
	
	Topics: Experience with scheduling policies on current systems;
	Performance implications of scheduling strategies; Fairness,
	priorities, and accounting issues; Workload characterization and
	classification; Support for different classes of jobs (e.g.
	interactive vs. batch); Static vs. dynamic partitioning; Time slicing
	and gang scheduling; Interaction of Scheduling with the model of
	computation; Interaction of scheduling with memory management and I/O;
	Load estimation and load balancing; Scheduling on heterogeneous nodes
	(e.g. with different amounts of memory); Performance metrics to
	compare scheduling schemes and others.
	
	Deadlines: Paper: 27th December 1996; Notification: 7th February
	1997; Final paper: 7th March 1997.
	
	See also <URL:http://cuiwww.unige.ch/~ipps97> (for IPPS97)

/parallel/events/ipps97-rand-par
	"2nd Workshop on Randomized Parallel Computing at IPPS'97"
	by Michael A. Palis <palis@crab.rutgers.edu>
	Call for papers for IPPS97 workshop being held on 5th April 1997 at
	Geneva, Switzerland.
	
	Topics: Network Algorithms; PRAM Algorithms; Architectures; I/O
	Systems; Scheduling; Network Fault Tolerance; Reconfigurable Networks;
	Optical Networks; Various Applications; Programming models and
	languages; Implementation Experience and others.
	
	Deadlines: Papers: 1st December 1996; Notification: 1st February
	1997; Camer ready manuscripts: 5th March 1997.
	
	See also <URL:http://cuiwww.unige.ch/~ipps97> (for IPPS97)

/parallel/events/spdp96-challenges
	"Challenges in Compiling for Scalable Parallel Systems workshop at
	SPDP'96"
	by Santosh Pande <santosh@ececs.uc.edu>
	Call for participation and advance program for workshop being held at
	8th IEEE Symposium on Parallel and Distributed Processing (SPDP '96)
	on 24th October 1996 at New Orleans, LA, USA.


7th October 1996

/parallel/occam/utilities/odt/
	The occam Design Tool (ODT) is an interactive graphical environment
	for the specification and partial implementation of concurrent systems
	and enables structured design of parallel systems through hierarchies
	of communicating process networks.
	
	ODT has knowledge of various paradigms for process synchronisation --
	whose correct application secures guarantees of freedom from deadlock,
	livelock and starvation -- and enforces that correct application by
	the designer.
	
	The tool also generates program skeletons that handle the parallel
	and communication aspects of the system implementation. The target
	programming language has to be one whose formal semantics respect the
	mathematical model (CSP) under which the design rules were able to be
	proven -- currently, this is occam. By targeting occam, we derive the
	further benefit of eliminating unguarded race hazards.

/parallel/occam/utilities/odt/odt-1.02.tar.gz
	"ODT 1.02 (source only)"
	by Dave Beckett <D.J.Beckett@ukc.ac.uk>
	Requires Tcl/Tk 7.5/4.1 (or better)

/parallel/occam/utilities/odt/odt-1.02-osf.tar.gz
	"ODT 1.02 with DEC Alpha OSF binaries of wish4.1"
	by Dave Beckett <D.J.Beckett@ukc.ac.uk>

/parallel/occam/utilities/odt/odt-1.02-sunos.tar.gz
	"ODT 1.02 with Sun Sparc SunOS 4.1.3U1 binaries of wish4.1"
	by Dave Beckett <D.J.Beckett@ukc.ac.uk>

http://ssd.sscc.ru/pact97/
	"Fourth International Conference on Parallel Computing Technologies" (PaCT-97)
	by Oleg Yu. Repin <repin@ssd.sscc.ru>,
	<URL:http://ssd.sscc.ru/repin/>
	Call for papers for conference being held from 8th-12th September
	1997 at Yaroslavl, Russia organised by the Computing Center of the
	Russian Academy of Sciences (Academgorodok, Novosibirsk) and the State
	University of Yaroslavl. Sponsored by the Russian Academy of Sciences,
	Russian Fund for Fundamental Research and the Ministry of Higher
	Education.
	
	Topics: All aspects of the applications of parallel computer systems;
	General methods and tools for the solution of large problems;
	Languages,environments and software tools supporting parallel
	processing; Operating systems, scheduling, mapping, load balancing;
	General architecture concepts, enabling technologies; Performance
	measurement and analysis and Teaching parallel processing.
	
	Deadlines: Extended abstract: 20th January 1997; Notification of
	acceptance: 15th March 1997; Camera-ready papers: 20th May 1997.


4th October 1996

/parallel/acronyms/hpccgloss/
	Updated HPCC Glossary to have Ken Hawick's new email address.

/parallel/libraries/communication/scotch/scotch_3.1.2A.tar.gz
	"The SCOTCH v3.1.2 Academic distribution"
	by Francois Pellegrini <pelegrin@labri.u-bordeaux.fr>
	Contains binaries for Sun Solaris 2 and SunOS 4; MIPS SGI IRIX 5 and
	6; Linux; PowerPC IBM AIX 4. Also contains sources, graphs and 3.1.2
	User Guide.

/parallel/standards/mpi/anl/
	Added HTML versions of the MPICH documentation:

/parallel/standards/mpi/anl/mpiuserguide.html.tar.gz
	"Users' Guide to mpich, a Portable Implementation of MPI"
	by William Gropp <gropp@mcs.anl.gov> and Ewing Lusk
	<lusk@mcs.anl.gov>. Mathematics and Computer Science Division, Argonne
	National Laboratory, USA.

/parallel/standards/mpi/anl/mpiinstall.html.tar.gz
	"Installation Guide to mpich, a Portable Implementation of MPI"
	by William Gropp <gropp@mcs.anl.gov> and Ewing Lusk
	<lusk@mcs.anl.gov>. Mathematics and Computer Science Division, Argonne
	National Laboratory, USA.

/parallel/architecture/communications/io/pario/papers/Kotz/nieuwejaar:jgalley.ps.Z
	"The Galley Parallel File System"
	by Nils Nieuwejaar <nils@cs.dartmouth.edu> and David Kotz
	<dfk@cs.dartmouth.edu>. Department of Computer Science, Dartmouth
	College, Hanover, NH 03755-3510, USA.
	To appear in Parallel Computing, 1996.
	ABSTRACT:
	Most current multiprocessor file systems are designed to use multiple
	disks in parallel, using the high aggregate bandwidth to meet the
	growing I/O requirements of parallel scientific applications. Many
	multiprocessor file systems provide applications with a conventional
	Unix-like interface, allowing the application to access multiple disks
	transparently. This interface conceals the parallelism within the file
	system, increasing the ease of programmability, but making it
	difficult or impossible for sophisticated programmers and libraries to
	use knowledge about their I/O needs to exploit that parallelism. In
	addition to providing an insufficient interface, most current
	multiprocessor file systems are optimized for a different workload
	than they are being asked to support. We introduce Galley, a new
	parallel file system that is intended to efficiently support realistic
	scientific multiprocessor workloads. We discuss Galley's file
	structure and application interface, as well as the performance
	advantages offered by that interface.

/parallel/architecture/communications/io/pario/papers/Kotz/kotz:flexibility2.ps.Z
	"Flexibility and Performance of Parallel File Systems"
	by David Kotz <dfk@cs.dartmouth.edu> and Nils Nieuwejaar
	<nils@cs.dartmouth.edu>. Department of Computer Science, Dartmouth
	College, Hanover, NH 03755-3510, USA.
	Copyright 1996 by Springer-Verlag. Appeared in the Third
	International Conference of the Austrian Center for Parallel
	Computation (ACPC), which was published in the Springer-Verlag Lecture
	Notes in Computer Science series.
	ABSTRACT:
	As we gain experience with parallel file systems, it becomes
	increasingly clear that a single solution does not suit all
	applications. For example, it appears to be impossible to find a
	single appropriate interface, caching policy, file structure, or
	disk-management strategy. Furthermore, the proliferation of
	file-system interfaces and abstractions make applications difficult to
	port. We propose that the traditional functionality of parallel file
	systems be separated into two components: a fixed core that is
	standard on all platforms, encapsulating only primitive abstractions
	and interfaces, and a set of high-level libraries to provide a variety
	of abstractions and applicationprogrammer interfaces (APIs). We
	present our current and next-generation file systems as examples of
	this structure. Their features, such as a three-dimensional file
	structure, strided read and write interfaces, and I/O-node programs,
	are specifically designed with the flexibility and performance
	necessary to support a wide range of applications.


2nd October 1996

/parallel/events/hpcn-europe97
	Updated: "High-Performance Computing and Networking, Europe 1997" (HPCN Europe 97)
	by Jaap Hollenberg <sondjaap@horus.sara.nl>
	Call for papers, posters and workshops and submission rules for
	conference being held from 28-30th April 1997 at Vienna, Austria.
	
	Topics: End-user HPCN applications, computational science and
	computer science research in HPCN.
	
	Deadlines: Extended Abstracts / Full Papers: 8th November 1996;
	Posters: 8th November 1996; Workshops: 8th November 1996;
	Notification: 1st February 1997.
	
	See also <URL:http://www.wins.uva.nl/research/HPCN97/>

http://www.cs.dartmouth.edu/iopads/
	Details of "Fifth Annual Workshop on I/O in Parallel and Distributed Systems" (IOPADS '97)
	by Rajeev Thakur <thakur@mcs.anl.gov>,
	<URL:http://www.mcs.anl.gov/people/thakur/>
	Call for papers for workshop being held at San Jose, California, USA in
	conjuction with SC97: High Performance Networking and Computing.
	Sponsored by ACM SIGACT, SIGARCH, and SIGOPS
	
	Topics: Storage-related parallel I/O; parallel file systems;
	real-time and multimedia I/O; database systems; I/O-intensive
	applications; I/O characterization; network-attached storage devices;
	devices including disk arrays and tape robots; and language, compiler,
	and run-time support for parallel I/O.
	
	Deadlines: Papers: March 25, 1997.

/parallel/events/hpcasia97
	Changed dates and location of HPC Asia'97 to April 28-May 2
	1997 and now at Seoul Hilton.

/parallel/occam/documentation/books
	Added details of "occam2" by John Galletly <john@buck.ac.uk>
	(1996, UCL Press)


