One of the major problems facing software developers is that of verification of their product For computational modelling this can be summarised as the question If I run this code for this problem, what degree of faith can I have that the answers will correspond to reality?. Deviations from reality can arise from a number of sources. The physical model and hence the equations being solved may be inappropriate. The numerical approach may be unstable or introduce spurious elements into the solution. And, of course, there may be bugs in the code.
Recognising that this is a particular problem in Computational Fluid Dynamics (CFD), where mistrust of CFD programs leads to a lack of acceptance of such techniques in the design environment, the CFD Community Club is gathering a body of test data against which existing and new programs and algorithms can be checked. The data is aimed mainly at the first two sources of problems but will, we hope, also be of use in flushing out bugs in software.
The test data comes in two parts. There is a catalogue, which is a list of published reports, and papers containing well-characterised and carefully carried out experiments and/or computations. In addition, data from some test problems is available in computer readable form. The Catalogue of Test Data Sets is organised by application area for ease of use into three main headings:
Non-application data makes a fourth section containing items such as analytic solutions to artificial problems. Within the three application sections there are sub-headings such as aerodynamics, combustion, ventilation, mixing, maritime structures and flow in ducts. Thus, if you wish to verify your program in a particular type of application it is simple to look up in the Catalogue to see if a dataset exists.
Release 1 of the Catalogue is now available. It is, of necessity, incomplete in that we do not believe that we have listed all the possible application areas for which data or a need for data exists, nor have we found suitable data for all those applications we have identified. The intention is to update the catalogue at regular intervals and we would be very pleased to hear of suggestions for inclusion in future releases. Any suggestions can be sent to the address below.
Currently the only machine readable data in the catalogue is a collection of results from turbulent flow experiments which were collected by experts for a pair of conferences on Complex Turbulent Flows held in Stanford in 1980 and 1981. This data was originally made available as tapes and latterly as IBM PC disks. It is now accessible via HENSA, the Higher Education National Software Archive. As well as the Data Sets the Community Club have extracted the abstracts of the experiments from the data and reformatted them in LaTeX and plain ASCII versions. We recommend that to save time and bandwidth anybody interested in accessing this data should first look at this abstract to check the data's suitability.
Again we would be very interested in offers of data which could be added to this archive. Please contact me at the address below.
To get a copy of the Catalogue the easiest way is to login over the JANET X25 network using the address unix.hensa.ac.uk and login name archive. You will be asked for your E-mail address. This will be used to send any files you request The hensa filestore is navigated using standard unix commands such as cd and ls. There is a chapter in the misc archive called cfd which contains two sub-directories, data and software. Once you have found what you are looking for in these type:
send <filename>
and the file will be E-mailed to you. Should you have any comments, suggestions (especially of data needed or available for the catalogue) or require any more information, please contact me.
Electronic exchange of product data will become increasingly important in the future. STEP is a significant contribution to engineering data exchange and data management. STEP (Standard for Exchange of Product Data) will be published as a member of Parts and is now rapidly approaching the moment at which the first Parts will become an international standard. The project for the development of STEP started in 1984--now 8 years ago. The long time scale between the start of the project and the emergence of the first products provides evidence for the complexity of the task. The original tasks, as envisaged in 1984, were the creation of standards for data exchange for geometry as it was used in Computer Aided Design systems of that date. Since then, the scope of the project has expanded considerably to include many areas of engineering, for instance:
The STEP standard addresses product data in its widest meaning, ie all data relating to a manufactured product during its entire life cycle.
In addition to those Parts which are focused on a particular subject area of engineering interest, there are a number of Parts which define the underlying structure of the standard. Examples are:
Several important Parts of STEP (those which define the overall STEP structure, geometry, draughting, Express language, physical file exchange and conformance testing) have been, or soon will be, submitted to the International Organisation for Standardisation for balloting on international standard status. A number of other Parts will follow shortly.
Each country has a single vote in the balloting process. The British vote is determined by the AMT/4 committee of BSI.
One important aspect of STEP is the integration between different disciplines. Because all data models are formulated in the same Express language, it is possible to present an integrated view of all models in STEP. The introduction of the concept of Application Protocol has made it possible to make actual implementations of this integrated view. Each Application Protocol defines a particular area of interest and all implementations of STEP are expected to be defined by an Application Protocol. The most important aspect of each Application Protocol is an Express scheme which can be implemented in an actual database or which can be used as a basis for exchanging data between two systems adhering to this Application Protocol.
For more information on STEP, please contact the author.
The netlib archive run from the University of Kent has been extended and now forms part of the Higher Education National Software Archive (HENSA) service.
HENSA is an ISC funded service to provide the academic and research communities with access to a number of software archives. The Kent archive, unix.hensa, currently shadows both the netlib archive and the uunet archive (Unix public domain software). An archive at Lancaster University, micros.hensa, provides access to pc based software.
Although the new netlib service will continue to be accessible from netlib@ukc.ac.uk for the near future, current users are encouraged to change to netlib@unix.hensa.ac.uk. In addition all netlib software may be obtained via both Internet ftp and Blue-book ftp (ni-ftp) and through an interactive browsing facility.
All the archive software is available via both DARPA ftp and Blue-book (ni-ftp).
For DARPA ftp, connect to unix.hensa.ac.uk, log in as anonymous and use your E-mail address as a password. The three main directories are netlib. uunet and misc.
The Blue-book request:
<ARCHIVE>path-of-file from uk.ac.hensa.unix Usemame: guest Password: <your E-mail address>
will retrieve the requested file.
There are two separate E-mail request facilities:
send index to netlib@unix.hensa.ac.uk
send path-of-file to archive@unix.hensa.an.uk
For more details on how to find out where things are, see the section on fbr below.
For example, more details of the HENSA service may be obtained using the request:
send uunet/intro
fbr is an interactive file browser which allows the use of Unix-like commands to move around the archive, list the contents of files and find files or directories by name. Files may be transferred via E-mail using the sendme command.
Connecting to unix.hensa.ac.uk (or 000049200900 if you do not have NRS) will produce the banner and login request:
Welcome to HENSA - the Higher Education National software Archive at the University of Kent at Canterbury Use the login name 'archive' to access the Unix archive login:archive Welcome to the UK Source Archive at the University of Kent at Canterbury Please enter your E-mail address: <your E-mail address>
You will then be placed inside fbr' s restricted shell. Use the help command for up to date details of what commands are available.
More details of how to access the new system may be obtained by Emailing the message:
send index send browser to netlib@unix.hensa.ac.uk
Any problems, comments or suggestions should be E-mailed to: netlib-admin@unix.hensa.ac.uk
Barcelona, the newest Olympic city, was the venue of PACTA '92 - Parallel Computing & Transputer Applications 1992 - Conference. This event brought together many of the leading researchers and companies representing the state of the art in parallel processing.
183 papers were presented. Parallel sessions included: Algorithms, Robotics/Control, Scientific Computing, Linear Algebra, Signal Processing, System Tuning, Neural Nets, Simulations, Image Processing, Finite Elements, CFD, Transputer Applications, OS and Databases.
The papers showed the emergence of parallel processing as a fundamental tool in many fields of computing.
As well as the technical presentations, six keynote lectures were presented by invited speakers. These talks provided a wider perspective of the current developments taking place to promote and engender the exploitation of parallel processing. A review of the High Performance Computing Initiatives being setup in both Europe and the USA was presented. The remaining talks featured Finite Elements, Weather Simulation, Oil Production and Automatic Parallelisation.
The general focus of the technical presentations was the "Teraflop Grand Challenge" - construction of a parallel supercomputer capable of achieving teraflop rates of calculation.
INTEL, IBM, CRAY, Thinking Machines, Fujitsu and Convex each presented their view of what a teraflop machine would be. It was interesting to note that based on current technologies a teraflop computer would cost hundreds of millions of pounds; need to be housed in a small warehouse; and require an extensive cooling system. To operate the machine would require upwards of three Megawatts, requiring its own electricity sub-station to down-load off the national grid. It was noted that switching such a machine on or off would require the approval of the local Electricity authority on each occasion! There was much debate on who could afford to run a teraflop machine and what this would mean to the engineer.
Running in parallel with the conference sessions was a large exhibition of the latest products in the parallel field. The products displayed ranged from large-scale supercomputers (Cray and Fujitsu) to workstation clusters (IBM) down to the current generations of chips from Texas Instruments (C40), Intel (i860) and INMOS (T8OO and T9OOO). It was fairly obvious that the existing transputer range is no longer reigning supreme amongst the stands and that the C40 is currently leading the way, with the i860 close behind.
A panel session on the final day considered how to move applications to teraflop architectures. The panel expressed the need for conformity across the developing architectures to enable easy migration between systems. The panel also stressed the need for the training of engineers to learn the mechanisms required to build and generate code to operate on large-scale multi-processing systems.
Overall, the conference was held in an exotic city and its seemed an appropriate location to discuss the increasingly exotic nature of parallel processing and its future directions.
Visualisation was identified as one of the key areas of the SERC Engineering Applications IT Support Programme (EASE) over the next few years. An effective and successful mechanism within EASE for bringing researchers with a common interest (frequently from different subject disciplines within the engineering community) together are community clubs. These clubs are not limited to either engineers or academics, which gives the members access to a broad range of experience. It was therefore a natural step to establish an EASE Community Club in Visualisation, which happened during the summer 1991.
The role of the Visualisation Community Club is to help researchers, developers and users interested in visualisation by:
The Club's Steering Group forms a regular point of contact between the community and EASE technical staff at Rutherford Appleton Laboratory (RAL). It is an essential link between the two, representing the views of the larger user community, and transmitting the results of the programme back to the community.
The Chairman of the Visualisation Community Club is Dr K W Brodlie (Leeds University) and the Secretary is Mrs R Popovic (RAL).
If you wish to join the Club, a membership form (including register of activity and interest) may be obtained by contacting me.
Organised by BCS Parallel Processing Specialist Group and Leeds University School of Computer Studies.
Following the first workshop on this theme in 1991, the aim is to provide a forum for the discussion of abstract machine models for highly parallel computers. Particular attention will be given to the specification, implementation and application of such models and to the identification of key issues for future research.
Invited speakers will give keynote presentations. A selection of submitted papers which highlight key themes will also be presented. Working groups will focus on major issues identified through submitted papers and position statements. The emphasis will be on providing maximum opportunity for discussion, interchange of ideas and identifying the most promising areas for progress.
The workshop will be restricted to 60 active researchers in relevant fields. In the event that more than 60 applications are received, selection will be on the basis of position statements or full papers received. All delegates will receive duplicated copies of these statements and papers. Selected papers will, after suitable revision, be included in an ensuing book from a leading publisher.
Invited Speakers include
Potential participants are asked to submit EITHER a short position statement (no more than 2 sides of A4) OR a full paper of up to 6000 words. Only full papers will be considered for presentation at the conference and for the subsequent book.
Suitable topics for papers include, but are not limited to: The role of abstraction in modelling parallel computation. Models for general purpose parallel and distributed computing. PRAM models. Performance modelling and verification. The role of abstract models in parallel and distributed program design. Hardware support for abstract machine models. Shared objects and object-oriented programming for parallel computers Portability and Open standards.
FIVE copies of full papers should be submitted no later than 16 November 1992. TWO copies of position statements should be submitted no later than December 31, 1992. All correspondence should be sent to:
The Abstract Machine Workshop Secretary
School of Computer Studies
The University LEEDS
Engineers and scientists would love to run their FORTRAN applications on faster machines. However, access time to supercomputers such as the CRA Y is limited and expensive. A new generation of high performance parallel architecture, based on microprocessors with high-speed interconnect, has the potential of providing an affordable desktop resource for computer intensive applications.
As part of the EASE programme, a course entitled FORTRAN for the transputer, i860 and C40, will be held at the Rutherford Appleton Laboratory on 27-28 January 1993. The aim of the course is for attendees to learn how to move their FORTRAN codes onto such parallel systems: from the general principles to particular techniques. The method will be a combination of presentations and practical sessions.
It is understandable that application developers are reluctant to move to new environments. Will support still be around for the latest parallel machine in 5 years time? Here standardisation is the key. The tutors will provide information on the standardisation efforts in progress and the best systems to support the portability and future-proofing of source code.
2nd School on Computational Fluid Dynamics, The Cosener's House, 4-8 January 1993
Following the success of its Summer School in Computational Fluid Dynamics (CFD) held in September 1991, the CFD Community Club has organised, in collaboration with the Royal Military College of Science (RMCS), a second School in CFD for January 1993.
The School will provide a foundation in the physical understanding of fluid flow and the consequences for the numerical solution of fluid flow problems. It is intended to be suitable for recently started graduate students or those returning to or starting in CFD research.
The topics to be covered in the School include:
The formal sessions will comprise a mixture of review lectures on aspects of these topics in addition to lectures which present current research. Lectures whose participation is confirmed are:
Some of the research topics that will be covered by the lecturing staff are:
In addition to the formal lectures, there will be practical sessions where all participants will tackle a set of test problems which will be solved using some of the methods presented in the lectures. A network of SUN3s, SUN4s, Stardent and Silicon Graphics superworkstations, will form the basis of the computing facilities for the School. Some commercial packages and the lecturers' own software will be mounted on the system for use in the practical sessions. A brief introduction to computing under Unix will be provided for those not familiar with this operating system.
RAL are organising an advanced workshop for developers of the next generation of User Interface Management Systems (UIMS) on behalf of the Department of Trade & Industry. This will take place at Queen Mary Westfield College on 2 December 1992. Attendance is restricted to 20 people and is by invitation only. The workshop is designed for UK researchers who are developing UIMS and other management systems to meet and discuss the area.
This conference will provide a forum for the presentation and discussion of recent advances in the development and application of numerical methods to the solution of heat transfer problems. Some key areas include:
The Organizing Committee welcome the submission of papers describing recent work within this general area. It is expected that most submitted papers will report on recently developed computational techniques, in particular finite difference and finite element methods. However, papers dealing with the comparison of standard numerical models with experimental data are also welcome. Papers involving innovative methods in thermal problems and industrial applications are also strongly encouraged as in the previous conferences.
The EASE Programme is organising a Workshop on Developing Parallel Engineering Applications to be held at The Cosener's House, Abingdon, on 23-24 February 1993.
Faced with the variety of parallel machines, programming models, parallel language extensions, development environments and tools, it is not surprising that application producers have yet to take up parallel processing technology in any wholehearted way despite its technical promise of affordable scalable performance. Perceived difficulties in developing applications to run on parallel systems has also limited exploitation of the technology to date.
The past two years have seen an increasing degree of commonality of approach across parallel systems, e.g. in message passing based on through-routing, in the emergence of systems with virtual shared memory, and in the growing convergence of parallel and distributed systems. Programming parallel applications has certainly become easier and more portable as a result, in most cases with little or no loss of performance.
The objectives of the Workshop are to:
Topics of interest include:
The programme will consist of selected presentations from position papers, a small number of invited presentations, discussion sessions in subgroups, and a closing plenary session. The number of places is limited (about 40) to foster active participation by all. A booklet of position papers will be distributed three weeks before the Workshop. A report will be produced containing the conclusions of the discussion sessions.
