"The EDSAC (electronic delay storage automatic calculator) is a serial electronic calculating machine working in the scale of two and using ultrasonic tanks for storage. The main store consists of 32 tanks, each of which is about 5 ft long and holds 32 numbers of 17 binary digits, one being a sign digit. This gives 1024 storage locations in all. It is possible to run two adjacent storage locations together so as to accommodate a number with 35 binary digits (including a sign digit); thus at any time the store may contain a mixture of long and short numbers. Short tanks which can hold one number only are used for accumulator and multiplier registers in the arithmetical unit, and for control purposes in various parts of the machine." [Paper by Wilkes & Renwick in 1, reprinted in 2]
The EDSAC contained some 3000 valves arranged in 12 racks and consumed about 12Kw of power. The delay lines (or "tanks") were arranged in two batteries providing 512 words each. The second battery came into operation in 1952. [3]
Input was on 5-track teleprinter paper tape via an electromechanical tape reader running at 6 2/3 characters per second. It was replaced by an improved version running at 16 char/sec in October 1949 and by a photoelectric reader running at 50 char/sec in early 1950. Output was delivered via a teleprinter at 6 2/3 char/sec. An output paper tape punch running at 16 char/sec was added in 1951. [3]
Magnetic tape storage was added in 1952 (but never worked sufficiently well to be of real use). [4]
The machine ran at 500kHz. Wilkes notes that he would have been happy to accept the challenge of working at 1MHz (as most other early designers were doing) but preferred a straightforward development to permit early experiments with the writing of programs to solve real problems. [5]
Instructions occupied a single 17-bit word but the whole delay line had to circulate before the next instruction could be executed, so the fastest that consecutive instructions could be executed was 500,000/((32+1).(17+1)) = 841 instructions per second. Average order times were 1.5ms (say 650 instructions per second), with a multiplication in 4.5ms. Division in software took about 200 ms. [3]
The operating system or "initial orders" consisted of 31 instructions which were hard-wired on uniselectors, a mechanical read-only memory. These instructions assembled programs in symbolic form from paper tape into the main memory and set them running. The second release of the initial orders was installed in August 1949. This occupied the full 41 words of read-only memory and included facilities for relocation or "coordination" to facilitate the use of subroutines (an important invention by D.J. Wheeler). [3]
Interest in programming methodology (as it would now be called) did not prevent the active use of EDSAC for solving genuine problems. One of the earliest of these was proposed by R.A. Fisher - the solution of a second order non-linear differential equation with two point boundary conditions. It was solved by D.J. Wheeler by April 1950 and was published with due acknowledgements later in the year in Biometrics, making some progress towards establishing the credibility of electronic computing. The first publication of results from EDSAC was in R.S. Scorer's note on a meteorological problem published in the Quarterly Journal of Mechanics early in 1950. This used a modification of M.V. Wilkes' program to compute values of Airy's integral. [5]
The preparation of programs for an electronic digital computer by M.V. Wilkes, D.J. Wheeler and S. Gill (the first book on programming) was published by Addison-Wesley in 1951. A film was made at the same time showing the EDSAC in use. [4]
The Diploma in Numerical Analysis and Automatic Computing (now the Diploma in Computer Science) began. This was a one-year postgraduate course, the first formal course leading to a university qualification in computing anywhere in the world. [4]
From 1953 the EDSAC was heavily used for (among other subjects) theoretical chemistry under S.F. Boys; X-ray molecular biology by J.C. Kendrew; numerical analysis by D.R. Hartree and J.C.P. Miller; atmospheric oscillations by Wilkes; early work on radioastronomy (much extended on EDSAC 2) by group under M.F. Ryle. [4]
M.V. Wilkes' paper, On the best way to design an automatic calculating machine, introduced the ideas of microprogramming and bit-slicing for the design of the EDSAC 2 in 1951. [4]
