Browse Prior Art Database

Revolving Storage Addresses

IP.com Disclosure Number: IPCOM000097041D
Original Publication Date: 1962-Apr-01
Included in the Prior Art Database: 2005-Mar-07
Document File: 3 page(s) / 40K

Publishing Venue

IBM

Related People

Bush, GH: AUTHOR [+2]

Abstract

To simultaneously read, compute and write in a stored program data processing system, it is usual to read data into a specified read area of memory, transfer this data to a compute area of memory for computation and finally transfer the results to an output area of memory for writing on tape. While computation is in process in the compute area, the next item of data is read into the input area and the computation result of the previous items is read out from the output area.

This text was extracted from a PDF file.
At least one non-text object (such as an image or picture) has been suppressed.
This is the abbreviated version, containing approximately 53% of the total text.

Page 1 of 3

Revolving Storage Addresses

To simultaneously read, compute and write in a stored program data processing system, it is usual to read data into a specified read area of memory, transfer this data to a compute area of memory for computation and finally transfer the results to an output area of memory for writing on tape. While computation is in process in the compute area, the next item of data is read into the input area and the computation result of the previous items is read out from the output area.

By revolving the memory addresses, simultaneous operation can take place using three areas of memory without actual transfer of data from area to area. Data is read into each of the three areas in sequence. Computation takes place within the area previously read into and write follows compute in the same area. In order for a single program to control processing in each of the three areas, some method of address modification is used. Indexing is usually used to accomplish the revolving, but at a cost in machine time, since each instruction must be indexed.

To avoid the necessity for indexing, the flow chart (left) is implemented according to the diagram (right). Three memory areas A, B, C are set aside. Read instructions always refer to a symbolic address alpha; compute instructions to beta; write instructions to gamma. After each cycle of simultaneous operation, a stored program operation code is executed which rotates the relationship of addresses alpha, beta and gamma to A, B, C. The revolve operation sets revolve switches which control referencing of A, B and C by addresses alpha, beta, and gamma as shown. After a first revolve operation, e.g., alpha may reference B. After the second alpha may reference C.

In the chart, the simultaneous read, compute,...