Page 135 - Neutrons for Sciences and Society

Chapter 5: The construction of the reactor and the scientific groups
5.3.1 Computing at the ILL84
The ILL was built at a time when computing was beginning slowly to take the importance which it now has adopted. In the work I did at Saclay when I was measuring the neutron time of flight from selectors I only had ten channels, which had to be read manually. I remember my amazement visiting the Brookhaven reactor and seeing each experiment coupled to a mini-computer (using PDP8s I think), which piloted the instrument and collected the data. The progress of colleagues at Brookhaven was a consequence of there being an industry capable of implementing such systems. Michael Taeschner was recruited in 1969 to deal with computerisation of the experiments. At that time the policy both in Germany and France was to support and use national industry. We had to take this into account.
Instrument control was based on two systems. The first, called CARINE, used two T2000 computers manufactured by the French company Télémécanique. The systems were used on 12 instruments with demanding control requirements, but simple data collection (diffractometers, triple-axis spectrometers.)
The realization was entrusted to the electronics and computing laboratory (LETI) of the CENG were Mathurin Le Sourne (recruited 1 January 1971) wrote a realtime FORTRAN for the instrument control. Once delivered it was supported by Walter Kaiser (recruited 1 July 1970) assisted by Gérard Pastor and Jean- Pierre Delacroix. The second system called NICOLE, was based on two Telefunken TR86 computers. This managed 6 instruments which had to treat a very large amount of data (e.g. small angle
84 I thank Alain Filhol and Ron Ghosh for their help in editing this section.
126