At the end of August 1975, the Director of the Atlas Computer Laboratory, Dr J Howlett CBE, retired. The responsibility for the Laboratory was transferred to the Director of the Rutherford Laboratory at the beginning of September 1975 when it became the Atlas Computing Division of the Rutherford Laboratory.
This Division provides computing facilities to university-based research workers who obtain approval from the specialist subject committees of the Science Research Council for the computational aspects of their work. The committees which make the awards of computer time deliberate on applications for research grants for which funds are allocated by the Science Research Council through its Science, Engineering, and Astronomy Space and Radio Boards. The Atlas Division provides computing facilities to the National Environmental Research Council (NERC) in consequence of the central agreement concluded between the two Councils concerning computing for NERC Institutes, and to other Government funded bodies.
The Atlas computing facilities are based on an ICL 1906A computer and on a 20% share of the Rutherford Laboratory's central IBM 360/195 computer. There is a complex system of connections for data communication to both computers from a large number of British Universities.
The Stromberg DatagraphiX 4020 microfilm recorder installed in 1968 has given users access to a specialised output device which records graphical and other information on microfilm or photosensitive paper. An application of the device is the production of cine film showing the time-dependent features of a model of a physical phenomenon being studied on the computer.
A decision was reached by SRC to replace the SD4020 machine by a more modern and sophisticated microfilm recorder, an FR80 manufactured by Information International Incorporated of Los Angeles; USA. The equipment was delivered in March and passed its acceptance trial in May. Three cameras arrived as ordered, a hard copy camera, a 16 mm precision camera, and a 35 mm camera. In addition, a microfiche camera was loaned by the manufacturer for a year, and the Atlas Computer Committee agreed to a proposal for its purchase in December.
The software available on the IBM 360/195 and ICL 1906A computers has been adapted or enhanced to permit the use of the FR80 in place of the SD4020. The ICL 1906A software was brought into service in June and that on the IBM 360/195 in November. In both cases the change was brought about smoothly and the very few cases of difficulty were quickly investigated and resolved.
The FR80 has improved features in addressability and resolution. The quality and the registration of its output is much superior to that of the SD4020, and users have already remarked on the fact.
Most of the activities with regard to Science were subsequently moved to the Daresbury Laboratory including the Microdensitometer service to focus Science Board there.
The objectives of the Atlas Computing Division Meeting House in Theoretical and Computational Physics and Chemistry are to provide a natural focus for activities in the general area of theoretical physics and chemistry for sponsoring projects and providing facilities for collaboration and information exchange among scientists from the universities and other research establishments. Project 1 - Electronic Correlation in Molecular Wave functions - has now been established. The Hartree-Fock approximation was a limiting factor in the computation of energy surfaces for use in studies of chemical reactions, and the work of the project will help alleviate this difficulty.
It was decided that the MUNICH-CI (Configuration Interaction) program, due to Professor Dr G M F Diercksen of the Max Planck Institut fur Astrophysik at Munich and Dr B T Sutcliffe of the University of York, should be mounted and interfaced with the ATMOL program on the IBM 360/ 195 computer. The particular interests of the Atlas group in quantum chemistry resulted in the work being done at Chilton. Many improvements and other programs were incorporated during the year so that by September work was substantially complete and the program was capable of working on general open shell systems. Particular attention had been paid to providing a tool for the study of excited states. Interest in the new program system is growing and could lead to substantial demands for time especially on the IBM 360/195 computer.
The working group for Project 1 met in October and December, when a programme of further work was discussed. The Meeting House was deemed to have been successful especially in highlighting the value of implementing and interfacing complex packages as a precursor to more extensive collaborative ventures.
The microdensitometer installed on the Atlas premises was purchased by the SRC Chemistry Committee to provide a service to university-based crystallographers. The machine consists of an Optronics P-1000 Photoscan interfaced to a Computer Automation Alpha-16 minicomputer with 16K of 16-bit word storage together with a 7-track magnetic tape drive. The photoscan system converts photometric data on film negatives or transparencies to digital form. The optical densities in the range 0-3D are converted to a range of 256 grey levels, and the intervals of raster for measurement are 25, 50 or 100 microns.
During 1975 nearly 50 chemical structures from 27 different crystallographic groups in the UK have been analysed with the aid of the microdensitometer. Each structure entails the digitisation of an average of 10 film packs each with an average of 5 films. The Weissenberg program has been designed to enable the digitisation and subsequent indexing of the spot data to be done almost automatically and has aroused worldwide interest. The results derived from the completed analysis of the crystal structures show that within the limitations imposed by the film method of recording X-ray data, the machine and the programming method are highly accurate.
During the year the service was extended by the successful introduction of a program to deal with precession films from both small molecules and proteins. At first the indexing program was run separately on the ICL 1906A but this has now been moved to the Alpha-16 where it reduces the delay in determining the success of the scan.
The microdensitometer has applications in areas other than crystallography, although its availability for these users is limited by the priority imposed by the crystallographic service. One such application is in pneumoconiosis screening. Here the objective is to determine which, if any, of a series of lung X-ray plates contain evidence of spots. Trial runs in digitising films for the Medical Research Council (MRC) indicate that a machine of this type could be used routinely to provide this valuable health service of particular importance to the mining industry. The machine also has potential in the optical character recognition field, particularly with reference to films of the printed page, and could enable manuscripts of early texts to be digitised.
An application during the year involved digitising 16 mm and 35 mm film strips which displayed the diffraction patterns of frog muscle fibre exposed to the light of a ruby laser. The experiment aimed to measure the differences in the diffraction pattern caused by muscular contraction and relaxation. The results of the experiment were reported by Dr L Noye of Hull University at a conference in Copenhagen, and aroused a great deal of interest.
Much of the work of the Atlas Computing Division lies in sustaining and developing the computing facilities used by university research workers. Many of the projects have considerable scientific importance.
The orbit of the Ariel 5 satellite is the concern of the control centre at Appleton Laboratory. The ICL 1906A computer provides a standby service for Appleton including a test run each day to check the data communications system. In addition, the Mullard Space Science Laboratory at Holmbury uses the IBM 360/195 to analyse the data collected by the satellite. The discovery of stellar X-Ray sources with rapidly varying intensity has provoked great interest in the data, and has stimulated a demand for rapid turnround when the computing facilities are used to investigate a suspected new source.
The work of the primary processing of the Ariel 4 satellite data was concluded early in the year. The Universities of Sheffield and Manchester have continued to analyse the scientific content of the data and the FR80 microfilm recorder plays a significant part in this work.
The data collected in the S2/68 experiment carried on the ESRO TDIA satellite had been almost completely scanned to form a base of information about stars and their positions, when the S2/68 Data Committee decided that much fainter objects should also be included in the base. A rescan of the whole set of magnetic tapes is now in progress, and will continue during 1976.
Other large scale projects being undertaken by university users on the Chilton computers include:
