WoTUG - The place for concurrent processes

BibTeX Proceedings details

BibTex html text dump of PaperDB database.

PaperDB is GPL Software, see: http://paperdb.sourceforge.net.

  title = "{V}irtual memory management for the transputer",
  author= "Bakkes, P. J.",
  editor= "Askew, Charlie",
  pages = "1--6",
  booktitle= "{OUG}-9: {O}ccam and the {T}ransputer -- {R}esearch and {A}pplications",
  isbn= "90 5199 010 3",
  year= "1988",
  month= "sep",
  abstract= "During the last decade a number of microprocessors appeared
     on the market with operating system mechanisms designed into
     the firmware e.g. the Intel 80286 which supports segmented
     memory management and the 80386 supporting segmentation and
     paging. The idea was very attractive as protection amongst
     users could be enforced by the hardware. The support of
     multi-users on one processor implied the conventional
     strategy of sharing the scarce processor resource amongst
     more than one user. The limitations in processing power of
     microprocessors did however limit their multi-user
  title = "{CDL} -- {A} distribution language for {HELIOS}",
  author= "Grimsdale, C. H. R.",
  editor= "Askew, Charlie",
  pages = "7--12",
  booktitle= "{OUG}-9: {O}ccam and the {T}ransputer -- {R}esearch and {A}pplications",
  isbn= "90 5199 010 3",
  year= "1988",
  month= "sep",
  abstract= "Helios is an Operating System designed to exploit highly
     distributed systems. The central goal of such a distributed
     system, is not only to provide increased fault tolerance,
     but also increased performance. This necessitates adequate
     support for parallelism at the application level. It is
     proposed that instead of extending an existing language, or
     producing a completely new language, parallelism can be
     defined at the distribution phase. A high level Component
     Distribution Language has been designed, which allows simple
     sequential components to be combined by simple parallel
     constructors to form more complex parallel structures. The
     implementation language is irrelevant thus providing
     consistent support for standard sequential languages in a
     distributed environment."
  title = "{A} concurrent approach to the {T}owers of {H}anoi",
  author= "Crowe, W. D. and Strain-Clark, P. E. D.",
  editor= "Askew, Charlie",
  pages = "13--22",
  booktitle= "{OUG}-9: {O}ccam and the {T}ransputer -- {R}esearch and {A}pplications",
  isbn= "90 5199 010 3",
  year= "1988",
  month= "sep",
  abstract= "The tower of Hanoi problem has long been known to have a
     closed (ie non-recursive) solution. In this paper we analyse
     two approaches to this solution which involve concurrency.
     This is not an end in itself, but serves to introduce the
     main ideas and notations of CAP (Communicating Asynchronous
     Processes) - a revision of CSP which the authors have used
     successfully to develop correct Occam programs. In turn, CAP
     is part of a wider development method (ODM) which is being
     studied with the intention of prototyping software tools to
     assist developers of concurrent software."
  title = "{A}n {OCCAM}@ implementation of prolog and its preliminary performance",
  author= "Zhang, Kang",
  editor= "Askew, Charlie",
  pages = "23--36",
  booktitle= "{OUG}-9: {O}ccam and the {T}ransputer -- {R}esearch and {A}pplications",
  isbn= "90 5199 010 3",
  year= "1988",
  month= "sep",
  abstract= "The paper presents an execution model for Prolog, which
     attempts to exploit the parallelism mainly at the argument
     level through the unification operation. The model,
     consisting of a number of virtual machine instructions, has
     been implemented in Occam2 on a Transputer Development
     System (TDS). The performance of the pure software
     implementation has been evaluated in real-time. The speed,
     as assessed by running a few hand compiled benchmark
     programs on the TDS with a single transputer, ranges from
     7-18 KLIPS. The paper gives some details of the performance,
     and then proposes a dataflow-based functionally distributed
     configuration of the multitransputer system."
  title = "{T}he meaning and implementation of {PRI} {ALT} in occam",
  author= "Barrett, Geoff and Goldsmith, Michael and Jones, Geraint and Kay, A.",
  editor= "Askew, Charlie",
  pages = "37--46",
  booktitle= "{OUG}-9: {O}ccam and the {T}ransputer -- {R}esearch and {A}pplications",
  isbn= "90 5199 010 3",
  year= "1988",
  month= "sep",
  abstract= "This paper describes what we believe to be an error in the
     implementation of PRI ALT on the Inmos transputer, and
     exhibits anomalous behaviour of that implementation. A
     correct implementation of comparable complexity is
     described, and some of the properties of the construct are
     described. Finally, an attempt is made to describe
     circumstances in which the behaviour of the existing
     implementation is adequate for the correctness of a program
     which uses it."
  title = "{S}imulation of gas pipeline networks",
  author= "Patel, Minesh and Bentley, Paul and Hughes, Clifton",
  editor= "Askew, Charlie",
  pages = "47--52",
  booktitle= "{OUG}-9: {O}ccam and the {T}ransputer -- {R}esearch and {A}pplications",
  isbn= "90 5199 010 3",
  year= "1988",
  month= "sep",
  abstract= "As part of the Alvey ParSiFal project, Logica Cambridge has
     been developing a model of a significant gas distribution
     network, in order to explore the potential for improvements
     in modelling speed using large numbers of transputers.There
     are two basic approaches to continuous simulation - the
     explicit and the implicit. The implicit approach models an
     entire system as a set of equations. Current implementations
     are well suited to conventional mono-processor machines but
     it is difficult to parallelize the algorithm and
     consequently to fully exploit the potential of
     transputers.The explicit approach models components such as
     pipes as individual objects which can be directly mapped
     into parallel processes, making it ideally suited to the
     transputer/occam architecture. However, explicit simulation
     is inherently expensive in terms of computational resources.
     To enhance performance we are also working on improvements
     to the model.We have developed a system using an explicit
     algorithm and have simulated networks ranging from 20 to
     500+ objects. Preliminary results indicate that the explicit
     algorithm can efficiently exploit high degrees of
  title = "{A}n extension of the processor farm using a tree architecture",
  author= "Green, S. A. and Paddon, Derek J.",
  editor= "Askew, Charlie",
  pages = "53--70",
  booktitle= "{OUG}-9: {O}ccam and the {T}ransputer -- {R}esearch and {A}pplications",
  isbn= "90 5199 010 3",
  year= "1988",
  month= "sep",
  abstract= "For problems that require a small database a processor farm
     model was proposed by May and Sheppard, and has subsequently
     had much success. Here, we assume that the database is too
     large to store completely at each node of the farm and
     consequently, data must be communicated throughout the
     network during the execution of a problem. The processor
     farm model is extended to allow work tasks to be completed
     when data items not resident locally at a node can be
     obtained by making a request for data across the network.
     The farm model is further extended by configuring the system
     as a tree topology. The extended farm topology is evaluated
     by using a set of graphics benchmark problems. Benchmark
     results are given for a tree farm of upto fifteen worker
  title = "{A} preprocessor to augment the description of occam processes for multitransputer machines",
  author= "Ohara, Hiromi and Iizuka, Hajime",
  editor= "Askew, Charlie",
  pages = "71--80",
  booktitle= "{OUG}-9: {O}ccam and the {T}ransputer -- {R}esearch and {A}pplications",
  isbn= "90 5199 010 3",
  year= "1988",
  month= "sep",
  abstract= "This paper describes a preprocessor which is designed to
     make the description of parallel processes of Occam easier.
     It takes a source program written in ordinary Occam language
     and a description of the configuration of the hardware on
     which it is intended to run. The programmer need not
     consider the particular hardware structure. The preprocessor
     assigns each process to an appropriate processor and
     produces the Occam program with process-to-processor
     allocation statements. Design philosophy, implementation
     details and some results are shown."
  title = "{R}andomised routing: \"{H}ot potato\" simulations",
  author= "Qiang, Xu Ming and Turner, Stephen J.",
  editor= "Askew, Charlie",
  pages = "81--90",
  booktitle= "{OUG}-9: {O}ccam and the {T}ransputer -- {R}esearch and {A}pplications",
  isbn= "90 5199 010 3",
  year= "1988",
  month= "sep",
  abstract= "In this paper, we describe experiments which simulate
     communication in a network of processors organised as a
     hypercube. A stochastic model o\pounds communication has
     been implemented in which each processor chooses at random
     another processor in the network with which to communicate.
     Deterministic and randomized routing algorithms are compared
     with respect to both the time taken for packets to reach
     their destinations and the space required for packet queues.
     A mixed strategy is discussed and evaluated which combines
     some of the advantages of each method. Questions of fault
     tolerance are also discussed and possible solutions
  title = "{M}apping a process network onto a processor network",
  author= "Lau, Francis C. M. and Shea, K. M.",
  editor= "Askew, Charlie",
  pages = "91--100",
  booktitle= "{OUG}-9: {O}ccam and the {T}ransputer -- {R}esearch and {A}pplications",
  isbn= "90 5199 010 3",
  year= "1988",
  month= "sep",
  abstract= "This paper addresses the problem of mapping an Occam program
     onto a transputer network. This mapping is essentially a
     graph embedding (or related) problem: i.e., find a function
     f: N -\> \{ 1, 2.....\textbar M \textbar\} such that
     certain performance criteria would be satisfied, where N is
     the set of processes in the Occam program (the graph) and M
     is the set of processors in the transputer network.
     Depending on the complexity of the graph as well as the
     target transputer network, many of these problems are
     NP-complete [1,2]. This paper addresses not the mapping
     algorithm and how it may achieve optimality, which is a
     policy matter, but the mechanism of mapping - that is, given
     a mapping algorithm, what exactly do we do to transform the
     program into an equivalent program that is ready to be
     downloaded onto the transputer network. Our answer is a
     precompiler which we have successfully implemented. We
     discuss several important problems that occurred in our
     construction of the mapping procedure, and briefly describe
     the actual implementation of the precompiler."
  title = "{S}upport for occam channels via dynamic switching in multi-transputer machines",
  author= "Jones, Peter and Murta, Alan",
  editor= "Askew, Charlie",
  pages = "101--112",
  booktitle= "{OUG}-9: {O}ccam and the {T}ransputer -- {R}esearch and {A}pplications",
  isbn= "90 5199 010 3",
  year= "1988",
  month= "sep",
  abstract= "The authors are motivated by the disparity between the
     arbitrarily large degree of inter-process connectivity
     allowed in the occam model of computation (by means of occam
     channels) and the implementation of that model on the Inmos
     transputer (under the familiar four links per processor
     constraint). A communication technique has been developed
     and implemented on the ParSiFal machine which addresses this
     problem. The method involves a novel use of switching
     hardware. In normal use, links between transputers in a
     variable topology array such as the ParSiFal machine are
     arranged using switching hardware to form a static network
     before the application is loaded. This study explores the
     run-time reconfiguration of the switch network under the
     direction of transputers executing user code, by means of a
     request-resource / relinquish-resource protocol."
  title = "{T}he computing tower: {A} supercomputer for real- time simulation of continuous systems",
  author= "Renterghem, Patrick van",
  editor= "Askew, Charlie",
  pages = "113--124",
  booktitle= "{OUG}-9: {O}ccam and the {T}ransputer -- {R}esearch and {A}pplications",
  isbn= "90 5199 010 3",
  year= "1988",
  month= "sep",
  abstract= "Computer simulation of systems (e.g. magnetic fusion,
     nuclear power plants, weather forecasting, ...) will have a
     strong impact on theoretical and experimental research in
     the future, because experiments on real systems are
     impossible due to the danger, the cost or the duration of
     the experiment. Computer simulation of a system is therefore
     an attractive alternative. Generally, large and complicated
     systems are simulated, so this requires an enormous amount
     of computing power. In this paper, we describe the
     possibility to use a home-built multitransputersystem, The
     Computing Tower, for system simulation and methods to
     implement a block-structured simulation language on this
     system the most efficient way."
  title = "{T}he application of transputers and occam to an industrial energy management system",
  author= "Sinclair, Andy and Kelly, Paul",
  editor= "Askew, Charlie",
  pages = "125--134",
  booktitle= "{OUG}-9: {O}ccam and the {T}ransputer -- {R}esearch and {A}pplications",
  isbn= "90 5199 010 3",
  year= "1988",
  month= "sep",
  abstract= "The work described here is part of a larger project to
     provide a knowledge-based decision support system to offer
     advice on the control and utilisation of energy within a
     large integrated steel works. It is concerned with the
     electricity power generation subsystem of the steelworks.
     The aim of this work is to produce a decision support system
     that will give advice on the steam distribution required to
     maximise the power output of three turbines when viewed as a
     system. This paper is concerned with the practical details
     of implementing such a system in an industrial environment"
  title = "{F}ast prototyping of architectural designs using transputers",
  author= "Skillicorn, David B.",
  editor= "Askew, Charlie",
  pages = "135--138",
  booktitle= "{OUG}-9: {O}ccam and the {T}ransputer -- {R}esearch and {A}pplications",
  isbn= "90 5199 010 3",
  year= "1988",
  month= "sep",
  abstract= "The existence of a processor with on-board communication
     facilities, such as the transputer, makes it possible to
     prototype new hardware designs by emulating their functional
     units on networks of transputers. This gives better
     information about performance and appearance than simulation
     and can be considerably easier to build. It also permits
     software to be built and executed with performance that
     approximates the final system-hence, software prototyping is
     also possible. We describe our experiences with two
     different systems: an object-based vision system and an
     implementation of the functional language Lucid."
  title = "{GECKO}: {A} graphical tool for the modelling and manipulation of occam software and transputer hardware topologies",
  author= "Stephenson, Marc and Boudillet, Olivier",
  editor= "Askew, Charlie",
  pages = "139--144",
  booktitle= "{OUG}-9: {O}ccam and the {T}ransputer -- {R}esearch and {A}pplications",
  isbn= "90 5199 010 3",
  year= "1988",
  month= "sep",
  abstract= "This paper outlines some of the work undertaken at the
     Polytechnic of Central London (PCL), within the framework of
     the Alvey supported Parsifal project. The aim of this work
     is to help the user with the problems of: conceptualising,
     designing, fine tunning and configuring occam applications."
  title = "{A} state-of-the-art radar pulse deiterleaver",
  author= "Turner, S. P. and Beton, Rick D. and Upstill, C.",
  editor= "Askew, Charlie",
  pages = "145--152",
  booktitle= "{OUG}-9: {O}ccam and the {T}ransputer -- {R}esearch and {A}pplications",
  isbn= "90 5199 010 3",
  year= "1988",
  month= "sep",
  abstract= "This paper discusses the conversion and development of a
     serial implementation of a state-of-the-art radar
     deinterleaving algorithm into a full-scale parallel
     implementation using occam2 and the transputer."
  title = "{T}echniques for rendering solid objects on a processor farm",
  author= "Dew, Peter M. and Holliman, Nick and Morris, David and Pennington, Alan de",
  editor= "Askew, Charlie",
  pages = "153--168",
  booktitle= "{OUG}-9: {O}ccam and the {T}ransputer -- {R}esearch and {A}pplications",
  isbn= "90 5199 010 3",
  year= "1988",
  month= "sep",
  abstract= "Collaborative research by Leeds University and the IBM UK
     Scientific Centre has resulted in an experimental parallel
     solid modeller called MISTRAL, based on the Constructive
     Solid Geometry (CSG) representation of solid objects. This
     paper gives an overview of the CSG rendering algorithms
     implemented in MISTRAL, a description of the MISTRAL
     modeller itself, and describes how several different kinds
     of parallelism may be exploited using a processor farm. It
     concludes by discussing some of the issues involved in
     writing complex parallel programs in OCCAM2."
  title = "{A} prototype simulator output movie system based on parallel processing technology",
  author= "Carmichael, N. and Hewson, D. and Vorst, J. van der",
  editor= "Askew, Charlie",
  pages = "169--175",
  booktitle= "{OUG}-9: {O}ccam and the {T}ransputer -- {R}esearch and {A}pplications",
  isbn= "90 5199 010 3",
  year= "1988",
  month= "sep",
  abstract= "New parallel processing technologies potentially offer
     immense computing power at a very reasonable price. However
     it is not yet clear how best this potential may be realised
     for the large class of applications of concern to the Shell
     computing community. Shell U.K. (Expro) Ltd., as one of that
     company's ninth licensing round initiatives, decided to
     purchase a parallel processing machine (a member of the
     \"computing surface\" range from Meiko Ltd.,a
     Bristol ,U.K., based company). In order to test the
     abilities of the machine and its suppliers, dynamic
     \"movie style\" display of oil or gas reservoir
     simulator output (ie. changing fluid distributions within
     the reservoir) was chosen as an application example. The
     presentation will give a brief description of the
     application and of how it is placed on the machine. These
     devices are now being routinely used by reservoir
     engineers.Operational usage of the system will be discussed
     (a video of an interactive session will be available) and
     some more general remarks about the placement of such
     technology within an extensive heterogeneous computing
     environment will be made."

If you have any comments on this database, including inaccuracies, requests to remove or add information, or suggestions for improvement, the WoTUG web team are happy to hear of them. We will do our best to resolve problems to everyone's satisfaction.

Copyright for the papers presented in this database normally resides with the authors; please contact them directly for more information. Addresses are normally presented in the full paper.

Pages © WoTUG, or the indicated author. All Rights Reserved.
Comments on these web pages should be addressed to: www at wotug.org

Valid HTML 4.01!