The Cyclone, was a vacuum tube computer, built by Iowa State College (later University) at Ames, Iowa. The machine was placed into operation in July 1959. It was based on the IAS architecture developed by John von Neumann. The prototype of this machine is ILLIAC, the University of Illinois Digital Computer. The Cyclone used 40-bit words, used two 20-bit instructions per word, and each instruction had an eight-bit op-code and a 12-bit operand or address field. In general IAS-based computers were not code compatible with each other, although originally math routines which ran on the ILLIAC would also run on the Cyclone.
Unfortunately, the Cyclone was completed just as the transistor was replacing the vacuum tube as an active computing element. The Cyclone had about 2,500 vacuum tubes, 1,521 of which were type 5844. (The IBM 1401 computer, announced the same year, was fully transistorized. About 15,000 IBM 1401 machines were produced.)
The Cyclone solved 40 equations with 40 unknowns in less than four minutes. This was the same type of problem that the Atanasoff–Berry Computer was designed to solve twenty years previously at the same institution.
The Cyclone computer was 10 feet tall, 12 feet long, 3 feet wide, and contained over 2,700 vacuum tubes. The Cyclone used 19 Kw of electric power and weighed 5 tons. "Good time" was about 40 hours per week.
The original Cyclone had:
- The input and output was five-hole paper tape.
- A model 28 Teleprinter, 10 characters per second, was also available for output.
- Memory was originally 1,024 40 bit words of Williams tube electrostatic memory.
The Cyclone had a major rebuild about 1961:
- Five-hole paper tape replaced by eight-hole paper tape.
- Console printer now right-hole Friden Flexowriter.
- 1024-word Williams memory replaced by four banks of magnetic core memory, 4096 words each bank.
Both versions above had some interesting ideas or limitations:
- All IAS machines used an Asynchronous CPU, no clock. Each unit generated an "answer-back" or "I'm ready" signal, which permitted the output to be used or the next step taken. Most machines designed since then are "synchronous" meaning after x clock pulses, the unit is finished with its operation, say an add is complete.
- There were no index registers, to access sequential data in a loop you used address modification in the instructions.
- This machine had a loudspeaker system attached to the sign bit of the accumulator. Operators and others could listen for an infinite loop or particular program. When a person was done with the machine, the memory exerciser program was started, which had a distinctive sound - signaling others the machine was available. Speakers were placed in offices and work areas for convenience.
The only input was a paper tape reader and the only outputs were the console printer and paper tape punch. As the paper tape punch was much faster than the printer, most output was punched, and then listed on an off-line printer.
1) "A Third Survey of Domestic Electronic Digital Computing Systems" Report No. 1115, March 1961 by Martin H. Weik, published by Ballistic Research Laboratories, Aberdeen Proving Ground, Maryland
2) "A Fourth Survey of Domestic Electronic Digital Computing Systems" Report No. 1227, January 1964 by Martin H. Weik, published by Ballistic Research Laboratories, Aberdeen Proving Ground, Maryland
3) LaFarr Stuart was an economics graduate student and also wrote test programs and utilities during and after the development of the second version of the Cyclone. LaFarr wrote the assembler in machine code as there was no machine for a cross assembler. Also wrote a music program, see External links.
- BRL Report No. 1115, March 1961
- BRL BRL Report No. 1227, January 1964
- Computer Music
- Iowa State University, Department of Statistics, Cyclone Computer Records, RS 13/24/5
- Iowa State University, CS Department History
- Iowa State University, Computer Technology Research