Browse Prior Art Database

A Method to Improve NVRAM Reliability

IP.com Disclosure Number: IPCOM000041154D
Original Publication Date: 1987-Oct-01
Included in the Prior Art Database: 2005-Feb-02
Document File: 1 page(s) / 11K

Publishing Venue

IBM

Related People

Fox, TH: AUTHOR [+2]

Abstract

A non-volatile random access memory (NVRAM) is divided into two parts. The first part contains the rarely changing values, the second part contains the frequently changing values and unused NVRAM memory. This second part can be used advantageously to increase the reliability of an NVRAM. Let M be the total memory in the second part and m is the total memory in the first part occupied by the rarely changing values; (unused memory = M - m). The frequently changing values are stored within a moving window of size m. This window can be anywhere within M and is surrounded by zeroes. It starts with a non-zero flag so that it can be located at power up. At power down, if any of the frequently changing values has changed, the window is written at a different location within M, so that a different part of the NVRAM is used.

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 100% of the total text.

Page 1 of 1

A Method to Improve NVRAM Reliability

A non-volatile random access memory (NVRAM) is divided into two parts. The first part contains the rarely changing values, the second part contains the frequently changing values and unused NVRAM memory. This second part can be used advantageously to increase the reliability of an NVRAM. Let M be the total memory in the second part and m is the total memory in the first part occupied by the rarely changing values; (unused memory = M - m). The frequently changing values are stored within a moving window of size m. This window can be anywhere within M and is surrounded by zeroes. It starts with a non-zero flag so that it can be located at power up. At power down, if any of the frequently changing values has changed, the window is written at a different location within M, so that a different part of the NVRAM is used. This results in uniform wear of M. If m is very much less than M, as is often the case, the NVRAM reliability resulting from this scheme approaches the reliability of an NVRAM that stores unchanging values. In conclusion, the method described improves the reliability of NVRAMs by intelligently managing the NVRAM storage, and can be applied whenever NVRAMs are used.

Disclosed anonymously.

1