Refer Proceedings details

%T A transputer implementation of a runtime timing constraint\-monitor
%A Fergus E. Fletcher
%E Patrick Nixon
%B Proceedings of WoTUG\-18: Transputer and occam Developments
%X A transputer implementation of a runtime constraint monitor
   is described. This monitor \-\- Sentinel \-\- uses
   Real\-Time Logic (RTL) for specifying a program\[rs]s timing
   constraints. Sentinel is a variant of an existing
   centralised uniprocessor monitor. In Sentinel,

%T PARAVER: A Tool to Visualize and Analyze Parallel Code
%A V. Pillet, J. Labarta, T. Cortes, S. Girona
%E Patrick Nixon
%B Proceedings of WoTUG\-18: Transputer and occam Developments
%X 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.

%T Parallel Imperative Functional Programming
%A Steve Hill
%E Patrick Nixon
%B Proceedings of WoTUG\-18: Transputer and occam Developments

%T A fast C Kernel for Portable occam Compilers
%A Barry M. Cook
%E Patrick Nixon
%B Proceedings of WoTUG\-18: Transputer and occam Developments
%X 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
   creation.

%T Porting the INMOS occam Compiler to the SPARC Architecture
%A K. Rothwell, G. Shaw, A Smith
%E Patrick Nixon
%B Proceedings of WoTUG\-18: Transputer and occam Developments
%X 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
   architecture.

%T Optimum Transputer Configurations for Real Applications Requiring Global Communication
%A Colin J. Burgess, Alan G. Chalmers
%E Patrick Nixon
%B Proceedings of WoTUG\-18: Transputer and occam Developments
%X 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.

%T Parallel application development with dynamo
%A A. Biriukov, D. Ulyanov
%E Patrick Nixon
%B Proceedings of WoTUG\-18: Transputer and occam Developments
%X 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.

%T Parallelisation for the progressive refinement radiosity method for the synthesis of realistic images
%A Peter Green, Ed Morgan
%E Patrick Nixon
%B Proceedings of WoTUG\-18: Transputer and occam Developments
%X 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.

%T Data Mining in Parallel
%A Sarabjot S. Anand, C. Mary Shapcott, David A. Bell, John G. Hughes
%E Patrick Nixon
%B Proceedings of WoTUG\-18: Transputer and occam Developments
%X 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
   implementation.

%T Linda implementation revisted
%A Andrew Douglas, Alan Wood, Antony Rowstron
%E Patrick Nixon
%B Proceedings of WoTUG\-18: Transputer and occam Developments
%X 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\[rs]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.

%T Real\-time Performance Evaluation Issues for Transputer Networks
%A M. Tokhi, M. Hossain, M. Baxter, P. J. Fleming
%E Patrick Nixon
%B Proceedings of WoTUG\-18: Transputer and occam Developments
%X 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.

%T Using records and pointers with occam 2.5
%A Michael D. Poole
%E Patrick Nixon
%B Proceedings of WoTUG\-18: Transputer and occam Developments
%X 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.

%T Mapping PCSC Design Model to Inmos ANSI C
%A S. Sahib, Gordon A. Manson
%E Patrick Nixon
%B Proceedings of WoTUG\-18: Transputer and occam Developments
%X 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.

%T Modelling Real\-Time Behaviour of Parallel Transputer Systems under Failure Conditions
%A Peter R. Croll, P. M. Griffiths
%E Patrick Nixon
%B Proceedings of WoTUG\-18: Transputer and occam Developments
%X 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 discussed.

%T Application of Transputer\-based Parallel Computation in Matching Real\-Time Control Models
%A Tushar K. Hazra
%E Patrick Nixon
%B Proceedings of WoTUG\-18: Transputer and occam Developments
%X Mathematical models describing physical processes can be
   \[rs]matched\[rs] 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\[rs]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.

%T Incorporating Use\-Case Analysis in PCSC Method
%A Chandran Elamvazuthi, Gordon A. Manson
%E Patrick Nixon
%B Proceedings of WoTUG\-18: Transputer and occam Developments
%X 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.

%T A Distributed I/O Communication Protocol for a Network of Transputers
%A Runhe Huang, Mike Morgan
%E Patrick Nixon
%B Proceedings of WoTUG\-18: Transputer and occam Developments
%X 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.

%T A Parallel Programming Methodology Based on Paradigms
%A Fethi A. Rabhi
%E Patrick Nixon
%B Proceedings of WoTUG\-18: Transputer and occam Developments
%X 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.