Semiconductor device on-line monitoring system

Abstract

The invention relates to the technical field of semiconductor monitoring, in particular to a semiconductor device online monitoring system which comprises a data acquisition module, a data preprocessing module, an operation analysis module, a data monitoring analysis module, a display and control module and a data storage module. According to the method, the data processing information is obtained by analyzing the data parameters corresponding to the operation data, the data volume can be optimized through the processing information, and then the computer control instruction is obtained by comprehensively analyzing the performance data and the optimized operation data volume, so that multi-calculation control can be performed according to the data volume; therefore, calculation power adjustment is carried out according to the data volume, and subsequent analysis and judgment on semiconductor device control are prevented from being influenced. In addition, monitoring state information of the semiconductor device is obtained by analyzing the optimized operation data, state monitoring is carried out according to the monitoring state information, the purpose of online monitoring of the semiconductor device is achieved, and the problem of the semiconductor device can be found in time.

Description

Technical Field

[0001] The present invention relates to the technical field of semiconductor monitoring, in particular to an online monitoring system for semiconductor devices. Background Art

[0002] In today's electronic information age, semiconductor devices, as core components of electronic equipment, are widely used in computers, communication equipment, consumer electronics, industrial control and other fields. Their performance and reliability are crucial. With the continuous advancement of science and technology, the integration of semiconductor devices is getting higher and higher, the operating frequency is getting faster and faster, and the power density is also increasing.

[0003] Currently, semiconductor devices generate a large amount of data information during operation, and the frequency of signal changes within them is extremely high. Therefore, when monitoring semiconductor devices online, due to the limitations of computer processing performance, the huge amount of data cannot be processed quickly, resulting in delays in data analysis, affecting the subsequent analysis and judgment of semiconductor device control. Summary of the invention

[0004] The present invention provides a semiconductor device online monitoring system for solving the above technical problems.

[0005] A first aspect of the present invention provides an online monitoring system for semiconductor devices, including a data acquisition module, a data preprocessing module, an operation analysis module, a data monitoring and analysis module, a display and control module, and a data storage module.

[0006] The data acquisition module acquires the operation data of the semiconductor device through the detection element, and acquires the performance data of the processing computer, and sends the operation data and performance data to the data storage module; the operation data includes optical parameters, thermal parameters and electrical parameters; the performance data includes the processing-related data, storage-related data and power consumption-related data of each processing computer.

[0007] The data preprocessing module is used to obtain the operating data and identify the operating data to obtain the corresponding data parameters, analyze the data parameters to obtain data processing information, and optimize the data according to the data processing information to obtain optimized operating data, and send the optimized operating data to the data storage module; data optimization includes but is not limited to removing invalid data, removing noise, compressing data, and screening key data.

[0008] As a further improvement of the present invention, the data parameters are analyzed, and the specific analysis method is as follows: Obtain the collection time, data volume and data feature information of each data information corresponding to the data parameters; obtain the pre-set unit detection time, divide the timeline into multiple analysis time periods according to the unit detection time, match each unit detection time period with the collection time of each data information to obtain the data information corresponding to each unit detection time period, count the data volume of each data information corresponding to each unit detection time period, and calculate and sum the data volume corresponding to each data information to obtain the total data volume; Identify the data feature information of each data information corresponding to each unit detection time period to obtain data distribution features, invalid data features and redundant data features; Analyze the data distribution characteristics, identify the types of data information to obtain corresponding category information, obtain the data values ​​corresponding to each category information, overlap and compare the data values ​​corresponding to each category information to obtain the trend data values ​​corresponding to each category information, and obtain a pre-set data fluctuation difference, calculate the trend data value and the data fluctuation difference to obtain the trend data interval, and mark the data information in the trend data interval corresponding to each category information as multi-distributed data; the trend data value is expressed as a relatively stable value of a majority data trend; the category information includes but is not limited to operating temperature, operating voltage, luminous intensity, and response time.

[0009] Analyze the invalid data features, which include out-of-range data and data that do not conform to physical laws; obtain the pre-set acquisition anomaly threshold for the data value corresponding to each data information, compare the data value of each data information with the corresponding acquisition anomaly threshold, and when the data value of each data information is greater than the acquisition anomaly threshold, mark the corresponding data information as out-of-range data information; obtain data information that does not belong to the physical principle in the data information based on each physical principle, and mark the corresponding information as data information that does not conform to the law; combine the out-of-range data information and the data information that does not conform to the law to obtain invalid data information; out-of-range data is represented by a measurement value that exceeds the upper limit of the device's working; data that does not conform to physical laws is represented by data that does not conform to universal physical laws. For example, when the voltage and current are measured at both ends of a semiconductor device with a fixed resistance, it is found that the voltage increases while the current decreases, which does not conform to Ohm's law, and the corresponding data is marked as data that does not conform to physical laws.

[0010] Analyze the redundant data characteristics, compare the data values ​​of each data information for overlap, mark the data information corresponding to the same data value as duplicate data, count the number of duplicate data and record it as the duplicate data amount, obtain a pre-set duplicate data amount threshold, and when the duplicate data amount is greater than the duplicate data amount threshold, calculate the difference between the duplicate data amount and the duplicate data amount threshold to obtain the corresponding duplicate data excess amount, and mark the data information corresponding to the duplicate data excess amount as redundant data information.

[0011] The multi-distributed data, invalid data information and redundant data information are collected to obtain data processing information; the amount of processed data corresponding to the data processing information is obtained, and the total amount of data is subtracted from the amount of processed data to obtain the optimized operation data amount.

[0012] The operation analysis module obtains the optimized operation data and identifies the optimized operation data to obtain the corresponding optimized operation data volume, obtains the performance data of each processing computer, and comprehensively analyzes the performance data and the optimized operation data volume to obtain the computer control instruction, and sends the computer control instruction to the display and control module.

[0013] As a further improvement of the present invention, a comprehensive analysis is performed on the performance data and the amount of optimized operation data, and the specific analysis steps are as follows: A1: Performance data includes processing-related data, storage-related data, and power consumption-related data of each processing computer; A2: Identify the processing-related data to obtain the CPU main frequency and MIPS value; divide the CPU main frequency into multiple frequency intervals, each frequency interval corresponds to a CPU capacity value, match the CPU main frequency of each processing computer with the multiple frequency intervals to obtain the corresponding CPU capacity value; obtain a pre-set instruction minimum limit value, compare the MIPS value of each processing computer with the instruction minimum limit value, and when the MIPS value is greater than the instruction minimum limit value, calculate the difference between the MIPS value and the instruction minimum limit value to obtain the MIPS capacity value; the MIPS value is expressed as the number of instructions that the computer can execute per second. The CPU capacity value and the MIPS capacity value corresponding to each processing computer are calculated and summed to obtain the corresponding processing capacity value, which is recorded as cfn.

[0014] A3: Identify storage-related capabilities to obtain memory capacity, hard disk capacity, and read / write speed; obtain memory usage of each processing computer, perform difference calculation between memory capacity and memory usage to obtain surplus memory, divide the surplus memory into multiple surplus memory intervals, each surplus memory interval corresponds to a memory impact value, and match the surplus memory of each processing computer with the multiple surplus memory intervals to obtain a corresponding memory impact value; A31: Obtain the hard disk capacity and the corresponding read / write speed of each processing computer, obtain the hard disk margin of each processing computer according to the hard disk capacity, and calculate the hard disk capacity value by multiplying the hard disk margin by the corresponding read / write speed of the hard disk; A32: Construct two circles with the memory impact value and hard disk capacity value as the radius, and use the centers of the two circles as the starting point and the end point. Draw a straight line perpendicular to the two circles with the length of the straight line as a fixed constant. Then construct a frustum with the two circles and the straight line, calculate the volume of the frustum and mark the volume value as the storage capacity value, which is recorded as cnl.

[0015] A4: Identify power consumption related data to obtain heat dissipation performance and process power consumption information; heat dissipation performance includes heat sink information and heat dissipation fan information, obtain the heat sink power range according to the heat sink information, obtain the maximum power based on the heat sink power range, and obtain the longest operating time corresponding to the heat sink information, and multiply the longest operating time by the maximum power to obtain the heat sink capacity value; obtain the fan speed according to the heat dissipation fan information, obtain a preset speed reference value, and when the fan speed is greater than the speed reference value, mark the part of the fan speed that exceeds the speed reference value as the fan capacity value; A41: obtaining the heating values ​​of the CPU and the graphics card according to the processing power consumption information of each processing computer, and calculating and summing the heating values ​​of the CPU and the graphics card of each processing computer to obtain the total heat of power consumption; A42: Construct a rectangle with the radiator capacity value and the fan capacity value as the long and short sides of the rectangle, then construct a circle with the center of gravity of the rectangle as the center and the total heat value of power consumption as the radius, remove the circle from the rectangle to obtain a closed figure, calculate the area of ​​the closed figure and mark the area value as the heat dissipation capacity value, and record it as srn.

[0016] A5: Normalize the processing capability, storage capability, and heat dissipation capability and take their values. Use the formula The computer performance value JS is obtained; wherein k1, k2 and k3 are all preset weight factors, and the values ​​are 3.33, 2.97 and 3.25 respectively.

[0017] A6: Obtain the optimized running data volume and divide the running optimized data volume into multiple data volume intervals, each data volume interval corresponds to a performance requirement value, match the current optimized data volume with the multiple data volume intervals to obtain the corresponding performance requirement values, and match the computer performance value of each processing computer with the performance requirement value to obtain the corresponding computer control instruction.

[0018] The data monitoring and analysis module is used to obtain the optimized operation data and analyze the optimized operation data to obtain the monitoring status information of the semiconductor device. When the monitoring status information indicates that the semiconductor device is abnormal, the corresponding semiconductor abnormality information is generated and the semiconductor abnormality information is sent to the display and control module.

[0019] As a further improvement of the present invention, the optimization operation data is analyzed, and the specific analysis steps are as follows: S1: Optimization operation data including optical parameters, thermal parameters and electrical parameters; S2: Obtain optical analysis items corresponding to the optical parameters, where the optical analysis items include but are not limited to luminous intensity, wavelength, and luminous efficiency; and obtain the upper limit threshold corresponding to each optical analysis item. When the value corresponding to each optical analysis item is greater than the upper limit threshold, obtain the part of the data corresponding to each optical analysis item that exceeds the upper limit threshold, and add and sum them to obtain the optical anomaly value.

[0020] S3: Obtain the operating temperature of the semiconductor device based on the thermal parameters, obtain the heat generation time corresponding to the operating temperature, calculate the operating temperature and the corresponding heat generation time to obtain the heat generation efficiency, obtain a pre-set heat generation efficiency threshold, compare the heat generation efficiency corresponding to the current thermal parameters with the heat generation efficiency threshold, and when the heat generation efficiency is greater than the heat generation efficiency threshold, record the corresponding heat generation efficiency as the thermal impact value.

[0021] S4: The electrical parameters include voltage value, current value and resistance value; obtain the voltage reference value corresponding to the semiconductor period, calculate the difference between the voltage value corresponding to the current electrical parameter and the voltage reference value to obtain the voltage difference; similarly, obtain the current difference and resistance difference corresponding to the current value and resistance value corresponding to the current electrical parameter; sum and calculate the voltage difference, current difference and resistance difference to obtain the electrical impact value.

[0022] S5: Take the values ​​of the optical anomaly value and the thermal impact value as the values ​​of the two vertical line segments, connect the end points of the two line segments to obtain a right triangle, take the midpoint of the hypotenuse of the right triangle as the starting point, draw a straight line perpendicular to the right triangle, the length of the straight line is equal to the value of the electrical impact value, and then construct a triangular pyramid with the straight line and the right triangle, calculate the volume of the triangular pyramid and mark the value of the volume as the semiconductor anomaly value; when the semiconductor anomaly value is greater than the set threshold, generate the corresponding monitoring status information as semiconductor anomaly.

[0023] The display and control module is used to receive computer control instructions and semiconductor abnormality information, perform computer operation control according to the computer control instructions, and display abnormal information and issue early warnings according to the semiconductor abnormality information.

[0024] The data storage module is used to store operation data and performance data, and store computer control instructions and semiconductor abnormality information.

[0025] Compared with the prior art, the technical solution provided by the present invention has the following beneficial effects: 1. The present invention obtains data processing information by analyzing data parameters corresponding to the operating data. The data volume can be optimized by processing the information. Then, computer control instructions are obtained by comprehensively analyzing the performance data and the optimized operating data volume. Therefore, multiple calculations can be controlled according to the data volume, so as to adjust the computing power according to the data volume to avoid affecting the subsequent analysis and judgment of the semiconductor device control.

[0026] 2. The present invention obtains monitoring status information of semiconductor devices by analyzing optimized operation data, and performs status monitoring based on the monitoring status information, thereby achieving the purpose of online monitoring of semiconductor devices and facilitating timely discovery of semiconductor device problems. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the following briefly introduces the drawings required for use in the description of the embodiments. The following drawings are not intentionally scaled to the actual size, and the focus is on illustrating the main purpose of the present application.

[0028] Figure 1 It is a principle block diagram of the present invention. DETAILED DESCRIPTION

[0029] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions 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.

[0030] For ease of understanding, the specific process of the embodiment of the present invention is described below. Figure 1 In an embodiment of the present invention, an embodiment of an online monitoring system for semiconductor devices includes: a data acquisition module, a data preprocessing module, an operation analysis module, a data monitoring and analysis module, a display and control module and a data storage module.

[0031] The data acquisition module acquires the operation data of the semiconductor device through the detection element, and acquires the performance data of the processing computer, and sends the operation data and performance data to the data storage module; the operation data includes optical parameters, thermal parameters and electrical parameters; the performance data includes processing-related data, storage-related data and power consumption-related data of each processing computer.

[0032] The data preprocessing module acquires the operating data and identifies the operating data to obtain corresponding data parameters, analyzes the data parameters to obtain data processing information, optimizes the data according to the data processing information to obtain optimized operating data, and sends the optimized operating data to the data storage module; data optimization includes but is not limited to removing invalid data, removing noise, compressing data, and screening key data.

[0033] The data parameters are analyzed in the following specific ways: Obtain the collection time, data volume and data feature information of each data information corresponding to the data parameters; obtain the pre-set unit detection time, divide the timeline into multiple analysis time periods according to the unit detection time, match each unit detection time period with the collection time of each data information to obtain the data information corresponding to each unit detection time period, count the data volume of each data information corresponding to each unit detection time period, and calculate and sum the data volume corresponding to each data information to obtain the total data volume; Identify the data feature information of each data information corresponding to each unit detection time period to obtain data distribution features, invalid data features and redundant data features; Analyze the data distribution characteristics, identify the types of data information to obtain corresponding category information, obtain the data values ​​corresponding to each category information, overlap and compare the data values ​​corresponding to each category information to obtain the trend data values ​​corresponding to each category information, and obtain a pre-set data fluctuation difference, calculate the trend data value and the data fluctuation difference to obtain the trend data interval, and mark the data information in the trend data interval corresponding to each category information as multi-distributed data; the trend data value is expressed as a relatively stable value of a majority data trend; the category information includes but is not limited to operating temperature, operating voltage, luminous intensity, and response time.

[0034] Analyze the invalid data features, which include out-of-range data and data that do not conform to physical laws; obtain the pre-set acquisition anomaly threshold for the data value corresponding to each data information, compare the data value of each data information with the corresponding acquisition anomaly threshold, and when the data value of each data information is greater than the acquisition anomaly threshold, mark the corresponding data information as out-of-range data information; obtain data information that does not belong to the physical principle in the data information based on each physical principle, and mark the corresponding information as data information that does not conform to the law; combine the out-of-range data information and the data information that does not conform to the law to obtain invalid data information; out-of-range data is represented by a measurement value that exceeds the upper limit of the device's working; data that does not conform to physical laws is represented by data that does not conform to universal physical laws. For example, when the voltage and current are measured at both ends of a semiconductor device with a fixed resistance, it is found that the voltage increases while the current decreases, which does not conform to Ohm's law, and the corresponding data is marked as data that does not conform to physical laws.

[0035] Analyze the redundant data characteristics, compare the data values ​​of each data information for overlap, mark the data information corresponding to the same data value as duplicate data, count the number of duplicate data and record it as the duplicate data amount, obtain a pre-set duplicate data amount threshold, and when the duplicate data amount is greater than the duplicate data amount threshold, calculate the difference between the duplicate data amount and the duplicate data amount threshold to obtain the corresponding duplicate data excess amount, and mark the data information corresponding to the duplicate data excess amount as redundant data information.

[0036] The multi-distributed data, invalid data information and redundant data information are collected to obtain data processing information; the amount of processed data corresponding to the data processing information is obtained, and the total amount of data is subtracted from the amount of processed data to obtain the optimized operation data amount.

[0037] The operation analysis module obtains the optimized operation data and identifies the optimized operation data to obtain the corresponding optimized operation data volume, obtains the performance data of each processing computer, and comprehensively analyzes the performance data and the optimized operation data volume to obtain computer control instructions, and sends the computer control instructions to the display and control module.

[0038] Comprehensively analyze the performance data and optimized operation data volume. The specific analysis steps are as follows: A1: Performance data includes processing-related data, storage-related data, and power consumption-related data of each processing computer; A2: Identify the processing-related data to obtain the CPU main frequency and MIPS value; divide the CPU main frequency into multiple frequency intervals, each frequency interval corresponds to a CPU capacity value, match the CPU main frequency of each processing computer with the multiple frequency intervals to obtain the corresponding CPU capacity value; obtain a pre-set instruction minimum limit value, compare the MIPS value of each processing computer with the instruction minimum limit value, and when the MIPS value is greater than the instruction minimum limit value, calculate the difference between the MIPS value and the instruction minimum limit value to obtain the MIPS capacity value; the MIPS value is expressed as the number of instructions that the computer can execute per second. The CPU capacity value and the MIPS capacity value corresponding to each processing computer are calculated and summed to obtain the corresponding processing capacity value.

[0039] A3: Identify storage-related capabilities to obtain memory capacity, hard disk capacity, and read / write speed; obtain memory usage of each processing computer, perform difference calculation between memory capacity and memory usage to obtain surplus memory, divide the surplus memory into multiple surplus memory intervals, each surplus memory interval corresponds to a memory impact value, and match the surplus memory of each processing computer with the multiple surplus memory intervals to obtain a corresponding memory impact value; A31: Obtain the hard disk capacity and the corresponding read / write speed of each processing computer, obtain the hard disk margin of each processing computer according to the hard disk capacity, and calculate the hard disk capacity value by multiplying the hard disk margin by the corresponding read / write speed of the hard disk; A32: Construct two circles with the memory impact value and hard disk capacity value as the radius, and use the centers of the two circles as the starting point and the end point. Draw a straight line perpendicular to the two circles with the length of the straight line as a fixed constant. Then construct a frustum with the two circles and the straight line, calculate the volume of the frustum and mark the volume value as the storage capacity value.

[0040] A4: Identify power consumption related data to obtain heat dissipation performance and process power consumption information; heat dissipation performance includes heat sink information and heat dissipation fan information, obtain the heat sink power range according to the heat sink information, obtain the maximum power based on the heat sink power range, and obtain the longest operating time corresponding to the heat sink information, and multiply the longest operating time by the maximum power to obtain the heat sink capacity value; obtain the fan speed according to the heat dissipation fan information, obtain a preset speed reference value, and when the fan speed is greater than the speed reference value, mark the part of the fan speed that exceeds the speed reference value as the fan capacity value; A41: obtaining the heating values ​​of the CPU and the graphics card according to the processing power consumption information of each processing computer, and calculating and summing the heating values ​​of the CPU and the graphics card of each processing computer to obtain the total heat of power consumption; A42: Construct a rectangle with the radiator capacity value and the fan capacity value as the long and short sides of the rectangle, then construct a circle with the center of gravity of the rectangle as the center and the total heat value of power consumption as the radius, remove the circle from the rectangle to obtain a closed figure, calculate the area of ​​the closed figure and mark the area value as the heat dissipation capacity value.

[0041] A5: Normalize the processing capability, storage capability, and heat dissipation capability and take their values. Use the formula The computer performance value JS is obtained; wherein cfn, cnl and srn represent the processing capability value, storage capability value and heat dissipation capability value respectively; k1, k2 and k3 are all preset weight factors, and the values ​​are 3.33, 2.97 and 3.25 respectively.

[0042] A6: Obtain the optimized running data volume and divide the running optimized data volume into multiple data volume intervals, each data volume interval corresponds to a performance requirement value, match the current optimized data volume with the multiple data volume intervals to obtain the corresponding performance requirement values, and match the computer performance value of each processing computer with the performance requirement value to obtain the corresponding computer control instruction.

[0043] The data monitoring and analysis module acquires the optimized operation data and analyzes the optimized operation data to obtain the monitoring status information of the semiconductor device. When the monitoring status information indicates that the semiconductor device is abnormal, the corresponding semiconductor abnormality information is generated and the semiconductor abnormality information is sent to the display and control module.

[0044] The optimization operation data is analyzed, and the specific analysis steps are as follows: S1: Optimization operation data including optical parameters, thermal parameters and electrical parameters; S2: Obtain optical analysis items corresponding to the optical parameters, where the optical analysis items include but are not limited to luminous intensity, wavelength, and luminous efficiency; and obtain the upper limit threshold corresponding to each optical analysis item. When the value corresponding to each optical analysis item is greater than the upper limit threshold, obtain the part of the data corresponding to each optical analysis item that exceeds the upper limit threshold, and add and sum them to obtain the optical anomaly value.

[0045] S3: Obtain the operating temperature of the semiconductor device based on the thermal parameters, obtain the heat generation time corresponding to the operating temperature, calculate the operating temperature and the corresponding heat generation time to obtain the heat generation efficiency, obtain a pre-set heat generation efficiency threshold, compare the heat generation efficiency corresponding to the current thermal parameters with the heat generation efficiency threshold, and when the heat generation efficiency is greater than the heat generation efficiency threshold, record the corresponding heat generation efficiency as the thermal impact value.

[0046] S4: The electrical parameters include voltage value, current value and resistance value; obtain the voltage reference value corresponding to the semiconductor period, calculate the difference between the voltage value corresponding to the current electrical parameter and the voltage reference value to obtain the voltage difference; similarly, obtain the current difference and resistance difference corresponding to the current value and resistance value corresponding to the current electrical parameter; sum and calculate the voltage difference, current difference and resistance difference to obtain the electrical impact value.

[0047] S5: Take the values ​​of the optical anomaly value and the thermal impact value as the values ​​of the two vertical line segments, connect the end points of the two line segments to obtain a right triangle, take the midpoint of the hypotenuse of the right triangle as the starting point, draw a straight line perpendicular to the right triangle, the length of the straight line is equal to the value of the electrical impact value, and then construct a triangular pyramid with the straight line and the right triangle, calculate the volume of the triangular pyramid and mark the value of the volume as the semiconductor anomaly value; when the semiconductor anomaly value is greater than the set threshold, generate the corresponding monitoring status information as semiconductor anomaly.

[0048] The display and control module receives computer control instructions and semiconductor abnormality information, performs computer operation control according to the computer control instructions, and performs abnormality information display and warning according to the semiconductor abnormality information.

[0049] The data storage module stores operation data and performance data, and stores computer control instructions and semiconductor abnormality information.

[0050] As described above, the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features thereof may be replaced by equivalents. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A semiconductor device online monitoring system, comprising a data acquisition module and a data storage module, characterized in that: Also includes: A data preprocessing module is used to obtain operation data and identify the operation data to obtain corresponding data parameters, analyze the data parameters to obtain data processing information, optimize the data according to the data processing information to obtain optimized operation data, and send the optimized operation data to the data storage module; An operation analysis module obtains optimized operation data and identifies the optimized operation data to obtain corresponding optimized operation data volume, obtains performance data of each processing computer, and comprehensively analyzes the performance data and the optimized operation data volume to obtain computer control instructions, and sends the computer control instructions to the display and control module; A data monitoring and analysis module is used to obtain optimized operation data and analyze the optimized operation data to obtain monitoring status information of the semiconductor device. When the monitoring status information indicates that the semiconductor device is abnormal, corresponding semiconductor abnormality information is generated and the semiconductor abnormality information is sent to the display and control module. The display and control module is used to receive computer control instructions and semiconductor abnormality information, perform computer operation control according to the computer control instructions, and display abnormal information and issue early warnings according to the semiconductor abnormality information.

2. A semiconductor device online monitoring system according to claim 1, characterized in that: The data acquisition module acquires the operation data of the semiconductor device through the detection element, acquires the performance data of the processing computer, and sends the operation data and the performance data to the data storage module; the operation data includes optical parameters, thermal parameters and electrical parameters; the performance data includes the processing-related data, storage-related data and power consumption-related data of each processing computer.

3. The semiconductor device online monitoring system according to claim 1, characterized in that: The specific analysis method of the data parameters is as follows: Obtain the collection time, data volume and data feature information of each data information corresponding to the data parameters; obtain the pre-set unit detection time, divide the timeline into multiple analysis time periods according to the unit detection time, match each unit detection time period with the collection time of each data information to obtain the data information corresponding to each unit detection time period, count the data volume of each data information corresponding to each unit detection time period, and calculate and sum the data volume corresponding to each data information to obtain the total data volume; Identify the data feature information of each data information corresponding to each unit detection time period to obtain data distribution features, invalid data features and redundant data features; Analyze the data distribution characteristics, identify the types of data information to obtain corresponding category information, obtain the data values ​​corresponding to each category information, compare the data values ​​corresponding to each category information to obtain the trend data values ​​corresponding to each category information, obtain the pre-set data fluctuation difference, calculate the trend data value and the data fluctuation difference to obtain the trend data interval, and mark the data information in the trend data interval corresponding to each category information as multi-distribution data; Analyze the invalid data features, which include out-of-range data and data that do not conform to physical laws; obtain the pre-set acquisition anomaly threshold for the data value corresponding to each data information, compare the data value of each data information with the corresponding acquisition anomaly threshold, and when the data value of each data information is greater than the acquisition anomaly threshold, mark the corresponding data information as out-of-range data information; obtain data information that does not belong to the physical principle in the data information based on each physical principle, and mark the corresponding information as data information that does not conform to the law; synthesize the out-of-range data information and the data information that does not conform to the law to obtain invalid data information; Analyze the redundant data features, compare the data values ​​of each data information, mark the data information corresponding to the same data value as duplicate data, count the number of each duplicate data and record it as the duplicate data amount, obtain a preset duplicate data amount threshold, and when the duplicate data amount is greater than the duplicate data amount threshold, calculate the difference between the duplicate data amount and the duplicate data amount threshold to obtain the corresponding duplicate data excess amount, and mark the data information corresponding to the duplicate data excess amount as redundant data information; The multi-distributed data, invalid data information and redundant data information are collected to obtain data processing information; the amount of processed data corresponding to the data processing information is obtained, and the total amount of data is subtracted from the amount of processed data to obtain the optimized operation data amount.

4. The semiconductor device online monitoring system according to claim 1, characterized in that: The specific analysis steps for the comprehensive analysis of performance data and optimized operation data are as follows: A1: Performance data includes processing-related data, storage-related data, and power consumption-related data of each processing computer; A2: Identify the processing-related data to obtain the CPU main frequency and MIPS value; divide the CPU main frequency into multiple frequency intervals, each frequency interval corresponds to a CPU capability value, match the CPU main frequency of each processing computer with the multiple frequency intervals to obtain the corresponding CPU capability value; obtain a preset instruction minimum limit value, compare the MIPS value of each processing computer with the instruction minimum limit value, and when the MIPS value is greater than the instruction minimum limit value, perform a difference calculation between the MIPS value and the instruction minimum limit value to obtain the MIPS capability value; A3: Identify storage-related capabilities to obtain memory capacity, hard disk capacity, and read / write speed; obtain memory usage of each processing computer, perform difference calculation between memory capacity and memory usage to obtain surplus memory, divide the surplus memory into multiple surplus memory intervals, each surplus memory interval corresponds to a memory impact value, and match the surplus memory of each processing computer with the multiple surplus memory intervals to obtain a corresponding memory impact value; A31: Obtain the hard disk capacity and the corresponding read / write speed of each processing computer, obtain the hard disk margin of each processing computer according to the hard disk capacity, and calculate the hard disk capacity value by multiplying the hard disk margin by the corresponding read / write speed of the hard disk; A32: Construct two circles with the memory impact value and hard disk capacity value as the radius, and use the centers of the two circles as the starting point and the end point, draw a straight line perpendicular to the two circles, and the length of the straight line is a fixed constant. Then construct a frustum with the two circles and the straight line, calculate the volume of the frustum and mark the volume value as the storage capacity value. A4: Identify power consumption related data to obtain heat dissipation performance and process power consumption information; The heat dissipation performance includes the radiator information and the cooling fan information. The radiator power range is obtained according to the radiator information, the maximum power is obtained based on the radiator power range, and the longest operating time corresponding to the radiator information is obtained. The maximum operating time is multiplied by the maximum power to obtain the radiator capacity value; the fan speed is obtained according to the cooling fan information, and a preset speed reference value is obtained. When the fan speed is greater than the speed reference value, the part of the fan speed that exceeds the speed reference value is marked as the fan capacity value; A41: obtaining the heating values ​​of the CPU and the graphics card according to the processing power consumption information of each processing computer, and calculating and summing the heating values ​​of the CPU and the graphics card of each processing computer to obtain the total heat of power consumption; A42: A rectangle is constructed with the radiator capacity value and the fan capacity value as the long and short sides of the rectangle, and then a circle is constructed with the center of gravity of the rectangle as the center and the value of the total heat of power consumption as the radius. The circle is removed from the rectangle to obtain a closed figure. The area of ​​the closed figure is calculated and the area value is marked as the heat dissipation capacity value. A5: The processing capacity value, storage capacity value and heat dissipation capacity value are calculated comprehensively to obtain the computer performance value; A6: Obtain the optimized running data volume and divide the running optimized data volume into multiple data volume intervals, each data volume interval corresponds to a performance requirement value, match the current optimized data volume with the multiple data volume intervals to obtain the corresponding performance requirement values, and match the computer performance value of each processing computer with the performance requirement value to obtain the corresponding computer control instruction.

5. The semiconductor device online monitoring system according to claim 1, characterized in that: The specific analysis steps for analyzing the optimization operation data are as follows: S1: Optimization operation data including optical parameters, thermal parameters and electrical parameters; S2: Obtain optical analysis items corresponding to the optical parameters, and obtain upper limit thresholds corresponding to each optical analysis item. When the value corresponding to each optical analysis item is greater than the upper limit threshold, obtain the portion of the data corresponding to each optical analysis item that exceeds the upper limit threshold, and add and sum them to obtain an optical abnormality value. S3: obtaining an operating temperature of the semiconductor device based on thermal parameters, obtaining a heat generation time corresponding to the operating temperature, calculating the operating temperature and the corresponding heat generation time to obtain a heat generation efficiency, obtaining a preset heat generation efficiency threshold, comparing the heat generation efficiency corresponding to the current thermal parameters with the heat generation efficiency threshold, and when the heat generation efficiency is greater than the heat generation efficiency threshold, recording the corresponding heat generation efficiency as a thermal impact value; S4: The electrical parameters include voltage value, current value and resistance value; obtaining a voltage reference value corresponding to the semiconductor period, and performing a difference calculation between the voltage value corresponding to the current electrical parameter and the voltage reference value to obtain a voltage difference value; Similarly, the current difference and resistance difference corresponding to the current value and resistance value of the current electrical parameters are obtained; The voltage difference, current difference and resistance difference are summed up to obtain the electrical influence value; S5: Take the values ​​of the optical anomaly value and the thermal impact value as the values ​​of the two vertical line segments, connect the end points of the two line segments to obtain a right triangle, take the midpoint of the hypotenuse of the right triangle as the starting point, draw a straight line perpendicular to the right triangle, the length of the straight line is equal to the value of the electrical impact value, and then construct a triangular pyramid with the straight line and the right triangle, calculate the volume of the triangular pyramid and mark the value of the volume as the semiconductor anomaly value; when the semiconductor anomaly value is greater than the set threshold, generate the corresponding monitoring status information as semiconductor anomaly.

Images