Error detection/correction and fault detection/recovery – Data processing system error or fault handling – Reliability and availability
Reexamination Certificate
2001-01-19
2004-01-13
Beausoliel, Robert (Department: 2184)
Error detection/correction and fault detection/recovery
Data processing system error or fault handling
Reliability and availability
C714S047300, C714S710000, C714S711000
Reexamination Certificate
active
06678836
ABSTRACT:
TECHNICAL FIELD
The present invention relates to fault tolerance for memory such as solid-state memory or other memory that may utilize flat memory space such as some disk drives, and more particularly to remapping of faulty memory addresses.
BACKGROUND
Whether memory be embodied in dedicated chips, be integrated into multi-purpose chips, or even be embodied in a disk drive utilizing flat memory space rather than cylinder/sector addressing, it may have locations that fail over time or are faulty at the time of manufacture. However, the majority of locations in the memory typically remain usable. Thus, methods have been devised to compensate for the faulty locations so that the use of the non-faulty locations may continue.
The conventional methods for fault tolerance include adding dedicated spare rows, columns, chips, modules, etc. that provide storage locations that substitute for the faulty locations. The dedicated spares are wasted unless there is a fault requiring a substitution. Other methods include permuting bits within words stored in the memory. These are complex schemes that do not allow memory components of more than one bit width to be used. Large granularity reconfiguration methods may be used whereby a group of locations containing one or more defective locations are disabled but more than the faulty memory area is left unusable.
Other methods include reconfiguring the faulty location to a predetermined location that is blocked. If access to the predetermined location is ever needed for additional storage, the reconfiguration fails. Methods involving arranging the circuitry of devices so that the faulty locations are avoided have been implemented. However, this is only performed during manufacturing and is inapplicable in the field. Other methods shuffle the address bits or fully permute the data, but these methods are more complex.
Accordingly, there is a need for simple fault tolerance for memory.
SUMMARY
The present invention may be viewed as a method for remapping locations in memory. The method involves generating a remapping value. The remapping value is logically combined with an intended address value to generate a remapped address value. A memory location having the remapped address value is then accessed.
The present invention may also be viewed as a system for remapping locations in memory. The system includes a first logic configured to generate a remapping value. A second logic is also included and is configured to combine the remapping value with an intended address value to generate a remapped address value. The system also includes a memory address input configured to access a memory location having the remapped address value.
REFERENCES:
patent: 3644902 (1972-02-01), Beausoleil
patent: 3765001 (1973-10-01), Beausoleil
patent: 3781826 (1973-12-01), Beausoleil
patent: 3897626 (1975-08-01), Beausoleil
patent: 4028539 (1977-06-01), Jacobs
patent: 4453248 (1984-06-01), Ryan
patent: 4461001 (1984-07-01), Bossen et al.
patent: 4489403 (1984-12-01), Bond
patent: 4506364 (1985-03-01), Aichelmann, Jr. et al.
patent: 4520453 (1985-05-01), Chow
patent: 4534029 (1985-08-01), Singh et al.
patent: 4649476 (1987-03-01), Sibigtroth
patent: 5067105 (1991-11-01), Borkenhagen et al.
patent: 5123101 (1992-06-01), Sindhu
patent: 5253350 (1993-10-01), Foster et al.
patent: 5253354 (1993-10-01), MacDonald et al.
patent: 5455834 (1995-10-01), Chang et al.
patent: 5541938 (1996-07-01), Di Zenzo et al.
patent: 5838893 (1998-11-01), Douceur
patent: 5937435 (1999-08-01), Dobbek et al.
patent: 5943283 (1999-08-01), Wong et al.
patent: 5991517 (1999-11-01), Harari et al.
patent: 5999480 (1999-12-01), Ong et al.
patent: 6338153 (2002-01-01), Sasaki et al.
patent: 6535995 (2003-03-01), Dobbek
patent: 0 146 891 (1985-07-01), None
Beausoliel Robert
Honeywell International , Inc.
Maskulinski Michael
Merchant & Gould
LandOfFree
Simple fault tolerance for memory does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Simple fault tolerance for memory, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Simple fault tolerance for memory will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3203296