Mathematical applications of the dynamic storage analog computer

Abstract

Many recent developments in analog computers have substantially increased their versatility and speed. Probably the two most important of these innovations is the automatic repetitive operation at frequencies of 60 cps and the development of high speed memory with a tracking error of less than 10 microseconds. With repetitive operation and the simultaneous solution of large matrices of ordinary differential equations at speeds of 60 cps, the analog is much faster in operation than the digital. However, an inherent disadvantage is the size of computer required for a complex problem. In general, the more complex or larger a problem, the more conventional analog computer equipment that has been required for its solution. The digital computer, on the other hand, has, from its earliest inception, employed memory to allow time sharing of components. In this manner, a large complex problem could be solved by time sharing the digital computer equipment and using the same equipment repeatedly to solve the larger scale problems. Therefore, large problems required longer solution times rather than larger computers. The disadvantage of larger sizes in the analog computer has been overcome by the development of high speed memory which is employed in a manner similar to that utilized in the digital computer for time sharing to increase the time of solution rather than increasing the size of the computer for the more complex problems. However, the order of magnitude increase in solution time is much less severe with the analog computer employing memory than it is with the conventional digital computer, since the analog simultaneously solves systems of equations, while the digital computer must use reiterative procedures for such a system.