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 = "{A} transputer implementation of a runtime timing constraint-monitor",
  author= "Fletcher, Fergus E.",
  editor= "Nixon, Patrick",
  pages = "1--15",
  booktitle= "{P}roceedings of {W}o{TUG}-18: {T}ransputer and occam {D}evelopments",
  isbn= "90 5199 222 X",
  year= "1995",
  month= "mar",
  abstract= "A transputer implementation of a runtime constraint monitor
     is described. This monitor -- Sentinel -- uses Real-Time
     Logic (RTL) for specifying a program's timing constraints.
     Sentinel is a variant of an existing centralised
     uniprocessor monitor. In Sentinel,"
  title = "{PARAVER}: {A} {T}ool to {V}isualize and {A}nalyze {P}arallel {C}ode",
  author= "Pillet, V. and Labarta, J. and Cortes, T. and Girona, S.",
  editor= "Nixon, Patrick",
  pages = "17--31",
  booktitle= "{P}roceedings of {W}o{TUG}-18: {T}ransputer and occam {D}evelopments",
  isbn= "90 5199 222 X",
  year= "1995",
  month= "mar",
  abstract= "PARAVER (PARallel Visualization and Events Representation)
     is a tool to visualize and analyze a parallel trace file.
     PARAVER can be used in two different ways: Creating its own
     trace file in accordance with the PARAVER trace format, or
     using a modified PVM library. PARAVER is based on a simple
     interface to manage several displaying windows. In fact,
     PARAVER can be seen as a tape recorder where the trace file
     is the tape. PARAVER provides many functionalities to see
     and analyze qualitatively the trace file. Moreover, the user
     can add his own analysis function to extend the
     functionality of the tool. This paper discusses the basic
     concept, design and use of PARAVER. We explain in detail
     each window. The different functionalities of PARAVER are
     also examined. Finally we present the future directions of
     our research."
  title = "{P}arallel {I}mperative {F}unctional {P}rogramming",
  author= "Hill, Steve",
  editor= "Nixon, Patrick",
  pages = "33--46",
  booktitle= "{P}roceedings of {W}o{TUG}-18: {T}ransputer and occam {D}evelopments",
  isbn= "90 5199 222 X",
  year= "1995",
  month= "mar",
  title = "{A} fast {C} {K}ernel for {P}ortable occam {C}ompilers",
  author= "Cook, Barry M.",
  editor= "Nixon, Patrick",
  pages = "47--65",
  booktitle= "{P}roceedings of {W}o{TUG}-18: {T}ransputer and occam {D}evelopments",
  isbn= "90 5199 222 X",
  year= "1995",
  month= "mar",
  abstract= "This paper describes an experiment in providing explicitly
     parallel constructs in the style of occam within a
     functional programming environment. The work is based upon
     the monadic style proposed and described by Moggi, Wadler,
     Peyton Jones and others. The approach is pragmatic and
     conservative in that parallelism is treated as an imperative
     feature implemented by monadic functions. Overloading is
     used to provide a class of channel protocols. As in occam,
     data transmitted over a channel is fully evaluated and must
     be of ground type. An algorithm for checking channel usage
     in the presence of dynamic process and channel creation is
     described. The work can be described as an extension of
     functional notations allowing for explicit imperative-style
     parallelism. Alternatively, one could view it as extending
     the vocabulary of the occam language to allow for recursion,
     higher-order constructions and dynamic process and channel
  title = "{P}orting the {INMOS} occam {C}ompiler to the {SPARC} {A}rchitecture",
  author= "Rothwell, K. and Shaw, G. and Smith, A",
  editor= "Nixon, Patrick",
  pages = "67--74",
  booktitle= "{P}roceedings of {W}o{TUG}-18: {T}ransputer and occam {D}evelopments",
  isbn= "90 5199 222 X",
  year= "1995",
  month= "mar",
  abstract= "occam is a relatively uncomplicated programming language
     designed to efficiently program parallel computer systems.
     It is not tied to a particular architecture but until
     recently has only been available for the transputer range of
     microprocessors. This paper describes a port of the standard
     INMOS occam 2 toolset Transputer compiler to the Sun SPARC
  title = "{O}ptimum {T}ransputer {C}onfigurations for {R}eal {A}pplications {R}equiring {G}lobal {C}ommunication",
  author= "Burgess, Colin J. and Chalmers, Alan G.",
  editor= "Nixon, Patrick",
  pages = "75--85",
  booktitle= "{P}roceedings of {W}o{TUG}-18: {T}ransputer and occam {D}evelopments",
  isbn= "90 5199 222 X",
  year= "1995",
  month= "mar",
  abstract= "If complex problems are to be solved in reasonable
     computation times, then large scale parallel processing is
     necessary. For many of these problems, the density of the
     global communications dominated the performance of the
     parallel implementation. In these cases, the design of the
     interconnection network for the processors is known to play
     a significant part in the efficient implementation of
     problems on a large T800 transputer systems. This paper
     presents a new genetic algorithm for generating optimal
     configurations, augmented by simulated annealing for
     selected refinement of difficult cases. These configurations
     have the further advantage that they satisfy the best known
     criteria for producing configurations that perform well on
     real applications. The paper concludes by describing the
     impact this might have on the design of future T9000
     transputer configurations."
  title = "{P}arallel application development with dynamo",
  author= "Biriukov, A. and Ulyanov, D.",
  editor= "Nixon, Patrick",
  pages = "87--95",
  booktitle= "{P}roceedings of {W}o{TUG}-18: {T}ransputer and occam {D}evelopments",
  isbn= "90 5199 222 X",
  year= "1995",
  month= "mar",
  abstract= "An approach to the development of parallel applications for
     the distributed memory computers based on the notion of
     configuration is presented. The configuration introduces a
     general context for the problems such as source program
     optimization, load balancing, mapping, processor network
     development. The paper contains the experimental results
     obtained with the DYNAMO system based on this approach."
  title = "{P}arallelisation for the progressive refinement radiosity method for the synthesis of realistic images",
  author= "Green, Peter and Morgan, Ed",
  editor= "Nixon, Patrick",
  pages = "97--112",
  booktitle= "{P}roceedings of {W}o{TUG}-18: {T}ransputer and occam {D}evelopments",
  isbn= "90 5199 222 X",
  year= "1995",
  month= "mar",
  abstract= "This paper discusses strategies for parallelising the
     radiosity method for image synthesis applicable to message
     passing architectures, in particular, transputer-based
     machines. The basic radiosity method, and a number of its
     developments are reviewed, as are previous parallelisation
     schemes. A number of parallelisation strategies are then
     discussed, with emphasis being placed on techniques for
     maximising the efficiency of the schemes by reducing
     communications overheads, scene structuring and scene
     partitioning. Results from the implementation of the various
     parallelisation approaches on a transputer-based Meiko
     Computing Surface are presented. The implications of the
     parallelisation strategies in terms of solution quality are
     also considered."
  title = "{D}ata {M}ining in {P}arallel",
  author= "Anand, Sarabjot S. and Shapcott, C. Mary and Bell, David A. and Hughes, John G.",
  editor= "Nixon, Patrick",
  pages = "113--124",
  booktitle= "{P}roceedings of {W}o{TUG}-18: {T}ransputer and occam {D}evelopments",
  isbn= "90 5199 222 X",
  year= "1995",
  month= "mar",
  abstract= "In this paper we discuss the efficient implementation of the
     STRIP (Strong Rule Induction in Parallel) algorithm in
     parallel using a transputer network. Strong rules are rules
     that are almost always correct. We show that STRIP is well
     suited for parallel implementation with scope at four
     different levels for the transputer network using different
     number of transputers. The choice of certain variables (the
     number and size of samples) in the STRIP algorithm affects
     the performance (speedup and efficiency) of the
  title = "{L}inda implementation revisted",
  author= "Douglas, Andrew and Wood, Alan and Rowstron, Antony",
  editor= "Nixon, Patrick",
  pages = "125--138",
  booktitle= "{P}roceedings of {W}o{TUG}-18: {T}ransputer and occam {D}evelopments",
  isbn= "90 5199 222 X",
  year= "1995",
  month= "mar",
  abstract= "Linda is a model for communication and co-ordination of
     parallel processes. The model provides a virtual shared
     memory called tuple space, and primitives to place tuples
     into and remove tuples from tuple space. The style of
     programming provided by Linda is quite different to the
     style of, say, occam. We describe a new implementation of
     Linda across a network of transputers. We provide the four
     Linda primitives, in, out, rd and eval, together with a new
     primitive, collect, developed at York. The implementation
     focusses on two issues. The first issue is that the ordering
     of out operations in a sequential process must be preserved
     if we want Linda to act as a co-ordination language. Our
     implementation provides this. The second issue is the
     implementation of eval, Linda's mechanism for spawning
     processes. We outline an implementation which provides
     arbitrary spawning of processes which execute concurrently,
     despite the restriction, enforced by the transputer
     architecture, of declaring a static number of processes at
     compile time. We provide a small example to show how Linda
     can be used to write parallel programs, then outline current
     work being undertaken at York, which focusses on
     interpretive environments for high level parallel
     programming techniques. A prototype Linda implementation and
     ISETL interpreter have already been developed."
  title = "{R}eal-time {P}erformance {E}valuation {I}ssues for {T}ransputer {N}etworks",
  author= "Tokhi, M. and Hossain, M. and Baxter, M. and Fleming, P. J.",
  editor= "Nixon, Patrick",
  pages = "139--150",
  booktitle= "{P}roceedings of {W}o{TUG}-18: {T}ransputer and occam {D}evelopments",
  isbn= "90 5199 222 X",
  year= "1995",
  month= "mar",
  abstract= "This paper presents an investigation into the critical
     issues involved in real-time performance evaluation of a
     transputer network. Issues such as algorithm partitioning,
     mapping, inter-processor communication, granularity,
     regularity and compilers efficiency for numerical
     computation are investigated and presented in this chapter.
     A finite difference simulation algorithm for a flexible beam
     in transverse vibration evaluation is made, demonstrating
     fast processing techniques for real-time implementations."
  title = "{U}sing records and pointers with occam 2.5",
  author= "Poole, Michael D.",
  editor= "Nixon, Patrick",
  pages = "151--162",
  booktitle= "{P}roceedings of {W}o{TUG}-18: {T}ransputer and occam {D}evelopments",
  isbn= "90 5199 222 X",
  year= "1995",
  month= "mar",
  abstract= "The language extensions from occam 2 to occam 2.1 are
     introduced. A programming project using the language is
     described. Specific examples of the use of the new named
     types and record structures including pointers, are
     described. Some specific and general conclusions are drawn."
  title = "{M}apping {PCSC} {D}esign {M}odel to {I}nmos {ANSI} {C}",
  author= "Sahib, S. and Manson, Gordon A.",
  editor= "Nixon, Patrick",
  pages = "163--181",
  booktitle= "{P}roceedings of {W}o{TUG}-18: {T}ransputer and occam {D}evelopments",
  isbn= "90 5199 222 X",
  year= "1995",
  month= "mar",
  abstract= "A methodology for parallel system design called Parallel
     Communicating Sequential Code (PCSC) is currently being
     developed. A CASE tool based on this methodology is also
     being built which automatically generates C code for the PVM
     environment. Work is currently being carried out to extend
     the code generator to automatically generate C code for the
     Inmos toolset. This paper discusses the mapping of the PCSC
     design model onto Inmos Ansi C to make the automatic code
     generation possible. Where relevant, comparisons are made
     with respect to the C-PVM implementation."
  title = "{M}odelling {R}eal-{T}ime {B}ehaviour of {P}arallel {T}ransputer {S}ystems under {F}ailure {C}onditions",
  author= "Croll, Peter R. and Griffiths, P. M.",
  editor= "Nixon, Patrick",
  pages = "183--195",
  booktitle= "{P}roceedings of {W}o{TUG}-18: {T}ransputer and occam {D}evelopments",
  isbn= "90 5199 222 X",
  year= "1995",
  month= "mar",
  abstract= "This paper describes work concerning the modelling of
     transputer The aims and problems of designing real-time
     systems are introduced. The key features of the model which
     are relevant to reliability are discussed. Algorithms for
     scheduling parallel and mono-processor real-time system are
     explained. Coloured Petri-nets in general are introduced,
     and then the key features of the CASE tool Design/CPN are
     covered. Details of how the model has been implemented, and
     how the model behaves are given. Conclusions from the work
     so far are presented, and the outstanding areas of work are
  title = "{A}pplication of {T}ransputer-based {P}arallel {C}omputation in {M}atching {R}eal-{T}ime {C}ontrol {M}odels",
  author= "Hazra, Tushar K.",
  editor= "Nixon, Patrick",
  pages = "197--212",
  booktitle= "{P}roceedings of {W}o{TUG}-18: {T}ransputer and occam {D}evelopments",
  isbn= "90 5199 222 X",
  year= "1995",
  month= "mar",
  abstract= "Mathematical models describing physical processes can be
     'matched' to discrete samples of processes by considering
     measured input-output data and adjusting coefficient
     parameters of the model to provide an optimal agreement
     between the system and model responses. This well known
     technique has been widely implemented in an off-line basis
     for many years. The process of matching can be based on
     standard optimization procedures such as Davidon Fletcher
     Powell, Fletcher Reeves or even simply Newton Raphson
     methods. Standard simulation routines such as Runge Kutta or
     Euler's methods can be used to generate the model responses.
     However, the procedure for model matching involves intensive
     computing. This paper offers a number of prospective
     approaches to the solution based on parallel computing
     methodologies. Using standardized model matching experiments
     and simulated noise-free data, performance comparisons for
     sequential and multiprocessor computations have been
     evaluated on a transputer-based system."
  title = "{I}ncorporating {U}se-{C}ase {A}nalysis in {PCSC} {M}ethod",
  author= "Elamvazuthi, Chandran and Manson, Gordon A.",
  editor= "Nixon, Patrick",
  pages = "213--226",
  booktitle= "{P}roceedings of {W}o{TUG}-18: {T}ransputer and occam {D}evelopments",
  isbn= "90 5199 222 X",
  year= "1995",
  month= "mar",
  abstract= "Use-case analysis is a simple but useful technique for
     driving the process of analysis in a meaningful manner. It
     is used, in one form or another, in various
     object-orientated software development methods notably OOSE,
     Booch and OMT. In this paper, we explain how use-case
     analysis has been incorporated into PCSC, a method based on
     the occam model for developing parallel software systems. To
     illustrate the usage of use-case analysis in PCSC, we
     present a simple case study of a Character Sorting System."
  title = "{A} {D}istributed {I}/{O} {C}ommunication {P}rotocol for a {N}etwork of {T}ransputers",
  author= "Huang, Runhe and Morgan, Mike",
  editor= "Nixon, Patrick",
  pages = "227--237",
  booktitle= "{P}roceedings of {W}o{TUG}-18: {T}ransputer and occam {D}evelopments",
  isbn= "90 5199 222 X",
  year= "1995",
  month= "mar",
  abstract= "This paper describes an asynchronous communication protocol
     for an embedded network of transputer I/O devices. The aim
     of the protocol was to allow a message routing
     infrastructure small enough to fit withing the internal
     memory of the inexpensize T2 transputer range and yet be
     flexible enough to cater for all control and message routing
     requirements of a complex assembly machine. It was also a
     requirement that the network was scalable without the need
     to recompile the source code."
  title = "{A} {P}arallel {P}rogramming {M}ethodology {B}ased on {P}aradigms",
  author= "Rabhi, Fethi A.",
  editor= "Nixon, Patrick",
  pages = "239--251",
  booktitle= "{P}roceedings of {W}o{TUG}-18: {T}ransputer and occam {D}evelopments",
  isbn= "90 5199 222 X",
  year= "1995",
  month= "mar",
  abstract= "Todays efforts are mainly concentrated on providing
     \"standard\" parallel languages to ensure the
     portability of programs across various architectures. It is
     now believed that the next level of abstraction that will be
     addresses is the application level. This paper argues that
     there is an intermediate level that consists of common
     parallel programming paradigms. It describes some of these
     paradigms and explains the basic principles behind a
     \"paradigm-orientated\" programming approach.
     Finally, it points to future directions which can make it
     feasable to build CASE tools that achieveautomatic parallel
     code generation."

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!