Dismiss
InnovationQ will be updated on Sunday, Oct. 22, from 10am ET - noon. You may experience brief service interruptions during that time.
Browse Prior Art Database

Fast Parity Generator

IP.com Disclosure Number: IPCOM000109654D
Original Publication Date: 1992-Sep-01
Included in the Prior Art Database: 2005-Mar-24
Document File: 2 page(s) / 53K

Publishing Venue

IBM

Related People

Handlogten, GH: AUTHOR

Abstract

This invention provides an alternative to parity generation or prediction. This article describes a method that may be used under several circumstances to attach good parity to data without using extra levels of logic. Basically, instead of making the parity match the data, the data is made to match the parity.

This text was extracted from an ASCII text file.
This is the abbreviated version, containing approximately 81% of the total text.

Fast Parity Generator

       This invention provides an alternative to parity
generation or prediction.  This article describes a method that may
be used under several circumstances to attach good parity to data
without using extra levels of logic.  Basically, instead of making
the parity match the data, the data is made to match the parity.

      Under certain circumstances, bits may be altered without
affecting the result of a computation.  Doing so makes parity
generation easier and in many cases trivial.  This can be done when
some of the data or control will not be used and may be considered a
"don't care."  Even though the bits are a "don't care," parity should
still be checked.  If half of a data field associated with 1 parity
bit is unused in a particular cycle, the unused field may be loaded
with the same value as the used field.  For example, an odd parity
bit is associated with two 4-bit fields.  If a particular operation
only uses one of the fields, instead of loading the unused field with
zeros and generating parity on the 4-bit field, both fields may be
loaded with the same value while the parity bit is forced to a binary
1 (see Fig. 1).  If it is advantageous, the unused field may instead
be loaded with the inverse of the data.  The parity bit would still
be forced to a binary 1.

      If there are more unused bits than used bits, the extra unused
bits would be forced to zero.  If there are more used bits than
unused bits, the unused...