CN108418589A - A kind of dynamic codec method of single layer nonvolatile storage - Google Patents
A kind of dynamic codec method of single layer nonvolatile storage Download PDFInfo
- Publication number
- CN108418589A CN108418589A CN201810179013.8A CN201810179013A CN108418589A CN 108418589 A CN108418589 A CN 108418589A CN 201810179013 A CN201810179013 A CN 201810179013A CN 108418589 A CN108418589 A CN 108418589A
- Authority
- CN
- China
- Prior art keywords
- data
- compression
- single layer
- nonvolatile storage
- coding
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Classifications
-
- H—ELECTRICITY
- H03—ELECTRONIC CIRCUITRY
- H03M—CODING; DECODING; CODE CONVERSION IN GENERAL
- H03M13/00—Coding, decoding or code conversion, for error detection or error correction; Coding theory basic assumptions; Coding bounds; Error probability evaluation methods; Channel models; Simulation or testing of codes
- H03M13/35—Unequal or adaptive error protection, e.g. by providing a different level of protection according to significance of source information or by adapting the coding according to the change of transmission channel characteristics
- H03M13/353—Adaptation to the channel
Abstract
The invention discloses a kind of dynamic codec methods of single layer nonvolatile storage, belong to data encoding and decoding technique field.The present invention first compresses the cache line data of write-in；The space saved again to compression calculates, according to section space-efficient size dynamic select, the coding mode best to single layer nonvolatile storage encoding efficiency encodes data using the space that compression is saved, and finally the data after coding are written in single layer non-volatile memory array again.The invention also achieves the corresponding coding/decoding methods of the application dynamic coding mode simultaneously.The method of the present invention is reduced under minimum space expense and is write to single layer nonvolatile storage, to reduce single layer nonvolatile storage write energy consumption, promoted the service life, thus solve the problems, such as in existing encoding and decoding technique for reduce memory energy consumption of writing increase space expense.
Description
Technical field
The invention belongs to data encoding and decoding technique fields, more particularly, to a kind of dynamic of single layer nonvolatile storage
Decoding method.
Background technology
Novel nonvolatile storage (NVM, Non-Volatile Memory), such as phase transition storage (PCM, Phase
Change Memory) memristor (RRAM, Resistive RAM), have data non-volatile, high compared to traditional DRAM technology
The advantage of integrated level, low read latency is expected to the memory for replacing traditional DRAM to be used as next generation computer system.PCM uses phase
Become material such as Ge2Sb2Te2(GST) crystalline state and amorphous variation carrys out logical value storage 0 and 1.When GST is in crystalline state,
PCM cell is in the state of low-resistance value, at this time PCM logical value storages 1.When GST is in amorphous state, PCM cell is in height
The state of resistance value, at this time PCM logical value storages 0.When to PCM cell apply a duration very short (about 40ns) but width
When being worth very high current impulse, PCM cell moment can be heated to 600 degrees Celsius, and PCM cell can be in height after material cooling
The state of resistance (logical value 0)；It is longer when applying a duration to PCM cell, and when the lower pulse of intensity, PCM
Unit can be in the state of low resistance (logical value 1).During repeated heating, PCM material can be damaged, when writing 106
To 109After secondary, PCM cell can be trapped in 0 or 1 state and cannot change.The energy consumption of writing of each units of PCM is about
20pJ is 2 times of DRAM.RRAM has the characteristic similar with PCM, and the abrasion number of RRAM device is 10 at present10Hereinafter, and
It to write energy consumption also be several times of DRAM.
PCM and RRAM all suffers from height and writes energy consumption, the problem of writing number or restricted lifetime.How novel non-volatile deposit is reduced
The service life of reservoir write energy consumption, while promoting them makes very crucial problem.
Existing method proposes that the bit flipping for writing NVM is reduced by way of data encoding (to be written as 1 from 0, write from 1
For 0).A kind of method proposed earliest is called Flip-N-Write.The core concept of this method is distributed to every N of data
One flag bit.Flag bit initial value is 0.If writing this N data and the required bit flipping of 1 flag bit being more than
(N+1)/2, then the data step-by-step by this N negates, while being 1 by mark position, be then written.Flip-N-Write can be with
Bit flipping is reduced, while the flag bit of its introducing will also result in the capacity overhead of 1/N.Flip-N-Write reduces bit flipping
Effect is limited to the expense of flag bit.When N is equal to 16, capacity overhead caused by flag bit is 1/16, while Flip-N-
Write can reduce by 14.6% bit flipping.When N is equal to 2, the capacity overhead that flag bit is brought is 50%, but Flip-N-
Write can reduce by 25% bit flipping.Also certain methods are that the data arrangement of cache lines is known as to the matrix of N rows M row, so
Flip-N-Write is used to row and column simultaneously afterwards.Also there are the more rules of number write for dsc data, by Flip-N-
Write is applied to reduce writing for dsc data, but these methods can all bring a large amount of additional spaces.In addition, FlipMin is encoded
Mode is encoded based on coset：Increase k redundant digits, while it includes 2 that data, which are mapped as one,kThe set of a element, often
It is secondary write before, from this 2kA kind of write-in that bit flipping is minimum is picked out in the set of a element, same this coding mode also can
Bring a large amount of space expense.
At present still nonvolatile storage can be reduced without a kind of coding mode that can be directed to single layer nonvolatile storage
Write energy consumption, promote the service life, and do not increase excessive space expense.
Invention content
For the disadvantages described above or Improvement requirement of the prior art, the present invention provides a kind of the dynamic of single layer nonvolatile storage
State decoding method, its object is to combine data compression and data encoding, it is suitable dynamically to be selected according to compression ratio
Coding mode, while will compress saved space be used as coding flag bit, under minimum space expense reduce pair
Single layer nonvolatile storage is write, to reduce single layer nonvolatile storage write energy consumption, promoted the service life, thus solve existing volume
In decoding technique the problem of energy consumption increases space expense is write for reduction memory.
To achieve the above object, described the present invention provides a kind of dynamic codec method of single layer nonvolatile storage
Method includes coding method and coding/decoding method：
The coding method includes the following steps：
(1) cache line data of write-in is compressed；
(2) if cache line data cannot compress, cache line data is sent directly to writing controller, is terminated；Otherwise into
Enter step (3)；
(3) if cache lines compression after saved space S >=(T-N × P)/2, select FlipMin coding modes pair
Compressed data is encoded；Otherwise selection Flip-N-Write coding modes encode compressed data；Wherein T indicates caching
Capable digit presets 512；N indicates the number of words of each cache lines, presets 8；P is the digit of prefix bit, presets 3；
(4) data after coding are written in non-volatile memory array.
Further, the step (1) specifically includes：
(11) cache line data is grouped as unit of word, the size of predetermined word is 64；And to each word respectively into
Row compression；
(12) whether setting compression flag bit record buffer memory row data can be compressed, if there is word that can compress in the cache lines,
The cache line data, which is denoted as, to be compressed, and note 1, otherwise the cache line data, which is denoted as, to compress, and note 0 compresses the size of flag bit
It is preset as 1；
(13) it is each word setting prefix bit recording compressed information in the cache lines that can compress, the size of prefix bit is preset
It is 3.
Further, the information of word includes after recording compressed in the step (13)：After compression before the size of word and compression
The pattern of word.
Further, the preferred FPC compression algorithms of the compression algorithm.
Further, flag bit, prefix bit will be compressed after step (1) the cache line data compression and encode later number
It is continuously stored according to big-endian.
Further, the step (2) if in cache line data cannot compress, will compression flag bit and cache line data
Big-endian is sent to writing controller after continuously storing.
Further, saved space after cache lines are compressed in the step (3)
Wherein, CiFor the digit of i-th of compression word.
Further, in the step (3)：
If 1≤S ＜ (T-N × P)/3, select Flip-N-Write coding modes to encode compressed data, per M
Data bit allocation one flag bit,
If (T-N × P)/3≤S ＜ (T-N × P)/2, Flip-N-Write coding modes is selected to carry out compressed data
Coding, the flag bit of every 2 data bit allocations one.
Further, the step (4) judges whether there is the legacy data stored in unit before data are specially written and will
The new data of write-in is the same, and the unit need not execute write operation if having, and otherwise execute write operation.
Further, the coding/decoding method, specifically includes：
S1, compression flag bit is read from non-volatile memory cells and compresses later cache line data；
S2, judge whether cache lines are compressed and encoded according to compression flag bit；
If S3, cache lines are uncompressed, the cache line data read out need not be decompressed or be decoded；If cache lines quilt
Compression, then read the prefix bit in cache lines, the saved space of compression is determined by prefix bit, so that it is determined that cache line data
Used coding method is decoded cache line data using corresponding coding/decoding method.
In general, through the invention it is contemplated above technical scheme is compared with the prior art, have following technology special
Sign and advantageous effect：
(1) present invention combines data compression and data encoding, will compress saved space and be used as coding, can
To promote the service life of single layer nonvolatile storage under minimum space expense (0.2%), while reducing single layer non-volatile memory
The energy consumption of device；
(2) present invention selects a kind of coding mode of best results according to the Spatial distributions that compression is saved, and makes full use of pressure
The best coding mode of spatial choice effect that casing coupling saves promotes the service life of single layer nonvolatile storage and reduces energy to greatest extent
Consumption；
(3) present invention employs the mode that reading comparison is write, writing for redundancy is can be avoided, can reduce and write energy consumption.
Description of the drawings
Fig. 1 is the implementation process flow chart of coding method of the present invention；
Fig. 2 is the arrangement mode of cache lines after data compression in the present invention；
Fig. 3 is data decoding schematic diagram of the present invention；
Fig. 4 is the schematic diagram of FlipMin codings.
Specific implementation mode
In order to make the purpose , technical scheme and advantage of the present invention be clearer, with reference to the accompanying drawings and embodiments, right
The present invention is further elaborated.It should be appreciated that described herein, specific examples are only used to explain the present invention, not
For limiting the present invention.As long as in addition, technical characteristic involved in the various embodiments of the present invention described below that
Conflict is not constituted between this to can be combined with each other.
Following embodiment combination attached drawing further introduces the present invention：
In this implementation column, higher level's caching system is write to memory as unit of cache lines, and typical cache line size is 64
Byte (64B), therefore this patent assumes that the size of cache lines is 64B.The size of word is 64 (64b), compression side in this patent
Method can select other arbitrary compression methods, and this patent embodiment is by taking FPC compression algorithms as an example.
As shown in Figure 1, the coding method of the embodiment of the present invention includes：
To every write request, it is first determined whether can compress.Writing controller is sent directly to if it cannot compress, and
It is final to be written to storage array, while being 0 by compression mark position.
If can compress, the cache line data of write request is compressed first, then calculates compressed data
Digit and the saved space digit of compression, while being 1 by compression mark position.Assuming that compressing saved space digit
For S, it is D to compress later data bits, then is encoded according to following rule to compressing later data：
If S<163, then it is encoded using Flip-N-Write to compressing later data bit, while to per D/S
Data bit allocation one flag bit；
If 163<=S<244, then it is encoded using Flip-N-Write to compressing later data bit, while to every
The flag bit of 2 data bit allocations one；
If 244=<S is then encoded using FlipMin to compressing later data bit.
Later data bit will be encoded and be sent to writing controller, and be finally written to storage array.
As shown in Fig. 2, for the cache lines that can be compressed, compression flag bit is 1, and each word uses 3 prefixes
Position identifies the type of compression.And prefix bit continuous storage since a high position for this cache lines.Later word is compressed from prefix
Position starts continuous storage, and for incompressible cache lines, compression flag bit is 0, and each word does not use prefix bit.
As shown in figure 3, coding/decoding method includes：
Judge whether the data of cache lines are compressed according to compression flag bit first, if compression flag bit position 0, then it represents that
Cache lines are uncompressed, need not be to caching and being decoded.If it is 1 to compress flag bit, then it represents that cache lines are compressible, need
Cache lines are decoded.
Decoded method is corresponding with the method for coding.Preceding 24 prefix bits are read first, can be counted according to prefix bit
It calculates the size of data bit after obtaining compression and compresses saved space.According to the size of data bit after compression and compression
Saved space can be determined that the coding mode that cache line data uses, and then use corresponding coding/decoding method Flip-N-
Write is decoded or FlipMin decodings are decoded data bit.Each compression can be obtained according to prefix bit after decoding
Word.The data of each word and entire cache lines can be obtained by finally being unziped it respectively to each word using FPC decompressions.
Separately introduce the other technologies occurred in embodiment：
Data compression technique, Frequent Pattern Compression (FPC) can be by data pressure in the present embodiment
It contracts to reduce the digit for needing to be written.It carrys out the rule of mark data using three prefix bits, to reduce data storage
Required space.FPC is compressed as unit of 64 words, can compress 7 kinds of different data patterns.As shown in table 1.
Table 1
The first pattern is the data of 64 full 0s, and such data can be indicated by 3 prefix bits.The first
64 full 0s are compressed to 3 prefix bits in pattern.In second of pattern, high 56 of a word are 0, and least-significant byte is variable,
This pattern is indicated by prefix bit 001.It is that prefix bit adds variable least-significant byte, therefore total number to compress later data
It it is 11 according to digit.In addition in addition both also have compressible data pattern different in 5.For different data patterns,
The space that FPC can be saved is variation.
Flip-N-Write encoding and decoding techniques：
The Flip-N-Write cataloged procedures are specially：
The flag bit being written into is initialized as 0, and the digit of digit and old flag bit is all N；It is written into N × K
K data of n-th in the cache line data of position are combined with N old flag bits, the data of composition K+1；Statistics wherein 0
Number, if 0 number is more than (K+1)/2, the nth position of the mark being written into is 0, the mark being otherwise written into
Nth position is 1；The flag bit to be written to all cache line datas and every executes above procedure successively.
The Flip-N-Write decoding process is specially：
Old cache line data N × K and old flag bit N are read, K data of n-th of cache lines correspond to N
Flag bit, if the flag bit is 0, corresponding K of data do not execute any operation, and the otherwise data execution to K negates
Operation；Above procedure is executed to all data successively.
FlipMin encoding and decoding techniques：
As shown in figure 4, the FlipMin cataloged procedures are specially：
Any one generator matrix G of RM (1,3) codings is found first, and finds out the left inverse matrix G# of G.Using it is all can
The information bit of energy, respectively binary zero 000-1111 are multiplied by generator matrix and obtain Coset respectively0(No. 0 coset) set pair
All code words answered.All code words are set as C0, C1... C15.To the information bit d that will be arbitrarily written, d is multiplied by G# first and is obtained
To corresponding Coset (the i.e. Coset of dd, d cosets) in an element.Then this element and the code word that has been written into is different
Or, obtaining transmission element.By this transmission element and Coset0Gather corresponding all code word exclusive or, so that it may to obtain transmission unit
The set of element.One 1 minimum element of number is found inside this set as Coset leader, then by Coset
The data exclusive or that leader has been written into can be obtained by coding vector later to be written.
Further, the FlipMin decoding process is specially：
The vector for encoding later is multiplied with generator matrix, obtained element is the element needed, so that it may to realize solution
Code.
The above content as it will be easily appreciated by one skilled in the art that the foregoing is merely illustrative of the preferred embodiments of the present invention,
Be not intended to limit the invention, all within the spirits and principles of the present invention made by all any modification, equivalent and improvement etc.,
It should all be included in the protection scope of the present invention.
Claims (10)
1. a kind of dynamic codec method of single layer nonvolatile storage, including coding method and coding/decoding method, which is characterized in that
The coding method includes the following steps：
(1) cache line data of write-in is compressed；
(2) if cache line data cannot compress, cache line data is sent directly to writing controller, is terminated；Otherwise enter step
Suddenly (3)；
(3) if cache lines compression after saved space S >=(T-N × P)/2, select FlipMin coding modes to compression
Data are encoded；Otherwise selection Flip-N-Write coding modes encode compressed data；Wherein T indicates cache lines
Digit；N indicates the number of words of each cache lines；P is the digit of prefix bit；
(4) data after coding are written in non-volatile memory array.
2. a kind of dynamic codec method of single layer nonvolatile storage according to claim 1, which is characterized in that described
Step (1) specifically includes：
(11) cache line data is grouped as unit of word；And each word is compressed respectively；
(12) whether setting compression flag bit record buffer memory row data can be compressed, if there is word that can compress in the cache lines, this is slow
It deposits row data and is denoted as and can compress, otherwise the cache line data, which is denoted as, to compress；
(13) it is each word setting prefix bit recording compressed information in the cache lines that can compress.
3. a kind of dynamic codec method of single layer nonvolatile storage according to claim 2, which is characterized in that described
The information of word includes after recording compressed in step (13)：The pattern of the size of word and the preceding word of compression after compression.
4. a kind of dynamic codec method of single layer nonvolatile storage according to claim 1 or 2, which is characterized in that
The compression algorithm is FPC.
5. a kind of dynamic codec method of single layer nonvolatile storage according to claim 1 or 2, which is characterized in that
It will compress flag bit, prefix bit after step (1) the cache line data compression and encode later data big-endian and connect
It renews and puts.
6. a kind of dynamic codec method of single layer nonvolatile storage according to claim 1 or 2, which is characterized in that
The step (2) if in cache line data cannot compress, will compression flag bit and cache line data big-endian it is continuous
Writing controller is sent to after storage.
7. a kind of dynamic codec method of single layer nonvolatile storage according to claim 1, which is characterized in that described
Saved space after cache lines compression in step (3)
Wherein, CiFor the digit of i-th of compression word.
8. a kind of dynamic codec method of single layer nonvolatile storage according to claim 1 or claim 7, which is characterized in that
In the step (3)：
If 1≤S ＜ (T-N × P)/3, select Flip-N-Write coding modes to encode compressed data, the number per M
According to one flag bit of bit allocation,
If (T-N × P)/3≤S ＜ (T-N × P)/2, select Flip-N-Write coding modes to encode compressed data,
The flag bit of every 2 data bit allocations one.
9. a kind of dynamic codec method of single layer nonvolatile storage according to claim 1, which is characterized in that described
Step (4) is specially to judge whether there is the legacy data stored in unit as the new data by write-in before data are written, if having
Then the unit need not execute write operation, otherwise execute write operation.
10. a kind of dynamic codec method of single layer nonvolatile storage according to claim 1, which is characterized in that institute
Coding/decoding method is stated, is specifically included：
S1, compression flag bit is read from non-volatile memory cells and compresses later cache line data；
S2, judge whether cache lines are compressed and encoded according to compression flag bit；
If S3, cache lines are uncompressed, the cache line data read out need not be decompressed or be decoded；If cache lines are pressed
Contracting, then read the prefix bit in cache lines, the saved space of compression is determined by prefix bit, so that it is determined that cache line data institute
The coding method of use is decoded cache line data using corresponding coding/decoding method.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810179013.8A CN108418589B (en) | 2018-03-05 | 2018-03-05 | Dynamic coding and decoding method for single-layer nonvolatile memory |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810179013.8A CN108418589B (en) | 2018-03-05 | 2018-03-05 | Dynamic coding and decoding method for single-layer nonvolatile memory |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN108418589A true CN108418589A (en) | 2018-08-17 |
| CN108418589B CN108418589B (en) | 2020-07-10 |
Family
ID=63130046
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810179013.8A Active CN108418589B (en) | 2018-03-05 | 2018-03-05 | Dynamic coding and decoding method for single-layer nonvolatile memory |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108418589B (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111538677A (en) * | 2020-04-26 | 2020-08-14 | 西安万像电子科技有限公司 | Data processing method and device |
Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US6301152B1 (en) * | 1999-05-12 | 2001-10-09 | Stmicroelectronics S.R.L. | Non-volatile memory device with row redundancy |
| US6477082B2 (en) * | 2000-12-29 | 2002-11-05 | Micron Technology, Inc. | Burst access memory with zero wait states |
| CN1499531A (en) * | 2002-10-28 | 2004-05-26 | Method and device for management of data integrity in non-volatile memory system | |
| EP1403879B1 (en) * | 2002-09-30 | 2010-11-03 | STMicroelectronics Srl | Method for replacing failed non-volatile memory cells and corresponding memory device |
| US20100329011A1 (en) * | 2009-06-30 | 2010-12-30 | Peter Wung Lee | Memory system having nand-based nor and nand flashes and sram integrated in one chip for hybrid data, code and cache storage |
| WO2011146364A3 (en) * | 2010-05-17 | 2012-01-26 | Seagate Technology Llc | Joint encoding of logical pages in multi-page memory architecture |
| CN103140894A (en) * | 2010-08-17 | 2013-06-05 | 技术研究及发展基金公司 | Mitigating inter-cell coupling effects in non volatile memory (nvm) cells |
| CN105427887A (en) * | 2014-09-15 | 2016-03-23 | Hgst荷兰有限公司 | Encoding scheme for 3D vertical flash memory |
| CN107622781A (en) * | 2017-10-12 | 2018-01-23 | 华中科技大学 | A kind of decoding method for lifting three layers of memristor write performance |
-
2018
- 2018-03-05 CN CN201810179013.8A patent/CN108418589B/en active Active
Patent Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US6301152B1 (en) * | 1999-05-12 | 2001-10-09 | Stmicroelectronics S.R.L. | Non-volatile memory device with row redundancy |
| US6477082B2 (en) * | 2000-12-29 | 2002-11-05 | Micron Technology, Inc. | Burst access memory with zero wait states |
| EP1403879B1 (en) * | 2002-09-30 | 2010-11-03 | STMicroelectronics Srl | Method for replacing failed non-volatile memory cells and corresponding memory device |
| CN1499531A (en) * | 2002-10-28 | 2004-05-26 | Method and device for management of data integrity in non-volatile memory system | |
| US20100329011A1 (en) * | 2009-06-30 | 2010-12-30 | Peter Wung Lee | Memory system having nand-based nor and nand flashes and sram integrated in one chip for hybrid data, code and cache storage |
| WO2011146364A3 (en) * | 2010-05-17 | 2012-01-26 | Seagate Technology Llc | Joint encoding of logical pages in multi-page memory architecture |
| CN103140894A (en) * | 2010-08-17 | 2013-06-05 | 技术研究及发展基金公司 | Mitigating inter-cell coupling effects in non volatile memory (nvm) cells |
| CN105427887A (en) * | 2014-09-15 | 2016-03-23 | Hgst荷兰有限公司 | Encoding scheme for 3D vertical flash memory |
| CN107622781A (en) * | 2017-10-12 | 2018-01-23 | 华中科技大学 | A kind of decoding method for lifting three layers of memristor write performance |
Non-Patent Citations (2)
| Title |
|---|
| ADAM N. JACOBVITZ等: ""Coset Coding to Extend the Lifetime of Memory"", 《2013 IEEE 19TH INTERNATIONAL SYMPOSIUM ON HIGH PERFORMANCE COMPUTER ARCHITECTURE (HPCA)》 * |
| 董伟: ""基于NVM的写操作优化策略研究与设计"", 《中国优秀硕士学位论文全文数据库•信息科技辑》 * |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111538677A (en) * | 2020-04-26 | 2020-08-14 | 西安万像电子科技有限公司 | Data processing method and device |
| CN111538677B (en) * | 2020-04-26 | 2023-09-05 | 西安万像电子科技有限公司 | Data processing method and device |
Also Published As
| Publication number | Publication date |
|---|---|
| CN108418589B (en) | 2020-07-10 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111326199B (en) | Multi-state programming for memory devices | |
| US11636039B2 (en) | Mapping for multi-state programming of memory devices | |
| JP5785330B2 (en) | Apparatus and method for operating for memory endurance | |
| US8650461B2 (en) | Adaptive over-provisioning in memory systems | |
| US20150067436A1 (en) | Nonvolatile Memory System Compression | |
| US20110138104A1 (en) | Multi-write coding of non-volatile memories | |
| KR20140060285A (en) | Systems and methods of storing data | |
| CN112486725B (en) | Method and device for carrying out error correction coding on compressed data | |
| WO2008145070A1 (en) | Flash memory data read/write processing method | |
| US11057060B1 (en) | Method and apparatus for matrix flipping error correction | |
| CN104778018A (en) | Broad-strip disk array based on asymmetric hybrid type disk image and storage method of broad-strip disk array | |
| US8161343B2 (en) | Nibble encoding for improved reliability of non-volatile memory | |
| US8463982B2 (en) | Method of storing and accessing error correcting code in NAND flash | |
| US11817883B2 (en) | Variable length ECC code according to value length in NVMe key value pair devices | |
| CN108418589A (en) | A kind of dynamic codec method of single layer nonvolatile storage | |
| CN113838510A (en) | Fast replication via controller | |
| CN107622781B (en) | Coding and decoding method for improving writing performance of three-layer memristor | |
| US11934264B2 (en) | ECC parity biasing for Key-Value data storage devices | |
| CN109189694B (en) | SCM data coding method and data storage method | |
| CN114203250B (en) | Data storage method and data reading method of solid-state memory and solid-state memory | |
| CN108073473A (en) | A kind of flash memory bad list compression method | |
| US11853160B2 (en) | Variable length ECC code according to data entropy in NVMe key value pair devices | |
| KR20180074561A (en) | Data mapping enabling fast read multi-level 3d nand to improve lifetime capacity | |
| CN116343890B (en) | Error correction unit management method, memory control chip and flash memory device | |
| CN115391240A (en) | Mapping table data compression method, controller and memory |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |