Growth of a Laboratory Computer System for Nuclear Physics

A computer system may typically be expected to progress through a cycle terminating in overloading. The experience with an early system at the author's nuclear physics laboratory serves as an example. The original computer and a similar machine later installed with it are now overloaded and a new system is under construction. The success of the interactive data analysis on the original system has made it desirable to enhance the display and light pen facilities while reducing the computer time involved in generating the displays. The use of a data storage/display disc effectively provides off-line displays but requires more manipulation in data acquisition. The solution is found in the large number of processing units economically feasible with third-generation equipment. Two linked computers will perform data acquisition and analysis, the smaller performing data acquisition under the control of the larger, which will run a fairly simple time-sharing system. Together with several I/O processors, this hierarchy of processors will provide ease of program development and a very high degree of computational power and data acquisition capability.