CN117574835A - Method, equipment and storage medium for simulating Nand false LLR distribution - Google Patents
Method, equipment and storage medium for simulating Nand false LLR distribution Download PDFInfo
- Publication number
- CN117574835A CN117574835A CN202311537535.8A CN202311537535A CN117574835A CN 117574835 A CN117574835 A CN 117574835A CN 202311537535 A CN202311537535 A CN 202311537535A CN 117574835 A CN117574835 A CN 117574835A
- Authority
- CN
- China
- Prior art keywords
- nand
- false
- llr
- data
- distribution
- 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.)
- Pending
Links
- 238000009826 distribution Methods 0.000 title claims abstract description 129
- 238000000034 method Methods 0.000 title claims abstract description 47
- 238000003860 storage Methods 0.000 title claims abstract description 27
- 238000004088 simulation Methods 0.000 claims abstract description 68
- 238000012549 training Methods 0.000 claims abstract description 42
- 238000013136 deep learning model Methods 0.000 claims abstract description 24
- 230000007613 environmental effect Effects 0.000 claims abstract description 12
- 238000004458 analytical method Methods 0.000 claims description 21
- 238000004590 computer program Methods 0.000 claims 1
- 238000012545 processing Methods 0.000 abstract description 4
- 230000008569 process Effects 0.000 description 12
- 238000012360 testing method Methods 0.000 description 9
- 230000006870 function Effects 0.000 description 7
- 238000005457 optimization Methods 0.000 description 6
- 238000013528 artificial neural network Methods 0.000 description 4
- 238000010586 diagram Methods 0.000 description 4
- 230000006872 improvement Effects 0.000 description 4
- 230000008439 repair process Effects 0.000 description 4
- 238000000342 Monte Carlo simulation Methods 0.000 description 3
- 238000004422 calculation algorithm Methods 0.000 description 3
- 238000004364 calculation method Methods 0.000 description 3
- 230000008859 change Effects 0.000 description 3
- 238000013461 design Methods 0.000 description 3
- 230000000694 effects Effects 0.000 description 3
- 238000010200 validation analysis Methods 0.000 description 3
- 230000004913 activation Effects 0.000 description 2
- 230000015556 catabolic process Effects 0.000 description 2
- 238000004891 communication Methods 0.000 description 2
- 238000013527 convolutional neural network Methods 0.000 description 2
- 238000006731 degradation reaction Methods 0.000 description 2
- 238000004519 manufacturing process Methods 0.000 description 2
- 210000002569 neuron Anatomy 0.000 description 2
- 230000003287 optical effect Effects 0.000 description 2
- 230000000306 recurrent effect Effects 0.000 description 2
- 230000008263 repair mechanism Effects 0.000 description 2
- ORILYTVJVMAKLC-UHFFFAOYSA-N Adamantane Natural products C1C(C2)CC3CC1CC2C3 ORILYTVJVMAKLC-UHFFFAOYSA-N 0.000 description 1
- 238000012937 correction Methods 0.000 description 1
- 238000013135 deep learning Methods 0.000 description 1
- 230000001934 delay Effects 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- 230000002349 favourable effect Effects 0.000 description 1
- 238000012544 monitoring process Methods 0.000 description 1
- 238000010606 normalization Methods 0.000 description 1
- 239000013307 optical fiber Substances 0.000 description 1
- 230000002085 persistent effect Effects 0.000 description 1
- 230000004044 response Effects 0.000 description 1
- 238000005070 sampling Methods 0.000 description 1
- 239000004065 semiconductor Substances 0.000 description 1
- 230000009466 transformation Effects 0.000 description 1
- 230000001131 transforming effect Effects 0.000 description 1
- 238000012795 verification Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/30—Circuit design
- G06F30/39—Circuit design at the physical level
- G06F30/392—Floor-planning or layout, e.g. partitioning or placement
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2117/00—Details relating to the type or aim of the circuit design
- G06F2117/02—Fault tolerance, e.g. for transient fault suppression
Landscapes
- Engineering & Computer Science (AREA)
- Computer Hardware Design (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Architecture (AREA)
- Evolutionary Computation (AREA)
- Geometry (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention relates to the field of data processing, and discloses a simulation method, equipment and a storage medium for Nand false LLR distribution. The method comprises the following steps: generating a deep learning model to be trained; taking preset environmental data and a preset threshold voltage distribution data set as training data, and executing model training operation on the deep learning model to obtain a vth offset prediction model; when threshold voltage distribution data of Nand false is detected, invoking a pre-trained vth offset prediction model to execute multi-condition simulation prediction operation on the threshold voltage distribution data to obtain multiple groups of simulation data; and acquiring the Nand false LLR distribution corresponding to the Nand false in the model data. In the embodiment of the invention, the simulation cost of the Nand false LLR distribution is reduced.
Description
Technical Field
The invention relates to the field of data processing, in particular to a simulation method, equipment and storage medium for Nand false LLR distribution.
Background
The LLR (Log-Likelihood Ratio) is an important indicator for bit error rate estimation and reflects the similarity between the received signal and each possible symbol value. By acquiring LLR distribution corresponding to Nand Flash, the error rate can be estimated more accurately, so that the performance of the communication system is optimized.
For simulation of LLR distribution, a Monte Carlo method, which is a numerical calculation method based on random sampling, can be used to generate two probability distributions conforming to the original assumption and the alternative assumption, such as Bernoulli distribution or normal distribution; generating an observation value according to the probability distribution of the original hypothesis or the alternative hypothesis for each sample; calculating LLR values of each sample, namely log-likelihood ratios of two probability distributions; repeating the above steps for a plurality of times to obtain the approximate distribution of LLR distribution. This method requires a lot of computing resources and time, and the simulation cost is high.
Disclosure of Invention
The invention mainly aims to solve the technical problem of high simulation cost of Nand false LLR distribution.
The first aspect of the present invention provides a method for simulating a Nand false LLR distribution, where the method for simulating a Nand false LLR distribution includes:
generating a deep learning model to be trained;
taking preset environmental data and a preset threshold voltage distribution data set as training data, and executing model training operation on the deep learning model to obtain a vth offset prediction model;
when threshold voltage distribution data of Nand false is detected, invoking a pre-trained vth offset prediction model to execute multi-condition simulation prediction operation on the threshold voltage distribution data to obtain multiple groups of simulation data;
and acquiring the Nand false LLR distribution corresponding to the Nand false in the model data.
Optionally, in a first implementation manner of the first aspect of the present invention, when threshold voltage distribution data of Nand false is detected, the step of calling a pre-trained vth offset prediction model to perform multi-condition simulation prediction operation on the threshold voltage distribution data to obtain multiple groups of simulation data includes:
when threshold voltage distribution data of Nand false is detected, invoking a pre-trained vth offset prediction model to generate a plurality of simulation conditions, wherein the time, temperature, PE or Read Disturb in the simulation conditions are different;
and executing multi-condition simulation prediction operation based on the threshold voltage distribution data and the simulation conditions to obtain a plurality of simulation data.
Optionally, in a second implementation manner of the first aspect of the present invention, the step of performing a model training operation on the deep learning model to obtain a vth offset prediction model using preset environmental data and a threshold voltage distribution data set corresponding to a sample Nand false as training data includes:
and taking the preset temperature, the preset time parameter, the preset PE, the preset Read Disturb and the threshold voltage distribution data set as training data, and executing model training operation on the deep learning model to obtain a vth offset prediction model.
Optionally, in a third implementation manner of the first aspect of the present invention, after the step of obtaining a Nand false LLR distribution corresponding to the Nand false LLR in the model data, the method further includes:
analyzing a plurality of groups of model data to obtain analysis results;
and obtaining the optimal Nand false LLR distribution corresponding to the Nand false LLR in the analysis result according to a preset LLR selection standard.
Optionally, in a fourth implementation manner of the first aspect of the present invention, after the step of obtaining the optimal Nand false LLR distribution corresponding to the Nand false LLR in the parsing result according to the preset LLR selection criterion, the method further includes:
obtaining a target condition corresponding to the optimal Nand false LLR;
and outputting prompt information capable of improving the LDPC decoding capability under the target condition.
Optionally, in a fifth implementation manner of the first aspect of the present invention, when threshold voltage distribution data of Nand false is detected, a pre-trained vth offset prediction model is called to perform a multi-condition simulation prediction operation on the threshold voltage distribution data, so as to obtain multiple groups of simulation data, and after the step of obtaining the multiple groups of simulation data, the method further includes:
analyzing a plurality of groups of model data to obtain analysis results;
and acquiring the RRT group corresponding to the Nand false from the analysis result.
Optionally, in a sixth implementation manner of the first aspect of the present invention, the step of obtaining, in the analysis result, the RRT group corresponding to the Nand false includes:
and acquiring the optimal RRT group corresponding to the Nand false from the analysis result according to a preset RRT group selection standard.
Optionally, in a seventh implementation manner of the first aspect of the present invention, after the step of obtaining a Nand false LLR distribution corresponding to the Nand false LLR in the model data, the method further includes:
outputting the Nand false LLR distribution.
The second aspect of the present invention provides a simulation apparatus for Nand false LLR distribution, comprising: a memory and at least one processor, the memory having instructions stored therein, the memory and the at least one processor being interconnected by a line; the at least one processor invokes the instructions in the memory to cause the simulation device of the Nand false LLR distribution to perform the simulation method of Nand false LLR distribution described above.
A third aspect of the present invention provides a computer-readable storage medium having instructions stored therein that, when run on a computer, cause the computer to perform the above-described method of simulating a Nand false LLR distribution.
In the embodiment of the invention, a deep learning model to be trained is generated; taking preset environmental data and a preset threshold voltage distribution data set as training data, and executing model training operation on the deep learning model to obtain a vth offset prediction model; when threshold voltage distribution data of Nand false is detected, invoking a pre-trained vth offset prediction model to execute multi-condition simulation prediction operation on the threshold voltage distribution data to obtain multiple groups of simulation data; and acquiring the Nand false LLR distribution corresponding to the Nand false in the model data. The simulation device of Nand false LLR distribution uses preset environment data and preset threshold voltage distribution data set as training data set to train the generated deep learning model. By training, the model will learn the relationship between the input data and the corresponding output, i.e. how to predict the shift in threshold voltage. After training, the deep learning model becomes a model capable of predicting threshold voltage shift. The model can predict corresponding offset conditions according to the input threshold voltage distribution data. When threshold voltage distribution data of Nand Flash is detected, a pre-trained vth offset prediction model can be called, and multi-condition simulation prediction operation is carried out on the threshold voltage distribution data. The model may generate sets of simulation data based on the input data. And in the generated simulation data, LLR distribution corresponding to Nand Flash can be obtained through analysis. The acquisition of LLR distribution using a deep learning model can help to improve understanding and optimization of Nand Flash performance and reliability. By training a model and simulating a prediction operation, a more accurate threshold voltage shift prediction result can be obtained, LLR distribution data of Nand Flash can be further analyzed and obtained, so that improvement of LDPC decoding capability is facilitated, compared with a Monte Carlo method, a large amount of calculation resources and time are not needed through model simulation, and simulation cost of Nand false LLR distribution is reduced.
Drawings
FIG. 1 is a schematic diagram of a first embodiment of a simulation method of the Nand false LLR distribution in an embodiment of the present invention;
FIG. 2 is a diagram showing a second embodiment of a simulation method of the Nand false LLR distribution in the embodiment of the present invention;
fig. 3 is a schematic diagram of an embodiment of a simulation apparatus for Nand false LLR distribution in an embodiment of the present invention.
Detailed Description
The embodiment of the invention provides a simulation method, equipment and storage medium for Nand false LLR distribution.
Embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. While the present disclosure has been illustrated in the drawings in some form, it is to be understood that the present disclosure may be embodied in various forms and should not be construed as limited to the embodiments set forth herein, but are provided to provide a more thorough and complete understanding of the present disclosure. It should be understood that the drawings and examples of the present disclosure are for illustrative purposes only and are not intended to limit the scope of the present disclosure.
In describing embodiments of the present disclosure, the term "comprising" and its like should be taken to be open-ended, i.e., including, but not limited to. The term "based on" should be understood as "based at least in part on". The term "one embodiment" or "the embodiment" should be understood as "at least one embodiment". The terms "first," "second," and the like, may refer to different or the same object. Other explicit and implicit definitions are also possible below.
For ease of understanding, a specific flow of an embodiment of the present invention is described below, referring to fig. 1, and an embodiment of a method for simulating Nand false LLR distribution in an embodiment of the present invention includes:
101. generating a deep learning model to be trained;
in the present embodiment, the type of the deep learning model may be selected from a neural network (Feedforward Neural Network), a convolutional neural network (Convolutional Neural Network), a recurrent neural network (Recurrent Neural Network), or the like. After selecting the proper model type, the number of layers of the model, the number of neurons of each layer and the connection mode are determined. The choice of model architecture can affect the representation capabilities and computational complexity of the model. A nonlinear transformation is introduced into each neuron, selection of activation functions including, but not limited to ReLU, sigmoid, tanh, etc. A measure of the difference between the model predicted output and the actual label is defined. The loss function selects a mean square error (Mean Squared Error) or Cross Entropy (Cross Entropy), etc. Selection of optimization algorithms, including but not limited to random gradient descent (Stochastic Gradient Descent), adam, adagard, and the like.
102. Taking preset environmental data and a preset threshold voltage distribution data set as training data, and executing model training operation on the deep learning model to obtain a vth offset prediction model;
specifically, the environmental data and threshold voltage distribution data sets are preset to ensure that the data sets have the proper labels or target values. For a threshold voltage distribution data set, each data sample is associated with a respective vth offset value. The data is preprocessed to enhance the training effect. Including but not limited to normalizing, normalizing or otherwise transforming the input data, and encoding the tag or target value. The data set is divided into a training set, a validation set and a test set. The training set is used for the training process of the model, the verification set is used for adjusting the super parameters and monitoring the performance of the model, and the test set is used for finally evaluating the generalization capability of the model.
The deep learning model is configured according to the model type, architecture, activation function and other choices of the previous step. The model objects may be created using an interface provided by the deep learning framework and corresponding parameters set. An appropriate loss function and optimizer are selected. The loss function is used to measure the difference between the model output and the training data, and the optimizer is used to update the model parameters to minimize the loss function. The training set is used to perform a training process of the model. By means of the back propagation algorithm and the optimizer, the model will gradually adjust the weights and biases according to the input data to minimize the loss function. In the training process, proper super parameters such as training iteration times (epochs), batch size (batch size) and the like are set.
After each training iteration period is completed, the performance of the model is evaluated using the validation set. Depending on the performance on the validation set, the hyper-parameters of the model (e.g., learning rate, regularization parameters, etc.) or other techniques (e.g., dropout, batch normalization, etc.) may be adjusted to improve the generalization ability of the model. After training is completed, the model is finally evaluated using the test set. The performance and predictive ability of the model are evaluated by calculating a difference metric (e.g., mean square error, accuracy, etc.) between the predicted result and the actual label.
If the model achieves a satisfactory training effect, the trained vth shift prediction model can be saved for subsequent use.
103. When threshold voltage distribution data of Nand false is detected, invoking a pre-trained vth offset prediction model to execute multi-condition simulation prediction operation on the threshold voltage distribution data to obtain multiple groups of simulation data;
specifically, by using the vth shift prediction model, the electrical characteristics can be predicted and simulated before the actual physical chip is manufactured, thereby saving a lot of time and resources. Traditionally, in order to evaluate electrical characteristics under different process conditions, multiple chip processes and tests are required, which is time consuming and costly. The pre-training model is used for simulation prediction, so that the chip performance under different conditions can be evaluated more quickly, and the production efficiency is improved. Through multi-condition simulation prediction operation, different process parameters, environmental factors and the like can be comprehensively evaluated and compared. This helps optimize the process design of Nand Flash, for example, by selecting appropriate process parameters or improving the production flow to improve chip performance and stability.
The pre-training model is used for simulation prediction, so that the requirement of experimental tests can be reduced, and the cost is reduced. Experimental testing requires a large number of chip samples and test equipment, while analog predictions can be made on a computer without the need for actual chips and test equipment. By reducing the number and scale of experimental tests, the cost can be significantly reduced.
Optionally, when threshold voltage distribution data of Nand false is detected, invoking a pre-trained vth offset prediction model to generate a plurality of simulation conditions, wherein time, temperature, PE or Read Disturb in the simulation conditions are different; and executing multi-condition simulation prediction operation based on the threshold voltage distribution data and the simulation conditions to obtain a plurality of simulation data. The performance change of Nand Flash under different process parameters and environmental conditions can be evaluated by simulating threshold voltage distribution under different conditions such as time, temperature, PE or Read Disturb. The method is favorable for optimizing process design, and proper parameter configuration is selected, so that the performance and reliability of the chip can be improved. Simulating the threshold voltage distribution under different conditions can help predict degradation of the chip over time, temperature, and number of uses. Through multi-condition simulation prediction operation, the performance change of Nand Flash under different process conditions can be better understood, so that the optimization and improvement can be carried out pertinently. By identifying potential problems in advance and taking corresponding measures, the quality and reliability of the product can be improved, and the risks of faults and degradation in the later period are reduced.
104. Obtaining a Nand false LLR distribution corresponding to the Nand false LLR in the model data;
specifically, in the decoding process, the LLR is used for decision judgment, and the most probable symbol value is selected according to the LLR size. Accurate LLR distribution can provide more reliable information, thereby improving decoding performance and reducing error rate.
Further, the LLR distribution can be used to analyze and evaluate the robustness of the system under different channel conditions. By observing the change of LLR distribution, the adaptability of the system to different channel distortion, noise and other conditions can be known, and the system can be adjusted and improved in a targeted manner.
Further, LLR distribution is an important basis for designing and optimizing communication chips such as modems. By acquiring LLR distribution corresponding to Nand Flash, reference can be provided for chip design, circuit parameters, algorithm selection and the like are optimized, and the performance and reliability of the chip are improved.
Furthermore, by analyzing LLR distribution corresponding to Nand Flash, optimization and improvement can be performed aiming at different coding schemes, and the error correction coding effect and fault tolerance are improved.
Optionally, analyzing a plurality of groups of model data to obtain analysis results; and obtaining the optimal Nand false LLR distribution corresponding to the Nand false LLR in the analysis result according to a preset LLR selection standard. The quality and stability of the chip can be improved by selecting the optimal Nand Flash LLR distribution. Optimization of the LLR distribution can reduce the probability of bit flipping and data loss, thereby improving the reliability of the chip.
Optionally, acquiring a target condition corresponding to the optimal Nand false LLR; and outputting prompt information capable of improving the LDPC decoding capability under the target condition.
Optionally, outputting the Nand false LLR distribution.
In the embodiment of the invention, a deep learning model to be trained is generated; taking preset environmental data and a preset threshold voltage distribution data set as training data, and executing model training operation on the deep learning model to obtain a vth offset prediction model; when threshold voltage distribution data of Nand false is detected, invoking a pre-trained vth offset prediction model to execute multi-condition simulation prediction operation on the threshold voltage distribution data to obtain multiple groups of simulation data; and acquiring the Nand false LLR distribution corresponding to the Nand false in the model data. The simulation device of Nand false LLR distribution uses preset environment data and preset threshold voltage distribution data set as training data set to train the generated deep learning model. By training, the model will learn the relationship between the input data and the corresponding output, i.e. how to predict the shift in threshold voltage. After training, the deep learning model becomes a model capable of predicting threshold voltage shift. The model can predict corresponding offset conditions according to the input threshold voltage distribution data. When threshold voltage distribution data of Nand Flash is detected, a pre-trained vth offset prediction model can be called, and multi-condition simulation prediction operation is carried out on the threshold voltage distribution data. The model may generate sets of simulation data based on the input data. And in the generated simulation data, LLR distribution corresponding to Nand Flash can be obtained through analysis. The acquisition of LLR distribution using a deep learning model can help to improve understanding and optimization of Nand Flash performance and reliability. By training a model and simulating a prediction operation, a more accurate threshold voltage shift prediction result can be obtained, LLR distribution data of Nand Flash can be further analyzed and obtained, so that improvement of LDPC decoding capability is facilitated, compared with a Monte Carlo method, a large amount of calculation resources and time are not needed through model simulation, and simulation cost of Nand false LLR distribution is reduced.
Referring to fig. 2, fig. 2 is a second embodiment of a simulation method of Nand false LLR distribution in the embodiment of the present invention, after step 104, the following steps may be performed:
105. analyzing a plurality of groups of model data to obtain analysis results;
106. acquiring RRT groups corresponding to the Nand false from the analysis result;
specifically, read Retry Table (RRT) is an error repair mechanism for NAND Flash memory. It records the repair measures taken in the case of a specific read error to improve the read reliability and reduce the probability of data corruption. In NAND Flash memory, errors may occur in read operations due to physical limitations of the device or environmental noise, etc. These errors may result in bit flipping, data loss, or incorrect data reading. To cope with these errors, NAND Flash chips are often equipped with error detection and repair mechanisms. Through RRT, NAND Flash can realize partial self-repairing ability, improves the reading reliability and reduces the risk of data damage.
Optionally, according to a preset RRT group selection criterion, the optimal RRT group corresponding to the Nand false is obtained from the analysis result. According to the preset RRT group selection standard, the optimal RRT group is obtained in the analysis result, and a more accurate and effective repairing strategy can be provided for specific reading error conditions. The optimal RRT group contains the verified optimal repair measures, so that the probability of reading errors and data damage can be remarkably reduced, and the reading reliability of the storage system is further improved. The optimal RRT group not only provides the optimal read retry strategy, but also can further optimize the read performance of the storage system by reasonably configuring parameters such as read delay, voltage and the like. This helps to increase the read speed and response time, meeting the need for fast data access. The optimal RRT group is verified under various reading error conditions, and has higher compatibility and applicability. By adopting the optimal RRT group, the compatibility of the storage system can be improved, and wider application scenes and data types are supported. The optimal RRT group not only can improve the reading reliability and performance of the storage system, but also can reduce the running cost of the system by reducing unnecessary reading retry times and delays.
In this embodiment, by acquiring the RRT group corresponding to the NAND Flash, an accurate repair policy may be provided for a specific read error condition. RRT records repair measures for different read error types, such as adjusting read voltage, delay or re-read times, etc. And selecting a proper RRT group according to the analysis result, so that the probability of reading errors and data damage can be effectively reduced, and the reading reliability of a storage system is improved.
Fig. 3 is a schematic structural diagram of a Nand false LLR distributed analog device 500 according to an embodiment of the present invention, where the Nand false LLR distributed analog device 500 may have a relatively large difference due to different configurations or performances, and may include one or more processors (central processing units, CPU) 510 (e.g., one or more processors) and a memory 520, and one or more storage media 530 (e.g., one or more mass storage devices) storing application programs 533 or data 532. Wherein memory 520 and storage medium 530 may be transitory or persistent storage. The program stored on the storage medium 530 may include one or more modules (not shown), each of which may include a series of instruction operations in the simulation apparatus 500 for Nand false LLR distribution. Still further, the processor 510 may be configured to communicate with the storage medium 530 to execute a series of instruction operations in the storage medium 530 on the Nand false LLR distributed analog device 500.
The Nand false LLR distribution-based simulation device 500 may also include one or more power supplies 540, one or more wired or wireless network interfaces 550, one or more input/output interfaces 560, and/or one or more operating systems 531, such as Windows service, mac OS X, unix, linux, free BSD, and the like. Those skilled in the art will appreciate that the configuration of the analog device for Nand false LLR distribution shown in fig. 3 is not limiting and that analog devices based on Nand false LLR distribution may include more or fewer components than shown, or may combine certain components, or may have a different arrangement of components.
The present invention also provides a computer readable storage medium, which may be a non-volatile computer readable storage medium, and may also be a volatile computer readable storage medium, where instructions are stored in the computer readable storage medium, when the instructions are executed on a computer, cause the computer to perform the steps of the method for simulating the Nand false LLR distribution.
In the context of this disclosure, a machine-readable medium may be a tangible medium that can contain, or store a program for use by or in connection with an instruction execution system, apparatus, or device. The machine-readable medium may be a machine-readable signal medium or a machine-readable storage medium. The machine-readable medium may include, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any suitable combination of the foregoing. More specific examples of a machine-readable storage medium would include an electrical connection based on one or more wires, a portable computer diskette, a hard disk, a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.
Moreover, although operations are depicted in a particular order, this should be understood as requiring that such operations be performed in the particular order shown or in sequential order, or that all illustrated operations be performed, to achieve desirable results. In certain circumstances, multitasking and parallel processing may be advantageous. Likewise, while several specific implementation details are included in the above discussion, these should not be construed as limiting the scope of the present disclosure. Certain features that are described in the context of separate embodiments can also be implemented in combination in a single implementation. Conversely, various features that are described in the context of a single implementation can also be implemented in multiple implementations separately or in any suitable subcombination.
Although the subject matter has been described in language specific to structural features and/or methodological acts, it is to be understood that the subject matter defined in the appended claims is not necessarily limited to the specific features or acts described above. Rather, the specific features and acts described above are example forms of implementing the claims.
Claims (10)
1. The method for simulating the Nand false LLR distribution is characterized by comprising the following steps:
generating a deep learning model to be trained;
taking preset environmental data and a preset threshold voltage distribution data set as training data, and executing model training operation on the deep learning model to obtain a vth offset prediction model;
when threshold voltage distribution data of Nand false is detected, invoking a pre-trained vth offset prediction model to execute multi-condition simulation prediction operation on the threshold voltage distribution data to obtain multiple groups of simulation data;
and acquiring the Nand false LLR distribution corresponding to the Nand false in the model data.
2. The method for modeling a Nand false LLR distribution as defined in claim 1, wherein the step of calling a pre-trained vth shift prediction model to perform a multi-condition modeling prediction operation on the threshold voltage distribution data when the threshold voltage distribution data of Nand false is detected, the step of obtaining a plurality of sets of modeling data includes:
when threshold voltage distribution data of Nand false is detected, invoking a pre-trained vth offset prediction model to generate a plurality of simulation conditions, wherein the time, temperature, PE or Read Disturb in the simulation conditions are different;
and executing multi-condition simulation prediction operation based on the threshold voltage distribution data and the simulation conditions to obtain a plurality of simulation data.
3. The method for simulating the Nand false LLR distribution according to any one of claims 1-2, wherein the step of performing model training operation on the deep learning model with the preset environmental data and the threshold voltage distribution data set corresponding to the sample Nand false as training data to obtain the vth shift prediction model comprises:
and taking the preset temperature, the preset time parameter, the preset PE, the preset Read Disturb and the threshold voltage distribution data set as training data, and executing model training operation on the deep learning model to obtain a vth offset prediction model.
4. The method of modeling a Nand false LLR distribution as defined in claim 1, wherein after the step of obtaining the Nand false LLR distribution corresponding to the Nand false LLR in the model data, the method further comprises:
analyzing a plurality of groups of model data to obtain analysis results;
and obtaining the optimal Nand false LLR distribution corresponding to the Nand false LLR in the analysis result according to a preset LLR selection standard.
5. The method for modeling a Nand false LLR distribution as defined in claim 4, wherein the root preset LLR selection criteria, after the step of obtaining the optimal Nand false LLR distribution corresponding to the Nand false LLR in the parsing result, further comprises:
obtaining a target condition corresponding to the optimal Nand false LLR;
and outputting prompt information capable of improving the LDPC decoding capability under the target condition.
6. The method for modeling a Nand false LLR distribution according to claim 1, wherein when the threshold voltage distribution data of Nand false is detected, invoking a pre-trained vth shift prediction model to perform a multi-condition modeling prediction operation on the threshold voltage distribution data, and obtaining a plurality of sets of modeling data, the method further comprises:
analyzing a plurality of groups of model data to obtain analysis results;
and acquiring the RRT group corresponding to the Nand false from the analysis result.
7. The method for modeling a Nand false LLR distribution as defined in claim 6, wherein the step of obtaining the RRT group corresponding to the Nand false LLR in the analysis result includes:
and acquiring the optimal RRT group corresponding to the Nand false from the analysis result according to a preset RRT group selection standard.
8. The method of modeling a Nand false LLR distribution as defined in claim 1, wherein after the step of obtaining the Nand false LLR distribution corresponding to the Nand false LLR in the model data, the method further comprises:
outputting the Nand false LLR distribution.
9. A simulation apparatus for Nand false LLR distribution, wherein the simulation apparatus for Nand false LLR distribution comprises: a memory and at least one processor, the memory having instructions stored therein, the memory and the at least one processor being interconnected by a line;
the at least one processor invoking the instructions in the memory to cause the simulation device of Nand false LLR distribution to perform the simulation method of Nand false LLR distribution as in any one of claims 1-8.
10. A computer readable storage medium having stored thereon a computer program, which when executed by a processor implements a method of simulating Nand false LLR distribution according to any of claims 1-8.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311537535.8A CN117574835A (en) | 2023-11-16 | 2023-11-16 | Method, equipment and storage medium for simulating Nand false LLR distribution |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311537535.8A CN117574835A (en) | 2023-11-16 | 2023-11-16 | Method, equipment and storage medium for simulating Nand false LLR distribution |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN117574835A true CN117574835A (en) | 2024-02-20 |
Family
ID=89892992
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311537535.8A Pending CN117574835A (en) | 2023-11-16 | 2023-11-16 | Method, equipment and storage medium for simulating Nand false LLR distribution |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117574835A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN118502677A (en) * | 2024-07-17 | 2024-08-16 | 成都佰维存储科技有限公司 | Memory read re-table optimization method and device, readable storage medium and electronic equipment |
-
2023
- 2023-11-16 CN CN202311537535.8A patent/CN117574835A/en active Pending
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN118502677A (en) * | 2024-07-17 | 2024-08-16 | 成都佰维存储科技有限公司 | Memory read re-table optimization method and device, readable storage medium and electronic equipment |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN112508243B (en) | Training method and device for multi-fault prediction network model of power information system | |
| CN111881023B (en) | Software aging prediction method and device based on multi-model comparison | |
| RU2720586C2 (en) | System for determining the allowable correlation limit using iterative cross validation and method of its implementation | |
| KR20190123040A (en) | Solar power generation prediction model management APPARATUS AND METHOD for setting the renewal cycle of the deep run-based solar power generation prediction model | |
| CN116307269B (en) | Photovoltaic power generation power prediction method and device based on artificial intelligence | |
| CN116148656B (en) | Portable analog breaker fault detection method | |
| CN117574835A (en) | Method, equipment and storage medium for simulating Nand false LLR distribution | |
| CN117688881B (en) | Integrated circuit verification method and device based on artificial intelligence and storage medium | |
| CN115017819A (en) | Engine remaining service life prediction method and device based on hybrid model | |
| CN115389743A (en) | Method, medium and system for predicting content interval of dissolved gas in transformer oil | |
| CN111930728A (en) | Method and system for predicting characteristic parameters and fault rate of equipment | |
| CN111612648A (en) | Training method and device of photovoltaic power generation prediction model and computer equipment | |
| CN117010723A (en) | Carbon emission prediction method, system, apparatus, and computer-readable storage medium | |
| CN111160419B (en) | Deep learning-based electronic transformer data classification prediction method and device | |
| CN115345376A (en) | Method and device for predicting oxygen content of boiler flue gas | |
| CN113139332A (en) | Automatic model construction method, device and equipment | |
| CN113569324A (en) | Slope deformation monitoring abnormal data analysis and optimization method | |
| CN112685957A (en) | Method for predicting relay protection defects | |
| CN113627088B (en) | Machine performance degradation evaluation method and system based on gene programming and data fusion | |
| CN115186597B (en) | Rotary multi-component degradation coupling influence assessment method and system based on transfer learning | |
| CN118034733B (en) | Automobile fault detection model updating and fault pre-judging method and device | |
| US20240362461A1 (en) | Anomaly detection using a pre-trained global model | |
| KR20240047570A (en) | Method of determining dictionary learning model for remaining life prediction transfer learning and apparatus performing thereof | |
| CN118966317A (en) | Prediction model correction method and device based on active migration learning and equipment medium | |
| CN118376892A (en) | Cable aging non-damage testing method and system |
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 |