1960-64: IBM 1620 to Uppsala

At a dinner in Stockholm during the spring of 1960, Professor of Physics Kai Siegbahn of Uppsala University met Count Gösta Lewenhaupt, the Swedish CEO of IBM.  Lewenhaupt asked why Swedish universities didn’t use computers as much as universities in other countries.  A computer should be of interest to the physics department, at least. “We can’t afford it,” Siegbahn answered laconically. Lewenhaupt then remarked that Siegbahn didn’t use the available free time with one of IBM’s mainframe computers in Paris. In addition, he could get a 60 percent rebate if he bought the computers from IBM. Siegbahn’s answer has become famous: “It’s the 40 percent we don’t have”.

However, Lewenhaupt wasn’t the Swedish CEO of IBM, with the responsibility to establish IBM computers in Sweden, for nothing.  He said, “But maybe the Professor has something else to offer which could be worth the 40 percent?”

Siegbahn said that Uppsala University, beginning 1960-61, would offer annual international seminars in physics and chemistry for promising young researchers from developing countries.  In the seminars there would be an item about the use of IBM’s computer.

Lewenhaupt was interested in getting international attention in that sense, and asked for time to think about it. He returned with three conditions for IBM to set up a computer for two years:

  • There had to be a qualified operations engineer.
  • There had to be computer courses for everyone interested at Uppsala University and the Swedish University of Agricultural Sciences, as well as for participants of international seminars.
  • Everyone at Uppsala University and the Swedish University of Agricultural Sciences had to be given access to the computer.

The first condition was met by Werner Schneider, and the second by Schneider and Ingvar Lindgren, who had just returned from a research stint at U.C.  Berkeley.  Between them, they also made sure the third condition was met.

The computer was scheduled to be installed by the fall of 1961. In order to achieve the best possible operations environment, the Committee of Mathematics Machinery at Uppsala University applied at the Research Council of Natural Sciences and the Council of Atomics Research for a grant to be used for a ‘mathematical machine group.’  The grant was given during the summer of 1961 and the computer was delivered and installed in November 1961.

When the hyper-modern IBM 1620 computer arrived from the U.S., it turned out to be quite a sizeable piece, and was to be put up on the second story of the department of Physics.  The only solution was to hoist it up and bring it in through a window.  However, no window was big enough, and consequently, a hole had to be cut through the wall.

Around ten o’clock on a winter’s day with heavy snowfall, a truck arrived with the miracle machine all wrapped up in plastic. The transportation crew disappeared quickly, and there in the snow sat the computer.

Not until that moment people realized the computer wasn’t insured.  What if it was ruined while it sat in the snow, or if it was dropped while hoisted up?

Kai Siegbahn was contacted, but he trusted that the government would cover any possible costs. The other people involved were not so sure about this, and eventually Werner Schneider got a private insurance policy.

No crane was available, of course.  Eventually it was found that an engineer, Mogren, had a suitable crane, but he had been called to Rimbo, where there had been a collision between a rail car and a bus.

Engineer Mogren didn’t arrive until four o’clock in the afternoon; the computer had by then been sitting in the snow for six hours. Hoisting it up was done without any mishaps, and six persons managed to get the machine in place.

IBM 1620 turned out be an incredible resource. It ran around the clock and wasn’t exclusively used by physicists, but also by sociologists, psychologists and everyone else who needed to process large amounts of data with many variables. For instance, the Swedish University of Agricultural Sciences was a big customer.

It was a problem that so many people used the computer, and that, at this time, most of them were beginners in the field of computers. As a result, Werner Schmidt often had to be on duty. However, gradually routines were developed to decrease the risk of mistakes. In accordance with one rule, the person who operated the machine had to have one hand in his back pocket in order to reduce the risk of causing a short-circuit.

The computer was fed punched tape, and the procedure was that first the tape was punched, then it was fed into the computer, and at last the result was printed by a separate type writer. Unfortunately it happened that the service of the computer – which was situated in Uppsala – was done by the IBM office in Stockholm. The service of the type writer, however, was done by the IBM office in Gävle. It was impossible for an engineer from one office to work on the computer without one from the other office being present.  In addition, it wasn’t always obvious if the fault was with the computer or with the type writer, or with both.  Eventually, it was approved that one technician was allowed to handle all the problems, and then the service work ran more smoothly.

The responsibility for the operation of the IBM 1620 was placed under the Committee of Mathematics Machinery at Uppsala University, and Professor Carleson was the president. The daily operation was done by the Mathematical Machine Group at the institution of Physics, led by Werner Schneider and three employees.

Naturally, the IBM 1620 was not enough for all the calculations to be done at Uppsala University, especially at the institution of Physics. Even before the IBM 1620 was installed, the need for big capacity computers was met by the free time offered by Count Lewenhaupt with the IBM mainframe computer in Paris, and then in New York. With the help of research colleagues, Werner succeeded to get free time with the mainframe computers in Madison, WI, and Cern in Geneva. The IBM 1620 was essential in Sweden to test and control the programs before they were sent off.

Beginning in 1962 (?), researchers at Swedish universities and colleges also got the ability to use the mainframe computer IBM 7090 at FOA (the Swedish Defence Research Establishment) free of charge (after they had applied for permission and carefully stated the purpose).

For a couple of years there was a relatively good supply of machine time. But as early as 1962 it was understood that Uppsala University would need its own mainframe computer, at the latest by 1964. After an application with the Committee of Mathematics Machinery, it suggested to the government that Uppsala University and the Swedish University of Agricultural Sciences should have 4,403,000 SEK in order to purchase a computer plant.  It was a considerable sum at that time.

However, it was turned down. The explanation by Sträng, the Secretary of the Treasury, was:

“I have previously, in connection with the proposal concerning the ADB-organization, suggested a special investigation of the organization of the mathematical computer processing and the connected research. Out of concern for that, at this point, I’m not prepared to approve the suggestion of the Committee of Mathematics Machinery about acquisition and means for next fiscal year for a computer plant for Uppsala University and the Swedish University of Agricultural Sciences.”

Sträng said that he had just appointed the so-called Council of Experts, whose task it was to create a plan for the future provision of all finance to the public sector except Defense. The result consequently had to be awaited.

The fact that the application was turned down was nothing short of a disaster for the work in Uppsala. However, Werner Schneider had helped the Committee of Mathematics Machinery to prepare the material for the application, and afterwards went through the explanation for why it was turned down. He then found out that one of the principal ‘adversaries’ was Åke Pernelid, Director of the Statistics for Sweden in Stockholm. Schneider contacted Pernelid, a former student at Uppsala. He had previously been the CEO of ADB at Sandviken and was considered to be very well informed about the status of computers in Sweden. However, Pernelid hadn’t at all realized what a disaster it was when Uppsala’s application for a grant for computer machinery was turned down.

Schneider’s rather upset speech  on a Saturday morning led to Pernelid suggesting  a meeting three days later in Uppsala, as he was going for a short vacation in his brother-in-law’s summer house outside Uppsala. During this meeting, Pernelid was informed about the catastrophic situation when it came to Uppsala University’s access to computer resources:

  • The access to the only mainframe computer, the IBM 7090 at FOA, had been withdrawn.
  • The Atomics Research Council and Research Council of Natural Sciences had turned down grants for the staff at the IBM 1620 facility at the Department of Physics.
  • Beginning November 1963 the free period for the IBM 1620 ended and no means had been granted for the payment of 60 percent of the rental cost.

All this, together with the fact that Uppsala University’s application for a powerful computer facility had been turned down, in reality meant that all computer operations in Uppsala had to shut down.

It turned out Pernelid wasn’t difficult to persuade that Uppsala needed more computer resources.

He did two things:

  1. Uppsala University received means enough to complete the IBM 1620 facility and to maintain the staff.
  2. The Council of Experts would prioritize Uppsala University in its work. The assembly had already concluded that the supply of computer power for Swedish universities would be a pilot project for all other governmental civil activity.

During the summer of 1964 the Council of Expert’s plan was ready enough for the Swedish Agency for Public Management to start the buying of a big computer facility for Uppsala University in late summer.

Because this computer facility was the first big computer resource in the planned gradual supply to all Swedish universities, the American computer supplier Control Data (CD) won, because of its unique central computer-satellite concept. This concept also gave the Swedish University of Agricultural Sciences in Uppsala, the KTH Royal Institute of Technology in Stockholm and Stockholm University together with the Swedish Agency for Public Management access to the facility at Uppsala University.  Satellite computers were installed in these three places, besides the central computer in the plant in Uppsala.

These satellites made input of programs and data possible by punched cards. Programs and data were sent to the central computer over the phone line, and the result from the central computer was printed on the line printer of the satellite computer.

The satellites could be used for minor tasks like processing data and testing programs, while simultaneously transmitting data, for instance while used for teaching students.

It is particularly interesting that a fourth satellite computer would be placed at the University Hospital in Uppsala.  It would be a breakthrough for applications in Swedish healthcare and an international role model.

The background was that during the fall of 1964 the planning manager, Erik Tynelius, of the Uppsala University Hospital, had been notified that the planning for a university computer center was under way. He immediately took the initiative to co-plan the hospital’s need for computers with Uppsala University. He thought it was especially important that a separate organization should not be built for the Uppsala University Hospital.

This meant that the computer work at the Clinical Chemical Central Laboratory would be integrated with UDAC’s work. In addition, he wanted a computer- based patient administrative system.

The choice of CD’s satellite concept was, of course, attractive: a satellite could be placed at the Clinical Chemistry Central Laboratory and be designed for co-use with the patient administrative office. In addition, these satellite computers could be used for development of applications in other departments within Uppsala University Hospital.

For all larger processing the central computer ….. [Werner will complete this paragraph]

Consequently there was a decision supported by the Swedish Agency for Public Management to buy such a satellite facility. However, it wasn’t delivered by Control Data, but by IBM.

During the spring of 1965, IBM Sweden showed a great interest in taking part of what had been achieved at the Clinical Chemistry Central Laboratory, and to contribute to further development in this field. Especially important in this context was the development of quality control routines. The computer support would also be expanded to include the communication between the clinical departments and the central laboratory.

Control Data and other providers didn’t show the same amount of interest.

As a consequence, a ‘joint study project’ was formed by Uppsala University Hospital, the computer central (UDAC) of the Uppsala universities and IBM Sweden. The first result was a prototype for CLS, the Clinical Chemistry Laboratory System. It was implemented on an IBM 1440 at the Central Laboratory at Uppsala University Hospital and was used beginning early 1966. The collection of data was still done using the system constructed by the hospital and delivered measurement data on punched cards via a modified card punch, IBM 026.