Harwell computer

Construction and use at Harwell

The computer, which weighs 2.5 metric tons (2.8 short tons), was built and used at the Atomic Energy Research Establishment in Harwell, then in Berkshire (now Oxfordshire). Construction started in 1949, and the machine became operational in April 1951. It was handed over to the computing group in May 1952 and remained in use until 1957.

It used 828 dekatrons for volatile memory, similar to RAM in a modern computer, and paper tape for input and program storage. A total of 480 relays were used for sequence control and 199 valves (electronic vacuum tube) for calculations. The computer stands 2 meters high, 6 meters wide, and 1 meter deep with a power consumption of 1.5 kW. Output was to either a Creed teleprinter or to a paper tape punch. The machine was decimal and initially had 20 eight-digit dekatron registers for internal storage, which was increased to 40 which appeared to be enough for nearly all calculations. It was assembled from components more commonly found in a British telephone exchange. The man who led the effort to rebuild the machine (see below) put it in perspective to the BBC: "All together, the machine can store 90 numbers. The closest analogy is a man with a pocket calculator," Delwyn Holroyd, who led the restoration effort, tells the BBC in a video about the restoration. Although it could on occasions act as a true stored-program computer, that was not its normal mode of operation. It had a multiplication time of between 5 and 10 seconds, very slow for an electronic computer.

As Ted Cooke-Yarborough wrote of his design in 1953 "a slow computer can only justify its existence if it is capable of running for long periods unattended and the time spent performing useful computations is a large proportion of the total time available". The design was noted for its reliability because in the period from May 1952 until February 1953 it averaged 80 hours per week running time. Dr Jack Howlett, Director of the Computer Laboratory at AERE 1948–61, said it "could be left unattended for long periods; I think the record was over one Christmas-New Year holiday when it was all by itself, with miles of input data on punched tape to keep it happy, for at least ten days and was still ticking away when we came back." It was the machine's untiring durability, rather than its speed, that was its main feature. Human mathematicians (a job role called a "hand-computer") could make calculations at a similar speed, but not continuously for the same lengths of time. Dr Howlett commented:

One day EB ‘Bart’ Fossey, an excellent hand-computer (still with what used to be called the Atlas Computer Laboratory), settled down beside the machine with his desk machine and attempted a race. He kept level for about half an hour working flat out, but had to retire, exhausted; the machine just ploughed on.