Browse Prior Art Database

Method of On-Line Self-Repair and Hard Error Detection for Memory Devices Disclosure Number: IPCOM000242942D
Publication Date: 2015-Sep-01
Document File: 3 page(s) / 452K

Publishing Venue

The Prior Art Database


Described is a new memory function "Auto-PPR" which enables the memory device to "self-repair" based on the results of completing a scrubbing activity.

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 51% of the total text.

Page 01 of 3

Method of On - -Line Self

Line Self -

-Repair and Hard Error Detection for Memory Devices

Repair and Hard Error Detection for Memory Devices

Memory technology continues to shrink, causing an increasing number of single cell errors. On-memory-die ECC has been demonstrated as a solution to mask this increasing number of errors from users. Additionally, post-package repair (PPR) has been introduced to enable users to selectively repair a small number of rows that are producing repeated errors. Prior inventions have additionally discussed the potential for an increased selection of ECC coverage over a user-defined subset of the memory. This enables selective 'repair' of a much larger amount of memory than PPR. In either case (normal PPR or extended ECC), the memory controller is required to direct the repair of specific portions of memory detected to be in error. This invention provides a method for the memory device to automatically repair problematic sections of memory using the available repair mechanisms. The invention is a logical method that the memory device uses to detect recurring (hard) errors and to subsequently take repair action. Recurring errors are detected by reading memory locations (either upon receipt of a command [called SCRUB] or on the expiration of a timer) and then using the on-die ECC logic to detect errors, re-write corrected data, then re-reading the newly written data and then using on-die ECC again to detect if an error remains. Post package repair (PPR) is an ability build into DDR4 that enables users to "repair" defective rows using built-in spares. Repair utilizing PPR is normally accomplished by the memory controller sending explicit commands to the memory device.

    Figure 1 illustrates the flow for a process whereby execution of a Scrub command automatically attempts to utilize PPR to immediately repair any hard error. Upon receipt of the Scrub command, the memory device executes a read. The on-die ECC checks for an error. If an error is detected, the device writes b...