A method for repairing defects in NAND flash memory devices

Abstract

A defect repair method for a NAND flash memory device includes: identifying defective blocks, data columns, and redundant columns from multiple blocks, multiple data columns, and multiple redundant columns of the NAND flash memory device, and generating a defect map; determining, based on the defect map, whether there are defective blocks and redundant columns that can be repaired in the NAND flash memory device; when there are defective blocks and redundant columns that can be repaired in the NAND flash memory device, determining whether there are defective blocks that can be eliminated due to bad columns; selecting defective blocks that are preferentially eliminated due to bad columns based on block weights calculated based on the defect map; and eliminating bad columns within the selected defective blocks; wherein the elimination of bad columns includes: replacing defective data columns in the selected defective blocks with redundant columns that are not bad columns in at least the selected defective blocks.

Description

Technical Field

[0001] This application relates to the field of data storage technology, and more specifically to methods and apparatus for repairing defects in NAND flash memory devices. Background Technology

[0002] NAND flash memory devices are a type of non-volatile memory increasingly used in modern electronic equipment. Compared to NOR flash memory devices, NAND flash memory devices are more widely used for high-capacity storage due to their higher storage density and capacity, as well as their higher sequential access speed.

[0003] NAND flash memory devices are typically organized into several blocks. Each block has the same number of pages, and each page has the same number of columns. Multiple memory cells are arranged in the columns. These columns can generally be divided into main columns for recording data, spare columns for providing backup to the main columns to increase the reliability of the device (the main columns and spare columns can be collectively referred to as data columns), and redundant columns for repairing defects in the NAND flash memory device. If one or more data columns develop defects during the manufacturing process, such as one or more memory cells in a data column being found to be faulty during testing, and if the defect severity is not beyond a certain range, the NAND flash memory device does not need to be scrapped directly. Instead, repair processing can be performed, allowing the defective data column to be disabled in subsequent operations and replaced by one of the redundant columns, thus enabling the NAND flash memory device to continue to function normally.

[0004] In existing technologies, various methods have been developed for repairing NAND flash memory devices by replacing defective data columns with redundant columns. One known method involves testing the NAND flash memory device to identify all defective data columns, then calculating the block weight of each defective block using a pre-defined algorithm. After calculating the block weight, the block with the highest repair weight is selected for repair by replacing the defective data column in that block with a completely intact (i.e., defect-free) redundant column. After repairing this block, the block weight of each remaining defective block is recalculated, and then the remaining completely intact redundant columns are used to repair the block with the highest block weight. This process is repeated until all defective blocks are repaired, or no more completely intact redundant columns are available.

[0005] In existing NAND flash memory device repair methods that use redundant columns, such as those described above, only completely intact redundant columns—that is, those without any defects—can typically replace defective data columns. If a redundant column also has some memory cells that fail during manufacturing, that redundant column is usually discarded in existing repair methods. Clearly, this approach completely discards defective redundant columns, even those with very minor defects, resulting in a waste of storage resources. Summary of the Invention

[0006] One embodiment of this application provides a defect repair method for a NAND flash memory device, which includes multiple blocks, multiple data columns, and multiple redundant columns. The method includes the following steps: identifying defective blocks, data columns, and redundant columns from the blocks, data columns, and redundant columns of the NAND flash memory device, and generating a defect map; determining whether there are defective blocks and redundant columns that can be repaired in the NAND flash memory device based on the defect map; when there are defective blocks and redundant columns that can be repaired in the NAND flash memory device, determining whether there are defective blocks that can be eliminated due to bad columns; selecting defective blocks that are preferentially eliminated due to bad columns based on the block weight calculated based on the defect map; and eliminating bad columns within the selected defective blocks; the elimination of bad columns includes: replacing the defective data columns in the selected defective blocks with redundant columns that are not bad columns in at least the selected defective blocks.

[0007] In some implementations, the redundant columns that are not bad columns in at least the selected defective blocks include completely intact redundant columns and redundant columns that are not bad columns in the selected defective blocks but are defective in at least one other block.

[0008] In some implementations, the bad column elimination process further includes discarding defective redundant columns in the selected defective block.

[0009] In some implementations, the block weight is calculated as follows: Block weight = Number of blocks covered ÷ Total number of bad columns in the block; wherein the number of blocks covered is the total number of all defective blocks whose defective columns are simultaneously eliminated when all defective columns in a defective block are eliminated; the total number of bad columns in the block is the total number of defective data columns and defective redundant columns in the defective block.

[0010] In some implementations, determining whether there are defective blocks that can be eliminated by bad column removal includes: determining whether there are defective blocks in which the number of bad columns is no more than the number of redundant columns in the block that are not bad columns.

[0011] In some implementations, selecting the defective block to be prioritized for bad column elimination based on the block weight calculated from the defect map includes: selecting the block with the largest block weight from among the defective blocks in which the number of bad columns in the block does not exceed the number of redundant columns that are not bad columns in the block as the block to be prioritized for bad column elimination.

[0012] In some implementations, the step of selecting the defective block to be prioritized for bad column elimination based on the block weight calculated based on the defect map further includes: among the defective blocks in which the number of bad columns in the block is no more than the number of redundant columns that are not bad columns in the block, when there are multiple blocks with the largest block weight, selecting the block with the largest number of block coverage from the blocks with the largest block weight as the block to be prioritized for bad column elimination.

[0013] In some implementations, replacing the defective data column in the selected defective block with a redundant column that is not a bad column in the selected defective block includes: when the number of redundant columns that are not bad columns in the selected defective block is greater than one, selecting the redundant columns that are not bad columns in the selected defective block in a preset order to replace the defective data column in the selected defective block.

[0014] In some implementations, replacing a defective data column in a selected defective block with a redundant column that is at least not a bad column in the selected defective block includes: when the number of redundant columns that are at least not bad columns in the selected defective block is greater than one, randomly selecting a redundant column that is at least not bad columns in the selected defective block to replace the defective data column in the selected defective block.

[0015] In some implementations, after the step of performing bad column elimination on the selected defective blocks within the block, the process returns to the step of determining whether there are defective blocks and redundant columns available for repair in the NAND flash memory device based on the defect map.

[0016] As described above, the defect repair method for NAND flash memory devices provided by the embodiments of this application can perform bad column elimination processing in NAND flash memory devices. This includes using redundant columns that are not bad within the block to cover defective data columns within the block to repair the defective data columns, and directly discarding redundant columns that are bad within the block, thereby repairing defective blocks in the NAND flash memory device. Compared to existing NAND flash memory device repair methods, the NAND flash memory device method provided by the embodiments of this application does not necessarily require the use of completely intact redundant columns when using redundant columns to repair defective blocks. Any redundant column that is not bad within at least the defective block to be repaired can be used. Thus, in the NAND flash memory device repair method provided by the embodiments of this application, not only can completely intact redundant columns be used to repair defective blocks, but redundant columns containing certain defects also have the opportunity to be used to repair defective blocks, instead of directly discarding any redundant columns containing defects as in existing methods. Thus, the NAND flash memory device method provided by the embodiments of this application improves the utilization rate of redundant columns and the repair rate of NAND flash memory devices, while also improving the fault tolerance rate in the NAND flash memory device manufacturing process. Attached Figure Description

[0017] To more clearly illustrate the technical solutions in the embodiments of this application, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0018] Figure 1 This is a flowchart of a defect repair method for NAND flash memory devices provided in an exemplary embodiment of this application.

[0019] Figure 2 This refers to an exemplary NAND flash memory device, according to Figure 1 The method shown is a schematic diagram of the defect map it generates.

[0020] Figure 3 It is based on Figure 1 The method shown is for those with Figure 2 The diagram shown illustrates the first round of bad column elimination for NAND flash memory devices with defect maps.

[0021] Figure 4 Yes, according to Figure 1 The method shown is for those with Figure 2 The diagram shown illustrates the second round of bad column elimination for NAND flash memory devices with defect maps. Detailed Implementation

[0022] The present application will now be described in further detail with reference to the accompanying drawings and embodiments. It should be particularly noted that the following embodiments are for illustrative purposes only and do not limit the scope of the application. Similarly, the following embodiments are only some, not all, embodiments of the present application, and all other embodiments obtained by those skilled in the art without inventive effort are within the scope of protection of the present application.

[0023] Please refer to Figure 1 One exemplary embodiment of this application provides a defect repair method for a NAND flash memory device. The method is applicable to a NAND flash memory device comprising multiple blocks and multiple data columns, each data column in the NAND flash memory device storing data in bytes using memory cells. The NAND flash memory device also includes multiple redundant columns for repairing defective data columns within the data columns.

[0024] like Figure 1 As shown, the method includes the following steps:

[0025] S1 identifies defective blocks, data columns, and redundant columns from the blocks, data columns, and redundant columns of the NAND flash memory device, and generates a defect map.

[0026] In step S1, the NAND flash memory device is tested to find all blocks, data columns, and redundant columns containing defective memory cells, and the defective memory cells are identified to form, for example... Figure 2 The defect map shown here, where each marker "R" represents a defective memory cell located at the position determined by the block containing that marker and the data / redundancy column. In this embodiment, for the sake of simplicity, Figure 2 The defect map shown represents a simplified, hypothetical scenario, depicting only eight defective blocks, each represented by a different block. Figure 2 The blocks are represented by eight rows numbered from 0 to 7 from top to bottom; each block includes only 16 data columns numbered from 0 to 15 and 4 redundant columns numbered from A to D. Those skilled in the art will understand that the disclosure of this application is not limited to these specific values, and in a defect map formed by testing actual flash memory devices, the blocks may also include other numbers of data columns and redundant columns.

[0027] like Figure 1As shown, in the eight defective blocks of the NAND flash memory device, for block 0, its corresponding data columns 0, 2, and redundant column A contain defective memory cells; for block 1, its corresponding data columns 4, 5, and 14 contain defective memory cells, and so on. The sum of the number of data columns and redundant columns identified as containing defective memory cells in each block is taken as the number of remaining bad columns in the block's full array corresponding to that block, as shown in... Figure 2 It is located in the column immediately to the right of this block.

[0028] S2: Determine whether there are defective blocks and redundant columns that can be repaired in the NAND flash memory device based on the defect map. If none are found, the process ends; otherwise, proceed to step S3.

[0029] Based on the obtained defect map, it can be directly determined whether there are defective blocks in the NAND flash memory device, and whether there are redundant columns that can be used to repair the defective data columns. If there are no defective blocks that need to be repaired, or no remaining redundant columns that can be used for repair, the method will end directly. If both defective blocks and redundant columns that can be used for repair exist, then proceed to step S3.

[0030] S3: When there are defective blocks and redundant columns available for repair in the NAND flash memory device, calculate the block weight of each defective block.

[0031] In this embodiment, the specific calculation method for the elimination value of a defective block is: Block weight = Block coverage number ÷ Total number of bad columns within the block. Here, the block coverage number of each defective block is the total number of all defective blocks whose defective columns (i.e., bad columns within the block) are eliminated simultaneously. The total number of bad columns within the block represents the total number of defective data columns and redundant columns within the defective block. In this embodiment, the specific method for eliminating defective columns in a defective block can include two types of operations: for defective data columns in the defective block, using redundant columns that are at least not bad columns within the defective block to repair the defective data columns; for defective redundant columns in the defective block, directly discarding the defective redundant columns, for example, marking the defective redundant columns as unusable.

[0032] This section explains why, when a defective column in a defective block is eliminated, defective data columns in other defective blocks can be repaired simultaneously. Specifically, based on the inherent principles of NAND flash memory devices, when a defective data column is replaced by a redundant column during the repair process, or when a defective redundant column is discarded, all blocks in the NAND flash memory device will be affected. For example, in... Figure 2In this context, if data column 0 is replaced by a non-defective redundant column B, then after repair, access to the memory cell at location (block 0, data column 0) will be addressed to the memory cell at location (block 0, redundant column B), and access to the memory cell at location (block 1, data column 0) will be addressed to the memory cell at location (block 1, redundant column B), and so on. Therefore, based on this principle, when using a non-defective redundant column to repair the data column of a specific defective block, in that redundant column, besides the memory cell used to repair the data column of that specific defective block, other intact memory cells may also simultaneously replace the defective memory cells in the data columns of other defective blocks, thus simultaneously repairing the defective data columns within those other defective blocks. Similarly, in... Figure 2 In this scenario, if redundant column A is directly discarded, all memory cells within it will be unused, meaning no block will use redundant column A. At this point, any block with defects in redundant column A will no longer fail due to calling redundant column A. In other words, discarding defective data columns from specific defective blocks will result in a higher "cost-effectiveness ratio."

[0033] The elimination value is a parameter reflecting the cost-effectiveness of repairing data columns in each defective block in a NAND flash memory device. The block weight for a specific block can be calculated based on two factors: the number of blocks covered and the total number of bad columns within the block. The number of blocks covered refers to the total number of defective blocks whose defective columns are simultaneously eliminated when eliminating all defective columns in that defective block. The total number of bad columns within the block represents the total number of defective data columns and defective redundant columns within that defective block. In an embodiment, the block weight can be determined by the following formula: Block weight = Number of blocks covered ÷ Total number of bad columns within the block. 。

[0034] For example, in Figure 2 In this context, if all defective columns in block 0 (including data columns 0 and 2 and redundant column A) are to be eliminated, then two redundant columns should be used to replace data columns 0 and 2 to eliminate the defective memory cells in block 0, and redundant column A should be discarded. At this point, since the defective columns in block 3 are data column 0 and redundant column A, and the defective data column in block 5 is only data column 2, if the two redundant columns used for replacement are completely intact, blocks 3 and 5 will obviously be repaired simultaneously due to the repair of block 0. That is, when all defective columns in block 0 are eliminated, all defective columns in the three defective blocks are also eliminated simultaneously. Therefore, according to the above definition, the block coverage of block 0 is 3. Since the number of defective columns (i.e., bad columns) in block 0 is 3, the block weight for block 0 can be calculated as 3 / 3 = 1.00. Figure 2 The diagram shows the total number of bad columns within a block, the number of blocks covered, and the block weight for all blocks in the defect map. It can be understood that for each defective block, a larger number of blocks covered indicates a better repair effect on the entire NAND flash memory device by eliminating all defective columns within that block; conversely, a smaller total number of bad columns within a block means less resources are needed to eliminate all defective columns within that block. Therefore, the ratio of the number of blocks covered to the total number of bad columns within a block is used as the block weight.

[0035] S4: Determine if there are any blocks that can be replaced by bad columns. If not, the process ends; otherwise, proceed to step S5.

[0036] In this embodiment, step S4 may specifically involve determining whether there exists a defective block where the number of bad columns (i.e., defective columns within the block) is no more than the number of redundant columns within the block that are not bad columns. For example, for... Figure 2 The NAND flash memory device shown determines whether there are any blocks (blocks 0 to 7) where the number of bad columns within a block is no more than the number of redundant columns that are not bad within that block. It is understood that if the number of bad columns in all defective blocks exceeds the number of redundant columns that are not bad within that block, then these defective blocks are unrepairable, and the method can terminate directly. However, in this embodiment, since the number of bad columns in each block (blocks 0 to 7) is at most 3, and the number of redundant columns that are not bad within each block is at least 3, all blocks (blocks 0 to 7) are blocks where the number of bad columns within a block is no more than the number of redundant columns that are not bad within that block, and can continue processing according to the subsequent step S5.

[0037] It is understood that the execution order of steps S3 and S4 is not limited to... Figure 1 As shown in the sequence, in various embodiments, step S3 may be executed before step S4, step S4 may be executed before step S3, or steps S3 and S4 may be executed simultaneously.

[0038] S5, select defective blocks for priority bad column elimination processing according to the block weight.

[0039] In this embodiment, step S5 may include the following sub-steps:

[0040] S51, among the defective blocks in which the number of bad columns in the block is no more than the number of redundant columns that are not bad columns in the block, select the block with the largest block weight as the block to be prioritized for bad column elimination.

[0041] S52, when there are multiple blocks with the largest block weight selected according to step S51, select the block with the largest number of block coverages from the multiple blocks with the largest block weights as the block to be prioritized for bad column elimination processing.

[0042] exist Figure 2 The block weights of blocks 0 through 7 have already been calculated. Among them, blocks 9, 5, and 7 have the largest block weights (all 1.00). Therefore, according to sub-step S51, blocks 0, 5, and 7 are first selected as the blocks to be prioritized for bad column elimination. Then, according to sub-step S52, block 0, which has the largest block coverage among blocks 0, 5, and 7, is selected as the block to be prioritized for bad column elimination.

[0043] S6: Eliminate the selected defective columns within the block, and then return to step S2.

[0044] As previously mentioned, bad columns within a defective block include defective data columns and defective redundant columns. Specific methods for discarding defective columns within this defective block can include at least one of two types of operations: for defective data columns within the defective block, repairing the defective data columns using redundant columns that are not bad columns within at least the defective block; or for defective redundant columns within the defective block, directly discarding the defective redundant columns, for example, marking the defective redundant columns as unusable.

[0045] The redundant columns that are not bad columns in at least the selected defective blocks can include completely intact redundant columns and redundant columns that are not bad columns in the selected defective blocks but are defective in at least one other block. For example Figure 2 As shown, in this embodiment, redundant columns B and D are completely intact redundant columns without any defects. Redundant column C has defects in block 6, but redundant columns B, C, and D have no defects in the selected block 0, that is, none of them are bad columns. Therefore, they can all be used to repair the data columns in block 0.

[0046] Specifically in this embodiment, due to the selection Figure 2 Block 0 in the block is prioritized for bad column replacement, therefore refer to Figure 3 As shown, any two of the redundant columns B, C, and D that are not considered bad columns in block 0 can be used to replace data columns 0 and 2 to repair them. Simultaneously, the defective redundant column A is discarded. This completes the bad column elimination process for block 0, thus repairing block 0. Furthermore, since redundant columns B, C, and D are also not considered bad columns in blocks 3 and 5, the bad column elimination process for blocks 0 is also performed simultaneously with that for blocks 3 and 5, repairing blocks 3 and 5.

[0047] In some implementations, when using redundant columns that are not bad columns in at least one selected defective block to repair a defective data column in that block, if the number of redundant columns that are not bad columns in at least one defective block is greater than one, the redundant columns that are not bad columns in at least one selected defective block can be selected in a preset order to replace the defective data column in the selected defective block. For example, in this implementation, the redundant columns that are not bad columns in at least one selected defective block can be selected in the order of B, C, D or in an order determined by other criteria (such as the integrity of the redundant columns) to replace the defective data column in the selected defective block; or the redundant columns that are not bad columns in at least one selected defective block can be randomly selected to replace the defective data column in the selected defective block.

[0048] In this embodiment, after the bad column elimination process for blocks 0, 3, and 5 is completed according to the above method, the process returns to step S2 for a second round of determination to determine whether there are defective blocks and redundant columns that can be repaired. (Refer to...) Figure 3 After repairing blocks 0, 3, and 5, defective blocks 1, 2, 4, 6, and 7 remain. Of the four redundant columns, A is discarded, and two of B, C, and D have been used for repair, leaving one redundant column available for repair. Therefore, the result of step S2 in the second round is still yes, and we proceed to step S3.

[0049] Reference Figure 3 Following the aforementioned step S3, a second round of block weight calculation is performed. Among the remaining defective blocks 1, 2, 4, 6, and 7, block 7 is found to have the highest block weight (note that in this embodiment, if...). Figure 3 In the scenario shown, all blocks except block 7 have been repaired or determined to be unrepairable. Therefore, only block 7 can still have its block weight calculated, and its block weight is directly determined to be the highest. In other embodiments, it is obvious that there may be multiple blocks that can have their block weights calculated in the second round of block weight calculation (in which case their block weights can be compared); before, after, or simultaneously with step S3, the second round of judgment is performed according to the aforementioned step S4; since the number of bad columns in block 7 is 1, and the number of available redundant columns that are not bad columns in block 7 is also 1, it is determined in the second round step S4 that there is a block (i.e., block 7) that can be processed for bad column elimination, and the process proceeds to the second round step S5, where block 7 is selected for bad column elimination. Then, the process proceeds to the second round step S6, where the remaining 1 available redundant column is used to cover data column 4 of block 7, thereby completing the bad column elimination process for block 7 and realizing the repair of block 7, as shown. Figure 4 As shown.

[0050] After repairing block 7, return to step S2 in the third round. Since there are no redundant columns available for repair at this point, the entire method ends.

[0051] It is understood that the above method execution process is only used as an example of a relatively simple NAND flash memory device structure to illustrate the working principle of the technical solution of this application. In other implementation methods and even in practical applications, the method described above can be implemented in NAND flash memory devices with larger scale and more complex structures. The corresponding defect map can have a larger scale, and the number of loop execution rounds of the method described above from step S2 to S6 can also be more, until it is determined in step S2 of a certain round that there are no defective blocks that need to be repaired or no remaining redundant columns that can be repaired, or it is determined in step S5 of a certain round that there are no blocks that can be repaired.

[0052] It is understood that in the method provided by the above embodiments, a bad column elimination quota in the NAND flash memory device can also be preset, and the maximum value of the bad column elimination quota does not exceed the total number of redundant columns. In some embodiments, after the above method reaches step S6, it can also check whether the total number of bad columns that have been eliminated in the NAND flash memory device (that is, the total number of defective data columns that have been repaired and the total number of defective redundant columns that have been discarded) has reached the maximum value of the preset bad column elimination quota, that is, whether the remaining bad column elimination quota is 0. If the remaining bad column elimination quota has reached 0, then there is no need to return to step S2 after step S6, and the process can end directly.

[0053] The NAND flash memory device method provided by the embodiments of this application can perform bad column elimination processing in NAND flash memory devices, including using redundant columns that are not bad columns in the block to cover defective data columns in the block to repair defective data columns in the block, and directly discarding redundant columns that are bad columns in the block, thereby repairing defective blocks in NAND flash memory devices.

[0054] Compared to existing NAND flash memory device repair methods, the NAND flash memory device repair method provided in this application does not necessarily require the use of completely intact redundant columns when using redundant columns to repair defective blocks. Any redundant column that is not a bad column in at least the defective block to be repaired can be used. Thus, in the NAND flash memory device repair method provided in this application, not only completely intact redundant columns can be used to repair defective blocks, but redundant columns with some defects also have the opportunity to be used, instead of discarding any redundant columns with defects as in existing methods. Therefore, the NAND flash memory device repair method provided in this application improves the utilization rate of redundant columns and the repair rate of NAND flash memory devices, while also improving the fault tolerance rate in the NAND flash memory device manufacturing process.

[0055] The above description is only a part of the embodiments of this application and does not limit the scope of protection of this application. Any equivalent device or equivalent process transformation made based on the content of this application specification and drawings, or direct or indirect application in other related technical fields, are similarly included in the patent protection scope of this application.

Claims

1. A defect repair method for NAND flash memory devices, used in NAND flash memory devices comprising multiple blocks, multiple data columns, and multiple redundant columns; characterized in that, The method includes the following steps: Defective blocks, data columns, and redundant columns are identified from the blocks, data columns, and redundant columns of the NAND flash memory device, and a defect map is generated. The defect map is used to determine whether there are defective blocks and redundant columns that can be repaired in the NAND flash memory device. When there are defective blocks and redundant columns available for repair in the NAND flash memory device, determine whether there are defective blocks that can be eliminated by bad column removal. Based on the block weights calculated from the defect map, defective blocks are selected for priority bad column elimination. The selected defective blocks are subjected to a bad column elimination process. The bad column elimination process includes replacing the defective data column in the selected defective block with a redundant column that is not a bad column in at least the selected defective block. The redundant column that is not a bad column in at least the selected defective block includes a completely intact redundant column and a redundant column that is not a bad column in the selected defective block but is defective in at least one other block.

2. The method as described in claim 1, characterized in that, The bad column elimination process also includes discarding the defective redundant columns in the selected defective block.

3. The method as described in claim 1, characterized in that, The block weight is calculated as follows: Block weight = Number of blocks covered ÷ Total number of bad columns in the block; wherein the number of blocks covered is the total number of all defective blocks whose defective columns are simultaneously eliminated when all defective columns in a defective block are eliminated; the total number of bad columns in the block is the total number of defective data columns and defective redundant columns in the defective block.

4. The method as described in claim 1, characterized in that, Determining whether there is a defective block that can be processed for bad column elimination includes: determining whether there is a defective block in which the number of bad columns is no more than the number of redundant columns that are not bad columns in the block.

5. The method as described in claim 1, characterized in that, The step of selecting defective blocks for priority bad column elimination based on the block weights calculated from the defect map includes: among defective blocks in which the number of bad columns does not exceed the number of redundant columns that are not bad columns in the block, selecting the block with the largest block weight as the block for priority bad column elimination.

6. The method as described in claim 4, characterized in that, The step of selecting defective blocks for priority bad column elimination based on the block weights calculated from the defect map further includes: In a defective block where the number of bad columns within the block is no more than the number of redundant columns that are not bad columns within the block, if there are multiple blocks with the highest block weight, the block with the largest number of block coverages is selected from the blocks with the highest block weights as the block to be prioritized for bad column elimination.

7. The method as described in claim 1, characterized in that, The step of replacing the defective data column in the selected defective block with a redundant column that is at least not a bad column in the selected defective block includes: When the number of redundant columns that are not bad columns in at least the selected defective block is greater than one, the redundant columns that are not bad columns in at least the selected defective block are selected in a preset order to replace the defective data columns in the selected defective block.

8. The method as described in claim 1, characterized in that, The step of replacing the defective data column in the selected defective block with a redundant column that is at least not a bad column in the selected defective block includes: When the number of redundant columns that are not bad columns in at least the selected defective block is greater than one, the redundant columns that are not bad columns in at least the selected defective block are randomly selected to replace the defective data columns in the selected defective block.

9. The method as described in claim 1, characterized in that, Also includes: After the step of performing bad column elimination on the selected defective blocks, return to the step of determining whether there are defective blocks and redundant columns that can be repaired in the NAND flash memory device based on the defect map.

Images