Difference between revisions of "IBM 1620"
(A pretty good start) |
(No difference)
|
Revision as of 16:13, 20 December 2023
The IBM 1620 (formally the IBM 1620 Data Processing System; informally known as the CADET, for 'Can't Add, Doesn't Even Try') was a relatively early small scientific computer (prior to the 360, IBM computers were specialized and separate for scientific and business applications) from IBM. It was one of the first IBM computers to be built with transistors, and used magnetic core for its main memory. It was announced in October, 1959, and withdrawn in November, 1970; around two thousand were produced.
Its main memory was divided into 'locations' (intended for storage of numbers in BCD); each location holds four bits of data, a 'wordmark' bit (principally used to mark the end of variable sized operands), and had a parity bit. Like its cousin, the business-oriented IBM 1401, it supported variable-length operands. The standard 'module' was 20K locations; up tom two additional modules could be added.
Instructions were 12 locations (72 bits) long, with a two-digit opcode, and two 5-digit addresses. All arithmetic operations (including addition) were performed by table lookup, using tables stored in main memory (hence its nickname).
It used paper tape for I/O, but also had a printing console device; a punched card reader/pounch was available as an option, as was a line printer, and eventually either disk or up to 6 magnetic tape drives for mass storage.
Standard software included a FORTRAN compiler.
See also
External links
- IBM 1620 Data Processing System - documentation at Bitsavers
- 1620 Data Processing System
- The IBM 1620 Model 1
- The IBM 1620 Data Processing System
- A review of the IBM 1620 Data Processing System - a brutal review of the 1620 by Edsger Dijkstra