A method, system, device and medium for processing memory threshold voltage data

The threshold voltage data of NAND Flash memory is processed through Gaussian fitting and least squares method, which solves the curve jitter and noise problems, optimizes the readability and accuracy of the curve, and provides better data support.

CN118644580BActive Publication Date: 2024-11-12合肥康芯威存储技术有限公司
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202411124781.5
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-08-16
Publication Date
2024-11-12
Estimated Expiration
2044-08-16

AI Technical Summary

Technical Problem

In the prior art, the threshold voltage curve chart of NAND Flash memory has severe jitter and a lot of data noise, which affects the readability and accuracy of the data.

Method used

By obtaining the threshold voltage source data of the flash memory chip, calculating the mean value and segmenting the process to generate voltage peaks, performing Gaussian fitting and least squares calculations, generating threshold voltage fitting data, and weighting the source data and fitting data to reduce noise , optimize the curve.

Benefits of technology

The smoothing of the threshold voltage curve is achieved, the readability of the curve is enhanced, and more accurate data support is provided, providing a basis for further analysis of the threshold voltage.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118644580B_ABST
    Figure CN118644580B_ABST
Patent Text Reader

Abstract

The present invention provides a method, system, device and medium for processing memory threshold voltage data, and relates to the field of static storage technology. The processing method includes: obtaining threshold voltage source data of a flash memory chip; calculating the mean value of the threshold voltage source data, and segmenting the threshold voltage source data according to the mean value to generate multiple voltage peaks, and saving the interval data of each voltage peak as peak data; performing Gaussian fitting processing on the peak data to generate a corresponding Gaussian fitting matrix; performing least squares calculation processing on the Gaussian fitting matrix to generate a parameter solution of the Gaussian fitting matrix, and substituting the parameter solution into the Gaussian fitting matrix to generate threshold voltage fitting data; performing weighted processing on the threshold voltage source data and the threshold voltage fitting data to generate threshold voltage weighted data. The present invention can perform fitting optimization on the VT curve, making the curve smoother and enhancing the readability of the VT curve graph.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of static storage technology, and in particular to a method, system, device and medium for processing memory threshold voltage data. Background Art

[0002] In NAND Flash memory, the state of each storage cell is characterized by its threshold voltage (Vt), which determines the gate voltage required to turn the transistor on or off. Since NAND Flash stores data by changing the number of electrons in the floating gate layer of the storage cell, different numbers of electrons will result in different threshold voltages, thus representing different data states.

[0003] The Vt curve shows the threshold voltage distribution of the memory cell, which is usually presented as a series of voltage peaks (representing concentrated voltage areas of different data states). The Vt curve uses the voltage level as the horizontal axis and the number of cells that are turned on at different voltages as the vertical axis. In the prior art, the Vt curve has severe jitter, and there is a lot of random noise and system noise in the measured data. Therefore, there is room for improvement. Summary of the invention

[0004] The present invention provides a method, system, device and medium for processing memory threshold voltage data, so as to solve the technical problems of serious jitter of Vt curve graph and much data noise in the prior art.

[0005] The present invention provides a method for processing memory threshold voltage data, comprising:

[0006] Acquire threshold voltage source data of a flash memory chip, wherein the threshold voltage source data represents the number of memory cells of the flash memory chip that are turned on at each voltage level;

[0007] Calculating the mean value of the threshold voltage source data, and segmenting the threshold voltage source data according to the mean value to generate a plurality of voltage peaks, and saving the interval data of each voltage peak as segmented peak data;

[0008] Performing Gaussian fitting processing on the peak data to generate a corresponding Gaussian fitting matrix;

[0009] Performing least squares calculation processing on the Gaussian fitting matrix to generate a parameter solution of the Gaussian fitting matrix, and substituting the parameter solution into the Gaussian fitting matrix to generate threshold voltage fitting data, wherein the parameter solution satisfies that the square value of the error between the threshold voltage fitting data and the threshold voltage source data is minimized;

[0010] The threshold voltage source data and the threshold voltage fitting data are weighted to generate threshold voltage weighted data.

[0011] In one embodiment of the present invention, the step of obtaining threshold voltage source data of a flash memory chip, wherein the threshold voltage source data represents the number of memory cells of the flash memory chip that are turned on at each voltage level, comprises:

[0012] Obtaining the threshold voltage of each storage cell in the flash memory chip;

[0013] All the threshold voltages are counted, and the number of memory cells turned on at each voltage level is saved as threshold voltage source data according to a preset format.

[0014] In one embodiment of the present invention, the steps of calculating the mean of the threshold voltage source data, segmenting the threshold voltage source data according to the mean to generate a plurality of voltage peaks, and saving the interval data of each voltage peak as peak segmentation data include:

[0015] Importing the threshold voltage source data into a voltage array, and calculating the unit conduction mean value of all the voltage levels;

[0016] According to the unit conduction mean value, the threshold voltage source data is segmented to generate a plurality of voltage peaks;

[0017] The left boundary value and the right boundary value of the interval where each voltage peak is located are found, and the peak separation data are generated based on the left boundary value and the right boundary value.

[0018] In one embodiment of the present invention, the step of segmenting the threshold voltage source data according to the cell conduction mean value to generate a plurality of voltage peaks comprises:

[0019] Acquire multiple intersection positions of a corresponding curve of the threshold voltage source data and a corresponding straight line of the unit conduction mean value;

[0020] All the intersection positions are traversed to divide the threshold voltage source data into a plurality of voltage peaks.

[0021] In one embodiment of the present invention, in the step of performing Gaussian fitting processing on the peak data to generate a corresponding Gaussian fitting matrix, the Gaussian fitting matrix satisfies the following formula:

[0022]

[0023] in, represents all voltage levels of the threshold voltage, represents the logarithmic value of the number of conducting units corresponding to each voltage level, represents the parameters of the Gaussian fitting matrix, Represents measurement error.

[0024] In one embodiment of the present invention, in the step of performing least squares calculation processing on the Gaussian fitting matrix to generate a parameter solution of the Gaussian fitting matrix, and substituting the parameter solution into the Gaussian fitting matrix to generate threshold voltage fitting data, wherein the parameter solution satisfies that the square value of the error between the threshold voltage fitting data and the threshold voltage source data is the smallest, the square value of the error L between the threshold voltage fitting data and the threshold voltage source data satisfies the following formula:

[0025]

[0026] in, represents the threshold voltage source data, represents the threshold voltage fitting data.

[0027] In one embodiment of the present invention, in the step of weighting the threshold voltage source data and the threshold voltage fitting data to generate threshold voltage weighted data, the threshold voltage weighted data satisfies the following formula:

[0028] Threshold voltage weighted data = threshold voltage fitting data × m + threshold voltage source data × n;

[0029] Wherein, m represents the preset weight of the threshold voltage fitting data, n represents the weight of the threshold voltage source data, and m+n=1 is satisfied.

[0030] The present invention also provides a memory threshold voltage data processing system, comprising:

[0031] A source data acquisition module, used to acquire threshold voltage source data of the flash memory chip, wherein the threshold voltage source data represents the number of memory cells that are turned on in each voltage level of the flash memory chip;

[0032] A multi-peak cutting module, used for calculating the mean value of the threshold voltage source data, and performing segmentation processing on the threshold voltage source data according to the mean value to generate multiple voltage peaks, and saving the interval data of each voltage peak as segmented peak data;

[0033] A Gaussian fitting module, used for performing Gaussian fitting processing on the peak data to generate a corresponding Gaussian fitting matrix;

[0034] a parameter calculation module, used for performing least squares calculation processing on the Gaussian fitting matrix to generate a parameter solution of the Gaussian fitting matrix, and substituting the parameter solution into the Gaussian fitting matrix to generate threshold voltage fitting data, wherein the parameter solution satisfies that the square value of the error between the threshold voltage fitting data and the threshold voltage source data is minimized;

[0035] The weighted processing module is used to perform weighted processing on the threshold voltage source data and the threshold voltage fitting data to generate threshold voltage weighted data.

[0036] The present invention also provides an electronic device, comprising a processor, a memory and a communication bus; the communication bus is used to connect the processor and the memory; the processor is used to execute a computer program stored in the memory to implement a method for processing memory threshold voltage data as described in any one of the above items.

[0037] The present invention further provides a computer-readable storage medium on which a computer program is stored, wherein the computer program is used to enable a computer to execute the method for processing memory threshold voltage data as described in any one of the above items.

[0038] As described above, a method, system, device and medium for processing memory threshold voltage data of the present invention have the following beneficial effects: the present invention can perform fitting optimization on the VT curve, reduce the noise in the data, make the curve smoother, and achieve the enhancement of the readability of the VT curve graph while ensuring that the data does not deviate from the actual situation, thereby providing data support for further analysis of the threshold voltage. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for describing the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative work.

[0040] Figure 1 It is a schematic diagram showing the steps of a method for processing memory threshold voltage data provided in one embodiment of the present invention.

[0041] Figure 2 Display as Figure 1 A schematic flow chart of a specific implementation of step S100.

[0042] Figure 3 FIG. 1 is a diagram showing a voltage distribution of an SLC memory according to an embodiment of the present invention.

[0043] Figure 4 FIG. 1 is a diagram showing a voltage distribution of an MLC memory according to an embodiment of the present invention.

[0044] Figure 5 FIG. 4 shows a voltage distribution diagram of a TLC memory according to an embodiment of the present invention.

[0045] Figure 6 Display as Figure 1A schematic flow chart of a specific implementation of step S200.

[0046] Figure 7 Display as Figure 3 A flowchart of a specific implementation of step S220 is shown in FIG.

[0047] Figure 8 The graph shows a comparison between a curve of threshold voltage source data and a curve of threshold voltage weighted data in one embodiment of the present invention.

[0048] Fig. 9 Shown is a schematic diagram of the structure of a memory threshold voltage data processing system provided in an embodiment of the present invention.

[0049] Fig.10 Shown is a schematic diagram of the structure of a computer device provided in one embodiment of the present invention. DETAILED DESCRIPTION

[0050] The following is an explanation of the embodiments of the present invention by specific examples, and those skilled in the art can easily understand other advantages and effects of the present invention from the contents disclosed in this specification. The present invention can also be implemented or applied through other different specific embodiments, and the details in this specification can also be modified or changed in various ways based on different viewpoints and applications without departing from the spirit of the present invention. It should be noted that the features in the following embodiments and the embodiments can be combined with each other without conflict. It should also be understood that the terms used in the embodiments of the present invention are intended to describe specific embodiments, rather than to limit the scope of protection of the present invention. The test methods for which specific conditions are not specified in the following examples are usually carried out under conventional conditions or according to the conditions recommended by the manufacturers.

[0051] It should be noted that the structures, proportions, sizes, etc. illustrated in the drawings of this specification are only used to match the contents disclosed in the specification for people familiar with this technology to understand and read, and are not used to limit the limiting conditions for the implementation of the present invention, so they have no substantial technical significance. Any modification of the structure, change in the proportional relationship, or adjustment of the size should still fall within the scope of the technical content disclosed by the present invention without affecting the effects and purposes that can be achieved by the present invention. At the same time, the terms such as "upper", "lower", "left", "right", "middle" and "one" quoted in this specification are only for the convenience of description, and are not used to limit the scope of the implementation of the present invention. The change or adjustment of their relative relationship should also be regarded as the scope of the implementation of the present invention without substantially changing the technical content.

[0052] First of all, it should be noted that in the NAND flash Vt curve graph, the number of conductive memory cells refers to the number of cells used to store data in the memory. Each NAND flash device consists of multiple memory cells, and each memory cell can store one or more bits of data (0 or 1). These memory cells represent different data states by being turned on and off. The number of conductive memory cells represents the number of cells available for storing data in a NAND flash device, that is, the number of bits that can be stored. For SLC (Single-Level Cell) memory, two voltages are required to write data, which can be set as V0 and V1, where V0 < V1. Theoretically, the number of electrons written at voltage V0 is y0, and the number of electrons written at voltage V1 is y1. Due to the quantum tunneling effect, it is actually difficult to ensure that the number of electrons written at voltage V0 is exactly y0 and the number of electrons written at voltage V1 is exactly y1. At voltage V0, the number of electrons written in most memory cells is near y0, that is, the observed values of the number of electrons in most of these memory cells gather near the central voltage peak (mean / average value) with the highest occurrence probability. They spread out on both sides of the central value. When reading with the voltage Vx corresponding to the number of electrons y0, the cell is conductive. Centered on voltage Vx, the number of conductive cells read at the voltages on both sides also spreads around y0, so it follows a normal distribution. Similarly, the Vt curve graph of TLC memory also follows a normal distribution, and the Vt curve graph of TLC memory has 8 voltage peaks. A memory cell in SLC memory only needs to represent one bit of data and needs to distinguish between 0 and 1, so only one reference voltage is required. Similarly, MLC memory needs to use 3 reference voltages for distinction, and TLC memory needs to use 7 reference voltages for distinction. The discrimination voltage can be understood as the valley voltage value in the Vt curve graph.

[0053] Please refer to Figures 1 to 10 , the present invention provides a method, system, device and medium for processing memory threshold voltage data. The present invention adopts the ideas of Gaussian fitting and the least squares method to perform curve smoothing and data noise reduction processing, which can make the Vt curve graph under the actual voltage approach the ideal value, the curve becomes smoother, and the curve jitter is reduced. The following will be described in detail through specific embodiments.

[0054] Please refer to Figure 1 , the present invention provides a method for processing memory threshold voltage data. Specifically, it may include the following steps:

[0055] Step S100: Obtain the threshold voltage source data of the flash chip, and the threshold voltage source data characterizes the number of conductive memory cells of the flash chip at each voltage level;

[0056] Step S200, calculating the mean value of the threshold voltage source data, and segmenting the threshold voltage source data according to the mean value to generate a plurality of voltage peaks, and saving the interval data of each voltage peak as segmented peak data;

[0057] Step S300, performing Gaussian fitting processing on the peak data to generate a corresponding Gaussian fitting matrix;

[0058] Step S400, performing least squares calculation processing on the Gaussian fitting matrix to generate a parameter solution of the Gaussian fitting matrix, and substituting the parameter solution into the Gaussian fitting matrix to generate threshold voltage fitting data, wherein the parameter solution satisfies that the square value of the error between the threshold voltage fitting data and the threshold voltage source data is minimized;

[0059] Step S500: weighting the threshold voltage source data and the threshold voltage fitting data to generate threshold voltage weighted data.

[0060] See also Figure 2 In one embodiment of the present invention, when executing step S100, the threshold voltage source data of the flash memory chip is obtained, and the threshold voltage source data represents the number of memory cells that are turned on in each voltage level of the flash memory chip. Specifically, the following steps may be included:

[0061] Step S110, obtaining the threshold voltage of each storage unit in the flash memory chip;

[0062] Step S120 , counting all threshold voltages, and saving the number of memory cells turned on at each voltage level as threshold voltage source data according to a preset format.

[0063] In one embodiment of the present invention, when executing step S110 to step S120, specifically, the Vt value (threshold voltage) of each storage cell of the page or block in the NAND flash memory can be first obtained through the Vt tool. Then, all Vt values ​​are counted to generate cells-Vt values. In this embodiment, the cells-Vt value represents the number of conductive storage cells corresponding to each voltage level under a set of increasing voltage levels. For example, the voltage level can be set as the horizontal axis, and the number of conductive cells corresponding to each voltage level can be set as the vertical axis. The obtained cells-Vt value is then saved as a preset format file. The preset format can be a csv file format.

[0064] See also Figure 3-Figure 6 In one embodiment of the present invention, when executing step S200, the mean value of the threshold voltage source data is calculated, and the threshold voltage source data is segmented according to the mean value to generate multiple voltage peaks, and the interval data of each voltage peak is saved as peak data. Specifically, step S200 may include the following steps:

[0065] Step S210, importing the threshold voltage source data into the voltage array, and calculating the cell conduction mean value of all voltage levels;

[0066] Step S220, segmenting the threshold voltage source data according to the cell conduction mean value to generate a plurality of voltage peaks;

[0067] Step S230, searching for the left boundary value and the right boundary value of the interval where each voltage peak is located, and generating peak data based on the left boundary value and the right boundary value.

[0068] In one embodiment of the present invention, when executing step S210, the threshold voltage source data is imported into the voltage array, and the average value of the cell conduction of all voltage levels is calculated. Specifically, first, the number of memory cells that are turned on at each voltage level in the threshold voltage source data can be stored in the array in the form of a vertical coordinate. For example, the array can be set , the array size can be adjusted as needed. Then, it can be seen that if the Vt curve is a single voltage peak, it is only necessary to fit all the data into a normal distribution curve. Figure 3-5 As shown, the Vt curve of the SLC memory has 2 voltage peaks, the Vt curve of the MLC memory has 4 voltage peaks, and the Vt curve of the TLC memory has 8 voltage peaks. In addition, if there is a double program problem, the voltage peak of the Vt curve will also have problems such as left bias or right bias. Therefore, the present invention can cut the threshold voltage source data into a corresponding number of voltage peaks according to the memory type. The voltage peak represents the concentrated voltage area of ​​the storage unit with different data states in the memory. In this embodiment, TLC is used as an example for explanation. When performing data processing, the average value of the unit conduction of all the voltage levels can be calculated, that is, the average value of the vertical coordinate of the array is calculated, and the average value of the conduction storage unit corresponding to all voltage levels can be set to avg0.

[0069] See also Figure 7 In one embodiment of the present invention, when executing step S220, the threshold voltage source data is segmented according to the unit conduction mean value to generate multiple voltage peaks. Specifically, the following steps may be included:

[0070] Step S221, obtaining multiple intersection positions of a corresponding curve of the threshold voltage source data and a corresponding straight line of the cell conduction mean value;

[0071] Step S222: traverse all intersection positions to divide the threshold voltage source data into a plurality of voltage peaks.

[0072] In one embodiment of the present invention, when executing step S221 to step S222, specifically, first, let the straight line y=avg0 to obtain the intersection points of the threshold voltage source data corresponding to the Vt curve and the straight line y=avg0, and record the horizontal coordinates of all the intersection points. All the horizontal coordinates can be marked as a global array. Then, loop through the array ,from Look to the right, the point with a value of 0 or 1 is the right boundary, and then Looking to the left, the point of 0 or 1 is the left boundary, so that the threshold voltage source data is divided into multiple voltage peaks. The voltage peak represents the concentrated voltage area of ​​the storage cells with different data states in the memory.

[0073] In one embodiment of the present invention, when step S230 is executed, the left boundary value and the right boundary value of the interval where each voltage peak is located are searched, and the peak separation data is generated based on the left boundary value and the right boundary value. Specifically, the left boundary value and the right boundary value of the interval where each voltage peak is located are searched in turn, and the interval data between the left boundary and the right boundary of each voltage peak is saved to generate the peak separation data. In this embodiment, the voltage peak separation data can represent the interval data of each voltage peak. It should be noted that for each set of points outside the boundary coordinates, they can be regarded as original points without processing, and the mapping relationship between the horizontal and vertical coordinates remains unchanged.

[0074] In one embodiment of the present invention, when step S300 is executed, Gaussian fitting is performed on the peak data to generate a corresponding Gaussian fitting matrix. Specifically, first, the parameter definition of the Gaussian fitting algorithm applied in the present invention is set. According to the NAND flash memory Vt tool acquisition principle, the generated threshold voltage source data characterizes the number of conduction cells under each voltage level, and the number of cells corresponding to these voltage levels is set as the y-coordinate, which can be expressed as y1, y2, ..., yn. The horizontal coordinate is the corresponding voltage level, which can be expressed as x1, x2, ..., xn. Because the voltage level in the original threshold voltage source data is a character string and does not increase linearly, the horizontal coordinate is optimized to: x1=1, x2=2, ..., xn=n. Therefore, the VT curve can satisfy the following normal distribution formula:

[0075] ;

[0076] Taking the logarithm of both sides of the above formula gives:

[0077] ;

[0078] make , , , , we can get:

[0079] ;

[0080] The corresponding matrix can be obtained from the above formula:

[0081] ;

[0082] The above matrix is ​​rewritten as Z=XB+E, where the vector E represents the measurement error each time, that is, the deviation between the number of on-cells and the Gaussian fitting value under the actual voltage corresponding to the horizontal axis.

[0083] In this embodiment, it can be known that x1, x2, ..., xn are known numbers, and z1, z2, ..., zn are ln logarithms obtained from actual Vt measurement data, which are also known numbers. Therefore, if the best b0, b1, b2 can be obtained, and a, b, c can be reversely deduced, the parameters of the normal distribution can be calculated, thereby fitting the ideal Vt curve.

[0084] Furthermore, in order to calculate the parameters b0, b1, and b2, the fitted curve should be the smallest with the actual measured value. According to the idea of ​​Gaussian fitting, when the square of the error between the measured value and the ideal value of the fit is the smallest, the fitted curve is the most consistent with the actual curve, and the square of the error L can satisfy the following formula:

[0085] .

[0086] In one embodiment of the present invention, when step S400 is executed, the least squares calculation process is performed on the Gaussian fitting matrix to generate a parameter solution of the Gaussian fitting matrix, and the parameter solution is substituted into the Gaussian fitting matrix to generate threshold voltage fitting data, wherein the parameter solution satisfies that the square value of the error between the threshold voltage fitting data and the threshold voltage source data is minimized. Specifically, the least squares calculation process is performed on the above-mentioned Gaussian fitting matrix to obtain a parameter solution of the Gaussian fitting matrix, and the parameter solution is substituted into the Gaussian fitting matrix to generate the threshold voltage fitting data.

[0087] ;

[0088] ;

[0089] From the above formula, we can see that L is about , , Therefore, finding the minimum value of L is to find the point where the value of L is the smallest ( , , ). , is a constant, L is respectively , , Find the partial derivative, the point where the derivative is 0 is the minimum value, and the following formula can be obtained:

[0090] ;

[0091] ;

[0092] ;

[0093] Transforming the above formula, we can get:

[0094] ;

[0095] ;

[0096] ;

[0097] Writing the above formula in matrix form, we get:

[0098] ;

[0099] make , , , then the matrix form satisfies the following formula:

[0100] ;

[0101] In the above formula, matrix B is the solution of the linear equation system. When , the above equation system has a unique solution. It can be seen that because any two rows of the matrix X are not proportional, then ,therefore, , so that the linear equations have a unique solution, , , is the only one.

[0102] because , and the number of rows and columns in the matrix is ​​relatively small, so Gaussian elimination method can be used to solve B and obtain the ideal fitting parameters. According to the above algorithm analysis, each set of values ​​is calculated according to Gaussian fitting to obtain the parameters , , , calculate the ideal value.

[0103] In one embodiment of the present invention, when step S500 is executed, the threshold voltage source data and the threshold voltage fitting data are weighted to generate threshold voltage weighted data. Specifically, there is already an original data csv file, and the mapping relationship between the horizontal and vertical coordinates is determined. Given the horizontal coordinate xi, use the fitted Gaussian function to calculate the corresponding vertical coordinate yi, which is the ideal fitting value. In this embodiment, for the ideal fitting value, it is necessary to take weights together with the actual value, and the threshold voltage weighted data satisfies the following formula:

[0104] Threshold voltage weighted data = threshold voltage fitting data × m + threshold voltage source data × n;

[0105] Wherein, m represents the preset weight of the threshold voltage fitting data, and n represents the weight of the threshold voltage source data, satisfying m+n=1. For example, threshold voltage weighted data = threshold voltage fitting data × 0.6 + threshold voltage source data × 0.4. Otherwise, the fitting result will be a smooth normal distribution curve, which deviates slightly from the actual situation. Write the threshold voltage weighted data into a new csv file, where the threshold voltage source data is written in one row, and the threshold voltage weighted data after fitting is written in the second row, and the fitting of each voltage peak is completed in turn. Compare the effects before and after fitting, such as Figure 8 As shown, after being processed by the method provided by the present invention, the fitted VT graph is compared with the original VT graph and it is found that the VT graph processed by the present invention achieves the fitting effect. On the basis of ensuring that the data does not deviate from the actual situation, the readability of the VT graph is enhanced.

[0106] See also Fig. 9 The present invention also provides a memory threshold voltage data processing system. In one embodiment of the present invention, the memory threshold voltage data processing system may include a source data acquisition module 901, a multi-peak cutting module 902, a Gaussian fitting module 903, a parameter calculation module 904 and a weighted processing module 905, and each functional module is described in detail as follows.

[0107] The source data acquisition module 901 can be used to obtain the threshold voltage source data of the flash memory chip, and the threshold voltage source data represents the number of memory cells that are turned on in each voltage level of the flash memory chip. Further, the source data acquisition module 901 can be specifically used to obtain the threshold voltage of each memory cell in the flash memory chip; count all the threshold voltages, and save the number of memory cells that are turned on in each voltage level as the threshold voltage source data according to a preset format.

[0108] The multi-peak cutting module 902 can be used to calculate the mean value of the threshold voltage source data, and segment the threshold voltage source data according to the mean value to generate multiple voltage peaks, and save the interval data of each voltage peak as the peak data. Further, the multi-peak cutting module 902 can be specifically used to import the threshold voltage source data into the voltage array, and calculate the unit conduction mean value of all voltage levels; segment the threshold voltage source data according to the unit conduction mean value to generate multiple voltage peaks; find the left boundary value and the right boundary value of the interval where each voltage peak is located, and save the left boundary value and the right boundary value as the peak data.

[0109] The Gaussian fitting module 903 can be used to perform Gaussian fitting processing on the peak data to generate a corresponding Gaussian fitting matrix. Furthermore, the Gaussian fitting module 903 can be specifically used to first set the parameter definition of the Gaussian fitting algorithm applied in the present invention. Then, the number of cells corresponding to the voltage gear is set to the y coordinate, expressed as y1, y2,..., yn. The horizontal coordinate is the corresponding voltage gear, which can be expressed as x1, x2,..., xn. Then the horizontal coordinate is optimized to: x1=1, x2=2,..., xn=n. The normal distribution formula of the VT curve is processed to generate a Gaussian fitting matrix:

[0110] .

[0111] The parameter calculation module 904 can be used to perform least squares calculation processing on the Gaussian fitting matrix to generate a parameter solution of the Gaussian fitting matrix, and substitute the parameter solution into the Gaussian fitting matrix to generate threshold voltage fitting data, wherein the parameter solution satisfies that the square value of the error between the threshold voltage fitting data and the threshold voltage source data is minimized. Further, the parameter calculation module 904 can be specifically used to perform least squares calculation processing on the above-mentioned Gaussian fitting matrix to obtain a parameter solution of the Gaussian fitting matrix, and substitute the parameter solution into the Gaussian fitting matrix to generate threshold voltage fitting data.

[0112] The weighted processing module 905 can be used to perform weighted processing on the threshold voltage source data and the threshold voltage fitting data to generate threshold voltage weighted data. Further, the weighted processing module 905 can be specifically used to take weights together with the actual value for the ideal fitting value, and the threshold voltage weighted data satisfies the following formula:

[0113] Threshold voltage weighted data = threshold voltage fitting data × m + threshold voltage source data × n;

[0114] Wherein, m represents the preset weight of the threshold voltage fitting data, and n represents the weight of the threshold voltage source data, satisfying m+n=1. For example, threshold voltage weighted data=threshold voltage fitting data×0.6+threshold voltage source data×0.4.

[0115] For the specific definition of the processing system, please refer to the definition of the processing method above, which will not be repeated here. Each module in the above processing system can be implemented in whole or in part by software, hardware and a combination thereof. The above modules can be embedded in or independent of the processor in the computer device in the form of hardware, or can be stored in the memory of the computer device in the form of software, so that the processor can call and execute the operations corresponding to the above modules.

[0116] See also Fig.10 , Fig.10 Shown is a structural schematic diagram of a computer device provided by the present invention. The computer device 1000 can be a server. The computer device 1000 includes a processor 1001, a memory 1002, a network interface and a database connected via a system bus. Among them, the processor 1001 of the computer device 1000 is used to provide computing and control capabilities. The memory 1002 of the computer device 1000 includes a non-volatile and / or volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The network interface of the computer device 1000 is used to communicate with an external client through a network connection. When the computer program is executed by the processor, the functions or steps on the server side of a method for processing three-dimensional vehicle model data are realized.

[0117] In one embodiment, a computer device 1000 is provided, including a memory 1002, a processor 1001, and a computer program stored in the memory and executable on the processor. When the processor 1001 executes the computer program, the following steps are implemented:

[0118] Acquire threshold voltage source data of the flash memory chip, where the threshold voltage source data represents the number of memory cells that are turned on at each voltage level of the flash memory chip;

[0119] Calculating the mean value of the threshold voltage source data, and segmenting the threshold voltage source data according to the mean value to generate a plurality of voltage peaks, and saving the interval data of each voltage peak as segmented peak data;

[0120] Perform Gaussian fitting on the peak data to generate a corresponding Gaussian fitting matrix;

[0121] The Gaussian fitting matrix is ​​processed by least squares calculation to generate a parameter solution of the Gaussian fitting matrix, and the parameter solution is substituted into the Gaussian fitting matrix to generate threshold voltage fitting data, wherein the parameter solution satisfies that the square value of the error between the threshold voltage fitting data and the threshold voltage source data is minimized;

[0122] The threshold voltage source data and the threshold voltage fitting data are weighted to generate threshold voltage weighted data.

[0123] In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored, and when the computer program is executed by a processor, the following steps are implemented:

[0124] Acquire threshold voltage source data of the flash memory chip, where the threshold voltage source data represents the number of memory cells that are turned on at each voltage level of the flash memory chip;

[0125] Calculating the mean value of the threshold voltage source data, and segmenting the threshold voltage source data according to the mean value to generate a plurality of voltage peaks, and saving the interval data of each voltage peak as segmented peak data;

[0126] Perform Gaussian fitting on the peak data to generate a corresponding Gaussian fitting matrix;

[0127] The Gaussian fitting matrix is ​​processed by least squares calculation to generate a parameter solution of the Gaussian fitting matrix, and the parameter solution is substituted into the Gaussian fitting matrix to generate threshold voltage fitting data, wherein the parameter solution satisfies that the square value of the error between the threshold voltage fitting data and the threshold voltage source data is minimized;

[0128] The threshold voltage source data and the threshold voltage fitting data are weighted to generate threshold voltage weighted data.

[0129] It should be noted that the functions or steps that can be implemented by the computer-readable storage medium or computer device can be found in the aforementioned method embodiments. To avoid repetition, they will not be described one by one here.

[0130] In summary, the present invention proposes a method, system, device and medium for processing memory threshold voltage data. The present invention adopts Gaussian fitting and least squares method to optimize the fitting of the VT curve, reduce the noise in the data, and make the Vt curve under the actual voltage close to the ideal value. The curve becomes smoother, reduces the curve jitter, enhances the readability of the VT curve, and provides data support for further analysis of the threshold voltage.

[0131] The above embodiments are merely illustrative of the principles and effects of the present invention, and are not intended to limit the present invention. Anyone familiar with the art may modify or alter the above embodiments without departing from the spirit and scope of the present invention. Therefore, all equivalent modifications or alterations made by a person of ordinary skill in the art without departing from the spirit and technical concept disclosed by the present invention shall still be covered by the claims of the present invention.

Claims

1. A method for processing memory threshold voltage data, characterized in that: include: Obtaining the threshold voltage of each storage cell in the flash memory chip; Counting all the threshold voltages, and saving the number of memory cells turned on at each voltage level as threshold voltage source data according to a preset format; Based on the threshold voltage source data, the voltage level is set as the horizontal coordinate, and the number of the conductive memory cells corresponding to each voltage level is set as the vertical coordinate, so as to construct a relationship curve; Calculating the mean value of the threshold voltage source data, and segmenting the corresponding relationship curve of the threshold voltage source data according to the mean value to generate a plurality of voltage peaks, and saving the interval data of each voltage peak as peak segmentation data; Performing Gaussian fitting processing on the peak data to generate a corresponding Gaussian fitting matrix; Performing least squares calculation processing on the Gaussian fitting matrix to generate a parameter solution of the Gaussian fitting matrix, and substituting the parameter solution into the Gaussian fitting matrix to generate threshold voltage fitting data, wherein the parameter solution satisfies that the square value of the error between the threshold voltage fitting data and the threshold voltage source data is minimized; The threshold voltage source data and the threshold voltage fitting data are weighted to generate threshold voltage weighted data.

2. The method for processing memory threshold voltage data according to claim 1, characterized in that: The step of calculating the mean value of the threshold voltage source data, segmenting the corresponding relationship curve of the threshold voltage source data according to the mean value to generate a plurality of voltage peaks, and saving the interval data of each voltage peak as peak segmentation data comprises: Importing the threshold voltage source data into a voltage array, and calculating the unit conduction mean value of all the voltage levels; According to the unit conduction mean value, segmenting the corresponding relationship curve of the threshold voltage source data to generate a plurality of voltage peaks; The left boundary value and the right boundary value of the interval where each voltage peak is located are found, and the peak separation data are generated based on the left boundary value and the right boundary value.

3. The method for processing memory threshold voltage data according to claim 2, characterized in that: The step of segmenting the corresponding relationship curve of the threshold voltage source data according to the unit conduction mean value to generate a plurality of voltage peaks includes: Acquire multiple intersection positions of the corresponding relationship curve of the threshold voltage source data and the corresponding straight line of the unit conduction mean value; All the intersection positions are traversed to divide the threshold voltage source data into a plurality of voltage peaks.

4. The method for processing memory threshold voltage data according to claim 1, characterized in that: In the step of performing Gaussian fitting processing on the peak data to generate a corresponding Gaussian fitting matrix, the Gaussian fitting matrix satisfies the following formula: in, represents all voltage levels of the threshold voltage, represents the logarithmic value of the number of conducting units corresponding to each voltage level, represents the parameters of the Gaussian fitting matrix, Represents measurement error.

5. The method for processing memory threshold voltage data according to claim 1, characterized in that: In the step of performing least squares calculation processing on the Gaussian fitting matrix to generate a parameter solution of the Gaussian fitting matrix, and substituting the parameter solution into the Gaussian fitting matrix to generate threshold voltage fitting data, wherein the parameter solution satisfies that the square value of the error between the threshold voltage fitting data and the threshold voltage source data is the smallest, the square value of the error L between the threshold voltage fitting data and the threshold voltage source data satisfies the following formula: in, represents the threshold voltage source data, represents the threshold voltage fitting data.

6. The method for processing memory threshold voltage data according to claim 1, characterized in that: In the step of performing weighted processing on the threshold voltage source data and the threshold voltage fitting data to generate threshold voltage weighted data, the threshold voltage weighted data satisfies the following formula: Threshold voltage weighted data = threshold voltage fitting data × m + threshold voltage source data × n; Wherein, m represents the preset weight of the threshold voltage fitting data, n represents the weight of the threshold voltage source data, and m+n=1 is satisfied.

7. A memory threshold voltage data processing system, characterized in that: include: A source data acquisition module is used to obtain the threshold voltage of each storage unit in the flash memory chip; Counting all the threshold voltages, and saving the number of memory cells turned on at each voltage level as threshold voltage source data according to a preset format; Based on the threshold voltage source data, the voltage level is set as the horizontal coordinate, and the number of the conductive memory cells corresponding to each voltage level is set as the vertical coordinate, so as to construct a relationship curve; A multi-peak cutting module, used for calculating the mean value of the threshold voltage source data, and segmenting the corresponding relationship curve of the threshold voltage source data according to the mean value to generate multiple voltage peaks, and saving the interval data of each voltage peak as peak data; A Gaussian fitting module, used for performing Gaussian fitting processing on the peak data to generate a corresponding Gaussian fitting matrix; a parameter calculation module, used for performing least squares calculation processing on the Gaussian fitting matrix to generate a parameter solution of the Gaussian fitting matrix, and substituting the parameter solution into the Gaussian fitting matrix to generate threshold voltage fitting data, wherein the parameter solution satisfies that the square value of the error between the threshold voltage fitting data and the threshold voltage source data is minimized; The weighted processing module is used to perform weighted processing on the threshold voltage source data and the threshold voltage fitting data to generate threshold voltage weighted data.

8. An electronic device, characterized in that: It comprises a processor, a memory and a communication bus; the communication bus is used to connect the processor and the memory; the processor is used to execute a computer program stored in the memory to implement the method for processing memory threshold voltage data as described in any one of claims 1-6.

9. A computer-readable storage medium, characterized in that: A computer program is stored thereon, and the computer program is used to enable a computer to execute the method for processing memory threshold voltage data as claimed in any one of claims 1 to 6.