A three-dimensional flash memory aging test method and system

Through the double row double-sided gold finger testing method, efficient multi-channel testing of three-dimensional flash aging test is achieved, solving the problems of low efficiency and low resource utilization in the existing technology, and providing more comprehensive aging situation evaluation and quality control.

CN118038958BActive Publication Date: 2024-11-08ZHENHE TECH (JIANGSU) CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202410280359.2
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-03-12
Publication Date
2024-11-08
Estimated Expiration
2044-03-12

AI Technical Summary

Technical Problem

The existing three-dimensional flash aging test methods are inefficient and have low resource utilization, and cannot effectively process and analyze large amounts of test data, lack accurate identification and targeted optimization of abnormal data, and lack flexibility and accuracy in the allocation of test resources.

Method used

Using the double row and double-sided gold finger testing method, through multi-channel aging test, obtain test data, calculate data deviation values, split test tasks, optimize resource allocation, improve test efficiency, ensure sufficient test of each channel, and pass abnormal data Identification and retesting improve test accuracy.

Benefits of technology

It significantly improves the efficiency and resource utilization of 3D flash aging tests, and can test 600 channels simultaneously, enhancing the parallelism and accuracy of the tests, reducing the connection failure rate, and providing more comprehensive aging situation evaluation and quality control.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118038958B_ABST
    Figure CN118038958B_ABST
Patent Text Reader

Abstract

The present invention proposes a three-dimensional flash memory aging test method and system, which relates to the technical field of three-dimensional flash memory aging test, performs double-row double-sided gold finger test connection, obtains aging test tasks, and performs multi-channel aging test on three-dimensional flash memory; obtains first test data, obtains data deviation values ​​of abnormal data types according to the first test data, splits the aging test tasks according to the abnormal data types, obtains test subtasks, performs retests according to the test subtasks, and obtains retest data; allocates test resources to the test subtasks according to the data deviation values, obtains second test data, calculates the test optimization rate, and then calculates the test error value, evaluates the test aging degree, and obtains the evaluation result. The present invention improves the efficiency and accuracy of the three-dimensional flash memory multi-channel aging test by optimizing the test process and configuration, and provides strong support for product quality control and life prediction.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The invention proposes a three-dimensional flash memory aging test method and system, and relates to the technical field of three-dimensional flash memory aging test, and specifically to the technical field of three-dimensional flash memory aging test with double-row and double-sided gold fingers. Background Art

[0002] Due to its large capacity, 3D-NAND FLASH usually requires a very long test time during electrical performance testing, especially aging testing. The general solution in the industry is to maximize the test density of aging testing and use high-density multi-channel aging testing to share the test cost of each chip. However, the aging test board commonly used in the market cannot have more than 300 gold finger contacts due to the limited width of the equipment furnace. Traditional aging test methods usually use single-channel testing, that is, only one channel is tested at a time. However, this method has problems such as low test efficiency and low test resource utilization. In order to improve test efficiency, multi-channel aging test technology came into being. Multi-channel aging testing can test multiple channels at the same time, thereby significantly improving the test speed. However, multi-channel aging testing also brings new challenges. Due to the substantial increase in the amount of test data, how to effectively process and analyze this data has become an urgent problem to be solved. At the same time, different channels may have different aging conditions, so it is necessary to perform refined testing and optimization for each channel. In existing technologies, although there are some methods for processing and analyzing test data, they often lack accurate identification and targeted optimization of abnormal data. In addition, the allocation of test resources is often based on experience or fixed rules, lacking flexibility and precision. Summary of the invention

[0003] The present invention provides a three-dimensional flash memory aging test method and system, which are used to solve the problem that the aging test board commonly used in the market cannot have more than 300 gold finger contacts due to the limited width of the equipment furnace. Due to the substantial increase in the amount of test data, how to effectively process and analyze the test data has become an urgent problem to be solved. At the same time, different channels may have different aging conditions, so it is necessary to perform refined testing and optimization for each channel. In the existing technology, although there are some methods for processing and analyzing test data, they often lack accurate identification and targeted optimization of abnormal data. In addition, the allocation of test resources is often based on experience or fixed rules, lacking flexibility and accuracy.

[0004] The present invention provides a three-dimensional flash memory aging test method and system, the method comprising:

[0005] a. Perform a double-row double-sided gold finger test connection to obtain an aging test task, wherein the double-row double-sided gold finger performs a multi-channel aging test on the three-dimensional flash memory according to the aging test task;

[0006] b. Obtain first test data, obtain data deviation values ​​of abnormal data types according to the first test data, split the aging test task according to the abnormal data types, obtain test subtasks, perform retests according to the test subtasks, and obtain retest data;

[0007] c. Allocate test resources to the test subtasks according to the data deviation value, obtain the second test data, calculate the test optimization rate, and then calculate the test error value, evaluate the test aging degree, and obtain the evaluation result.

[0008] Further, the a includes:

[0009] The gold fingers in the PCB are staggered in double rows of contacts, the whole PCB is inserted into the connector, each row of gold fingers is contacted with its corresponding contact, and then the electrical connection is obtained, and the electrical signal pins are matched with the mechanical position to obtain the double-row double-sided gold finger test connection;

[0010] A preset aging test task is obtained, and a multi-channel aging test is performed on the three-dimensional flash memory according to the aging test task.

[0011] Further, said b includes:

[0012] Acquire test data in a multi-channel aging test, obtain first test data, pre-process the first test data, classify the first test data according to data types, obtain multiple types of data, compare the multiple types of data with corresponding thresholds respectively, obtain abnormal data types according to the comparison results, calculate the difference between the type data and the corresponding abnormal data, obtain data deviation values, sort the data deviation values, and obtain abnormal sorting of the type data;

[0013] A corresponding test subtask is obtained according to the type of abnormal data, a corresponding channel or a closest channel is retrieved in the multi-channel aging test according to the test subtask, and a targeted retest is performed on the abnormal data to obtain retest data.

[0014] Further, said c includes:

[0015] Acquire test resources, allocate test resources to test subtasks from most to least based on data deviation values ​​combined with exception sorting, and obtain resource allocation information;

[0016] After the test resource allocation is completed, the test data in the test process is acquired again to obtain second test data, and the test optimization rate is calculated according to the first test data, the retest data and the second test data to obtain the test optimization result.

[0017] Further, the calculating of the test error value, evaluating the test aging degree, and obtaining the evaluation result includes:

[0018] A test error value is calculated according to the test optimization rate, the second test data and the first test data, and the test error value is analyzed for the test situation to obtain an analysis result; the proportion of the test error value in the total test value is obtained to obtain an error proportion result.

[0019] The test aging degree is evaluated according to the analysis result to obtain an evaluation result.

[0020] Furthermore, the system comprises:

[0021] An aging test connection module is used to perform a double-row double-sided gold finger test connection and obtain an aging test task. The double-row double-sided gold finger performs a multi-channel aging test on the three-dimensional flash memory according to the aging test task.

[0022] A splitting and retesting module is used to obtain first test data, obtain data deviation values ​​of abnormal data types according to the first test data, split the aging test task according to the abnormal data types, obtain test subtasks, retest according to the test subtasks, and obtain retest data;

[0023] The resource allocation and evaluation module is used to allocate test resources to the test subtasks according to the data deviation value, obtain the second test data, calculate the test optimization rate, and then calculate the test error value, evaluate the test aging degree, and obtain the evaluation result.

[0024] Furthermore, the aging test connection module includes:

[0025] The connection test module is used to set the double-row contact staggered positions of the gold fingers in the PCB, insert the whole PCB into the connector, make each row of gold fingers contact with its corresponding contact, and then obtain electrical connection, match the electrical signal pins with the mechanical position, and obtain the double-row double-sided gold finger test connection;

[0026] The task acquisition module is used to acquire a preset aging test task and perform a multi-channel aging test on the three-dimensional flash memory according to the aging test task.

[0027] Furthermore, the splitting and retesting module includes:

[0028] an abnormal data acquisition module, used to acquire test data in a multi-channel aging test, acquire first test data, pre-process the first test data, classify the first test data according to data types, acquire multiple types of data, compare the multiple types of data with corresponding thresholds respectively, acquire the type of abnormal data according to the comparison result, perform difference calculation between the type data and the corresponding abnormal data, acquire data deviation values, sort the data deviation values, and obtain abnormal sorting of the type data;

[0029] The subtask extraction module is used to obtain the corresponding test subtask according to the type of abnormal data, call the corresponding channel or the closest channel in the multi-channel aging test according to the test subtask, perform targeted retest on the abnormal data, and obtain retest data.

[0030] Furthermore, the resource allocation and evaluation module includes:

[0031] The resource allocation module is used to obtain test resources, allocate test resources to test subtasks from more to less according to data deviation values ​​combined with abnormal sorting, and obtain resource allocation information;

[0032] The optimization calculation module is used to obtain the test data in the test process again after the test resource allocation is completed, obtain the second test data, calculate the test optimization rate according to the first test data, the retest data and the second test data, and obtain the test optimization result.

[0033] Furthermore, the optimization calculation module also includes:

[0034] The error analysis module is used to calculate the test error value according to the test optimization rate, the second test data and the first test data, analyze the test situation of the test error value, and obtain the analysis result; obtain the proportion of the test error value in the total test value, and obtain the error proportion result.

[0035] The aging evaluation module is used to evaluate the test aging degree according to the analysis result to obtain an evaluation result.

[0036] Beneficial effects of the present invention:

[0037] The present invention proposes a three-dimensional flash memory aging test method and system. The present invention designs a special layout of the gold finger of the aging test board, and utilizes the staggered layout of the contacts in the same plug-in stroke within the limited PCB width to double the number of contacts while keeping the electrical contact spacing unchanged, while not reducing the load current of a single contact. The special design of the present invention is suitable for large-capacity, high-density, multi-channel aging tests of three-dimensional flash memory (3D-NAND FLASH).

[0038] The "double-row double-sided gold finger" layout design of the present invention completely changes the single-row gold finger design of the general aging test board. Within the same PCB connector width, the number of electrical channels is doubled while the current carrying capacity of a single contact is not reduced. The nearly doubled test signal channel and power supply channel can be introduced into the furnace body for high-parallel aging test of three-dimensional flash memory, thereby greatly increasing the number of simultaneous tests of mass production aging tests, greatly improving the test efficiency of users, and greatly saving user testing costs. At the same time, the electrical signal pins defined in the present invention match the mechanical position, which is the optimal configuration after several rounds of improvements in the optimization of PCB wiring and the matching of test equipment resources. The present invention has been put into actual production product application of our company, and its electrical characteristics are in line with the production process parameters and the long-term test stability requirements after actual measurement of high and low temperature products. The actual measurement data show that the double-row double-sided gold finger design of the present invention strengthens the power supply stability of product testing and also significantly helps to improve the product test yield. The present invention doubles the test digital channel, improves the test efficiency of high-end aging test equipment by about 1.6 times, and can truly achieve cost-effectiveness for the aging test link of the mass production line. Through the setting and connection test of double-row double-sided gold fingers, the number of test interface contacts can be doubled to a maximum of 600 channels while the overall aging board width remains unchanged. This is crucial for multi-channel aging tests, which can significantly improve test efficiency, increase the stability and reliability of gold fingers, and reduce the connection failure rate during the test. Multi-channel aging tests can perform aging tests on multiple channels of three-dimensional flash memory at the same time, improving the parallelism and efficiency of the test. This method can more comprehensively detect the aging of three-dimensional flash memory and ensure that each channel can be fully tested. According to the first test data, the data deviation value of the abnormal data type is obtained, and the abnormal situation in the aging test can be accurately identified, which provides a basis for subsequent test optimization. The aging test task is split according to the abnormal data type, and the complex test task is decomposed into multiple simple test subtasks, which reduces the complexity and difficulty of the test. At the same time, unnecessary test tasks are greatly reduced, saving test time. Through retesting, the test subtasks can be tested and verified more finely, improving the accuracy and reliability of the test. The allocation of test resources to the test subtasks according to the data deviation value can ensure that the test resources are reasonably allocated, avoiding resource waste and bottleneck problems. Calculating the test optimization rate can quantitatively evaluate the effect of test process improvement and provide data support for subsequent test optimization. Calculating the test error value can accurately evaluate the error conditions that occur during the test, which helps to locate problems and make improvements. By evaluating the test aging degree, we can fully understand the aging of 3D flash memory, which provides an important basis for product quality control and life prediction. The evaluation results can be used to guide subsequent test optimization and product improvement, and improve product reliability and stability. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] Figure 1 A schematic diagram of a three-dimensional flash memory aging test method;

[0040] Figure 2 It is a general design schematic diagram and physical picture of the gold finger area of ​​the universal three-dimensional flash memory aging test board;

[0041] Figure 3 It is a schematic diagram and a physical diagram of the special double-row gold finger plug layout design of the present invention;

[0042] Figure 4 It is a schematic diagram of the electrical signal definition (540 channels) of the dedicated double-row gold finger;

[0043] Figure 5 Define the schematic diagram for the design of a dedicated double-row gold finger mechanical structure. DETAILED DESCRIPTION

[0044] The preferred embodiments of the present invention are described below in conjunction with the accompanying drawings. It should be understood that the preferred embodiments described herein are only used to illustrate and explain the present invention, and are not used to limit the present invention.

[0045] In one embodiment of the present invention, a three-dimensional flash memory aging test method and system are provided, and the method includes:

[0046] a. Perform a double-row double-sided gold finger test connection to obtain an aging test task, wherein the double-row double-sided gold finger performs a multi-channel aging test on the three-dimensional flash memory according to the aging test task;

[0047] b. Obtain first test data, obtain data deviation values ​​of abnormal data types according to the first test data, split the aging test task according to the abnormal data types, obtain test subtasks, perform retests according to the test subtasks, and obtain retest data;

[0048] c. Allocate test resources to the test subtasks according to the data deviation value, obtain the second test data, calculate the test optimization rate, and then calculate the test error value, evaluate the test aging degree, and obtain the evaluation result.

[0049] The working principle of the above technical solution is: perform a double-row double-sided gold finger test connection to obtain an aging test task, and the double-row double-sided gold finger performs a multi-channel aging test on the three-dimensional flash memory according to the aging test task; obtain the first test data, obtain the data deviation value of the abnormal data type according to the first test data, split the aging test task according to the abnormal data type, obtain the test subtask, perform a retest according to the test subtask, and obtain retest data; allocate test resources to the test subtask according to the data deviation value, obtain the second test data, calculate the test optimization rate, and then calculate the test error value, evaluate the test aging degree, and obtain the evaluation result.

[0050] The technical effect of the above technical solution is: through the setting and connection test of double-row double-sided gold fingers, the number of test interface contacts can be doubled to a maximum of 600 channels under the condition that the overall aging board width remains unchanged, which is crucial for multi-channel aging test, can significantly improve test efficiency, increase the stability and reliability of gold fingers, and reduce the connection failure rate during the test. Multi-channel aging test can perform aging test on multiple channels of three-dimensional flash memory at the same time, improving the parallelism and efficiency of the test. This method can more comprehensively detect the aging of three-dimensional flash memory and ensure that each channel can be fully tested. According to the first test data, the data deviation value of the abnormal data type is obtained, and the abnormal situation in the aging test can be accurately identified, which provides a basis for subsequent test optimization. The aging test task is split according to the abnormal data type, and the complex test task is decomposed into multiple simple test subtasks, which reduces the complexity and difficulty of the test. At the same time, unnecessary test tasks are greatly reduced, saving test time. Through retesting, the test subtasks can be tested and verified more finely, improving the accuracy and reliability of the test. Allocating test resources to test subtasks based on data deviation values ​​can ensure that test resources are allocated reasonably, avoiding resource waste and bottleneck problems. Calculating the test optimization rate can quantitatively evaluate the effect of test process improvement and provide data support for subsequent test optimization. Calculating the test error value can accurately evaluate the errors that occur during the test, which helps to locate problems and make improvements. By evaluating the test aging degree, the aging of 3D flash memory can be fully understood, providing an important basis for product quality control and life prediction. The evaluation results can be used to guide subsequent test optimization and product improvement to improve product reliability and stability.

[0051] In one embodiment of the present invention, the a comprises:

[0052] The gold fingers in the PCB are staggered in double rows of contacts, the whole PCB is inserted into the connector, each row of gold fingers is contacted with its corresponding contact, and then the electrical connection is obtained, and the electrical signal pins are matched with the mechanical position to obtain the double-row double-sided gold finger test connection;

[0053] A preset aging test task is obtained, and a multi-channel aging test is performed on the three-dimensional flash memory according to the aging test task.

[0054] The working principle of the above technical solution is: double-row contact staggering is performed on the gold fingers in the PCB, and the staggering spacing is set to 1.27mm, so that each row of gold fingers has a 2.54mm spacing and a corresponding copper foil width, the entire PCB is inserted into the connector, each row of gold fingers is contacted with its corresponding contact, and then an electrical connection is obtained, and the electrical signal pins are matched with the mechanical position to obtain a double-row double-sided gold finger test connection;

[0055] A preset aging test task is obtained, and a multi-channel aging test is performed on the three-dimensional flash memory according to the aging test task.

[0056] In the prior art, such as Figure 2 As shown, the general design of the gold finger area of ​​the general three-dimensional flash memory aging test board is characterized. In the figure, Size A indicates the width of a gold finger connector PCB; Size B indicates that the maximum width of the general aging test board is 450mm. Due to the limitations of the aging furnace structure design of this general design aging test board, the design width of the BIB is generally a fixed value of 450MM. Due to this width limitation, the maximum width of a single connector of "Size A" is also limited. Because the width of the contact must be guaranteed to ensure the current carrying capacity of the contact, the best gold finger contact spacing obtained after practical optimization in the industry is 2.54mm. Obviously, designing contacts with a spacing of 2.54mm within a certain width range results in a limited number of contacts on the maximum gold finger interface. Generally, the maximum number of connections of the gold finger interface contacts of the aging test board with a width of 450mm is designed to be 300 channels. In the context of this universal aging board design, since the maximum number of channels of this electrical connector is limited to less than 300, and these channels must be classified into ground contacts, test power contacts, ID coding contacts, digital test contacts, etc., different hardware resources of the test machine are connected to the test socket area of ​​the aging test board for electrical performance testing. Therefore, the number of channels on this interface directly determines the maximum simultaneous testing capability of the overall aging test system. Obviously, when running the same test program, the greater the number of parallel tests, the higher the test efficiency and the lower the test cost of a single chip.

[0057] In the present invention, Figure 3As shown in the figure, the dedicated double-row gold finger plug layout design is characterized. When the overall aging board width remains unchanged, the number of test interface contacts can be doubled to a maximum of 600 channels.

[0058] The "double-row double-sided gold finger" layout design of the present invention has completely changed the single-row gold finger design of the general aging test board. By utilizing the 1.27mm spacing design of the upper and lower double-row contacts, each row of gold fingers has the original 2.54mm spacing and its corresponding copper foil width, which is the basis for ensuring that the current carrying capacity and contact characteristics of a single contact remain unchanged. After the overall PCB is inserted into the connector, each row of gold fingers will contact its corresponding contact to achieve an electrical connection. In this way, within the same PCB connector width, the number of electrical channels can be doubled while the current carrying capacity of a single contact can be maintained. The nearly doubled test signal channels and power supply channels can be introduced into the furnace body for high-parallelism aging tests of three-dimensional flash memories, thereby greatly increasing the number of simultaneous tests for mass production aging tests, greatly improving the user's testing efficiency, and greatly saving the user's testing costs. At the same time, the electrical signal pins defined in the present invention match the mechanical position, which is the optimal configuration after several rounds of improvements in the optimization of PCB wiring and the matching of test equipment resources.

[0059] The technical effect of the above technical scheme is as follows: the method is to design a special layout test method for the gold finger of the aging test board, and utilize the staggered layout mode of the contacts in the same plug-in stroke within the limited PCB width to double the number of contacts while keeping the electrical contact spacing unchanged, and at the same time not reducing the load current of a single contact. The special design of the present invention is suitable for large-capacity, high-density, multi-channel aging test of three-dimensional flash memory (3D-NAND FLASH). The "double-row double-sided gold finger" layout design of the present invention completely changes the single-row gold finger design of the general aging test board, and within the same PCB connector width, the number of electrical channels is doubled while the current carrying capacity of a single contact is not reduced. The nearly doubled test signal channel and power supply channel can be introduced into the furnace body for high-parallel aging test of the three-dimensional flash memory, thereby greatly increasing the number of simultaneous tests of mass production aging tests, greatly improving the test efficiency of users, and greatly saving the test cost of users. At the same time, the matching of the electrical signal pins and mechanical positions defined in the present invention is the optimal configuration after several rounds of improvements in the optimization of PCB wiring and the matching of test equipment resources. This method has been put into actual production product application by our company, and its electrical characteristics have been tested on high and low temperature products and are in line with the production process parameters and the long-term test stability requirements. The measured data results show that the double-row double-sided gold finger design of the present invention enhances the power supply stability of product testing and also significantly helps improve the product test yield. The present invention doubles the test digital channels, which improves the test efficiency of high-end aging test equipment by about 1.6 times, and can truly achieve cost-effectiveness for the aging test link of mass production lines.

[0060] In one embodiment of the present invention, said b comprises:

[0061] Acquire test data in a multi-channel aging test, obtain first test data, pre-process the first test data, classify the first test data according to data types, obtain multiple types of data, compare the multiple types of data with corresponding thresholds respectively, obtain abnormal data types according to the comparison results, calculate the difference between the type data and the corresponding abnormal data, obtain data deviation values, sort the data deviation values, and obtain abnormal sorting of the type data;

[0062] A corresponding test subtask is obtained according to the type of abnormal data, a corresponding channel or a closest channel is retrieved in the multi-channel aging test according to the test subtask, and a targeted retest is performed on the abnormal data to obtain retest data.

[0063] The working principle of the above technical solution is: obtain the test data in the multi-channel aging test, obtain the first test data, pre-process the first test data, classify the first test data according to the data type, obtain multiple types of data, compare the multiple types of data with the corresponding thresholds respectively, obtain the type of abnormal data according to the comparison result, calculate the difference between the type data and the corresponding abnormal data, obtain the data deviation value, sort the data deviation value, and obtain the abnormal sorting of the type data; compare each type of data with the corresponding type threshold, when the specific data is greater than its corresponding type threshold, the specific data is determined as abnormal data, and its corresponding type is the abnormal data type. The type data includes contact resistance data, insulation resistance data and temperature data, etc. The data deviation value is equal to |type data-corresponding threshold|.

[0064] According to the type of abnormal data, a corresponding test subtask is obtained. According to the test subtask, a corresponding channel or the closest channel is called in the multi-channel aging test, and the abnormal data is retested in a targeted manner to obtain retest data. The tasks corresponding to the type of abnormal data in the aging test task are extracted and split to obtain test subtasks. The multi-channel aging test includes multiple test channels, and the corresponding channel called is a channel in the multi-channel that can obtain retest data corresponding to the abnormal data. When there is no corresponding channel, the channel that can obtain the most retest data, that is, the closest channel, is called.

[0065] The technical effects of the above technical scheme are as follows: preprocessing the first test data can eliminate noise and interference in the data and improve the accuracy and reliability of the data. Classifying according to the data type makes the subsequent analysis and processing more targeted and efficient. Comparing the type data with the corresponding threshold value can quickly and accurately identify the type of abnormal data, providing a clear direction for subsequent optimization and improvement. Calculating and sorting the data deviation value can quantitatively evaluate the degree of data deviation, providing a basis for subsequent test resource allocation and test subtask splitting. Splitting the test subtask according to the type of abnormal data can decompose complex test tasks into multiple specific subtasks, thereby improving the pertinence and efficiency of the test. In the multi-channel aging test, calling the corresponding channel or the closest channel for targeted retesting can ensure in-depth analysis and verification of the abnormal data, thereby improving the accuracy and reliability of the test. The design of the multi-channel aging test allows multiple channels to be tested simultaneously, thereby improving the test throughput. During retesting, the channel that can directly obtain the corresponding abnormal data is given priority. When such a channel does not exist, the closest channel is selected, which ensures the maximum utilization of test resources and reduces resource waste. At the same time, flexible allocation of resources is achieved. By identifying abnormal data types and retesting them in a targeted manner, problems can be quickly located and improved, reducing unnecessary test steps and time. By splitting test subtasks and allocating resources rationally, the test process is made more efficient and smoother. Through refined testing and data processing, the performance and aging of products can be more fully understood, improving product quality and reliability. Evaluating the degree of test aging helps to promptly identify potential problems and make improvements, enhancing product competitiveness.

[0066] In one embodiment of the present invention, said c comprises:

[0067] Acquire test resources, allocate test resources to test subtasks from most to least based on data deviation values ​​combined with exception sorting, and obtain resource allocation information;

[0068] After the test resource allocation is completed, the test data in the test process is acquired again to obtain second test data, and the test optimization rate is calculated according to the first test data, the retest data and the second test data to obtain the test optimization result.

[0069] The working principle of the above technical solution is: obtain test resources, allocate test resources from more to less to test subtasks according to the data deviation value combined with the abnormal sorting, and obtain resource allocation information; the test resources are allocated from more to less according to the abnormal sorting, the higher the abnormal sorting, the more resources are allocated, and the resource allocation amount is the proportion of the data deviation value in the total data deviation value, and the proportion of the resource allocation amount between the total allocation amount is obtained according to the proportion, and the resource allocation amount is obtained. The test resources include test time, test rate, and test tools. After the test resource allocation is completed, the test data in the test process is obtained again to obtain the second test data, and the test optimization rate is calculated according to the first test data, the retest data and the second test data to obtain the test optimization result. When the test optimization rate is greater than the preset optimization threshold, the test state is a qualified state, when the test optimization rate is less than or equal to the preset optimization threshold, the test state is an unqualified state, when the test state is an unqualified state, the preset parameters are adjusted according to the preset strategy, and retesting is performed until the test state is a qualified state. When the test state is a qualified state, the test error value is calculated.

[0070] The calculation formula of the test optimization rate is:

[0071] ,

[0072] in, To test the optimization rate, is the total data deviation value of the retest data, is the total data deviation value of the first test data, is the total data deviation value of the second test data, is the total amount of the first test data, is the total amount of retest data, is the total data volume of the second test data.

[0073] The technical effect of the above technical solution is: by combining the data deviation value with the anomaly ranking, test resources are allocated to the test subtasks, ensuring the rationality and effectiveness of resource allocation. The subtasks with higher anomaly rankings, that is, the subtasks with larger data deviation values, will obtain more test resources, which helps to prioritize key issues and improve test efficiency. The amount of resource allocation is determined based on the proportion of data deviation values ​​in the total data deviation values, making resource allocation more accurate and reasonable. This allocation method can ensure that test resources are fully utilized, avoid resource waste, and meet the resource needs of different subtasks.

[0074] By calculating the test optimization rate, we can quantitatively evaluate the improvement effect of the test process and provide clear data support for subsequent test optimization. The proportion of the total data deviation value of the retest data in the total data deviation value of the first test data can be calculated. Relatively speaking, The bigger, The bigger, The smaller, The smaller the value, the larger the value. The smaller the value, the smaller the optimization rate, and vice versa. By comparing the test optimization rate with the preset optimization threshold, it is possible to determine whether the test status is qualified, so that problems can be discovered and adjusted in a timely manner. When the test status is unqualified, the preset parameters are adjusted according to the preset strategy and the test is re-performed until the test status is qualified, which ensures the continuous improvement of the test process and the reliability of the test results. By continuously optimizing the test process and resource allocation, the accuracy and efficiency of the test can be improved, thereby improving the quality and reliability of the product. When the test status is qualified, calculating the test error value helps to further analyze the problems and potential risks that occur during the test process. By analyzing the test error value, targeted suggestions and directions can be provided for subsequent product improvements and test optimization.

[0075] In one embodiment of the present invention, the step of calculating the test error value, evaluating the test aging degree, and obtaining the evaluation result includes:

[0076] A test error value is calculated according to the test optimization rate, the second test data and the first test data, and the test error value is analyzed for the test situation to obtain an analysis result; the proportion of the test error value in the total test value is obtained to obtain an error proportion result.

[0077] The test aging degree is evaluated according to the analysis result to obtain an evaluation result.

[0078] The working principle of the above technical solution is: calculate the test error value according to the test optimization rate, the second test data and the first test data, analyze the test situation based on the test error value, and obtain the analysis result; obtain the proportion of the test error value in the total test value, and obtain the error proportion result. Evaluate the test aging degree according to the analysis result to obtain the evaluation result. The error proportion result is the evaluation result.

[0079] The calculation formula of the test error value is: ,

[0080] in, To test for error values, is the error value of the first test data, To retest the data error value, is the error value of the second test data, To test the optimization rate, is the total amount of the first test data, is the total amount of retest data, is the total data volume of the second test data.

[0081] The greater the proportion of the test error values ​​in the total test values, the worse the aging performance of the 3D flash memory is.

[0082] The technical effect of the above technical solution is: combining the test optimization rate, the second test data and the first test data to calculate the test error value. This calculation method not only considers the degree of optimization during the test, but also combines the changes in the actual test data, thereby obtaining a more accurate and comprehensive error value. The sum of the error-free values ​​of the three types of data can be calculated. The bigger, The smaller the The smaller it is, the smaller the test error value is, and vice versa. Similarly. Accurate error values ​​can more accurately reflect the problems in the test process, providing strong data support for subsequent problem analysis and optimization. By analyzing the test error values, we can gain an in-depth understanding of abnormal situations and potential problems in the test process, providing direction for the optimization and improvement of the test process. The analysis results can help identify bottlenecks and weaknesses in the test, and provide targeted guidance for subsequent testing work. By calculating the error ratio results, we can intuitively understand the proportion of test errors in the total test value, which helps to quickly evaluate the overall quality and reliability of the test.

[0083] The error percentage results are concise and easy to understand and compare, providing decision makers with an important basis for quickly judging the test status. By evaluating the test aging degree based on the analysis results, an accurate judgment of the aging of the 3D flash memory can be obtained. The evaluation results are helpful to predict the product life and performance change trend, providing an important basis for product quality control and after-sales service.

[0084] By analyzing the test error values ​​and evaluating the test aging degree, problems in the test process can be discovered in time and improvements can be made, thereby optimizing the test process and improving test efficiency. Continuous test optimization and error analysis help improve product quality and reliability and enhance product market competitiveness.

[0085] In one embodiment of the present invention, the system comprises:

[0086] An aging test connection module is used to perform a double-row double-sided gold finger test connection and obtain an aging test task. The double-row double-sided gold finger performs a multi-channel aging test on the three-dimensional flash memory according to the aging test task.

[0087] A splitting and retesting module is used to obtain first test data, obtain data deviation values ​​of abnormal data types according to the first test data, split the aging test task according to the abnormal data types, obtain test subtasks, retest according to the test subtasks, and obtain retest data;

[0088] The resource allocation and evaluation module is used to allocate test resources to the test subtasks according to the data deviation value, obtain the second test data, calculate the test optimization rate, and then calculate the test error value, evaluate the test aging degree, and obtain the evaluation result.

[0089] The working principle of the above technical solution is as follows: the aging test connection module is used to perform double-row double-sided gold finger test connection and obtain the aging test task, and the double-row double-sided gold finger performs multi-channel aging test on the three-dimensional flash memory according to the aging test task; the splitting and retesting module is used to obtain the first test data, obtain the data deviation value of the abnormal data type according to the first test data, split the aging test task according to the abnormal data type, obtain the test subtask, retest according to the test subtask, and obtain retest data; the resource allocation and evaluation module is used to allocate test resources to the test subtask according to the data deviation value, obtain the second test data, calculate the test optimization rate, and then calculate the test error value, evaluate the test aging degree, and obtain the evaluation result.

[0090] The technical effect of the above technical solution is: through the setting and connection test of double-row double-sided gold fingers, the number of test interface contacts can be doubled to a maximum of 600 channels under the condition that the overall aging board width remains unchanged, which is crucial for multi-channel aging test, can significantly improve test efficiency, increase the stability and reliability of gold fingers, and reduce the connection failure rate during the test. Multi-channel aging test can perform aging test on multiple channels of three-dimensional flash memory at the same time, improving the parallelism and efficiency of the test. This method can more comprehensively detect the aging of three-dimensional flash memory and ensure that each channel can be fully tested. According to the first test data, the data deviation value of the abnormal data type is obtained, and the abnormal situation in the aging test can be accurately identified, which provides a basis for subsequent test optimization. The aging test task is split according to the abnormal data type, and the complex test task is decomposed into multiple simple test subtasks, which reduces the complexity and difficulty of the test. At the same time, unnecessary test tasks are greatly reduced, saving test time. Through retesting, the test subtasks can be tested and verified more finely, improving the accuracy and reliability of the test. Allocating test resources to test subtasks based on data deviation values ​​can ensure that test resources are allocated reasonably, avoiding resource waste and bottleneck problems. Calculating the test optimization rate can quantitatively evaluate the effect of test process improvement and provide data support for subsequent test optimization. Calculating the test error value can accurately evaluate the errors that occur during the test, which helps to locate problems and make improvements. By evaluating the test aging degree, the aging of 3D flash memory can be fully understood, providing an important basis for product quality control and life prediction. The evaluation results can be used to guide subsequent test optimization and product improvement to improve product reliability and stability.

[0091] In one embodiment of the present invention, the aging test connection module includes:

[0092] The connection test module is used to set the double-row contact staggered positions of the gold fingers in the PCB, insert the whole PCB into the connector, make each row of gold fingers contact with its corresponding contact, and then obtain electrical connection, match the electrical signal pins with the mechanical position, and obtain the double-row double-sided gold finger test connection;

[0093] The task acquisition module is used to acquire a preset aging test task and perform a multi-channel aging test on the three-dimensional flash memory according to the aging test task.

[0094] The working principle of the above technical solution is: the connection test module is used to set the double-row contact offset of the gold fingers in the PCB, and the offset spacing is set to 1.27mm, so that each row of gold fingers has a 2.54mm spacing and its corresponding copper foil width, the entire PCB is inserted into the connector, and each row of gold fingers is contacted with its corresponding contact to obtain an electrical connection, and the electrical signal pins are matched with the mechanical position to obtain a double-row double-sided gold finger test connection; the task acquisition module is used to obtain a preset aging test task, and a multi-channel aging test is performed on the three-dimensional flash memory according to the aging test task.

[0095] The technical effect of the above technical scheme is as follows: the method is to design a special layout test method for the gold finger of the aging test board, and utilize the staggered layout mode of the contacts in the same plug-in stroke within the limited PCB width to double the number of contacts while keeping the electrical contact spacing unchanged, and at the same time not reducing the load current of a single contact. The special design of the present invention is suitable for large-capacity, high-density, multi-channel aging test of three-dimensional flash memory (3D-NAND FLASH). The "double-row double-sided gold finger" layout design of the present invention completely changes the single-row gold finger design of the general aging test board, and within the same PCB connector width, the number of electrical channels is doubled while the current carrying capacity of a single contact is not reduced. The nearly doubled test signal channel and power supply channel can be introduced into the furnace body for high-parallel aging test of the three-dimensional flash memory, thereby greatly increasing the number of simultaneous tests of mass production aging tests, greatly improving the test efficiency of users, and greatly saving the test cost of users. At the same time, the matching of the electrical signal pins and mechanical positions defined in the present invention is the optimal configuration after several rounds of improvements in the optimization of PCB wiring and the matching of test equipment resources. This method has been put into actual production product application by our company, and its electrical characteristics have been tested on high and low temperature products and are in line with the production process parameters and the long-term test stability requirements. The measured data results show that the double-row double-sided gold finger design of the present invention enhances the power supply stability of product testing and also significantly helps improve the product test yield. The present invention doubles the test digital channels, which improves the test efficiency of high-end aging test equipment by about 1.6 times, and can truly achieve cost-effectiveness for the aging test link of mass production lines.

[0096] In one embodiment of the present invention, the splitting and retesting module includes:

[0097] an abnormal data acquisition module, used to acquire test data in a multi-channel aging test, acquire first test data, pre-process the first test data, classify the first test data according to data types, acquire multiple types of data, compare the multiple types of data with corresponding thresholds respectively, acquire the type of abnormal data according to the comparison result, perform difference calculation between the type data and the corresponding abnormal data, acquire data deviation values, sort the data deviation values, and obtain abnormal sorting of the type data;

[0098] The subtask extraction module is used to obtain the corresponding test subtask according to the type of abnormal data, call the corresponding channel or the closest channel in the multi-channel aging test according to the test subtask, perform targeted retest on the abnormal data, and obtain retest data.

[0099] The working principle of the above technical solution is as follows: an abnormal data acquisition module is used to acquire test data in a multi-channel aging test, obtain first test data, pre-process the first test data, classify the first test data according to data type, obtain multiple types of data, compare the multiple types of data with corresponding thresholds respectively, obtain the type of abnormal data according to the comparison result, calculate the difference between the type data and the corresponding abnormal data, obtain the data deviation value, sort the data deviation value, and obtain the abnormal sorting of the type data; compare each type of data with the corresponding type threshold, when the specific data is greater than its corresponding type threshold, the specific data is determined to be abnormal data, and its corresponding type is the abnormal data type. The type data includes contact resistance data, insulation resistance data and temperature data, etc. The data deviation value is equal to |type data-corresponding threshold|.

[0100] A subtask extraction module is used to obtain a corresponding test subtask according to the type of abnormal data, and to call a corresponding channel or a closest channel in the multi-channel aging test according to the test subtask, to perform targeted retesting on the abnormal data, and to obtain retest data. The tasks corresponding to the type of abnormal data in the aging test task are extracted and split to obtain test subtasks. The multi-channel aging test includes multiple test channels, and the corresponding channel called is a channel in the multi-channel that can obtain retest data corresponding to the abnormal data. When there is no corresponding channel, the channel that can obtain the most retest data, that is, the closest channel, is called.

[0101] The technical effects of the above technical scheme are as follows: preprocessing the first test data can eliminate noise and interference in the data and improve the accuracy and reliability of the data. Classifying according to the data type makes the subsequent analysis and processing more targeted and efficient. Comparing the type data with the corresponding threshold value can quickly and accurately identify the type of abnormal data, providing a clear direction for subsequent optimization and improvement. Calculating and sorting the data deviation value can quantitatively evaluate the degree of data deviation, providing a basis for subsequent test resource allocation and test subtask splitting. Splitting the test subtask according to the type of abnormal data can decompose complex test tasks into multiple specific subtasks, thereby improving the pertinence and efficiency of the test. In the multi-channel aging test, calling the corresponding channel or the closest channel for targeted retesting can ensure in-depth analysis and verification of the abnormal data, thereby improving the accuracy and reliability of the test. The design of the multi-channel aging test allows multiple channels to be tested simultaneously, thereby improving the test throughput. During retesting, the channel that can directly obtain the corresponding abnormal data is given priority. When such a channel does not exist, the closest channel is selected, which ensures the maximum utilization of test resources and reduces resource waste. At the same time, flexible allocation of resources is achieved. By identifying abnormal data types and retesting them in a targeted manner, problems can be quickly located and improved, reducing unnecessary test steps and time. By splitting test subtasks and allocating resources rationally, the test process is made more efficient and smoother. Through refined testing and data processing, the performance and aging of products can be more fully understood, improving product quality and reliability. Evaluating the degree of test aging helps to promptly identify potential problems and make improvements, enhancing product competitiveness.

[0102] In one embodiment of the present invention, the resource allocation and evaluation module includes:

[0103] The resource allocation module is used to obtain test resources, allocate test resources to test subtasks from more to less according to data deviation values ​​combined with abnormal sorting, and obtain resource allocation information;

[0104] The optimization calculation module is used to obtain the test data in the test process again after the test resource allocation is completed, obtain the second test data, calculate the test optimization rate according to the first test data, the retest data and the second test data, and obtain the test optimization result.

[0105] The working principle of the above technical solution is as follows: a resource allocation module is used to obtain test resources, and allocate test resources from more to less to test subtasks according to the data deviation value combined with the abnormal ranking to obtain resource allocation information; test resources are allocated from more to less according to the abnormal ranking, and the higher the abnormal ranking, the more resources are allocated, and the resource allocation amount is the proportion of the data deviation value in the total data deviation value. According to the proportion, the proportion of the resource allocation amount in the total allocation amount is obtained to obtain the resource allocation amount. The test resources include test time, test rate, test tools, etc.

[0106] The optimization calculation module is used to obtain the test data in the test process again after the test resource allocation is completed, obtain the second test data, calculate the test optimization rate according to the first test data, the retest data and the second test data, and obtain the test optimization result. When the test optimization rate is greater than the preset optimization threshold, the test state is a qualified state, when the test optimization rate is less than or equal to the preset optimization threshold, the test state is an unqualified state, when the test state is an unqualified state, adjust the preset parameters according to the preset strategy, and perform retesting until the test state is a qualified state. When the test state is a qualified state, calculate the test error value.

[0107] The calculation formula of the test optimization rate is:

[0108]

[0109] in, To test the optimization rate, is the total data deviation value of the retest data, is the total data deviation value of the first test data, is the total data deviation value of the second test data, is the total amount of the first test data, is the total amount of retest data, is the total data volume of the second test data.

[0110] The technical effect of the above technical solution is: by combining the data deviation value with the anomaly ranking, test resources are allocated to the test subtasks, ensuring the rationality and effectiveness of resource allocation. The subtasks with higher anomaly rankings, that is, the subtasks with larger data deviation values, will obtain more test resources, which helps to prioritize key issues and improve test efficiency. The amount of resource allocation is determined based on the proportion of data deviation values ​​in the total data deviation values, making resource allocation more accurate and reasonable. This allocation method can ensure that test resources are fully utilized, avoid resource waste, and meet the resource needs of different subtasks.

[0111] By calculating the test optimization rate, the improvement effect of the test process can be quantitatively evaluated, providing clear data support for subsequent test optimization. By comparing the test optimization rate with the preset optimization threshold, it is possible to determine whether the test status is qualified, so that problems can be discovered and adjusted in a timely manner. When the test status is unqualified, the preset parameters are adjusted according to the preset strategy and the test is re-performed until the test status is qualified, which ensures the continuous improvement of the test process and the reliability of the test results. By continuously optimizing the test process and resource allocation, the accuracy and efficiency of the test can be improved, thereby improving the quality and reliability of the product. When the test status is qualified, calculating the test error value helps to further analyze the problems and potential risks that arise during the test process. By analyzing the test error value, targeted suggestions and directions can be provided for subsequent product improvements and test optimization.

[0112] In one embodiment of the present invention, the optimization calculation module further includes:

[0113] The error analysis module is used to calculate the test error value according to the test optimization rate, the second test data and the first test data, analyze the test situation of the test error value, and obtain the analysis result; obtain the proportion of the test error value in the total test value, and obtain the error proportion result.

[0114] The aging evaluation module is used to evaluate the test aging degree according to the analysis result to obtain an evaluation result.

[0115] The working principle of the above technical solution is as follows: the error analysis module is used to calculate the test error value according to the test optimization rate, the second test data and the first test data, analyze the test situation with the test error value, and obtain the analysis result; obtain the proportion of the test error value in the total test value, and obtain the error proportion result. The aging evaluation module is used to evaluate the test aging degree according to the analysis result, and obtain the evaluation result. The error proportion result is the evaluation result.

[0116] The calculation formula of the test error value is:

[0117]

[0118] in, To test for error values, is the error value of the first test data, To retest the data error value, is the error value of the second test data, To test the optimization rate, is the total amount of the first test data, is the total amount of retest data, is the total data volume of the second test data.

[0119] The technical effect of the above technical solution is: combining the test optimization rate, the second test data and the first test data to calculate the test error value. This calculation method not only takes into account the degree of optimization in the test process, but also combines the changes in the actual test data to obtain a more accurate and comprehensive error value. Accurate error values ​​can more accurately reflect the problems in the test process and provide strong data support for subsequent problem analysis and optimization. Through the analysis of the test error values, we can have an in-depth understanding of the abnormal situations and potential problems in the test process, and provide direction for the optimization and improvement of the test process. The analysis results can help identify bottlenecks and weaknesses in the test and provide targeted guidance for subsequent testing work. By calculating the error ratio results, you can intuitively understand the proportion of test errors in the total test value, which helps to quickly evaluate the overall quality and reliability of the test.

[0120] The error percentage results are concise and easy to understand and compare, providing decision makers with an important basis for quickly judging the test status. By evaluating the test aging degree based on the analysis results, an accurate judgment of the aging of the 3D flash memory can be obtained. The evaluation results are helpful to predict the product life and performance change trend, providing an important basis for product quality control and after-sales service.

[0121] By analyzing the test error values ​​and evaluating the test aging degree, problems in the test process can be discovered in time and improvements can be made, thereby optimizing the test process and improving test efficiency. Continuous test optimization and error analysis help improve product quality and reliability and enhance product market competitiveness.

[0122] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include these modifications and variations.

Claims

1. A three-dimensional flash memory aging test method, characterized in that: The method comprises: a. Perform a double-row double-sided gold finger test connection to obtain an aging test task, wherein the double-row double-sided gold finger performs a multi-channel aging test on the three-dimensional flash memory according to the aging test task; Wherein, the a includes: The gold fingers in the PCB are staggered in double rows of contacts, the whole PCB is inserted into the connector, each row of gold fingers is contacted with its corresponding contact, and then the electrical connection is obtained, and the electrical signal pins are matched with the mechanical position to obtain the double-row double-sided gold finger test connection; Acquire a preset aging test task, and perform a multi-channel aging test on the three-dimensional flash memory according to the aging test task; b. Obtain first test data, obtain data deviation values ​​of abnormal data types according to the first test data, split the aging test task according to the abnormal data types, obtain test subtasks, perform retests according to the test subtasks, and obtain retest data; Wherein, said b includes: Acquire test data in a multi-channel aging test, obtain first test data, pre-process the first test data, classify the first test data according to data types, obtain multiple types of data, compare the multiple types of data with corresponding thresholds respectively, obtain abnormal data types according to the comparison results, calculate the difference between the type data and the corresponding abnormal data, obtain data deviation values, sort the data deviation values, and obtain abnormal sorting of the type data; Obtain a corresponding test subtask according to the type of abnormal data, call a corresponding channel or a closest channel in the multi-channel aging test according to the test subtask, perform targeted retest on the abnormal data, and obtain retest data; c. Allocate test resources to the test subtasks according to the data deviation value, obtain the second test data, calculate the test optimization rate, and then calculate the test error value, evaluate the test aging degree, and obtain the evaluation result; Wherein, the c includes: Acquire test resources, allocate test resources to test subtasks from most to least based on data deviation values ​​combined with exception sorting, and obtain resource allocation information; After the test resource allocation is completed, the test data in the test process is acquired again to obtain second test data, and the test optimization rate is calculated according to the first test data, the retest data and the second test data to obtain the test optimization result; The calculation formula of the test optimization rate is: Among them, Y o is the test optimization rate, ΔCP is the total data deviation value of the retest data, ΔYP is the total data deviation value of the first test data, ΔRP is the total data deviation value of the second test data, YC is the total data volume of the first test data, CE is the total data volume of the retest data, and RC is the total data volume of the second test data; The step of calculating the test error value, evaluating the test aging degree, and obtaining the evaluation result includes: Calculate a test error value according to the test optimization rate, the second test data and the first test data, analyze the test error value on the test situation, and obtain an analysis result; The calculation formula of the test error value is: Among them, W cz is the test error value, Y Cc is the error value of the first test data, C Ec is the error value of the retest data, R Cc is the error value of the second test data, Y o is the test optimization rate, YC is the total data volume of the first test data, CE is the total data volume of the retest data, and RC is the total data volume of the second test data; The test aging degree is evaluated according to the analysis result to obtain an evaluation result.

2. A three-dimensional flash memory aging test system, characterized in that: The system comprises: An aging test connection module is used to perform a double-row double-sided gold finger test connection and obtain an aging test task. The double-row double-sided gold finger performs a multi-channel aging test on the three-dimensional flash memory according to the aging test task. Wherein, the aging test connection module includes: The connection test module is used to set the double-row contact staggered positions of the gold fingers in the PCB, insert the whole PCB into the connector, make each row of gold fingers contact with its corresponding contact, and then obtain electrical connection, match the electrical signal pins with the mechanical position, and obtain the double-row double-sided gold finger test connection; A task acquisition module, used for acquiring a preset aging test task, and performing a multi-channel aging test on the three-dimensional flash memory according to the aging test task; A splitting and retesting module is used to obtain first test data, obtain data deviation values ​​of abnormal data types according to the first test data, split the aging test task according to the abnormal data types, obtain test subtasks, retest according to the test subtasks, and obtain retest data; The splitting and retesting module includes: an abnormal data acquisition module, used to acquire test data in a multi-channel aging test, acquire first test data, pre-process the first test data, classify the first test data according to data types, acquire multiple types of data, compare the multiple types of data with corresponding thresholds respectively, acquire the type of abnormal data according to the comparison result, perform difference calculation between the type data and the corresponding abnormal data, acquire data deviation values, sort the data deviation values, and obtain abnormal sorting of the type data; A subtask extraction module, used for obtaining a corresponding test subtask according to the type of abnormal data, calling a corresponding channel or a closest channel in the multi-channel aging test according to the test subtask, and performing a targeted retest on the abnormal data to obtain retest data; A resource allocation and evaluation module is used to allocate test resources to the test subtasks according to the data deviation value, obtain the second test data, calculate the test optimization rate, and then calculate the test error value, evaluate the test aging degree, and obtain the evaluation result; Wherein, the resource allocation and evaluation module includes: The resource allocation module is used to obtain test resources, allocate test resources to test subtasks from more to less according to data deviation values ​​combined with abnormal sorting, and obtain resource allocation information; The optimization calculation module is used to obtain the test data in the test process again after the test resource allocation is completed, obtain the second test data, calculate the test optimization rate according to the first test data, the retest data and the second test data, and obtain the test optimization result; The calculation formula of the test optimization rate is: Among them, Y o is the test optimization rate, ΔCP is the total data deviation value of the retest data, ΔYP is the total data deviation value of the first test data, ΔRP is the total data deviation value of the second test data, YC is the total data volume of the first test data, CE is the total data volume of the retest data, and RC is the total data volume of the second test data; Wherein, the optimization calculation module also includes: An error analysis module, used to calculate a test error value according to the test optimization rate, the second test data and the first test data, analyze the test error value on the test situation, and obtain an analysis result; The calculation formula of the test error value is: Among them, W cz is the test error value, Y Cc is the error value of the first test data, C Ec is the error value of the retest data, R Cc is the error value of the second test data, Y o is the test optimization rate, YC is the total data volume of the first test data, CE is the total data volume of the retest data, and RC is the total data volume of the second test data; The aging evaluation module is used to evaluate the test aging degree according to the analysis result to obtain an evaluation result.