INTERNATIONAL Computers will have to rely on its own, abilities rather than Government support to ensure its prosperity. This is the tenor of the Government's policy towards the British computer industry, which was explained by the Minister for Industry, Sir John Eden last week. As such it marks a significant if unobtrusive change from the Labour Government's policy, which offered a large umbrella of protection to ICL.
The previous Minister of Technology, Mr. Wedgwood Benn, argued persuasively that computer technology is one of the areas in which Britain should concentrate its efforts. He sees Britain's future in terms of computers, supersonic transport and atomic power stations rather than cars, textiles or ships. It is an argument which Mr. Wedgwood Benn was prepared to press with Government loans, favourable purchasing policies and a good deal of Ministerial pressure on private and autonomous buyers of computers.
Sir John Eden takes a different but equally logical view. He holds that ICL is successful enough to stand on its own feet and that Government grants and purchases are seen to be politically based cannot help the company's image in the market. Where Mr. Wedgwood Benn would clinch his arguments by saying that the Germans and Japanese would dearly like to emulate the British computer industry, Sir John would retort that unless it can be seen to be profitable and successful, it is of no value to the country. In essence, perhaps, these two views demonstrate the gulf between Labour and Conservative industrial philosophy. ICL itself welcomes the new policy. Sir John Wall, the chairman, has always maintained that his company needs research grants less than orders. The ending of the £13½m. research and development grant which the Government has paid to ICL is not regretted, particularly since there is a promise of development contracts to meet specific Government needs. The Government is and will continue to be ICL's largest customer. Central government bodies such as the Department of Health and Social Security, or the Ministry of Defence, account for at least 10 per cent. by value of all the computer orders placed in Britain. In the past over 90 per cent have been placed with ICL. Thus the ½16m. worth of commercial computer orders which central government agencies placed with ICL last year (out of total orders of £18m. represented a third of ICL's British business.
The nationalised industries, hospitals, universities and local government and other authorities are even more significant. They account for a quarter of all commercial computer orders - their business was worth about £40m. last year. ICL cannot rely on such a large share of this business since some of the authorities, particularly in local government, have considerable freedom of action. But even on the conservative estimate that ICL wins only two thirds of it, Government and quasi-government business represents a major part of ICL's domestic sales.
ICL, indeed, has not been as successful in penetrating the general commercial and industrial computer market in Britain as IBM. Most large computer users outside the Government, such as ICI, Ford, Shell and Unilever, use IBM equipment.
In part, this is the result of deliberate ICL strategy. The cost of developing a full set of software suitable for the largest and most sophisticated computer users in Britain would have strained the Company's resources. And so would the development effort needed to produce the terminals, communications equipment and large memories required for the computer networks, operated by companies like BOAC and Rolls-Royce.
ICL wisely chose to concentrate on the local. government market, which required less sophisticated equipment, and where success in a few pioneering installations would quickly result in widespread acceptance. This has broadly been the case. ICL has developed some advanced programs such as its Project Evaluation and Review Technique specifically for local Government users and it has dominated the market. Its share of this business is far larger than IBM's, while the other U.S. companies have scarcely penetrated it at all.
ICL's greatest success in the commercial market has been its small 1901A computer, the bottom of its current range. This has easily outsold the competing IBM 360/20 and has given ICL a base of customers using computers for the first time. With this security, ICL has been able to watch the sophisticated industrial market turning to IBM, the banks turning to Burroughs and the oil companies to Univac, with something approaching equanimity.
There, is, however, some danger in ICL's position. It lacks both the developed equipment and the expertise in software to challenge IBM, Burroughs or Univac in the increasingly important real-time, communications-orientated computer systems. Its only guaranteed customers who have the sophistication to make use of such equipment are the Government and a few commercial organisations like Plessey and the British Aircraft Corporation, which have financial links with ICL.
For this reason ICL has fought hard to win any Government orders which might advance its expertise in these more sophisticated areas. The contracts for the Giro and for the London Airport cargo acceptance system are particularly valuable because they provide real time and communications experience. The American companies have generally learnt this art first through U.S. defence contracts such as the air defence systems which Burroughs installed, the world-wide communications systems built by Univac or IBM's space-flight systems. ICL has not had this advantage because there are no British defence systems of comparable size, complexity or cost.
It is generally accepted that the previous British Government would have given all large computer orders to ICL unless the company's tender was too expensive or it could not guarantee to complete the work in time. The only large systems ordered from American manufacturers in the past couple of years were for the Meteorological Office, the police and the Army Pay Corps.
Weather forecasting requires: very powerful, fast computers: of a type which ICL is unlikely, to market in the next five years. The Meteorological Office's choice was effectively limited to IBM and Control Data, both of which import U.S.-built machines. Police identification networks also require large machines with a fail-safe capability and extensive special software. Burroughs had such systems working successfully in New York and Michigan and was a natural choice. Each of these systems is likely to cost over £4m. and they are the largest ordered by the British Government.
The Army Pay Corps, order, which ICL lost, and the London Airport cargo order, which it won, were less clear-cut. The Army needed to replace its old IBM 705 computers before its new pay structures came into effect last April. ICL could not meet this time-scale with its 4/70 computers since, software development had slipped on these machines. The IBM 360/50s were cheaper, offered better performance, and could be delivered sooner, although the first did not arrive until a few months after the change in pay systems. This order created considerable embarrassment for ICL in Moscow, where the Russians joked about the fact that the British Army would not buy ICL.
The Airport order was won by ICL despite the preference of the airlines for Univac. Univac had installed similar systems elsewhere and had made a speciality of the airline market. It had suitable equipment already operational whereas ICL had not. This was one case where the Government chose to buy ICL equipment despite a higher price because the tender came to within the 25 per cent. margin allowed.
The present Government has ended the margin of preference under which ICL could be up to 25 per cent more expensive. It will still issue single tender orders for large ICL computers where appropriate; but expects ICL to match its competitors on performance. For the bulk of general purpose commercial systems such as those processing social security or health data, ICL should win orders on merit. For the more complicated systems such as those in education and hospitals where American companies have sometimes developed special software at high cost, ICL is unlikely to be competitive. But the Government will probably issue development contracts for just this type of system, and buy several at a time from one supplier.
The Government's use of computers, which has lagged behind that of the U.S. and Sweden in the past, will accelerate rapidly. In terms of computer power, the Government's stock of administrative computers doubled last year and will double again this year and next.
By the end of 1972 it should be using computers with power equivalent to 12 of the very large ICL 1908As or 200 of the medium sized ICL 1905Es. Within ten years, the present computer power employed should have grown tenfold with 108 very large computers employed instead of one and 50 large ones in place of today's 25. Among the very large computer systems contemplated, in addition to those of the police and the weather forecasters, are a central stores control for the Ministry of Defence, a short-term social security benefits system and National Health, education and transport planning data banks.
If the Government presses ahead with its own computer programme with the vigour it has displayed in the past couple of years, ICL should benefit significantly. The universities and hospitals which have been deprived of the large computers regarded as essential in the U.S. or Scandinavia are at last beginning to receive them - the Rutherford Laboratory and Cambridge University have both been allowed to order large IBM systems.
ICL's task is to demonstrate that it can produce large, powerful computers with adequate software in time for Government needs. In the past it has sometimes not been warned of Government requirements until too late - as with the police network. But it has also failed to produce a credible giant computer plan. Its aims have changed several times, relying at first on developments of the System 4 and the 1908A and later switching to an unspecified Project 52. ICL's collaboration with Control Data, the most technically competent designer of giant computers, should help to accelerate its effort and smooth the difficulties. This link, in fact, has come at an opportune moment for it is clear that ICL will not be able to rely on so much Government support in the next five years as it has enjoyed in the past.
THOSE who find their minds wandering from payroll and applications programs will be interested to learn that the Science Research Council is holding a one-day symposium on computer animation at the Rutherford High Energy Laboratory, Berkshire, on July 30. John Halas, the well-known animator, will be making an introductory speech, and other speakers will outline the requirements of animators and educationalists. Computer animation facilities under development at the Atlas Computer Laboratory will be examined together with the work of the Education Research Centre, MIT, and there will be a discussion on applications of computer animated films.
It's hoped to include a varied selection of recent computer generated films from both Britain and the U.S.
Since the day of Felix the Cat, the producers of animated films have worked in a world of paradox. They bring artistic reality to creative thought but do so by a process of intense labour that eschews the very spontaneity of art. The production of an animated feature film can involve over half a million separate drawings, taking 100 people three years to produce. It is a repetitive labour for which the computer might well have been exclusively invented.
The use of the computer in producing animated films has made great strides in the last five years. A novel idea explored by a few pioneers has already become a complex of science, art and craft with practitioners all over the world. Television programmes transmitted by the Open university are making regular use of computer animation, notably in teaching mathematics. The technique has played an important role in NASA's space programme - especially in simulation work. And as a research tool, computer animation. is providing a new sensory device for the scientist.
The technique involves a means of programming a computer with data representing individual drawings. This data can be retrieved on a cathode ray tube and filmed in a more or less conventional camera. With the aid of the computer. Such individual drawings representing individual frames of the animation can be changed instantly, without the need for additional artwork. Thus one drawing of a ship on the horizon can, via the computer; be progressively enlarged on the monitor display without any further work by the animator; it can be moved up and down; it can even be turned round to show a new perspective, if its geometry is simple. With more elaborate motion, such as a walking man, the computer can produce new drawings for in-between phases in the movement.
The labour saving offered by this technology is, of course, enormous - even if present systems do not readily accept elaborate forms of artwork. The artist merely draws or traces the work on to the flat surface of a digitiser; and this equipment translates the shape of the drawing into data which the computer can process. Additionally the computer is also being used in more conventional animation to programme the often complicated movements of the camera rostrum, which must sometimes zoom in and out, pan and perform other gyrations.
The subject of computer animation flared into prominence in the last fortnight when a one-day symposium was held by the Science research Council at the Atlas Computer Laboratory in Didcot. This seems to be the spiritual home of the craft in Britain - an odd mecca where clean-shaven physicists and hirsute artists find themselves in total harmony. At any rate the symposium was packed and many potential delegates had to be turned away.
A dichotomy is, however, discernible in the objectives of the scientists on the one hand and the animators on the other. The peer of international animation - John Halas - has expressed a concern that if the film-makers leave too much to the scientists, the visual possibilities of computer animation may be lost. Halas has become a leading authority on the subject and is preparing a book about it for Focal Press. With another British animator, Stan Haywood, the two at least guarantee that Britain will not be left behind in this new technology.
In 1969 Hayward won a Honeywell competition for new uses for the computer - the prize £1,000 worth of computer time and a terminal in his own bedroom. This he has used indefatigably in exploring the technique. Hayward, whose fame has been based on brilliantly funny scriptwriting for cartoon films, is now expanding his ideas into the field of management studies.
At the Didcot symposium, the possibilities of computer animation were demonstrated in some mind-bending examples. A Halas film on which Holywood collaborated, showed how science, art and humour could combine to produce what for me is the best educational film I have ever seen - Topology. This is about the science of shapes, and was designed for computer production, but ironically had to be made by conventional means because computer time was not available at the time it was needed.
One American speaker at Didcot, Professor J. Schwartz, underlined the importance of such facilities when, he said you cannot think intuitively without experience - and you cannot gain experience without sensory aids. It means that the scientist has been given a new dimension in which to develop his knowledge - a way of connecting intellectual and abstract values with visual perception in real time. As Schwartz pointed out, it offered a way to escape from Descartes and the graph which presented the concept of flux in a static mode.
This may mean that computer animation will develop into two quite separate, disciplines - one as a production aid for animators, the other as a research tool; The danger which Halas sees must lie somewhere in the middle. For example educationists may try to use computer animation merely to visualise scientific concepts, instead of teaching them with all the creative processes that teaching demands.
What the animators hope to see, perhaps, is scope for greater creativity. Unshackled from the labour intensive methods of traditional animation, a new freedom could emerge. The animators envisage a central program bank into which they can plug their own terminals. The tedium of calculating leg movements for a trotting horse will be gone - the program bank will supply it instead. With an international computer grid, animators of the world could really unite, borrowing complex movements from each other as needed.
In art, Leonardo da Vinci saw the potential of the camera obscura - and through photography, Fox Talbot realised new freedom for the painter. For the animator, history is only repeating itself.
The final study is from A. G. Bell and R. A. Lawes from Rutherford Laboratory, Harwell, which illustrates the use of a minicomputer as a data concentrator. At Rutherford, the main computer is an IBM 360/195, with 2 megabytes of core store and a cycle time of one microsecond. At present there are four satellites connected to the 360. The satellites buffer from and concentrate within the big machine the activities of a variety of peripherals.
About 250 people attended an informal Families Day held in the laboratory on Saturday 4 September. They had plenty to see - our new Computer Block, Atlas, S2 and SD4020 operational on normal work, the PDP 15 being run up after its recent move, 1906A undergoing commissioning trials, demonstrations by AERE Fire Brigade of rescue from a car crash, and mouth to mouth resuscitation by Ricky Eaton of the Rutherford Laboratory Safety Section attracted many people, so did the SRC film 'Insight'. At the end of their tours everybody appreciated the refreshments prepared and served by Gillian Keats, Trude Trewin, Joan Markham and Synolda Butler.
Thank you Robbie and all the 'Adminers' for a very successful day.
Two new Council members have been appointed from October 1 1971. They are:
The retiring members were Professor Sir Ronald Nyholm, FRS and Professor P. A. Sheppard, CBE, FRS.
Current members of Council are:
THE Science Research Council's Atlas Computing Laboratory at Chilton, set up nine years ago and still the mainstay of universities needing larger scale computing than they themselves possess, is to find its future somewhat different from its past, though exactly what the future will be is not finally decided.
During the life of the laboratory, the Computer Board's provision of computer hardware has transformed the situation in universities, and enough universities already have or will have soon, computers of equivalent or greater power than the Chilton Atlas and the ICL 1906A now being installed with it. So the laboratory's role as the provider of large scale computing is not now as crucial as it once was, and there have been suggestions that the laboratory has fulfilled its usefulness.
But the SRC has been deliberating for months on what the laboratory should do in the future, and it now seems that a policy will emerge which will give it a future at least as viable as its past.
Continuing for the most part to be tied up with the work of the universities, the laboratory is in a unique position as part of the Science Research Council, to guarantee to provide as much computing time as is needed by large scale research projects put up to the council for its support.
There are few research projects of any size nowadays which do not use a computer at some stage during their progress, and for some of the bigger ones the availability of sufficient computing time throughout their duration becomes one of the most important single factors in considering whether they will be viable or not.
A research project supported by the SRC in a particular university must necessarily be in competition for computer time on that university's machine, and if several hours computing is involved the time may have to be booked weeks ahead. But if SRC projects can get time on SRC computers, the element of competition is considerably lessened.
The laboratory will probably use the 1906A in this way, and continue to use Atlas on a first come first served basis until it is finally run down late next year. Twenty per cent of the time on the IBM 370/195 destined for the Rutherford High Energy Laboratory next door will also be available to the Atlas Laboratory, and this will probably also be used to provide guaranteed time.
Apart from this slightly altered use of its hardware, which has been quadrupled in power by the arrival of the 1906A, the Atlas Laboratory will continue work in specific areas of interest in which it has built up some expertise. Large scale computing development will continue, and the laboratory may also become more active in the field of computer graphics.
Last year the laboratory made several films for the Open University's foundation course in mathematics, and this computer animation work will continue to be one of the laboratory's main projects.
Another big project was the setting up of four geological data banks with associated retrieval systems as a feasibility study for the Institute of Geological Sciences. The Atlas Laboratory has continued to be involved in X-ray crystallography. A new system developed at the University of Maryland is now implemented on Atlas, a successor to the X-Ray 63 and X-Ray 67 crystallography programs also run on Atlas.
The SRC report indicates that during the year 1970-71, the laboratory processed 116,000 jobs for a total of 900 users.
Apart from deliberations on the future of the Atlas Laboratory, the SRC report also mentions that the council's computing science committee has been taking a wide angle look at the future of computing science in the long term. Having up to now provided substantial support in quite a general way for the setting up of research and training programs in computing, the computing science committee is turning its attention to giving support in more specific areas and to particular expertise.
The report states: The (Computing Science) Committee has particularly examined the university role from the aspects of long-range research which will provide the theoretical foundations to the subject, key problems in the areas of application of computers and national requirements for the training of high calibre personnel. In the latter aspect the committee believes that despite present indications there is a long-term trend requiring very substantial expansion in the number of computing scientists and considers that there is a strong case for major growth in computing science in universities.
What this appears to mean is that the SRC will in future be more selective in its choice of computing science projects to support.
WITH no fanfares or fuss the second IBM 360/195 computer in the UK has been installed at the Science Research. Council's Rutherford High Energy Laboratory, at Chilton, Berks, and was successfully commissioned only two weeks after delivery. The first of these machines was delivered last month to the Meteorological Office at nearby Bracknell.
The new hardware supersedes a 360/75 system which has been in use for four years, and so far only a Model 195 central processor and two megabytes core have been installed. with existing peripherals. But early next year an 800 megabyte disc store and high speed tape units will be added.
A capital sum of £40,000 has also been provided to upgrade the central processor during 1973 from a 360/195 to a 370/195, which apparently is not very different except for some additional registers. This will ensure that the Rutherford Lab can take advantage of IBM's software. developments for the 370 series.
