Selective Sequence Electronic Calculator

The Selective Sequence Electronic Calculator was IBM's intermediate step between the ground-breaking Automatic Sequence Controlled Calculator, and its first electronic computer, the Defense Calculator. It was a physically very large, but functionally limited, stored-program computer, which used a combination of electro-mechanical and electronic technologies; it contained 12,500 vacuum tubes, 21,400 relays, and no less than 66 paper tape readers.

Its architects were Wallace Eckert, who had been prominent in pre-WWII use of punched card machinery to perform scientific calculations, and Robert Seeber (who had worked on the ASCC). The engineering was done by a consortium of people, principally Frank Hamilton, with assistance from Byron Phelps, Ernest Hughes, and James Troy, with oversight from John McPherson and Halsey Dickinson. Construction was approved in March, 1946, and the bulk of the construction was complete by the end of December; preliminary testing occurred in mid-1947; and it was revealed to the public in January, 1948.

Like the ASCC, it was decimal internally (using BCD encoding); it was novel in that data and instructions were represented internally in the same way, and could be interchanged, and stored in any location. All computations were performed electronically (an approach explored previously by the Type 603 predecessor of the IBM 604 Electronic Calculating Punch); multiplication and division were available, as well as addition and subtraction. For memory, it used a combination of tube flip-flops (holding eight 20-digit numbers), relays (holding one hundred and fifty numbers), and paper tape (twenty thousand).

It was programmable (again like the ASCC), also using a mixture of technologies; mostly wide paper tape, but also any memory, including the eight electronic registers. Instructions contained six fields: typically the addresses of two source, and one destination operands (some of which had side-effects, like several other early machines); the address of the next instruction; the operation to be performed; and a mathematical control field. However, it was also possible to do a limited amount of configuration, using plug-boards. Input was from paper tape, or two punched card readers; output could be produced on cards or tapes, or on two printers.

People at IBM, not yet the giant it later became, had been annoyed at Howard Aiken for claiming the bulk of the credit for the ASCC, whose engineering had been performed by IBM personnel; the SSEC, in which Aiken was not involved, was IBM's response. It was installed on the ground floor of an IBM building in New York, visible from the street through large plate-glass windows, and it attracted many curious passers. Uniquely among the early machines, time on it could be rented by commercial entities; no charge was made for use on scientific problems.

It was not directly very influential, as neither its architecture nor its engineering were particularly novel. Its main importance was more indirect; it got IBM further involved with computing applications, which led to IBM's first computer. It was thus a major step on the path to the IBM System/360, which consolidated IBM's place as the world's largest computer company. Many of the people who had worked on the SSEC went on to work on the early IBM 650 computer, which shares a number of details with its predecessor.

Further reading

• Charles J. Bashe, Lyle R. Johnson, John H. Palmer, Emerson W. Pugh, IBM's Early Computers, MIT Press, Cambridge, 1986 - covers the SSEC in some detail

External links

• 2,000 Tube Electron Brain - article in Radio Craft, May 1948
• Arup Bhattacharya, The IBM Selective Sequence Electronic Calculator - a detailed description, prepared recently from the only source now readily available, the patent (which is included)