Semiconductor test system and method based on artificial intelligence

Abstract

The invention discloses a semiconductor test system and method based on artificial intelligence, and relates to the technical field of semiconductor test, and the system comprises a special processor which is in signal connection with a data collection module for collecting semiconductor test data and a time relay for limiting the data collection time. And the time relay is in signal connection with a time management module which is in signal connection with the central processing unit. The test system has the advantages that the time relay is arranged in the test system to limit the duration of data acquisition, the timing module is arranged to time the transfer time and the sorting time of the semiconductor, and the time management module is used for comprehensively managing the total consumed time of testing, transferring and sorting of the semiconductor; the total consumed time of each test of the semiconductor is calculated, the fault position in the test process can be judged through the time difference of the consumed time in the semiconductor test process, and the fault position in the test system can be accurately and efficiently locked.

Description

Technical Field

[0001] The present invention relates to the field of semiconductor testing technology, and in particular to an artificial intelligence-based semiconductor testing system and method. Background Art

[0002] Semiconductor devices are widely used in information communication, computers, consumer electronics, aerospace, automotive electronics and other fields. The design and manufacture of semiconductor devices is a very complex process. In the semiconductor device manufacturing process, testing is an important link to ensure the quality of devices before they leave the factory. Through testing, some defective products or products with unqualified performance produced in the manufacturing process can be selected, or through testing, the performance parameters of the device can be known, and the product grades can be differentiated. Probe testing is a very important test item in silicon wafer testing. Probe testing uses a probe card to contact the pads on the wafer to transmit electrical signals. The probe on the probe card contacts the test contact pads on the wafer. The test machine sends the test electrical signal through the probe and the pads in contact with the probe to the grains on the wafer and obtains test data.

[0003] However, the existing semiconductor testing system is not convenient for judging the time consumption of the test during the testing process. The test time may be extended due to problems such as test system failure during the testing process, affecting the test efficiency. It is also not convenient for judging the quality of the tested semiconductor, which is not conducive to improving the semiconductor production process.

[0004] Based on the above problems, a semiconductor testing system and method based on artificial intelligence is proposed. Summary of the invention

[0005] 1. Technical issues to be resolved

[0006] In view of the shortcomings of the prior art, the present invention provides a semiconductor testing system and method based on artificial intelligence, which has the advantages of judging the fault location in the test process according to the test time difference, which is conducive to accurately and efficiently locking the fault location in the test system, and solves the problem that the test time may be extended due to problems such as test system failure during the test process, affecting the test efficiency.

[0007] (II) Technical solution

[0008] To achieve the above purpose, the present invention is implemented through the following technical solutions: a semiconductor testing system based on artificial intelligence, including a central processing unit as the core of the system, the central processing unit signal is connected to an intelligent database storing standard values ​​of various semiconductor test data and a number of special processors for extracting database data, the special processor signal is connected to a data acquisition module for collecting semiconductor test data and a time relay for limiting the data acquisition time, the time relay signal is connected to a time management module connected to the central processing unit signal, the time management module signal is connected to a semiconductor transport module for transporting semiconductors, the semiconductor transport module signal is connected to a timing module for monitoring the transport time, and the time management module signal is connected to a sorting and exporting module for sorting unqualified semiconductors;

[0009] Specifically, semiconductors to be tested are placed on a semiconductor transfer module, which is used to transfer the semiconductors to a test location for testing. The semiconductor test data is collected through a data acquisition module and transmitted to the corresponding special processor. The special processor extracts standard data and test data from an intelligent database for comparison, detects whether the test data is located within the standard data, and transmits the comparison result to a central processing unit. The central processing unit sends a control signal to the semiconductor transfer module and a sorting and exporting module to transfer qualified semiconductors to corresponding locations and sort and export unqualified semiconductors to corresponding locations. During data collection, the collection time is controlled by a time relay, the time for transferring semiconductors is detected by a timing module, and the time management module manages the time for transferring semiconductors. During continuous testing, each process interval of the semiconductor test can be timed, and during individual testing, the time consumption of various semiconductor tests can be integrated.

[0010] The dedicated processor is internally provided with a metering module for counting processed data, the metering module signal is connected to a data statistics module, the data statistics module signal is connected to a display module which is connected to a central processing unit signal.

[0011] Preferably, the several special processors correspond to different data acquisition modules, which include temperature and humidity data acquisition modules, electrical signal data acquisition modules, vibration data acquisition modules, storage life data acquisition modules and thermal shock data acquisition modules, wherein each special processor is correspondingly provided with a time relay, a semiconductor transfer module and a sorting export module, which limits the test time for each test data and transfers and sorts the semiconductors after the test.

[0012] Specifically, semiconductors can be transported to temperature and humidity test locations, electrical signal test locations, vibration test locations, storage life test locations, and thermal shock test locations through semiconductor transport modules. Each test corresponds to a semiconductor transport module and a sorting and export module, which facilitates the transport of semiconductors in each test and the export of semiconductors that fail the test.

[0013] Preferably, several of the semiconductor transfer modules are connected to each other, so that the semiconductors can be tested for continuity, and each semiconductor transfer module is provided with a semiconductor input component, so that the semiconductors can be tested individually. At the same time, each semiconductor transfer module is connected to a sorting and exporting module to export unqualified semiconductors.

[0014] Specifically, during the semiconductor testing process, a semiconductor can be transported between multiple test locations through a semiconductor transport module, that is, multiple different tests can be performed on a semiconductor. A new semiconductor can also be tested in each test item and the semiconductor can be tested individually. During the continuity test of a semiconductor, the test time and transport time of the semiconductor continuous test can be counted through the time management module. During the individual test of the semiconductor, the time consumption of each test can be counted through the time management module. The time of the same test can be used to determine whether there is a fault in the semiconductor testing process based on the difference in test time.

[0015] Preferably, the metering module is arranged inside each special processor and the central processing unit, wherein the metering module inside the central processing unit counts the number of times the central processing unit makes data judgments, the metering module counts the amount of data processed by each special processor, and transmits the counted data to the data statistics module, the data statistics module counts the amount of data processed inside each special processor and constructs a data model to perform data difference analysis, and displays the analysis structure on the display module.

[0016] Specifically, by counting each test in each test, and combining the number of sorted semiconductors, the yield and defective rates of the semiconductors after testing can be calculated and displayed, and the yield and defective rates of the semiconductors in each test can be compared to determine the items that need to be improved in the production process, which can effectively optimize production and analyze and determine the number of semiconductor production tests.

[0017] A testing method for a semiconductor testing system based on artificial intelligence, characterized in that it comprises the following steps:

[0018] S1. Store various standard value data of semiconductors into a database to form an intelligent database for testing, providing data support for the central processor, the brain of artificial intelligence;

[0019] S2. Collect various test data of semiconductors within the time limit of the time relay through various data acquisition modules;

[0020] 1) Set the time relay to control the data collection module to collect data;

[0021] 2) The timing module counts the time taken by the semiconductor transport module to transport semiconductors, and transmits the timing data to the time management module;

[0022] 3) The time management module comprehensively measures the test time and transportation time of each test item, analyzes the time consumed by each test item, and calculates the consumed time and transmits it to the central processor for determining the test duration;

[0023] S3, the special processor receives the corresponding test data and extracts the corresponding test data from the intelligent database for comparison, and transmits the comparison result to the central processor. At the same time, the metering module calculates the number of test data inside the special processor;

[0024] S4, the central processor determines the comparison result of the special processor, and controls the operation of the semiconductor transport module sorting and exporting module according to the determination result;

[0025] 1) Each time the special processor processes the test data, the metering module calculates the number of times by one, and accumulates the number of times the special processor processes the test data;

[0026] 2) Each metering module transmits the accumulated numbers to the data statistics module after each accumulation is completed. The data statistics module combines the corresponding data judgment results of the central processor to make statistics on the quality of each semiconductor data, and analyzes and compares the accumulated numbers in the same period;

[0027] 3) Observe and judge the yield rate and yield advantage items of semiconductor production in the same period on the display module through the analysis results of the data statistics module.

[0028] (III) Beneficial effects

[0029] Compared with the prior art, the present invention provides a semiconductor testing system and method based on artificial intelligence, which has the following beneficial effects:

[0030] 1. This semiconductor testing system and method based on artificial intelligence sets a time relay inside the testing system to limit the duration of data collection, and sets a timing module to time the semiconductor transportation time and sorting time. The total time consumed for semiconductor testing, transportation and sorting is comprehensively managed through the time management module, and the total time consumed for various semiconductor tests is calculated. During the semiconductor testing process, the fault location during the test process can be judged by the time difference consumed, which is conducive to accurately and efficiently locking the fault location in the test system. At the same time, a metering module is set inside the data processor of the testing system to count each group of data processed by the processor and the processing result of each group of data. According to the total number of processed data, the number of output semiconductors and the number of sorted semiconductors, the yield and defective rate of semiconductors in various tests can be effectively judged. According to the yield and defective rate in various tests, it is easy to understand the advantages and disadvantages of various performances in semiconductor production, which is conducive to improving the production process of semiconductors. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] Figure 1 A schematic diagram of the system structure layout of a semiconductor testing system based on artificial intelligence of the present invention;

[0032] Figure 2 The present invention is a schematic flow chart of the steps of a semiconductor testing method based on artificial intelligence. DETAILED DESCRIPTION

[0033] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0034] See also Figure 1-2 The present invention provides a technical solution: a semiconductor test system based on artificial intelligence, comprising a central processing unit as the core of the system, the central processing unit signal connected to an intelligent database storing standard values ​​of various semiconductor test data and a number of special processors for extracting database data, the special processor signal connected to a data acquisition module for collecting semiconductor test data and a time relay for limiting the data acquisition time, the time relay signal connected to a time management module connected to the central processing unit signal, the time management module signal connected to a semiconductor transport module for transporting semiconductors, the semiconductor transport module signal connected to a timing module for monitoring the transport time, and the time management module signal connected to a sorting and exporting module for sorting unqualified semiconductors;

[0035] Specifically, semiconductors to be tested are placed on a semiconductor transfer module, which is used to transfer the semiconductors to a test location for testing. The semiconductor test data is collected through a data acquisition module and transmitted to the corresponding special processor. The special processor extracts standard data and test data from an intelligent database for comparison, detects whether the test data is located within the standard data, and transmits the comparison result to a central processing unit. The central processing unit sends a control signal to the semiconductor transfer module and a sorting and exporting module to transfer qualified semiconductors to corresponding locations and sort and export unqualified semiconductors to corresponding locations. During data collection, the collection time is controlled by a time relay, the time for transferring semiconductors is detected by a timing module, and the time management module manages the time for transferring semiconductors. During continuous testing, each process interval of the semiconductor test can be timed, and during individual testing, the time consumption of various semiconductor tests can be integrated.

[0036] The dedicated processor is internally provided with a metering module for counting processed data, the metering module signal is connected to a data statistics module, the data statistics module signal is connected to a display module which is connected to a central processing unit signal.

[0037] Preferably, the several special processors correspond to different data acquisition modules, which include temperature and humidity data acquisition modules, electrical signal data acquisition modules, vibration data acquisition modules, storage life data acquisition modules and thermal shock data acquisition modules, wherein each special processor is correspondingly provided with a time relay, a semiconductor transfer module and a sorting export module, which limits the test time for each test data and transfers and sorts the semiconductors after the test.

[0038] Specifically, semiconductors can be transported to temperature and humidity test locations, electrical signal test locations, vibration test locations, storage life test locations, and thermal shock test locations through semiconductor transport modules. Each test corresponds to a semiconductor transport module and a sorting and export module, which facilitates the transport of semiconductors in each test and the export of semiconductors that fail the test.

[0039] Preferably, several of the semiconductor transfer modules are connected to each other to perform continuity testing on semiconductors, and each semiconductor transfer module is provided with a semiconductor input component. At the same time, each semiconductor transfer module is connected to a sorting and exporting module to export unqualified semiconductors.

[0040] Embodiment 1: The various test items of a semiconductor are numbered in sequence, and during the semiconductor test, a semiconductor is sequentially transferred to multiple test positions through a semiconductor transfer module, and multiple different tests are performed on a semiconductor, and data of the semiconductor in different tests, as well as the time of each test of the semiconductor and the transfer time between each test are collected until the semiconductor is damaged or the semiconductor is sorted out, so as to understand the time consumed by a semiconductor to complete each test, and judge the time consumed for each test to be performed continuously;

[0041] Embodiment 2: Test new semiconductors in each test item, and transfer new semiconductors in each test item through the semiconductor transfer module, so that multiple semiconductors can undergo different tests at the same time, and the semiconductors can be tested individually. In the process of individually testing the semiconductors, the time consumption of each test is counted through the time management module, the time consumption difference of each test is determined, and the efficiency of each test is analyzed.

[0042] Preferably, the metering module is arranged inside each special processor and the central processing unit, wherein the metering module inside the central processing unit counts the number of times the central processing unit makes data judgments, the metering module counts the amount of data processed by each special processor, and transmits the counted data to the data statistics module, the data statistics module counts the amount of data processed inside each special processor and constructs a data model to perform data difference analysis, and displays the analysis structure on the display module.

[0043] Specifically, by counting each test in each test, and combining the number of sorted semiconductors, the yield and defective rates of the semiconductors after testing can be calculated and displayed, and the yield and defective rates of the semiconductors in each test can be compared to determine the items that need to be improved in the production process, which can effectively optimize production and analyze and determine the number of semiconductor production tests.

[0044] A testing method for a semiconductor testing system based on artificial intelligence, characterized in that it comprises the following steps:

[0045] S1. Store various standard value data of semiconductors into a database to form an intelligent database for testing, providing data support for the central processor, the brain of artificial intelligence;

[0046] S2. Collect various test data of semiconductors within the time limit of the time relay through various data acquisition modules;

[0047] 1) Set the time relay to control the data collection module to collect data;

[0048] 2) The timing module counts the time taken by the semiconductor transport module to transport semiconductors, and transmits the timing data to the time management module;

[0049] 3) The time management module comprehensively measures the test time and transportation time of each test item, analyzes the time consumed by each test item, and calculates the consumed time and transmits it to the central processor for determining the test duration;

[0050] S3, the special processor receives the corresponding test data and extracts the corresponding test data from the intelligent database for comparison, and transmits the comparison result to the central processor. At the same time, the metering module calculates the number of test data inside the special processor;

[0051] S4, the central processor determines the comparison result of the special processor, and controls the operation of the semiconductor transport module sorting and exporting module according to the determination result;

[0052] 1) Each time the special processor processes the test data, the metering module calculates the number of times by one, and accumulates the number of times the special processor processes the test data;

[0053] 2) Each metering module transmits the accumulated numbers to the data statistics module after each accumulation is completed. The data statistics module combines the corresponding data judgment results of the central processor to make statistics on the quality of each semiconductor data, and analyzes and compares the accumulated numbers in the same period;

[0054] 3) Observe and judge the yield rate and yield advantage items of semiconductor production in the same period on the display module through the analysis results of the data statistics module.

[0055] In summary, the semiconductor testing system and method based on artificial intelligence sets a time relay inside the testing system to limit the duration of data collection, and sets a timing module to time the semiconductor transportation time and sorting time. The total time consumed for semiconductor testing, transportation and sorting is comprehensively managed through the time management module, and the total time consumed for various semiconductor tests is calculated. During the semiconductor testing process, the fault location during the test process can be judged by the time difference consumed, which is conducive to accurately and efficiently locking the fault location in the test system. At the same time, a metering module is set inside the data processor of the testing system to count each group of data processed by the processor and the processing results of each group of data. According to the total number of processed data, the number of output semiconductors and the number of sorted semiconductors, the yield and defective rate of semiconductors in various tests can be effectively judged. According to the yield and defective rate in various tests, it is easy to understand the advantages and disadvantages of various performances in semiconductor production, which is conducive to improving the production process of semiconductors.

[0056] The above description is only a preferred specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can make equivalent replacements or changes according to the technical scheme and inventive concept of the present invention within the technical scope disclosed by the present invention, which should be covered by the protection scope of the present invention.

Claims

1. A semiconductor testing system based on artificial intelligence, characterized in that: The system comprises a central processing unit as the core of the system, wherein the central processing unit is signal-connected to an intelligent database storing standard values ​​of various semiconductor test data and several special processors for extracting database data, wherein the special processor is signal-connected to a data acquisition module for acquiring semiconductor test data and a time relay for limiting the data acquisition duration, wherein the time relay is signal-connected to a time management module connected to the central processing unit, wherein the time management module is signal-connected to a semiconductor transport module for transporting semiconductors, wherein the semiconductor transport module is signal-connected to a timing module for monitoring the transport time, and wherein the time management module is signal-connected to a sorting and exporting module for sorting unqualified semiconductors; The dedicated processor is internally provided with a metering module for counting processed data, the metering module signal is connected to a data statistics module, the data statistics module signal is connected to a display module which is connected to a central processing unit signal.

2. The semiconductor testing system based on artificial intelligence according to claim 1, characterized in that: Several of the special processors correspond to different data acquisition modules, which include temperature and humidity data acquisition modules, electrical signal data acquisition modules, vibration data acquisition modules, storage life data acquisition modules and thermal shock data acquisition modules. Each special processor is correspondingly provided with a time relay, a semiconductor transfer module and a sorting export module, which limits the test time for each test data and transfers and sorts the semiconductors after the test.

3. The semiconductor testing system based on artificial intelligence according to claim 2, characterized in that: Several of the semiconductor transfer modules are connected to each other to perform continuity testing on semiconductors, and each semiconductor transfer module is provided with a semiconductor input component. At the same time, each semiconductor transfer module is connected to a sorting and exporting module to export unqualified semiconductors.

4. The semiconductor testing system based on artificial intelligence according to claim 3, characterized in that: The metering module is set inside each special processor and the central processing unit, wherein the metering module inside the central processing unit counts the number of times the central processing unit makes data judgments, the metering module counts the amount of data processed by each special processor, and transmits the counted data to the data statistics module, the data statistics module counts the amount of data processed inside each special processor and constructs a data model to perform data difference analysis, and displays the analysis structure on the display module.

5. A testing method for a semiconductor testing system based on artificial intelligence, characterized in that: The following steps are involved: S1. Store various standard value data of semiconductors into a database to form an intelligent database for testing, providing data support for the central processor, the brain of artificial intelligence; S2. Collect various test data of semiconductors within the time limit of the time relay through various data acquisition modules; 1) Set the time relay to control the data collection module to collect data; 2) The timing module counts the time taken by the semiconductor transport module to transport semiconductors, and transmits the timing data to the time management module; 3) The time management module comprehensively measures the test time and transportation time of each test item, analyzes the time consumed by each test item, and calculates the consumed time and transmits it to the central processor for determining the test duration; S3, the special processor receives the corresponding test data and extracts the corresponding test data from the intelligent database for comparison, and transmits the comparison result to the central processor. At the same time, the metering module calculates the number of test data inside the special processor; S4, the central processor determines the comparison result of the special processor, and controls the operation of the semiconductor transport module sorting and exporting module according to the determination result; 1) Each time the special processor processes the test data, the metering module calculates the number of times by one, and accumulates the number of times the special processor processes the test data; 2) Each metering module transmits the accumulated numbers to the data statistics module after each accumulation is completed. The data statistics module combines the corresponding data judgment results of the central processor to make statistics on the quality of each semiconductor data, and analyzes and compares the accumulated numbers in the same period; 3) Observe and judge the yield rate and yield advantage items of semiconductor production in the same period on the display module through the analysis results of the data statistics module.

Images