Computers & Business Machines
Imagine the loss, 100 years from now, if museums hadn't begun preserving the artifacts of the computer age. The last few decades offer proof positive of why museums must collect continuously—to document technological and social transformations already underway.
The museum's collections contain mainframes, minicomputers, microcomputers, and handheld devices. Computers range from the pioneering ENIAC to microcomputers like the Altair and the Apple I. A Cray2 supercomputer is part of the collections, along with one of the towers of IBM's Deep Blue, the computer that defeated reigning champion Garry Kasparov in a chess match in 1997. Computer components and peripherals, games, software, manuals, and other documents are part of the collections. Some of the instruments of business include adding machines, calculators, typewriters, dictating machines, fax machines, cash registers, and photocopiers
- This circular device was an aid to programming the UNIVAC solid-state computer. It consists of a paper disc with equal divisions running from 1 to 200 near the edge, and a clear plastic rotating disc that are pivoted together at the center. The upper disc is marked in red with two perpendicular diameters. The solid state computer had a magnetic storage drum on which locations were specified numerically. The latency calculator allowed programmers to write code for the machine to make the most efficient possible use of the drum memory. The back of the instrument gives a list of instruction codes and corresponding word times. Recieved in bag. Reference: Sperry Rand Corporation, Programming: Simple Transition to Electronic Processing UNIVAC Solid-State 80, 18-26.
- Compare 2005.0271.01. Date based on date of documents 2015.3097.03 and 2015.3097.04.
- According to Kirk Lubbes, who programmed the Univac Solid State Computer:
- "The SS90 had a drum memory, i.e. memory was not random accessible. One had slow memory and fast memory. The slow memory had only a single read/write head per track on the drum and fast memory had four read/write heads spaced at 90 degrees, so therefore the drum had to rotate a full revolution to access a memory word in slow memory and only a quarter turn to access fast memory.
- The trick in programming the SS90 was to have the instruction and its operand accessible at an optimal time so that the instruction could access its operand without waiting for the drum very far. As one started a program, this was not much a problem. The programmer new how much time that a given instruction would take to execute and the speed of the drum. Therefore, he calculated the position of the next instruction, based these two parameters. The minimum latency calculator was a mechanical device to help in this calculation. The problem was that as the programmer progressed, collisions occurred, i.e. the optimal location of an instruction or an operand was already taken by a previous instruction or operand. Since the drum was arranged in bands and the read/write heads were at the same location on each band, if one had a collision, you could put the necessary instruction or operand in a parallel band at the same position. This worked the bands all filled up.
- The basic approach was to get a program working using the best latency that you could. Then the programmer would go back and start rearranging instructions and operand locations to achieve minimum latency. In those early times, machine time was expensive and memory severely limited. So it was important that production programs were efficient."
- Nonccession file 2015.3097.
- Currently not on view
- date made
- ca 1960
- Remington Rand Univac
- ID Number
- nonaccession number
- catalog number
- Data Source
- National Museum of American History