Intelligent AI computing power system working state control system

Abstract

The invention discloses an intelligent AI computing power system working state control system, and relates to the technical field of AI computing power system management, and the system comprises a data collection module, a computing power analysis module, an equipment prediction module and an equipment planning module: the data collection module is used for collecting the data of the intelligent AI computing power system during the operation of the last year, and classifying the collected data; according to the technical key points, data of the intelligent AI computing power system during operation in the last year are analyzed and processed, then prediction is carried out, the total computing power Zsl needed by operation of the intelligent AI computing power system at the future time point of the current year is predicted, and then the number of needed devices is analyzed and judged according to the total computing power Zsl. Related personnel can be helped to know the number of devices needed at different time points in the future, the devices can be reasonably distributed, the situation that too many devices are idle or standby devices are insufficient is avoided, the overall using effect is good, and the good using prospect is achieved.

Description

Technical Field

[0001] The present invention relates to the field of AI computing power system management technology, and specifically to an intelligent AI computing power system working status control system. Background Art

[0002] Artificial intelligence is an important driving force for the new round of scientific and technological revolution and industrial transformation. It is a new technical science that studies and develops theories, methods, technologies and application systems for simulating, extending and expanding human intelligence.

[0003] Artificial intelligence is an important part of the intelligent discipline. It attempts to understand the essence of intelligence and produce a new intelligent machine that can respond in a similar way to human intelligence. Artificial intelligence is a very broad science, including robotics, language recognition, image recognition, natural language processing, expert systems, machine learning, computer vision, etc.

[0004] With the rapid development of artificial intelligence (AI) technology, the penetration of AI applications in various fields continues to deepen, and the demand for high-performance computing resources is also increasing rapidly.

[0005] The existing patent for invention entitled "A method and system for calling an application server based on Nginx" with the authorization announcement number CN109788008B records the following: using a pre-written shell script to detect the operating parameters of application server A, when the operating parameters are greater than or equal to a preset threshold, obtaining the operating parameters of other application servers in the system, comparing and obtaining the most idle application server B among the other application servers, and determining that the operating parameters of application server B are greater than the operating parameters of application server A; sending a start command to application server B so that the shell script of application server B starts the same application pre-deployed on application server B; modifying the Nginx configuration of application server A through a shell script, and using the asynchronous non-blocking event processing mechanism of Nginx to switch the preset percentage of visits on application server A to the application server B. The present invention can start an idle server to bear part of the visits when the performance of the application server is reduced, so as to achieve the purpose of reasonable use of resources.

[0006] However, based on the above content and the prior art, the central idea of ​​the solution recorded in the above patent is mainly to judge the operating status of the device, and then judge whether it can support the computing power required for the use of the software based on the operating status of the existing device.

[0007] However, the above scheme has great defects in actual use. The above scheme is mainly suitable for simple usage scenarios with little data fluctuation. However, most industries have peak and off-peak seasons. The amount of data that needs to be processed in different periods is very large, and the types of data to be processed are many. The above method is usually not prepared, and it is easy for the reserved equipment to be insufficient for the software to run or too many equipment to be reserved, resulting in low overall economic benefits and failing to meet people's usage requirements. For this reason, we have developed an intelligent AI computing system working status control system. Summary of the invention

[0008] 1. Technical issues to be resolved

[0009] In view of the deficiencies in the prior art, the present invention provides a working status control system for an intelligent AI computing power system, which analyzes and processes the data of the intelligent AI computing power system during its operation in the previous year, and then makes a prediction to predict the total computing power Zsl required for the operation of the intelligent AI computing power system at a future time point in this year, and then analyzes and determines the number of required devices based on the total computing power Zsl. This can help relevant personnel understand the number of devices required at different time points in the future, and can reasonably allocate equipment to avoid too many idle devices or insufficient spare equipment. The overall use effect is good and it has a good prospect for use.

[0010] (II) Technical solution

[0011] To achieve the above objectives, the present invention is implemented through the following technical solutions:

[0012] An intelligent AI computing system working state control system, including a data acquisition module, a computing power analysis module, an equipment prediction module and an equipment planning module:

[0013] Data collection module: used to collect the data of the intelligent AI computing system during its operation in the previous year and classify the collected data;

[0014] Computing power analysis module: used to process and analyze the classified data and calculate the total computing power Zsl required for the operation of the intelligent AI computing power system at a future time point;

[0015] Equipment prediction module: obtains the historical operation data of the equipment, analyzes the state value Ztz of the equipment at different time points, and calculates the computing power Slx that the equipment can provide at different times based on the state value Ztz;

[0016] Equipment planning module: obtains the total computing power Zsl required for the operation of the intelligent AI computing system at a future time point and the computing power Slx that the equipment can provide at different times, analyzes the list of equipment required at a future time point, and generates an adjustment plan based on the analysis results.

[0017] The present invention provides a working status control system for an intelligent AI computing power system, which analyzes and processes the data of the intelligent AI computing power system during its operation in the previous year, and then makes a prediction to predict the total computing power Zsl required for the operation of the intelligent AI computing power system at a future time point in this year, and then analyzes and determines the number of required devices based on the total computing power Zsl. This can help relevant personnel understand the number of devices required at different time points in the future, and can reasonably allocate devices to avoid the situation where too many devices are idle or there is insufficient spare equipment. The overall use effect is good and the use prospect is good.

[0018] The present invention provides an intelligent AI computing power system working status control system, which obtains historical operation data of a device and analyzes the status of the device. It can directly analyze the status value Ztz of the device at different time points, and further analyze to obtain the computing power that the device can provide at different time points, and analyze the influence of the loss of the device on the computing power. It can effectively avoid the situation where the device provides insufficient computing power due to the loss of the device, causing the device to run at high power for a long time, resulting in a short service life of the device. Moreover, after calculation, the allocation of subsequent equipment tasks is more reasonable, the use effect is good, and it has a good prospect for use.

[0019] Preferably, the data collected by the data acquisition module includes the number of model parameters of the intelligent AI computing system, the time set for model operation and the amount of data that the system needs to process, and the collected data is classified in chronological order.

[0020] Preferably, the calculation formula for the total computing power Zsl required for the operation of the intelligent AI computing power system at a future time point is as follows:

[0021]

[0022] Where Mxcs is the number of model parameters, A is the ratio of the number of model parameters to the complexity of model operation, Sjl is the amount of data that the system needs to process, and B is the annual growth rate of data.

[0023] Preferably, the steps of obtaining the historical operation data of the device and analyzing the status values ​​of the device at different time points are as follows:

[0024] S1 obtains the initial working data of the equipment, the current working data and the working time of the equipment;

[0025] S2 calculates the loss value of the device and determines the loss stage of the device according to the attenuation value;

[0026] S3 calculates the state value Ztz of the device at different time points according to the wear stage of the device.

[0027] Preferably, the initial working data and current working data of the device in step S1 are the maximum computing power of the tested device.

[0028] Preferably, the calculation formula for calculating the equipment loss value in step S2 is as follows:

[0029]

[0030] In the formula, Shsz is the equipment loss value, Sl c is the initial maximum computing power of the device, Sl x This is the current maximum computing power of the device.

[0031] Preferably, the loss stage of the device in step S2 includes a stable stage and a rapid loss stage, the stable stage corresponds to an interval of (0, D), the rapid loss stage corresponds to an interval of (D, 100), and the value range of D is (30, 40);

[0032] When the calculated Shsz is in the stable stage, the formula for calculating the state value Ztz of the equipment at different time points is as follows:

[0033]

[0034] In the formula, sysc is the working time of the device, Sjd is the time point, and Dqsj is the current time;

[0035] When the calculated Shsz is in the rapid loss stage, mark and replace with new equipment.

[0036] Preferably, the formula for calculating the computing power Slx that the device can provide at different times according to the state value Ztz is as follows:

[0037] Slx=Sl c ×Ztz×Aqyz

[0038] Where Aqyz is the ratio of the computing power when the device is initially working normally to the initial maximum computing power of the device.

[0039] Preferably, the steps for analyzing the list of equipment required at a future time point are as follows:

[0040] Obtain the total computing power Zsl required for the operation of the intelligent AI computing system at the closest time point in the future and the computing power Slx that the device can provide at different times;

[0041] Calculate the total computing power Slh of the devices currently in working state, and then determine the size of the total computing power Slh and the total computing power Zsl required for the operation of the intelligent AI computing system;

[0042]

[0043] In the formula, n is the computing power of the device currently in working state, Slx i is the computing power of the i-th device in working state;

[0044] If Slh<Zsl, increase the number of devices and recalculate the total computing power until Slh n ≥Zsl, and obtain the added device names, and summarize the added device names and the names of the devices currently in working state to generate a device list;

[0045] If Slh>Zs, reduce the device with the minimum computing power and recalculate the total computing power until Slh n <Zsl, at this time, obtain the equipment that has been reduced n-1 times, and remove the reduced equipment names from the equipment currently in working state, and summarize and generate an equipment list.

[0046] Preferably, when the analysis result generates the adjustment plan, a list of devices at all time points is obtained, and then the devices added or reduced in the list of devices are marked.

[0047] (III) Beneficial effects

[0048] The present invention provides an intelligent AI computing system working state control system, which has the following beneficial effects:

[0049] 1. The present invention records a working status control system for an intelligent AI computing power system, which analyzes and processes the data of the intelligent AI computing power system during its operation in the previous year, and then makes a prediction to predict the total computing power Zsl required for the operation of the intelligent AI computing power system at a future time point in this year, and then analyzes and determines the number of required devices based on the total computing power Zsl. This can help relevant personnel understand the number of devices required at different time points in the future, and can reasonably allocate equipment to avoid too many idle devices or insufficient spare equipment. The overall use effect is good and has a good prospect for use.

[0050] 2. The present invention records an intelligent AI computing power system working status control system, which obtains the historical operation data of the equipment and analyzes the status of the equipment. It can directly analyze the status value Ztz of the equipment at different time points, and further analyze to obtain the computing power that the equipment can provide at different time points. The impact of equipment loss on the computing power is analyzed, which can effectively avoid the situation where the equipment provides insufficient computing power due to equipment loss, causing the equipment to run at high power for a long time, resulting in a short service life of the equipment. Moreover, after calculation, the allocation of subsequent equipment tasks is more reasonable, the use effect is good, and it has a good prospect for use. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] Figure 1 This is a flow chart of a working state control system of an intelligent AI computing power system of the present invention. DETAILED DESCRIPTION

[0052] 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.

[0053] Research reasons

[0054] The central idea of ​​the existing intelligent AI computing system working status control system solution is to judge the operating status of the equipment, and then determine whether it can support the computing power required for the software based on the operating status of the existing equipment.

[0055] However, this solution has major flaws in actual use. This solution is mainly suitable for simple usage scenarios with little data fluctuation. However, most industries have peak and off-peak seasons, and the amount of data that needs to be processed in different periods is very large, and the types of data to be processed are many. The above method is usually not prepared, and it is easy for the reserved equipment to be insufficient for the software to run or too much equipment to be reserved, resulting in low overall economic benefits.

[0056] Moreover, the tasks of existing equipment are evenly distributed, but the equipment has different service lives and the wear and tear caused by daily operation, so the computing power that can be provided is different. In order to avoid damage to the equipment due to high-power operation caused by excessive task allocation, this method usually allocates a small amount of tasks to the equipment, resulting in waste of equipment.

[0057] Study plan

[0058] An intelligent AI computing system working state control system, such as Figure 1 As shown, it includes data acquisition module, computing power analysis module, equipment prediction module and equipment planning module:

[0059] In order to better allocate equipment, it is necessary to understand the computing power required when the intelligent AI computing system is running. In order to understand the computing power when the intelligent AI computing system is running, it is necessary to collect past data and analyze the past data. This process is based on the data collection module.

[0060] The data collection module is used to collect the data of the intelligent AI computing system during its operation in the previous year and classify the collected data;

[0061] The data collected by the data acquisition module includes the number of model parameters of the intelligent AI computing system, the time set for model operation, and the amount of data that the system needs to process. The collected data is classified in chronological order.

[0062] The above data can be directly obtained through the enterprise database.

[0063] After obtaining the required data, further analysis and calculation of the obtained data is required to obtain the required data, and further analysis and calculation of the obtained data is based on the computing power analysis module.

[0064] Computing power analysis module: used to process and analyze the classified data and calculate the total computing power Zsl required for the operation of the intelligent AI computing power system at a future time point;

[0065] The calculation formula for the total computing power Zsl required for the operation of the intelligent AI computing system at a future time point is as follows:

[0066]

[0067] Where Mxcs is the number of model parameters, A is the ratio of the number of model parameters to the complexity of model operation, Sjl is the amount of data that the system needs to process, and B is the annual growth rate of data.

[0068] The annual growth rate of data is provided by the enterprise's plan, and special plans are adopted for special activities. This patent only aims at the calculation of the total computing power Zsl for daily use.

[0069] The present invention provides an intelligent AI computing power system working status control system, which obtains historical operation data of a device and analyzes the status of the device. It can directly analyze the status value Ztz of the device at different time points, and further analyze to obtain the computing power that the device can provide at different time points, and analyze the influence of the loss of the device on the computing power. It can effectively avoid the situation where the device provides insufficient computing power due to the loss of the device, causing the device to run at high power for a long time, resulting in a short service life of the device. Moreover, after calculation, the allocation of subsequent equipment tasks is more reasonable, the use effect is good, and it has a good prospect for use.

[0070] After calculating the total computing power Zsl, it is necessary to allocate the total computing power Zsl. For better allocation, it is necessary to understand the working status of the equipment, and understanding the working status of the equipment is based on the equipment prediction module.

[0071] The device prediction module obtains the historical operation data of the device, analyzes the state value Ztz of the device at different time points, and calculates the computing power Slx that the device can provide at different times based on the state value Ztz;

[0072] The steps to obtain the historical operation data of the device and analyze the status values ​​of the device at different time points are as follows:

[0073] S1 obtains the initial working data of the equipment, the current working data and the working time of the equipment;

[0074] The initial working data and current working data of the device are the maximum computing power of the tested device;

[0075] S2 calculates the loss value of the device and determines the loss stage of the device according to the attenuation value;

[0076] The calculation formula for calculating the equipment loss value is as follows:

[0077]

[0078] In the formula, Shsz is the equipment loss value, Sl c is the initial maximum computing power of the device, Sl x The maximum computing power of the device at present;

[0079] The loss stage of the equipment includes the stable stage and the rapid loss stage. The interval corresponding to the stable stage is (0, D), and the interval corresponding to the rapid loss stage is (D, 100). The value range of D is (30, 40).

[0080] S3 calculates the state value Ztz of the device at different time points according to the wear stage of the device.

[0081] When the calculated Shsz is in the stable stage, the formula for calculating the state value Ztz of the equipment at different time points is as follows:

[0082]

[0083] In the formula, sysc is the working time of the device, Sjd is the time point, and Dqsj is the current time;

[0084] When the calculated Shsz is in the rapid loss stage, mark and replace the new device, because the rapid loss stage provides less computing power and is prone to failure, so it needs to be replaced with a new device at this stage, or it can be replaced with a state value Ztz in equipment.

[0085] The formula for calculating the computing power Slx that the device can provide at different times based on the state value Ztz is as follows:

[0086] Slx=Sl c ×Ztz×Aqyz

[0087] Where Aqyz is the ratio of the computing power when the device is initially working normally to the initial maximum computing power of the device.

[0088] Equipment planning module: obtains the total computing power Zsl required for the operation of the intelligent AI computing system at a future time point and the computing power Slx that the equipment can provide at different times, analyzes the list of equipment required at a future time point, and generates an adjustment plan based on the analysis results.

[0089] The steps to analyze the list of equipment required at a future point in time are as follows:

[0090] Obtain the total computing power Zsl required for the operation of the intelligent AI computing system at the closest time point in the future and the computing power Slx that the device can provide at different times;

[0091] Calculate the total computing power Slh of the devices currently in working state, and then determine the size of the total computing power Slh and the total computing power Zsl required for the operation of the intelligent AI computing system;

[0092]

[0093] In the formula, n is the computing power of the device currently in working state, Slx i is the computing power of the i-th device in working state;

[0094] If Slh<Zsl, increase the number of devices and recalculate the total computing power until Slh n ≥Zsl, and obtain the added device names, and summarize the added device names and the names of the devices currently in working state to generate a device list;

[0095] If Slh>Zs, reduce the device with the minimum computing power and recalculate the total computing power until Slh n <Zsl, at this time, obtain the equipment that has been reduced n-1 times, and remove the reduced equipment names from the equipment currently in working state, and summarize and generate an equipment list.

[0096] When analyzing the results and generating an adjustment plan, obtain the device list at all time points, and then mark the devices that are added or reduced in the device list.

[0097] The present invention provides a working status control system for an intelligent AI computing power system, which analyzes and processes the data of the intelligent AI computing power system during its operation in the previous year, and then makes a prediction to predict the total computing power Zsl required for the operation of the intelligent AI computing power system at a future time point in this year, and then analyzes and determines the number of required devices based on the total computing power Zsl. This can help relevant personnel understand the number of devices required at different time points in the future, and can reasonably allocate devices to avoid the situation where too many devices are idle or there is insufficient spare equipment. The overall use effect is good and the use prospect is good.

[0098] The above formula is a formula obtained by collecting a large amount of data and performing software simulation to obtain the most recent real situation. The preset parameters in the formula are set by technicians in this field according to actual conditions.

[0099] The above embodiments may be implemented in whole or in part by software, hardware, firmware or any other combination thereof. When implemented using software, the above embodiments may be implemented in whole or in part in the form of a computer program product. A person of ordinary skill in the art may appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein may be implemented in electronic hardware, or in a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution.

[0100] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, and may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0101] The above description is only a specific implementation manner of the present application, but the protection scope of the present application is not limited thereto. Any technician familiar with the technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application.

Claims

1. An intelligent AI computing system working state control system, characterized in that: include: Data collection module: used to collect the data of the intelligent AI computing system during its operation in the previous year and classify the collected data; Computing power analysis module: used to process and analyze the classified data and calculate the total computing power Zsl required for the operation of the intelligent AI computing power system at a future time point; Equipment prediction module: obtains the historical operation data of the equipment, analyzes the state value Ztz of the equipment at different time points, and calculates the computing power Slx that the equipment can provide at different times based on the state value Ztz; Equipment planning module: obtains the total computing power Zsl required for the operation of the intelligent AI computing system at a future time point and the computing power Slx that the equipment can provide at different times, analyzes the list of equipment required at a future time point, and generates an adjustment plan based on the analysis results.

2. The intelligent AI computing system working state control system according to claim 1, characterized in that: The data collected by the data acquisition module includes the number of model parameters of the intelligent AI computing system, the time set for model operation and the amount of data that the system needs to process, and the collected data is classified in chronological order.

3. The intelligent AI computing system working state control system according to claim 2, characterized in that: The calculation formula for the total computing power Zsl required for the operation of the intelligent AI computing system at a future time point is as follows: Where Mxcs is the number of model parameters, A is the ratio of the number of model parameters to the complexity of model operation, Sjl is the amount of data that the system needs to process, and B is the annual growth rate of data.

4. The intelligent AI computing system working state control system according to claim 3, characterized in that: The steps to obtain the historical operation data of the device and analyze the status values ​​of the device at different time points are as follows: S1 obtains the initial working data of the equipment, the current working data and the working time of the equipment; S2 calculates the loss value of the device and determines the loss stage of the device according to the attenuation value; S3 calculates the state value Ztz of the device at different time points according to the wear stage of the device.

5. The intelligent AI computing system working state control system according to claim 4, characterized in that: The initial working data and current working data of the device in step S1 are the maximum computing power of the tested device.

6. The intelligent AI computing system working state control system according to claim 5, characterized in that: The calculation formula for calculating the equipment loss value in step S2 is as follows: In the formula, Shsz is the equipment loss value, Sl c is the initial maximum computing power of the device, Sl x This is the current maximum computing power of the device.

7. The intelligent AI computing system working state control system according to claim 6, characterized in that: The loss stage of the device in step S2 includes a stable stage and a rapid loss stage. The interval corresponding to the stable stage is (0, D), the interval corresponding to the rapid loss stage is (D, 100), and the value range of D is (30, 40); When the calculated Shsz is in the stable stage, the formula for calculating the state value Ztz of the equipment at different time points is as follows: In the formula, sysc is the working time of the device, Sjd is the time point, and Dqsj is the current time; When the calculated Shsz is in the rapid loss stage, mark and replace with new equipment.

8. The intelligent AI computing system working state control system according to claim 7, characterized in that: The formula for calculating the computing power Slx that the device can provide at different times based on the state value Ztz is as follows: Slx=Sl c ×Ztz×Aqyz Where Aqyz is the ratio of the computing power when the device is initially working normally to the initial maximum computing power of the device.

9. The intelligent AI computing system working state control system according to claim 8, characterized in that: The steps to analyze the list of equipment required at a future point in time are as follows: Obtain the total computing power Zsl required for the operation of the intelligent AI computing system at the closest time point in the future and the computing power Slx that the device can provide at different times; Calculate the total computing power Slh of the devices currently in working state, and then determine the size of the total computing power Slh and the total computing power Zsl required for the operation of the intelligent AI computing system; In the formula, n is the computing power of the device currently in working state, Slx i is the computing power of the i-th device in working state; If Slh<Zsl, increase the number of devices and recalculate the total computing power until Slh n ≥Zsl, and obtain the added device names, and summarize the added device names and the names of the devices currently in working state to generate a device list; If Slh>Zs, reduce the device with the minimum computing power and recalculate the total computing power until Slh n <Zsl, at this time, obtain the equipment that has been reduced n-1 times, and remove the reduced equipment names from the equipment currently in working state, and summarize and generate an equipment list.

10. The intelligent AI computing system working state control system according to claim 9, characterized in that: When analyzing the results and generating an adjustment plan, obtain the device list at all time points, and then mark the devices that are added or reduced in the device list.

Images