The experimental facilities available to users of the Rutherford Laboratory for scientific studies are capable of generating very considerable quantities of data. To allow the analysis of such data, some of the most powerful computing facilities in Great Britain are installed at the Laboratory. These are used not only for the analysis of the very large volumes of data, which are collected in Particle Physics experiments and aboard space satellites, but also in computations employing complex mathematical models developed for the theoretical study of physical and chemical phenomena.
The use of the computing facilities is implicit in much of he work reported in other sections. The work reported in the following sub-sections, however, has taken place mainly in the two Divisions of the Laboratory responsible for the development of the computing facilities. During the year the work and structure of the two Divisions was rationalised. Broadly, the Computing and Automation Division assumed the responsibility for running the existing IBM 360/195 and ICL 1906A computers and the FR80 microfilm recorder, for the provision of computing services based on this equipment to a widely distributed community of users, and for the development of the systems and applications software and the telecommunications facilities necessary to sustain a high level of service. The Atlas Computing Division, again broadly, was assigned responsibility for the detailed implementation of the outline proposals on the SRC's Interactive Computing Facility and for continuing work on applications programming in such specialised fields as Quantum Chemistry, Crystallography, and Databases.
The Atlas Computing Division has two major roles to play in support of computing within the Science Research Council.
The implementation of the recommendations of the Technical Group Report on Engineering Computing Requirements (April 1976) has been pursued vigorously during the year. In the early phase of the project, considerable effort was put into the tender exercise for the large central interactive computer to be installed at Chilton. After full discussion with six manufacturers, a comprehensive tender document was prepared containing both batch and interactive benchmarks. Unfortunately, due to lack of funds generally within the SRC, the decision was taken in September 1976 not to buy a large central machine, but instead, to proceed more quickly with a programme for developing multi-user minicomputer systems as the main method for providing interactive facilities within SRC in future.
After a tender exercise involving thirty manufacturers, orders were placed for a Prime 400 and GEC 4070. The Prime 400 has been installed, and is intended to provide interactive facilities for up to 100 new users after appropriate software development. It is intended that the Prime 400 remain at Chilton and give the Laboratory local experience of managing an interactive system. The GEC 4070 will be placed in a university department once the assessment and development of the system is complete. Both machines have 192K bytes of memory, card reader, lineprinter and communications lines for about 8 users. The Prime 400 has two 70M byte discs, while the GEC 4070 has two 35M byte discs. Initially, these systems will be assessed by mounting and running existing interactive software packages at present used in university engineering departments. This assessment programme will be a joint exercise between the laboratory and the universities of Cambridge, Glasgow and Leeds.
The proposed enhancements of the PDP 10 systems at Edinburgh and UMIST have been completed, and an extended service at both sites is planned. The programme of enhancements of these PDP10 machines has been co-ordinated by the Division. Interactive services are to be operated by the two universities under contract with SRC, to operate from December 1976, Arrangements have been made to connect the three main sites; Edinburgh, UMIST and Chilton, by 9.6K bytes Post Office communications lines, and further connections to the individual star networks are planned as well as a connection between the Interactive Computing Facility network and the IBM 360/195 processors at the Laboratory, thus providing access to powerful batch computing facilities.
In addition to the general meeting held on 26 April 1976 to inform the user population in the universities and polytechnics of the plans for the Interactive Facility, the Laboratory has organised five meetings with special subject groups, including finite element techniques, artificial intelligence, electromagnetic studies, and circuit design (digital and continuous). Those who attended these special meetings have assisted us greatly in defining suitable software and associated development programme which the Interactive Computing Facility can usefully support.
The past year has marked the introduction of a new version of the Laboratory's quantum chemistry program package, ATMOL3. This new version provides a wider range of Hartree-Fock methods, more powerful molecular integral evaluation codes, and can handle larger cases than the previous version, ATMOL2. The new version is now well documented, and in production use by numerous university groups.
To provide facilities for more accurate computation of molecular wave functions, a general purpose configuration interaction program constructed in Munich in a collaboration between workers at the University of York and the Max Planck Institut, Munich, has been obtained as part of the Meeting House project, and in collaboration with the originators, numerous corrections and extensions have been incorporated; particularly new procedures for configuration selection have been coded, and new procedures for the diagonalization of large sparse matrices have been incorporated, to provide powerful facilities for the extraction of excited state wave functions. The system is interfaced to ATMOL3, and now referred to by the acronym SPLICE. SPLICE is now in production, thus providing the first easily available large scale general purpose configuration interaction program for UK workers.
The SRC microdensitometer service to university and polytechnic crystallographers, based on an Optronics P1000 photoscan interfaced to a Computer Automation Alpha-16 minicomputer, has continued to run successfully and has served approximately thirty university groups. The intensity data for some 75 structures have been collected during the year, an increase of 50% on last year. There are now some 30 publications in the crystallographic literature which describe crystal structure analyses based on data obtained from this service. Accuracy remains excellent; the R-factor, which is a measure of the overall accuracy of the densitometric and film taking operations and the theoretical model, for about 60 fully refined structure analyses averages 7.0%, which compares well with factors of 10% and 4-5% obtainable using the human eye or a diffractometer respectively.
A minor, though not unimportant application of the microdensitometer continued to be in the routine digitization of film negatives to magnetic tape. Such jobs were performed for 15 different groups, and for a total of over 200 films. The range of disciplines covered included pneumoconiosis studies of lung X-ray photographs for the Medical Research Council, X-ray fibre diffraction patterns for Exeter University and proton beam pictures of metal edges for UKAEA Harwell.
An important project was the implementation on the Interactive Computing Facility's PDP10 computer at Edinburgh of an on-line interactive program for the retrieval of crystal structure information compiled by Dr 0 Kennard's group at Cambridge. The work was performed at the request of the SRC Data Compilation Committee, and will provide a country-wide service to crystallographers and chemists from early in 1977. The system allows the rapid retrieval and display of bibliographic and structural information from the complete organic and organometallic crystallographic literature of over 14,000 compounds.
The work for this project has been logically divided into two components:
All code for this project was received from the originator, R J Feldmann of the National Institutes of Health, Bethesda, USA, but considerable work has since been put into correcting and enhancing the original code.
The Information International Inc FR80 microfilm record has run for the past year using the basic software supplied by the manufacturer, and an accounting system develop by Atlas Computing Division. Modifications of the manufacturer's software have resulted in some 10,000 lines of code being added to the standard software. These additions have had to be installed as binary patches because the original source code has not been made available by the manufacturer, and these patches have to be re-applied with each release of the manufacturer's software. In consequence, it has become increasingly difficult to maintain and develop the existing software, and therefore a totally new system called FR80 DRIVER is under construction.
FR80 DRIVER will provide all the facilities of the existing software and additionally includes a number of features to improve the performance of the FR80 and extend its facilities in an upwards compatible mode. These new capabilities include performance measurement, powerful text processing orders and a sophisticated colour selection mechanism. FR80 DRIVER has been specifically designed to provide a flexible, maintainable basis for future developments. The prototype was completed in April 1976, and the initial production version is due for release in February 1977 .
Since the arrival of the FR80 microfilm recorder, interest has grown steadily in computerised text processing, to improve the quality and economics of certain publications. Particularly suitable candidates for this work are large collections of text and/or numeric data which are already in machine-readable form (on tape or disc). Much of the time consuming activity involved in manual typesetting of tables and indexes followed by the necessary proof reading can be eliminated by using suitable programs on a large computer (IBM 360/195 or ICL 1906A) and the FR80 to handle the layout. Additionally, the FR80 allows one to mix text and numeric tables with diagrams, and to vary the output medium (eg microfiche, 35 mm film or hardcopy).
One example of such work at the Laboratory has been a project to produce the print masters for the indexes to publications in the field of X-ray crystallography. This has been undertaken in collaboration with staff from the Cambridge Crystallographic Data Centre and members of the Data Bases, Information Retrieval and Text Processing Section of the Atlas Computing Division under the sponsorship of the SRC Data Compilation Committee. The preparation and validation of the data is performed at Cambridge and the data is then sent to Chilton on magnetic tape. The tape is subsequently processed by programs developed by the Laboratory, to control the choice of font, tabulation, line-splitting, justification (vertical and horizontal), pagination etc, before generating the appropriate orders to the FR80 to output the result at high speed. In addition to the yearly indexes, a 20-year cumulative index of about 14,000 compounds is in the course of preparation.
It is anticipated that the experience gained from this project will enable more complex programs to be developed. A further publication that the Cambridge group hope to produce by this method requires the economical formatting of text mixed with tables of various sizes and accurately drawn stereoscopic pictures for about 1700 molecular structures.
This twelve-minute 16mm computer-produced colour film, complete with optical sound-track, was the award winning Great Britain entry in the International Technical Films Competition held in Moscow, October 1976. Produced by the Division in collaboration with the Royal College of Art, Finite Elements introduces the theory behind this engineering computational technique, and presents a selection of real engineering problems which have been solved by the finite elements method.
Although intended for an audience of practising engineers, the film is eminently suitable, because of its visually attractive computer-animated screenplay, for a non-scientific audience. The computer-generated title music adds the finishing touch to what must be counted as the first entirely computer-produced engineering film to be made in the UK. It is now being marketed by Compeda Ltd., a subsidiary of the National Research Development Corporation, London.
The data collected by the S2/68 ultra-violet sky survey telescope aboard the TD1-A satellite is still the subject of much activity. Workers at University College, London and the Royal Observatory, Edinburgh supported by staff of the Atlas Computing Division have used the data banks extracted by the first scan of the raw data tapes; the main topics of interest have been the distribution and properties of interstellar dust and ultra-violet objects. The FR80 has been used extensively to plot the spectra.
The second scan of the raw data is almost complete. It is building up a new data-bank containing all the useful spectra detected by the experiment - about 200,000 spectra of 30,000 stars, many of them very faint. A star catalogue is to be published on microfiche, using the FR80 microfilm recorder, to achieve a suitable level of data compression, containing a summary of the mean spectra based on this data-bank.
