Method and apparatus for determining aging time

Abstract

The embodiment of the present application provides a kind of determination method and device of aging time, by dividing each wafer particle in wafer into multiple test units, sample capacity is expanded, by dividing test time into multiple time periods, after the end of aging experiment in each time period, the test result of test unit is acquired, according to the test result of the tested time period, determine the failure distribution model of test unit in current time, current time is the end time of last tested time period, according to the failure distribution model of current time, determine the predicted aging time corresponding to failure distribution model, if predicted aging time is not less than current time, then enter next time period and carry out aging test, if predicted aging time is less than current time, then determine the predicted aging time corresponding to failure distribution model as final aging time. The predicted aging time determined by failure distribution model is more accurate, avoid the problems of overaging and underaging.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of wafer testing, and in particular to a method and device for determining burn-in time. BACKGROUND

[0002] Failure problems are inevitable in the development, production and use of chips. After the production of a wafer is completed, the wafer will be tested to check the quality of the wafer. This mainly includes a chip probe test (CP test) and a final test (FT). The wafer that passes the CP test will be subjected to the FT, in which a burn-in (BI) test will be performed. The purpose of the BI test is to provide a high-temperature, high-voltage and high-current environment for the wafer particles, so that the particles of poor quality can be damaged as soon as possible in the environment, and the stable and reliable particles are retained.

[0003] The prior art determines a fixed time as the total time of the burn-in test in advance through a large number of batch experimental results. Due to the differences between batches, the fixed burn-in time cannot effectively distinguish the differences between batches or the abnormalities in the production process, which may result in over-burn-in or under-burn-in, and finally lead to uneven product quality. SUMMARY

[0004] The present application provides a method and device for determining burn-in time, which can determine the burn-in time of a wafer according to a failure distribution model. The burn-in time obtained by different failure distribution models is different, which improves the accuracy of the burn-in time.

[0005] In a first aspect, the present application provides a method for determining burn-in time, comprising:

[0006] dividing each wafer particle in a wafer into a plurality of test units, the wafer comprising a plurality of wafer particles;

[0007] dividing a test time into a plurality of time periods, after the end of the burn-in test in each time period, obtaining a test result of the test unit, the test result being pass or failure;

[0008] determining a failure distribution model of the test unit at a current time according to the test result of the tested time period;

[0009] determining a predicted burn-in time corresponding to the failure distribution model according to the failure distribution model at the current time;

[0010] if the predicted burn-in time corresponding to the failure distribution model is not less than the current time, entering the next time period for burn-in test;

[0011] If the predicted aging time corresponding to the failure distribution model is less than the current time, the predicted aging time corresponding to the failure distribution model is determined as the final aging time.

[0012] Optionally, the failure distribution model comprises any one of the following models: a Weibull distribution model, a normal distribution model, or a lognormal distribution model.

[0013] Optionally, the determining of the predicted aging time corresponding to the failure distribution model according to the failure distribution model at the current time comprises:

[0014] calculating a failure rate of the test unit at the current time according to the parameters of the failure distribution model at the current time;

[0015] when the failure rate at the current time is less than a preset failure rate, the current time is determined as the predicted aging time.

[0016] Optionally, the determining of the predicted aging time corresponding to the failure distribution model according to the failure distribution model at the current time comprises:

[0017] calculating failure rates of the test unit at at least three times respectively according to the parameters of the failure distribution model at the at least three times, the at least three times comprising the current time;

[0018] fitting a failure distribution curve according to the failure rates at the at least three times, the horizontal axis of the failure distribution curve being time and the vertical axis being failure rate;

[0019] determining a tangent line of the failure distribution curve;

[0020] determining a time corresponding to an intersection of the tangent line and the horizontal axis as the predicted aging time.

[0021] Optionally, the determining of the predicted aging time corresponding to the failure distribution model according to the failure distribution model at the current time comprises:

[0022] calculating failure rates of the test unit at at least three times respectively according to the parameters of the failure distribution model at the at least three times, the at least three times comprising the current time;

[0023] fitting a failure distribution curve according to the failure rates at the at least three times, the horizontal axis of the failure distribution curve being time and the vertical axis being failure rate;

[0024] determining a tangent line of the failure distribution curve and an intersection of the tangent line and the horizontal axis;

[0025] determining a time T2 corresponding to the intersection and a point on the failure distribution curve closest to the intersection;

[0026] determining the T2 as the predicted aging time, or multiplying the T2 by a preset coefficient to obtain the predicted aging time, the preset coefficient being greater than 0.

[0027] Optionally, when the failure distribution model is a Weibull distribution model, the calculating the failure rate of the test unit at the current time according to the parameter of the failure distribution model corresponding to the current time comprises:

[0028] The failure rate at the current time is calculated by any one of the following formulas:

[0029]

[0030]

[0031]

[0032] wherein λ(t) is the failure rate in the current time period, β and η are the parameters of the Weibull distribution model, b is the intermediate parameter of the Weibull distribution model, t is the current time, and X = ln t.

[0033] In a second aspect, the present application provides a device for determining an aging time, the device comprising functional modules required for performing the method of the first aspect of the present application.

[0034] In a third aspect, the present application provides an electronic device, comprising at least one processor and a memory.

[0035] The memory stores computer-executable instructions.

[0036] The at least one processor executes the computer-executable instructions stored in the memory, so that the at least one processor performs the method of the first aspect of the present application.

[0037] In a fourth aspect, the present application provides a computer-readable storage medium, the computer-readable storage medium storing computer-executable instructions, the computer-executable instructions being executed by a processor to implement the method of the first aspect of the present application.

[0038] In a fifth aspect, the present application provides a computer program product, comprising a computer program, the computer program being executed by a processor to implement the method of the first aspect of the present application.

[0039] The embodiment of the present application provides a method and device for determining aging time. Each wafer particle in a wafer is divided into a plurality of test units, so that the sample capacity is expanded. The test time is divided into a plurality of time periods. After the aging experiment in each time period ends, the test result of the test unit is obtained. According to the test result of the tested time period, the failure distribution model of the test unit at the current time is determined. The current time is the end time of the last tested time period. According to the failure distribution model at the current time, the predicted aging time corresponding to the failure distribution model is determined. If the predicted aging time is not less than the current time, the aging test in the next time period is performed. If the predicted aging time is less than the current time, the predicted aging time corresponding to the failure distribution model is determined as the final aging time. The predicted aging time determined by the failure distribution model is more accurate, and the problems of overaging and underaging are avoided. BRIEF DESCRIPTION OF DRAWINGS

[0040] Figure 1 A flow chart of the method for determining aging time provided by the first embodiment of the present application;

[0041] Figure 2 A schematic diagram of the fitted failure distribution curve;

[0042] Figure 3 Another schematic diagram of the fitted failure distribution curve;

[0043] Figure 4 A structure schematic diagram of the device for determining aging time provided by the second embodiment of the present application;

[0044] Figure 5 A structure schematic diagram of the electronic device provided by the third embodiment of the present application. DETAILED DESCRIPTION

[0045] The exemplary embodiments will be described in detail herein with reference to the attached drawings. In the following description, the same numbers are used to indicate the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all the embodiments consistent with the present disclosure. Instead, they are merely examples of devices and methods consistent with some aspects of the present disclosure, as detailed in the appended claims.

[0046] In the prior art, a fixed time is determined as the aging time through a large number of experimental results. On the one hand, a large number of product batches of wafers are required to determine the aging time of the wafer, the test period is long, and the cost is high. The relationship between the number of wafer samples and the confidence of failure rate is as follows: C=1-R n, R = 1-F, C is the confidence, F is the failure rate, n is the sample number, F*1000000 gets the million failure rate (PPM), PPM is the number of failed samples in one million samples. If the calculated PPM value is 1000 (F = 0.001, R = 0.999), if the confidence is 0.95, 2999 die of wafers are needed.

[0047] If only one wafer is used, and the die in the wafer is used as a test unit, when n = 900 (the largest wafer at present), in the same R, F, C can only reach 0.59, obviously, such results are inaccurate.

[0048] In order to solve the problem of sample capacity, each die in a wafer is divided into multiple test units, and the division rule can be divided by bank. Generally, a die has 16 banks, so n reaches 16*900, which can fully meet the test sample amount. Of course, 2, 3 or 4 banks can also be divided into a test unit, and the embodiment of the present application does not limit this.

[0049] On the other hand, the fixed aging time cannot effectively distinguish the differences between product batches or the abnormalities in the production process due to the differences between product batches, which may cause overaging or underaging problems, ultimately leading to uneven product quality.

[0050] In order to solve the problem of the prior art, the present application provides a method for determining aging time, which solves the problem of sample capacity by dividing the die into multiple test units, and additionally divides the test time into multiple time periods, calculates the instantaneous failure rate after the end of each time period according to the failure distribution mode, and predicts the aging time according to the failure rate.

[0051] Figure 1 The flow chart of the method for determining the aging time provided by the embodiment of the present application is shown in Figure 1 The method provided by the present embodiment comprises the following steps.

[0052] S101, divide each die in the wafer into multiple test units, and the wafer comprises multiple die.

[0053] The method of the embodiment can use the aging test results of one wafer to predict the aging time. By dividing each wafer die into multiple test units and taking the test unit as the minimum test unit, the sample capacity is expanded. Moreover, when the same wafer is used as the sample, the wafer itself has less performance fluctuation, thereby reducing the uncertainty of the aging time itself.

[0054] S102, divide the test time into multiple time periods, and obtain the test result of the test unit after the aging test of each time period ends, the test result being pass or fail.

[0055] The test time is the total length of the aging test of the wafer. The test time is divided into multiple time periods, and the size of each time period can be the same or different, which is not limited by the embodiment. For example, the test time is 1 minute, and the test time can be divided into 12 time periods, and the size of each time period is 5 seconds (s). Or, the test time is divided into 10 time periods, and the size of each time period is 6 s.

[0056] The wafer is aged in each time period, and then the test result of each test unit is obtained, and the test result of each test unit is pass or fail.

[0057] S103, determine the failure distribution model of the test unit at the current time according to the test result of the tested time period.

[0058] The aging test is sequentially performed on each time period in chronological order, and the test result is obtained. The tested time period is relative to the current time, the current time is the end time of the last tested time period, and the tested time period can include one or more time periods. For example, the test time is divided into 12 time periods, if the tested time period is the first 3 time periods, then the current time is the end time of the third time period, if the tested time is 5 time periods, then the current time is the end time of the fifth time period. It can also be considered that the current time is a cumulative time, and the initial value of the current time is 0, and the current time gradually accumulates changes from the start of the aging test.

[0059] The failure distribution model is used to reflect the distribution diagram of the failure rate of the test unit with time, and the failure distribution model can be represented by a failure distribution function. The failure distribution model includes any one of the following models: Weibull distribution model, normal distribution model or lognormal distribution model. Among them, with the accumulation of time, the failure distribution model obtained at different times is different.

[0060] Taking the Weibull distribution model as an example, the parameters of the Weibull distribution model are determined according to the test results of the tested time periods. The Weibull distribution model includes two forms of three parameters and two parameters, and the three-parameter Weibull distribution model can be converted into the two-parameter Weibull distribution model. In this embodiment, the three-parameter Weibull distribution model can be used or the two-parameter Weibull distribution model can be used.

[0061] The parameters of the three-parameter Weibull distribution model include β, η and γ, wherein β is a shape parameter, the value of β is greater than 0, determines the shape of the distribution curve, η is a scale parameter, the value of η is greater than 0, and plays a role in amplifying and reducing the horizontal coordinate scale, and γ is a position parameter, the value of γ is greater than or equal to 0, and determines the starting position of the distribution curve. When the value of γ is 0, the three-parameter Weibull distribution model is converted into the two-parameter Weibull distribution model.

[0062] The failure distribution function of the two-parameter Weibull distribution model is shown in formula (1).

[0063]

[0064] The regression estimation method is used for parameter estimation, formula (1) is deformed, and formula (2) is obtained.

[0065]

[0066] Formula (3) is obtained by taking the natural logarithm of formula (2).

[0067]

[0068] Formula (4) is obtained by taking the natural logarithm of formula (3) again.

[0069] ln[-1n(1-F(t)]=βlnt-βlnη (4)

[0070] Let x=lnt, y=ln[-ln(1-F(t)], and F(t)=1-R(t) is brought into, wherein R(t) is a test passing ratio of a test unit, and the ratio of the number of test units with a test result of pass to the total amount of test units is taken as R(t), t is the current time.

[0071] After the x and y corresponding to each tested time period are calculated, the parameters β and η of the Weibull distribution model are obtained through the following formulas (5)-(7), wherein a set of x and y corresponding to each time period.

[0072]

[0073]

[0074]

[0075] wherein x is the x corresponding to the current time period, y is the y corresponding to the current time period, the current time period is the last tested time period, is the average of the x corresponding to the tested time periods, is the average of the y corresponding to the tested time periods, and n is the number of the tested units.

[0076] After obtaining β and η, the parameters β and η are brought into formula (1) or formula (2) to obtain the failure distribution function F(t), and the failure distribution model is expressed by the failure distribution function.

[0077] It can be understood that the failure distribution functions of different failure distribution models are different, and here only the Weibull distribution is taken as an example for illustration, and the calculation of other failure distribution functions is not listed one by one.

[0078] S104, determining the predicted aging time corresponding to the failure distribution model according to the current time.

[0079] For example, the failure rate of the current time corresponding to the failure distribution model of the tested unit can be obtained according to the parameters of the failure distribution model, and it is determined whether to end the aging experiment according to the failure rate of the current time, and the time of ending the aging experiment is the final aging time. Or, according to the failure distribution model, the failure distribution curve corresponding to the failure distribution function is obtained, and the predicted aging time is determined according to the failure distribution curve.

[0080] Method one, according to the parameters of the failure distribution model corresponding to the current time, the failure rate of the tested unit at the current time is calculated, and when the failure rate of the current time is less than the preset failure rate, the current time is determined as the predicted aging time.

[0081] Taking the Weibull distribution model as an example, formula (8), formula (9) or formula (10) can be used to calculate the failure rate corresponding to the failure distribution model, and the failure rate corresponding to the failure distribution model is the failure rate of the current time.

[0082]

[0083]

[0084]

[0085] δ in formula (10) is a constant.

[0086] Optionally, the failure rate can also be PPM, and the failure rate calculated by formula (8), (9) or (10) multiplied by one million can obtain PPM.

[0087] The failure rate corresponding to the failure distribution model is compared with the preset failure rate. If the failure rate corresponding to the failure distribution model is less than the preset failure rate, it is determined that the current time is the predicted aging time, and the aging test experiment is ended. If the failure rate corresponding to the failure distribution model is not less than (i.e., greater than or equal to) the preset failure rate, the aging test of the next time period is continued until the test time ends.

[0088] In the second mode, the failure rates of the test unit at the respective times are calculated according to the parameters of the failure distribution models corresponding to the at least three times, and a failure distribution curve is fitted according to the failure rates corresponding to the at least three times. The horizontal axis of the failure distribution curve is time, and the vertical axis is failure rate. A tangent line of the failure distribution curve is determined, and the intersection point of the tangent line and the X-axis corresponds to the predicted aging time.

[0089] Taking three time points as an example, the three times correspond to three time periods. The parameters of three failure distribution models are obtained from the data of the three time periods selected from the tested time periods, and then the failure rates of the three times are obtained according to the parameters of the failure distribution models. A failure distribution curve is fitted according to the three times and the failure rates of the three times. The horizontal axis of the failure distribution curve is time, and the vertical axis is failure rate. At least 3 points (corresponding to three times) of data are required when the failure distribution curve is fitted. Of course, the more data points taken, the more accurate the failure distribution curve fitted.

[0090] Similarly, a straight line is fitted using the data of the at least three times, which is represented as y=kx+b, where k is the slope of the straight line, x is time t, y is the failure rate corresponding to time t, and k and b are constants. After obtaining the slope k, a point on the failure distribution curve with a slope of k is found as a tangent point, and a tangent line is drawn through the tangent point. The tangent line has an intersection point with the horizontal axis of the failure distribution curve.

[0091] Optionally, the data of the first three time periods after the start of the test time can be selected to fit the failure distribution curve and the straight line.

[0092] Figure 2 For a schematic diagram of the fitted failure distribution curve, a tangent line L of the failure distribution curve is determined, and the tangent line L has an intersection point A with the horizontal axis. In this embodiment, the time corresponding to the intersection point A can be taken as the predicted aging time, and the predicted aging time is T1 in the figure.

[0093] In the third mode, the failure distribution curve is fitted by the method of the second mode, and then a tangent line of the failure distribution curve and an intersection point of the tangent line and the horizontal axis are determined. The time T2 corresponding to the intersection point and the point on the failure distribution curve with the shortest distance is determined as the predicted aging time.

[0094] The determination of the tangent of the failure distribution curve in this mode is the same as that in mode two, which will not be described here.

[0095] Figure 3 For another schematic diagram of the obtained failure distribution curve, a tangent L of the failure distribution curve is determined, the tangent L has an intersection point A with the horizontal axis, and a point B on the failure distribution curve that is closest to the intersection point A is found, and the time corresponding to the point B is the predicted aging time, that is, the predicted aging time is T2 in the figure.

[0096] Mode four, the difference between mode four and mode three is that after T2 is determined, T2 is multiplied by a preset coefficient to obtain the predicted aging time, and the preset coefficient is greater than 0, for example, 2.

[0097] S105, if the predicted aging time corresponding to the failure distribution model is not less than the current time, the next time period is entered for aging test, and if the predicted aging time corresponding to the failure distribution model is less than the current time, the predicted aging time corresponding to the failure distribution model is determined as the final aging time.

[0098] For each time period, the corresponding predicted aging time is obtained, and the current time is compared with the corresponding predicted aging time. Among them, the predicted aging time obtained by different failure distribution models is also different, compared with the fixed aging time in the prior art, the method of the embodiment can obtain the aging time matched with the product according to the actual difference of the product, and avoid the problems of over-aging or insufficient aging.

[0099] In the embodiment, each wafer particle in the wafer is divided into a plurality of test units, the sample capacity is expanded, the test time is divided into a plurality of time periods, the test results of the test units are obtained after the aging experiment in each time period ends, the failure distribution model of the test units at the current time is determined according to the test results of the tested time period, the current time is the end time of the last tested time period, the predicted aging time corresponding to the failure distribution model is determined according to the failure distribution model at the current time, if the predicted aging time is not less than the current time, the next time period is entered for aging test, and if the predicted aging time is less than the current time, the predicted aging time corresponding to the failure distribution model is determined as the final aging time. The predicted aging time determined by the failure distribution model is more accurate, and the problems of over-aging and insufficient aging are avoided.

[0100] Figure 4 The structure schematic diagram of the aging time determination device provided in the second embodiment of the application is shown in FIG. 1, the device 100 provided in the embodiment includes the following modules. Figure 4

[0101] ​The dividing module 11 is configured to divide each wafer particle in a wafer into a plurality of test units, the wafer comprising a plurality of wafer particles;

[0102] The aging module 12 is configured to divide a test time into a plurality of time periods, and obtain a test result of the test unit after an aging experiment in each time period ends, the test result being pass or failure;

[0103] The first determining module 13 is configured to determine a failure distribution model of the test unit at a current time according to the test result of the tested time period.

[0104] The predicting module 14 is configured to determine a predicted aging time corresponding to the failure distribution model at the current time according to the failure distribution model.

[0105] If the predicted aging time corresponding to the failure distribution model is not less than the current time, the next time period is entered for aging test.

[0106] If the predicted aging time corresponding to the failure distribution model is less than the current time, the predicted aging time corresponding to the failure distribution model is determined as a final aging time.

[0107] Optionally, the failure distribution model comprises any one of the following models: a Weibull distribution model, a normal distribution model or a lognormal distribution model.

[0108] Optionally, the predicting module 14 is specifically configured to calculate a failure rate of the test unit at the current time according to a parameter of the failure distribution model corresponding to the current time, and determine the current time as the predicted aging time when the failure rate at the current time is less than a preset failure rate.

[0109] Optionally, the predicting module 14 is specifically configured to calculate failure rates of the test unit at at least three times respectively according to parameters of failure distribution models corresponding to the at least three times, the at least three times comprising the current time; fit a failure distribution curve according to the failure rates corresponding to the at least three times, the horizontal axis of the failure distribution curve being time and the vertical axis being failure rate; determine a tangent line of the failure distribution curve; and determine a time corresponding to an intersection of the tangent line and the horizontal axis as the predicted aging time.

[0110] Optionally, the prediction module 14 is specifically used to: calculate the failure rate of the test unit at each time according to the parameters of the failure distribution model corresponding to at least three times, wherein the at least three times include the current time; fit a failure distribution curve according to the failure rates corresponding to the at least three times, wherein the horizontal axis of the failure distribution curve is time and the vertical axis is failure rate; determine the tangent of the failure distribution curve and the intersection point of the tangent with the horizontal axis; determine the time T2 corresponding to the point with the shortest distance between the intersection point and the point on the failure distribution curve; determine T2 as the predicted aging time, or multiply T2 by a preset coefficient to obtain the predicted aging time, wherein the preset coefficient is greater than 0.

[0111] Optionally, when the failure distribution model is a Weibull distribution model, the prediction module 14 is specifically used to: calculate the failure rate at the current time using any of the following formulas.

[0112]

[0113]

[0114]

[0115] Where λ(t) is the failure rate in the current time period, β and η are the parameters of the Weibull distribution model, b is the intermediate parameter of the Weibull distribution model, t is the current time, and X = lnt.

[0116] The apparatus in this embodiment can be used to execute the method described in Embodiment 1 above. The specific implementation and technical effects are similar, and will not be repeated here.

[0117] It should be noted that the above embodiments are only illustrative examples of the division of the above functional modules when the device performs the method described in the embodiments. In actual applications, the above functions can be assigned to different functional modules as needed, that is, the internal structure of the device can be divided into different functional modules to complete all or part of the functions described above.

[0118] Figure 5 This is a schematic diagram of the structure of an electronic device provided in Embodiment 3 of the present invention, as shown below. Figure 5 As shown, the electronic device 200 includes a processor 21, a memory 22, and a transceiver 23. The memory 22 is used to store instructions, the transceiver 23 is used to communicate with other devices, and the processor 21 is used to execute the instructions stored in the memory so that the electronic device 200 performs the method described in Embodiment 1 above. The specific implementation and technical effects are similar, and will not be described again here.

[0119] Embodiment four of the present application provides a computer readable storage medium, the computer readable storage medium stores computer execution instructions, the computer execution instructions are executed by a processor to implement the method of embodiment one, and the specific implementation manners and technical effects are similar, and details are not described herein.

[0120] Embodiment five of the present application provides a computer program product, including a computer program, the computer program is executed by a processor to implement the method of embodiment one, and the specific implementation manners and technical effects are similar, and details are not described herein.

[0121] Other embodiments of the present disclosure will be apparent to those skilled in the art from consideration of the specification and practice of the application disclosed herein. This application is intended to cover any variations, uses or adaptations of the present disclosure that follow, in general, the principles of the present disclosure and include such features that are evident to those skilled in the art or are known in the art and that include such features that are not disclosed herein. The specification and examples are to be regarded as illustrative only, and the true scope and spirit of the present disclosure are indicated by the following claims.

[0122] It should be understood that the present disclosure is not limited to the precise structures as herein described and illustrated in the drawings, and that various modifications and changes can be made without departing from its scope. The scope of the present disclosure is limited only by the claims that follow.

Claims

1. A method of determining the age time, characterized in that, The method comprises the following steps: dividing each wafer particle in a wafer into a plurality of test units, the wafer comprising a plurality of wafer particles; dividing a test time into a plurality of time periods, obtaining a test result of the test unit after the end of the aging experiment in each time period, the test result being pass or failure, and the test time being the total duration of the aging experiment of the wafer; determining a failure distribution model of the test unit at a current time according to the test result of the tested time period; determining a predicted aging time corresponding to the failure distribution model according to the failure distribution model at the current time; if the predicted aging time corresponding to the failure distribution model is not less than the current time, entering the next time period for aging test; if the predicted aging time corresponding to the failure distribution model is less than the current time, determining the predicted aging time corresponding to the failure distribution model as a final aging time.

2. The method of claim 1, wherein, The failure distribution model comprises any one of the following models: Weibull distribution model, normal distribution model or lognormal distribution model.

3. The method of claim 2, wherein, The method comprises the following steps: calculating a failure rate of the test unit at the current time according to the parameters of the failure distribution model corresponding to the current time; when the failure rate at the current time is less than a preset failure rate, determining the current time as the predicted aging time.

4. The method of claim 2, wherein, The method comprises the following steps: calculating a failure rate of the test unit at each of at least three times according to the parameters of the failure distribution model corresponding to the at least three times, the at least three times comprising the current time; fitting a failure distribution curve according to the failure rates corresponding to the at least three times, the horizontal axis of the failure distribution curve being time and the vertical axis being failure rate; determining a tangent line of the failure distribution curve; determining a time corresponding to the intersection of the tangent line and the horizontal axis as the predicted aging time.

5. The method of claim 2, wherein, The method comprises the following steps: calculating a failure rate of the test unit at each of at least three times according to the parameters of the failure distribution model corresponding to the at least three times, the at least three times comprising the current time; fitting a failure distribution curve according to the failure rates corresponding to the at least three times, the horizontal axis of the failure distribution curve being time and the vertical axis being failure rate; determining a tangent line of the failure distribution curve and the intersection of the tangent line and the horizontal axis; determining a time T2 corresponding to the intersection and the point on the failure distribution curve closest to the intersection; determining the T2 as the predicted aging time, or multiplying the T2 by a preset coefficient to obtain the predicted aging time, the preset coefficient being greater than 0.

6. The method of claim 3, wherein, When the failure distribution model is a Weibull distribution model, the method comprises the following steps: calculating the failure rate at the current time by any one of the following formulas: Wherein, λ(t) is the failure rate in the current time period, β and η are parameters of the Weibull distribution model, b is the median parameter of the Weibull distribution model, δ is a constant, t is the current time, and X = ln t.

7. An apparatus for determining an aging time, characterized in that The method comprises the steps of: a dividing module, configured to divide each wafer particle in a wafer into a plurality of test units, the wafer comprising a plurality of wafer particles; an aging module, configured to divide a test time into a plurality of time periods, and obtain test results of the test units after an aging experiment in each time period ends, the test results being pass or failure, the test time being a total duration of the aging experiment of the wafer; a first determining module, configured to determine a failure distribution model of the test units at a current time according to the test results of the tested time periods; a predicting module, configured to determine a predicted aging time corresponding to the failure distribution model at the current time according to the failure distribution model at the current time; if the predicted aging time corresponding to the failure distribution model is not less than the current time, proceed to the next time period for aging test; if the predicted aging time corresponding to the failure distribution model is less than the current time, determine the predicted aging time corresponding to the failure distribution model as a final aging time.

8. The apparatus of claim 7, wherein, The failure distribution model comprises any one of the following models: a Weibull distribution model, a normal distribution model, or a lognormal distribution model.

9. The apparatus of claim 8, wherein, The predicting module is specifically configured to: calculate a failure rate of the test units at the current time according to parameters of the failure distribution model corresponding to the current time; and determine the current time as the predicted aging time when the failure rate at the current time is less than a preset failure rate.

10. The apparatus of claim 8, wherein, The predicting module is specifically configured to: calculate failure rates of the test units at at least three times respectively according to parameters of the failure distribution models corresponding to the at least three times, the at least three times including the current time; fit a failure distribution curve according to the failure rates corresponding to the at least three times, the horizontal axis of the failure distribution curve being time and the vertical axis being failure rate; and determine a tangent line of the failure distribution curve, and a time corresponding to an intersection of the tangent line and the horizontal axis as the predicted aging time. The predicting module is specifically configured to:

11. The apparatus of claim 8, wherein, calculate failure rates of the test units at at least three times respectively according to parameters of the failure distribution models corresponding to the at least three times, the at least three times including the current time; fit a failure distribution curve according to the failure rates corresponding to the at least three times, the horizontal axis of the failure distribution curve being time and the vertical axis being failure rate; and determine a tangent line of the failure distribution curve, and an intersection of the tangent line and the horizontal axis; determine a time T2 corresponding to a point on the failure distribution curve closest to the intersection; and determine the T2 as the predicted aging time, or multiply the T2 by a preset coefficient to obtain the predicted aging time, the preset coefficient being greater than 0. When the failure distribution model is a Weibull distribution model, the predicting module is specifically configured to:

12. The apparatus of claim 9, wherein, calculate the failure rate at the current time by any one of the following formulas: ​ Wherein, λ(t) is the failure rate in the current time period, β and η are parameters of the Weibull distribution model, b is the median parameter of the Weibull distribution model, δ is a constant, t is the current time, and X = ln t.

13. An electronic device, comprising: Comprising: at least one processor and a memory; the memory stores computer-executable instructions; the at least one processor executes the computer-executable instructions stored in the memory, so that the at least one processor executes the method in any one of claims 1 to 6.

14. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer-executable instructions, and the computer-executable instructions are executed by the processor to implement the method in any one of claims 1 to 6.

15. A computer program product comprising a computer program, characterized in that, The computer program is executed by the processor to implement the method in any one of claims 1 to 6.

Images