Method, apparatus and system for handling data faults

A system comprising:

• multiple dies each including a respective one of multiple memory arrays each to store a respective one of vertical error correction (VEC) codes, wherein for each of the VEC codes, the VEC code is specific to data of the memory array which stores the VEC code;
• a first die coupled to the multiple dies to store a first Bose, Ray-Chaudhuri, Hocquenghem (BCH) code for respective data of each of the multiple memory arrays;
• and
• control logic to monitor for a die failure of the multiple dies, the control logic further to monitor for a request to perform error correction for the multiple memory arrays;
• and
• decoder logic coupled to the control logic, wherein, if the request is detected while each of the multiple dies is operative, the decoder logic to perform the error correction, including the decoder logic to perform a first horizontal BCH decode operation based on the first BCH code, and the decoder logic to perform first VEC decode operations based on a result of the BCH decode operation and the VEC codes;
• wherein, if the die failure is indicated, the decoder logic to perform a second horizontal BCH decode operation based on a first BCH code for recovery of data of a failed die.

A method comprising:

• monitoring for a die failure of multiple dies each including a respective one of multiple memory arrays;
• monitoring for a request to perform error correction for the multiple memory arrays;
• if the request is detected while each of the multiple dies is operative, performing the error correction, including:
• performing a first horizontal Bose, Ray-Chaudhuri, Hocquenghem (BCH) decode operation based on a first BCH code of a first die coupled to the multiple dies, the first BCH code for respective data of each of the multiple memory arrays;
• and
• performing first vertical error correction (VEC) decode operations based on a result of the BCH decode operation and VEC codes each of a respective memory array of the multiple memory arrays, wherein, for each of the VEC codes, the VEC code is specific to data of the respective memory array which includes the VEC code;
• and
• if the die failure is indicated, performing a second horizontal BCH decode operation based on a first BCH code for recovery of data of a failed die.

A non-transitory computer-readable storage medium having stored thereon instructions which, when executed by one or more processing units, cause the one or more processing units to perform a method comprising:

• monitoring for a die failure of multiple dies each including a respective one of multiple memory arrays;
• monitoring for a request to perform error correction for the multiple memory arrays;
• if the request is detected while each of the multiple dies is operative, performing the error correction, including:
• performing a first horizontal Bose, Ray-Chaudhuri, Hocquenghem (BCH) decode operation based on a first BCH code of a first die coupled to the multiple dies, the first BCH code for respective data of each of the multiple memory arrays;
• and
• performing first vertical error correction (VEC) decode operations based on a result of the BCH decode operation and VEC codes each of a respective memory array of the multiple memory arrays, wherein, for each of the VEC codes, the VEC code is specific to data of the respective memory array which includes the VEC code;
• and
• if the die failure is indicated, performing a second horizontal BCH decode operation based on a first BCH code for recovery of data of a failed die.