TWI567745B - So that the anti-gate flash memory has a multi-performance method - Google Patents

Description

Method for making the anti-gate flash memory have read performance

The present invention relates to a method for making a reverse flash memory having read performance, and more particularly to a method for increasing the number of readable times and improving data access reliability.

Please refer to FIG. 4, which is an embodiment of a conventional technology NAND flash memory. The anti-gate flash memory 1 includes a reverse flash memory controller 2 and a plurality of anti-gate flash memory chips 3, wherein each of the anti-gate flash memory chips includes a plurality of data storage areas. Block 31, each data storage block includes a plurality of data storage pages (Page) 311. The gate flash memory module 1 is read or programmed in units of a data storage page 311 when reading data or data programming. However, when erasing the data, all the data storage pages 311 in the entire data storage block 31 must be erased at one time. On the other hand, the anti-gate flash memory has a limit on the number of programming erase cycles, and the general commercial single-level storage unit (SLC) and the gate flash memory, the number of programming erase cycles The limit is about 50,000 times; the number of times the multi-level cell (MLC) and the gated flash memory programming erase cycle is about 3,000 times.

In addition, the anti-gate flash memory has the characteristics of read interference and programming interference. For example, when the same data storage page is read multiple times, not only will the page be stored for the data. As a result, other data storage pages in the same data storage block as the data storage page may also be disturbed, resulting in the generation of data stored in the same data storage block or in other data storage pages. error. For the general commercial single-stage storage unit and the gate flash memory, the interference of the reading may be caused by a permanent error after reading 1,000,000 times; and the multi-level storage unit is opposite to the gate flash memory. The interference of reading may be caused by a permanent error after about 100,000 readings. Generally, the general anti-gate flash memory controller will have an error checking and correction system. When the stored data is wrong, the error checking and correction system can calculate each data and obtain an error correction code (Error- Correcting Codes) to check for errors and correct errors. However, the error checking and correction system can tolerate a certain degree of error in the data. When the fault tolerance limit is exceeded, the data will not be corrected by the error checking and correction system. The data errors caused by the interference of the reading and the interference of the programming must also be corrected before the fault tolerance limit is reached, otherwise the data will be permanently wrong.

In addition, if the anti-gate flash memory is used for long-term data storage, it is not clear how long the data stored on the anti-gate flash memory can be stored. Especially when there is a part that is not frequently programmed, or if it is not read for a long time, the data may be damaged and not found for a long time, which makes the reliability of reading the data lower. .

In view of this, the inventors have developed a method for making the anti-gate flash memory have read performance, which can increase the number of data readable times and reduce the frequency at which data storage blocks are erased, so as to enhance data reading. Reliability, and avoiding the data being damaged for a long time without being read, avoiding the above disadvantages, and thus the present invention.

One problem to be solved by the present invention is that the number of times of programming erase cycles of the gate flash memory and the interference of reading and programming may cause permanent errors in the data, so as to increase the number of times the data can be read as much as possible. And thereby reducing the frequency of data storage blocks being erased.

Another problem to be solved by the present invention is to correct the error data before the data error has exceeded the fault tolerance limit, so as to enhance the reliability of data reading.

Another problem to be solved by the present invention is to prevent the data that has not been read for a long time from being permanently damaged before it is discovered, so as to achieve the purpose of enhancing the reliability of data reading for long-term data storage purposes.

In order to solve the foregoing problems, in order to achieve the expected effect, the present invention provides a method for making a reverse flash memory having read performance, wherein the inverse gate flash memory includes a plurality of data storage blocks, each of which The data storage block includes a plurality of data storage pages, and the method is a reading method, comprising the following steps: A1: reading one of the data storage pages in a data storage block, wherein the data storage page is The number of times of reading after being read this time is one read number; and A2: when an update condition is established, the step A2 is ended; otherwise, the steps including B1 to B4 are performed, wherein the update condition is: The number of readings is less than the threshold of the number of readings, wherein the steps B1 to B4 are as follows: B1: copying the entire block data stored in the data storage block to another data storage block in the blank, wherein the full block The data is all the data stored in all the data storage pages in the data storage block; B2: updating a flash conversion layer mapping table, so that the storage address of the full block data is linked to the other data storage Block; B3: erase the whole block of the data stored in the data storage block; B4: The read frequency is set to 0.

In one embodiment, the foregoing makes the anti-gate flash memory have read performance. The method, wherein the step A2 further comprises calculating one of the error correction codes of the data, wherein the update condition is that the error correction code is smaller than an error correction code threshold and the number of read times is less than the number of read times.

In addition, the present invention further provides a read-back performance for the anti-gate flash memory. The method includes a plurality of data storage blocks, each data storage block includes a plurality of data storage pages, and each of the data storage blocks has a block number, and the method is The method for reading the screening includes the following steps: C1: the data storage block having a starting block number is a data storage block in the screening; C2: the data storage block in the screening includes D1~D2 Step: D1: reading one of the data storage pages of one of the data storage blocks in the screening, wherein the data storage page is read a number of times after being read; D2: When an update condition is met, the C3 step is performed; otherwise, the steps including E1~E4 are performed, wherein the update condition is: the read count is less than a read count threshold, wherein the E1~E4 steps are as follows: E1: will be stored in The entire block data of one of the data storage blocks in the screening is copied to another data storage block in the blank, wherein the full block data is stored in all data storage pages of the data storage block in the screening. All information; E2: a new flash conversion layer mapping table, such that the storage address of the full block data is linked to the other data storage block; E3: erasing the full block data stored in the data storage block of the screening; E4: setting the number of read times to 0 and performing step C3; C3: adding 1 to the block number of the data storage block in the screening; and C4: when the current block number is equal to When the block number is ended, the C1 step is performed. Otherwise, the data storage block having the next block number is the data storage block in the screening and the C2 step is performed.

In one embodiment, the foregoing makes the anti-gate flash memory have read performance. The method, wherein the step D2 further comprises calculating one of the error correction codes of the data, wherein the update condition is that the error correction code is smaller than an error correction code threshold and the number of read times is less than the number of read times.

In one embodiment, the foregoing makes the anti-gate flash memory have read performance. The method wherein the screening data storage page of the data storage block in the screening is sequentially or randomly selected in the step D1.

In one embodiment, the foregoing makes the anti-gate flash memory have read performance. The method, wherein the anti-gate flash memory system is a redundant array of independent data storage, the redundant array of independent data storage being replaced by a standard independent hard disk redundant array (RAID) with at least one data storage body a plurality of hard disk drives, wherein the data storage system is a solid state hard disk, a reverse flash memory hard disk or a flash memory, the solid state hard disk, the anti-gate flash memory hard disk Or the flash memory includes at least one anti-gate flash memory chip, and the plurality of data storage blocks are included in each of the anti-gate flash memory chips.

In one embodiment, the foregoing makes the anti-gate flash memory have read performance. The method, wherein the anti-gate flash memory system is a redundant array of independent data storage, the redundant array of independent data storage being replaced by a standard independent hard disk redundant array (RAID) with at least one data storage body The data storage system is composed of at least one anti-gate flash memory chip and at least one data storage controller, and the plurality of data storage blocks are included in each of the at least one anti- And the gate flash memory chip.

In addition, the present invention further provides a read-back performance of the anti-gate flash memory. The method, wherein the anti-gate flash memory comprises a plurality of data storage blocks, each data The storage block includes a plurality of data storage pages, and the method includes the steps of: reading a data by the foregoing reading method; and checking each of the data storage blocks one by one and cyclically by the aforementioned read screening method.

In one embodiment, the foregoing makes the anti-gate flash memory have read performance. The method, wherein the anti-gate flash memory system is a redundant array of independent data storage, the redundant array of independent data storage being replaced by a standard independent hard disk redundant array (RAID) with at least one data storage body a plurality of hard disk drives, wherein the data storage system is a solid state hard disk, a reverse flash memory hard disk or a flash memory, the solid state hard disk, the anti-gate flash memory hard disk Or the flash memory includes at least one anti-gate flash memory chip, and the plurality of data storage blocks are included in each of the anti-gate flash memory chips.

In one embodiment, the foregoing makes the anti-gate flash memory have read performance. The method, wherein the anti-gate flash memory system is a redundant array of independent data storage, the redundant array of independent data storages replacing at least one data storage body with a plurality of hard disks in a standard independent hard disk redundant array The data storage system is composed of at least one anti-gate flash memory chip and at least one data storage controller, and the plurality of data storage blocks are included in each of the at least one reverse gate In a flash memory chip.

In order to further understand the present invention, the specific embodiments of the present invention and the effects achieved thereby are described in detail below with reference to the drawings and drawings.

1‧‧‧Anti-gate flash memory

2‧‧‧Anti-gate flash memory controller

3‧‧‧Anti-gate flash memory chip

31‧‧‧Data storage block

311‧‧‧Data storage page

A1, A2‧‧‧ steps

B1, B2, B3, B4‧‧‧ steps

C1, C2, C3‧‧‧ steps

D1, D2‧‧‧ steps

E1, E2, E3, E4‧‧‧ steps

F1, F2, F3‧‧‧ steps

BRIEF DESCRIPTION OF THE DRAWINGS Figure 1 is a flow diagram of a first embodiment of the method of the present invention for enabling the read-back performance of the anti-gate flash memory.

Figure 1A is a flow diagram of a second embodiment of the method of the present invention for enabling the read-back performance of the anti-gate flash memory.

Fig. 2 is a flow chart showing a third embodiment of the method for making the gate flash memory have read performance.

Fig. 2A is a flow chart showing a fourth embodiment of the method for making the anti-gate flash memory have read performance.

Figure 3 is a flow chart showing a fifth embodiment of the method for making the anti-gate flash memory have read performance.

Figure 4 is an embodiment of a conventional technique against gate flash memory.

Please refer to FIG. 1 , which is a flowchart of a first embodiment of a method for enabling a gate flash memory to have read performance, wherein the gate flash memory includes a plurality of data storage blocks. Each data storage block includes a plurality of data storage pages. The method is a reading method, comprising the following steps: A1: reading one of the data storage pages in a data storage block, wherein the data storage page is read after the current reading is A number of times read; and A2: when an update condition is established, the process of step A2 is ended; otherwise, the steps including B1 to B4 are performed, wherein the update condition is: the number of times read is less than the threshold of a number of read times, wherein steps B1 to B4 As follows: B1: Copy all the block data stored in the data storage block to another data storage block in the blank, wherein the whole block data is stored in all data storage pages in the data storage block. All the data; B2: update a flash conversion layer mapping table, so that the storage address of the entire block data link to another data storage block; B3: erase the data stored in the data storage block Full block data; B4: Set the number of times read to 0.

Please refer to FIG. 1A, which is a view of the present invention for making the anti-gate flash memory read. A flow chart of a second embodiment of the multi-performance method, which is a reading method, which is substantially the same as the first embodiment, except that step A2 is replaced by step A2', wherein step A2' : Calculate one of the error correction codes. When the update condition is met, the step A2' is ended. Otherwise, the steps including B1 to B4 are performed, wherein the update condition is: the error correction code is smaller than an error correction code threshold and the read count is less than The threshold has been read.

In step B4, the data storage page in the data storage block may be further included. The number of times read is set to 0 and the number of times of reading of all data storage pages in another data storage block is set to zero.

Every time you read a piece of data, if there is only an error correction code to check the data If the error correction code threshold has been reached, the following situation may occur: the data has not reached the error correction code threshold, but the number of times the data has been read has already exceeded the threshold of the number of readings. As a result, the information will be permanently wrong and will no longer be able to respond. In addition, if it is only checked whether the number of readings of the data has reached the threshold of reading, it is possible that the data has not yet reached the threshold of reading, and the error correction code of the data has already exceeded the threshold of the error correction code. As a result, this information will also cause permanent errors. Therefore, through the method of the present invention, the anti-gate flash memory has a multi-reading performance, and each time a piece of data is read, the number of times the read data has been read has been checked to the level of the read count and the data is checked. Whether the error correction code has reached the error correction code threshold. When the number of times the data has been read has not reached the threshold of the number of readings and the error correction code of the data has not reached the threshold of the error correction code, it indicates that the data is still within the safe range, and can continue to be safely read. This is as much as possible to increase the information in The number of times the data storage block can be read, so that the anti-gate flash memory has read performance and thereby reduces the frequency at which the data storage block is erased. Once the number of times the data has been read reaches the threshold of the number of readings or the error correction code of the data reaches the threshold of the error correction code, it indicates that the data is in danger of a permanent error of the data, so the data can still be corrected. When replying, the data is re-written in another data storage block of the blank, so that the number of times the data has been read is reset to zero, and the error correction code of the data is reset to zero.

However, it is not enough to check the data to be read every time. Many are stored The data in the same data storage block may not be read after a long period of time after being stored. The data may be damaged and not found for a long time. In order to avoid the above situation, please refer to FIG. 2, which further provides a third embodiment of a method for enabling the read-back performance of the anti-gate flash memory, wherein the anti-gate flash memory includes a plurality of data. The storage block, each data storage block includes a plurality of data storage pages, and each of the data storage blocks has a block number. The method is a read screening method, comprising the following steps: C1: making a data storage block having a starting block number a data storage block in a screening; C2: performing a data storage block in the screening Including D1~D2 steps: D1: reading one of the data storage pages in one of the data storage blocks in the screening, wherein the number of times the data storage page is read after being read this time is one read number; D2 : When an update condition is met, the C3 step is performed; otherwise, the steps including E1~E4 are performed, wherein the update condition is: the number of read times is less than the threshold of the number of read times, wherein the steps of E1~E4 are as follows: E1: will be stored in The entire block data of one of the data storage blocks in the screening is copied to another data storage block in the blank. The whole block data is all the data stored in all the data storage pages of the data storage block in the screening. ; E2: update a flash conversion layer mapping table, so that the storage address of the entire block data link to another data storage block; E3: erase and store in the screening The total block data of the middle data storage block; E4: set the number of read times to 0 and perform the C3 step; C3: increase the block number of the data storage block in the screening by one for one block number; and C4 : When the next block number is equal to an end block number, the C1 step is performed; otherwise, the data storage block having the next block number is the data storage block in the screening and the C2 step is performed. Through this read screening method, all data storage blocks are checked one by one and cyclically while the system is idle. In this way, each data storage block is read and checked one by one for a period of time to avoid the situation that the data is not damaged after being read for a long time.

Please refer to FIG. 2A, which is a method for making the anti-gate flash memory read. A flow chart of a fourth embodiment of the multi-performance method, which is a read screening method, which is substantially the same as the third embodiment, except that step D2 is replaced by step D2', wherein the step D2': Calculate one of the error correction codes. When the update condition is met, the D2' step is ended. Otherwise, the steps including E1~E4 are performed, wherein the update condition is: the error correction code is smaller than an error correction code threshold and has been read. The number of times is less than the threshold of the number of readings.

The E4 step may further include storing all data in the data storage block in the screening. The number of times the page is read is set to 0 and the number of times the data storage page of another data storage block is read is set to zero.

In the D1 step, the screening of the data storage block in the screening can be selected sequentially or randomly. Data storage page.

In addition, referring to FIG. 3, the present invention further provides an anti-gate flash memory device. A fifth embodiment of the method for reading multi-performance, wherein the anti-gate flash memory comprises a plurality of data storage blocks, each of the data storage blocks comprising a plurality of data storage pages, the method comprising the following steps: Reading the first embodiment of the invention or the reading method of the second embodiment A data is taken; and each data storage block is inspected one by one and cyclically in the background with the third embodiment of the present invention or the read screening method of the fourth embodiment when the system is idle. When the data is generally required to be read, the reading method is used to read the data; and when the system is idle, each of the data storage blocks is checked one by one and cyclically in the background by the reading screening method. Thereby, the number of times the data can be read in the data storage block is increased as much as possible, so as to improve the reliability of data reading, and each data storage block is read one by one for a period of time. And check to avoid the situation that the data is not damaged after being read for a long time.

The first method for making the anti-gate flash memory have read performance The fifth embodiment can also be applied to a standard hard disk redundancy array (RAID) architecture. The conventional standard hard disk redundant array uses a plurality of general hard disks, and the plurality of general hard disk systems can be replaced by at least one data storage to form a redundant array of independent data storage. Each of the data stores includes a plurality of data storage blocks. The use of a method of the present invention to enable the anti-gate flash memory to have read performance on the architecture of the redundant array of independent data storages further enhances the reliability and integrity of data access.

The above-mentioned data storage system for replacing a general hard disk may be a solid state hard disk or a reverse And the flash memory hard disk or a flash memory. The solid state hard disk, the anti-gate flash memory hard disk or the flash memory includes at least one anti-gate flash memory chip, and each of the anti-gate flash memory chips includes a plurality of data storage areas. Piece.

The above-mentioned data storage system replacing the general hard disk may also use at least one The gate flash memory chip and at least one data storage controller are replaced, wherein each of the anti-gate flash memory chips includes a plurality of data storage blocks.

The aforementioned general standard independent hard disk redundant array architecture includes, for example, RAID 0 Or the architecture of RAID 5.

The above is a specific embodiment of the present invention and the technical means applied thereto, Many variations and modifications can be derived from the present disclosure or the teachings of the present invention. The equivalents of the present invention are still considered to be equivalent to the spirit of the present invention. In the technical scope of the present invention, it is first known.

In summary, the present invention can achieve the pre-invention according to the above disclosure. The purpose is to provide a method for making the anti-gate flash memory have read performance, increase the number of data readable, reduce the frequency of data storage blocks being erased, enhance the reliability of data reading and avoid data length. The time has not been read and damaged, and it is very expensive to use in the industry.

A1, A2‧‧‧ steps

B1, B2, B3, B4‧‧‧ steps

Claims (10)

A method for making a reverse flash memory having read performance, wherein the inverse flash memory comprises a plurality of data storage blocks, each data storage block comprising a plurality of data storage pages, wherein the method is A reading method comprises the following steps: A1: reading one of the data storage pages in a data storage block, wherein the data storage page is read one time after being read. The number of times; and A2: when an update condition is established, the process of step A2 is ended; otherwise, the steps including B1 to B4 are performed, wherein the update condition is: the number of read times is less than the threshold of a number of read times, wherein the steps B1 to B4 are as follows :B1: Copying the entire block data stored in the data storage block to another data storage block in the blank, wherein the full block data is all data stored in the data storage block. All the information in the page; B2: update a flash conversion layer mapping table, so that the storage address of the full block data is linked to the other data storage block; B3: erasing and storing in the data storage block The whole block Material B4: Set the number of readings to 0. For example, in the method of claim 1, the method for making the anti-gate flash memory have read performance, wherein the step A2 further comprises calculating an error correction code of the data, wherein the update condition is: the error correction code It is less than an error correction code threshold and the number of read times is less than the threshold of the read number. A method for making a reverse flash memory having read performance, wherein the inverse flash memory comprises a plurality of data storage blocks, each data storage block comprising a plurality of data storage pages, each data storage The block has a block number, and the method is a read screening method. The method includes the following steps: C1: the data storage block having a starting block number is a data storage block in the screening; C2: performing the D1~D2 step in the data storage block in the screening: D1: reading One of the data storage pages of the data storage block in the screening, wherein the data storage page is read a number of times after being read; D2: when an update condition is established, Then, the step C3 is performed; otherwise, the steps including E1~E4 are performed, wherein the update condition is: the read count is less than a read count threshold, wherein the E1~E4 steps are as follows: E1: the data stored in the screening is stored. The entire block data of one of the blocks is copied to another data storage block of the blank, wherein the full block data is all the data stored in all the data storage pages of the data storage block in the screening; E2: Updating a flash translation layer mapping table to link the storage address of the full block data to the another data storage block; E3: erasing the full block data stored in the data storage block of the screening; E4: Set the number of readings to 0 and execute Step C3: C3: Add 1 to the block number of the data storage block in the screening; and C4: When the current block number is equal to an ending block number, perform the C1 step. Otherwise, the data storage block having the current block number is the data storage block in the screening and the C2 step is performed. For example, in the method of claim 3, the anti-gate flash memory has a read multi-performance method, wherein the D2 step further comprises calculating an error correction code of the data, wherein the update condition is: the error correction code It is less than an error correction code threshold and the number of read times is less than the threshold of the read number. For example, in the method of claim 3, the anti-gate flash memory has a read multi-performance method, wherein the data storage page of the data storage block in the screening is sequentially or randomly selected in the D1 step. The method for making the anti-gate flash memory have read performance as described in any one of claims 1 to 5, wherein the anti-gate flash memory system is an independent data storage redundancy. The array, the redundant array of independent data storages is formed by replacing at least one data storage body with a plurality of hard disks in a standard independent hard disk redundant array, wherein the data storage system is a solid state hard disk, a reverse gate a flash memory hard disk or a flash memory, the solid state hard disk, the reverse flash memory hard disk or the flash memory body including at least one anti-gate flash memory chip, the plurality of data A storage block is included in each of the anti-gate flash memory chips. The method for making the anti-gate flash memory have read performance as described in any one of claims 1 to 5, wherein the anti-gate flash memory system is an independent data storage redundancy. The array, the redundant array of independent data storages is formed by replacing at least one data storage body with a plurality of hard disks in a standard independent hard disk redundant array, wherein the data storage system is composed of at least one anti-gate flash memory The chip and the at least one data storage controller are configured, and the plurality of data storage blocks are included in each of the at least one anti-gate flash memory chips. A method for making a reverse flash memory having read performance, wherein the inverse flash memory comprises a plurality of data storage blocks, each data storage block comprising a plurality of data storage pages, the method comprising the following Step: reading a data by a reading method as described in claim 1 of the patent application; and checking each data storage block one by one and cyclically by a reading screening method as described in claim 4 of the patent application scope . For example, in the method of claim 8, the anti-gate flash memory has a read multi-performance method, wherein the anti-gate flash memory system is a redundant array of independent data storage, and the independent data storage is redundant. The remaining array is formed by replacing at least one data storage body with a plurality of hard disks in a standard independent hard disk redundant array, wherein the data storage system is a solid state hard disk, a reverse flash memory hard disk or a The flash memory, the solid-state hard disk, the anti-gate flash memory hard disk or the flash memory includes at least one anti-gate flash memory chip, and the plurality of data storage blocks are included in each The reverse gate flash memory chip. For example, in the method of claim 8, the anti-gate flash memory has a read multi-performance method, wherein the anti-gate flash memory system is a redundant array of independent data storage, and the independent data storage is redundant. The remaining array is formed by replacing at least one data storage body with a plurality of hard disks in a standard independent hard disk redundant array, wherein the data storage system is controlled by at least one anti-gate flash memory chip and at least one data storage body The plurality of data storage blocks are included in each of the at least one anti-gate flash memory chips.