NIRNS was set up in 1957 and, as with the Atlas Laboratory, many staff came initially from Harwell and transfered to Rutherford in January 1961. NIMROD, the 7GeV proton synchrotron, was due to begin working at the Rutherford Laboratory in 1962. The initial estimate was that the information processing requirements would be equivalent to between one and two shifts of an IBM 704. A suitable UK machine was the Ferranti ORION, due for release in mid-1961. This was initially estimated as having the power of an IBM 709, and would cost about £400,000. ORIONs started being delivered in 1961 and the Rutherford machine was delivered early in 1964.
The ORION was designed to be a multi-programming machine and it was front-ended by a DDP 224 (May 1965) to attach:
Several devices could be run together with a control program managing the I/O to and from the ORION. The HPD took precedence over the other devices. In practice, the size of the HPD program was such that when it was running the other programs rarely got any access to the ORION.
During the early part of 1964 a considerable amount of effort and computing time was used in preparing master programs which determined the way in which the machine handled the computational programs. Once this was completed, more time became available for the analysis of data from bubble chamber film. The load grew rapidly during 1963 and 1964. In consequence, it became necessary to buy time on other computers, leaving Orion available for short calculations and data processing from counter experiments running on Nimrod.
The ORION's performance was never as great as originally expected. It performed at between 0.05 and 0.1 of an Atlas. In consequence, as the Rutherford load built up, most of the batch computing was carried out on Atlas, the IBM 7050 at Imperial and the IBM 7044 at Glasgow leaving the ORION to handle the HPD. The DDP acted as a data collector and analyser for the CRT film scanner and the physics experiments. A decision was soon made to replace the ORION with an IBM 360/75 capable of handling both the on-line and batch processing. The ORION was removed in September 1966 and the IBM system was delivered in December 1966. The ORION was in operation for less than 3 years.
More details of the Orion computer can be found at http://ferranti-orion.co.uk/.
