Equipment power consumption control method, product, equipment and medium

Abstract

The invention discloses an equipment power consumption control method, a product, equipment and a medium, and relates to the technical field of power consumption control, and the method comprises the steps: judging whether a task operated on current target equipment is a periodic task, and if the task operated on the current target equipment is the periodic task, determining a corresponding operation time period; predicting the calculated amount of a processor in the target equipment in the running time period to obtain a corresponding target calculated amount predicted value; determining a target processor frequency required for the processor to complete an actual calculation amount consistent with the target calculation amount predicted value based on the minimum power consumption; and performing operation control on the target equipment based on the target processor frequency, so that the target equipment adopts the target processor frequency as the actual working frequency of the processor in the next operation time period, and power consumption control on the target equipment is realized. According to the scheme, services can be normally executed, and the power consumption of the server can be reasonably controlled.

Description

Technical Field

[0001] The present invention relates to the technical field of power consumption control, and in particular to a method, product, device and medium for controlling power consumption of a device. Background Art

[0002] Nowadays, the scale of newly built data centers is getting larger and larger, and power consumption is also increasing. Therefore, the management of server power consumption in data center management has become increasingly important. For example, when a large number of servers are overloaded, the server power consumption may be too high, approaching or even exceeding the rated power consumption of the data center design, increasing the power burden of the data center server. For example, the business demand for some servers in the data center is not high, and it is necessary to reduce the server power consumption to reduce power consumption, save data center operating costs, and reduce data center carbon emissions. After the server runs a single cycle according to the current processor power consumption, the server power consumption can be obtained, and the processor power consumption is proportional to the processor frequency. Therefore, controlling the processor frequency can effectively control the server power consumption, and then control the server power consumption.

[0003] Currently, the optimal processor frequency curve is fitted according to the external temperature of the server, and the processor frequency is adaptively adjusted according to the curve to solve the problem of processor heating. This cannot take into account performance and cannot guarantee the normal execution of business. In addition, frequency adjustment based on temperature is not completely suitable for power consumption control. In other words, the current optimal frequency can reduce the processor temperature, but the server power consumption is not necessarily the minimum power consumption.

[0004] It can be seen that how to not only execute services normally but also reasonably control the power consumption of the server is a problem that technical personnel in this field need to solve. Summary of the invention

[0005] The purpose of the embodiments of the present invention is to provide a device power consumption control method, product, device and medium, which can not only perform services normally, but also reasonably control server power consumption. The specific scheme is as follows:

[0006] In a first aspect, the present invention discloses a method for controlling power consumption of a device, comprising:

[0007] Determine whether the task currently running on the target device is a periodic task, and if the task currently running on the target device is a periodic task, determine the corresponding running time period;

[0008] Predicting the amount of computation of the processor in the target device during the running time period to obtain a corresponding target amount of computation prediction value;

[0009] Determining a target processor frequency required for the processor to complete an actual amount of computation consistent with a predicted value of a target amount of computation based on minimum power consumption;

[0010] The target device is controlled to operate based on the target processor frequency, so that the target device uses the target processor frequency as the actual operating frequency of the processor in the next operating time cycle, thereby achieving power consumption control of the target device.

[0011] Optionally, determining the corresponding operating time period includes:

[0012] Collecting the utilization rate of each target processor of the target device at different times within a preset time period, and determining the utilization rate of each target processor as original time series data;

[0013] The original time series data is periodically analyzed to determine the corresponding operating time period.

[0014] Optionally, the periodic analysis of the original time series data to determine the corresponding operating time period includes:

[0015] Performing Fourier transformation on the original time series data to obtain a frequency spectrum, and determining a corresponding operating time period based on a main frequency component of the frequency spectrum;

[0016] Optionally, the periodic analysis of the original time series data to determine the corresponding operating time period includes:

[0017] The correlation between each time series data in the original time series data is analyzed, and the corresponding operation time period is determined according to the correlation.

[0018] Optionally, determining a target processor frequency required by the processor to complete an actual amount of computation consistent with a predicted value of a target amount of computation based on minimum power consumption includes:

[0019] Obtain a pre-built initial processor energy consumption model with processor frequency as an independent variable and processor power consumption as a dependent variable;

[0020] Inputting the target computing amount prediction value into the initial processor energy consumption model as the processor computing amount in the initial processor energy consumption model to obtain a target processor energy consumption model;

[0021] The target processor energy consumption model is derived, and based on the derived model, the target processor frequency required for the processor to complete the actual computing amount consistent with the target computing amount prediction value based on the minimum power consumption is determined.

[0022] Optionally, the initial processor energy consumption model includes a first component and a second component, wherein the first component is determined based on the product of the processor computing amount and the processor frequency, and the second component is determined based on the ratio of the processor computing amount to the processor frequency.

[0023] Optionally, the first component and the second component in the initial processor energy consumption model are respectively configured with corresponding constant coefficients, and the constant coefficients are coefficients determined based on the pre-acquired reference processor frequency and the corresponding power consumption of each reference processor.

[0024] Optionally, obtaining a reference processor frequency and corresponding power consumption of each reference processor includes:

[0025] Acquire each benchmark processor frequency, and control the target device to run smoothly according to each benchmark processor frequency, so as to collect the benchmark processor power consumption of the target device at different benchmark processor frequencies;

[0026] Optionally, determining the constant coefficient based on the reference processor frequency and the corresponding power consumption of each reference processor includes:

[0027] Performing power function fitting on each of the reference processor frequencies and the reference processor power consumption to obtain a target power consumption frequency relationship function;

[0028] Based on the coefficients in the target power consumption frequency relationship function, constant coefficients corresponding to the first component and the second component in the initial processor energy consumption model are determined.

[0029] Optionally, controlling the target device to run smoothly according to each of the reference processor frequencies to collect reference processor power consumption of the target device at different reference processor frequencies includes:

[0030] Controlling the target device to run smoothly according to the current reference processor frequency to collect the reference processor power consumption of the target device at the current reference processor frequency;

[0031] The sum of the preset frequency increase and the current reference processor frequency is determined as the next reference processor frequency, and the next reference processor frequency is updated to the current reference processor frequency, and the process is then jumped again to the step of controlling the target device to run smoothly according to the current reference processor frequency.

[0032] Optionally, the performing operation control on the target device based on the target processor frequency so that the target device adopts the target processor frequency as the actual operating frequency of the processor in the next operation time period to achieve power consumption control of the target device includes:

[0033] Determine the target processor power consumption corresponding to the target device when the target device is running in a single running time period based on the current processor power consumption of the target device, and determine whether the target processor power consumption is greater than a preset threshold;

[0034] If the power consumption of the target processor is greater than the preset threshold, the target device is controlled to operate based on the target processor frequency so that the target device uses the target processor frequency as the actual operating frequency of the processor in the next operating time cycle to achieve power consumption control of the target device.

[0035] Optionally, after determining whether the power consumption of the target processor is greater than a preset threshold, the method further includes:

[0036] If the power consumption of the target processor is not greater than the preset threshold, the priority of the current task being processed by the target device is determined; if the priority of the current task is lower than the preset priority, the operation of the target device is controlled based on the target processor frequency, so that the target device uses the target processor frequency as the actual operating frequency of the processor in the next operating time cycle to achieve power consumption control of the target device.

[0037] In a second aspect, the present invention discloses a computer program product, including a computer program / instruction, which implements the steps of the aforementioned device power consumption control method when executed by a processor.

[0038] In a third aspect, the present invention discloses an electronic device, comprising:

[0039] Memory, used to store computer programs;

[0040] The processor is used to execute the computer program to implement the steps of the device power consumption control method disclosed above.

[0041] In a fourth aspect, the present invention discloses a computer-readable storage medium for storing a computer program; wherein, when the computer program is executed by a processor, the steps of the aforementioned device power consumption control method are implemented.

[0042] It can be seen that the present invention determines whether the task running on the current target device is a periodic task. If the task running on the current target device is a periodic task, the corresponding running time period is determined; the computing amount of the processor in the target device within the running time period is predicted to obtain the corresponding target computing amount prediction value; the target processor frequency required for the processor to complete the actual computing amount consistent with the target computing amount prediction value based on the minimum power consumption is determined; and the operation of the target device is controlled based on the target processor frequency, so that the target device adopts the target processor frequency as the actual operating frequency of the processor in the next running time period to achieve power consumption control of the target device.

[0043] Beneficial effect: When the task running on the current target device of the present application is a periodic task, the corresponding running time cycle is determined, the computing amount of the processor in the target device within the running time cycle is predicted to obtain the corresponding target computing amount prediction value, and the target processor frequency required for the processor to complete the actual computing amount consistent with the target computing amount prediction value based on the minimum power consumption is determined. That is to say, the determined target processor frequency can ensure that the power consumption of the processor is minimized when completing the actual workload matching the target computing amount prediction value. In this way, the target device is controlled to run according to the target processor frequency, and the target device uses the target processor frequency as the actual operating frequency of the processor in the next running time cycle. Then, the power consumption of the target device within the running time cycle can be controlled, and there is no need to reduce the processor computing amount of the target device within the running time cycle. The processor computing amount can reflect the business execution status. That is to say, keeping the processor computing amount stable and unchanged means that the target device can execute the business normally. That is, the present application can not only execute the business normally, but also reasonably control the power consumption of the device. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] In order to more clearly illustrate the embodiments of the present invention, the following briefly introduces the drawings required for use in the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0045] Figure 1 A flow chart of a method for controlling device power consumption provided by an embodiment of the present invention;

[0046] Figure 2 A schematic diagram of a specific power consumption control structure layout provided by an embodiment of the present invention;

[0047] Figure 3 A specific schematic diagram of determining an optimal frequency provided by an embodiment of the present invention;

[0048] Figure 4 A flow chart of a specific device power consumption control method provided by an embodiment of the present invention;

[0049] Figure 5 A schematic diagram of the structure of a device power consumption control device provided by an embodiment of the present invention;

[0050] Figure 6 A structural diagram of an electronic device provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0051] 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.

[0052] Nowadays, the scale of newly built data centers is getting larger and larger, and power consumption is also increasing. Therefore, the management of server power consumption in data center management has become increasingly important. For example, when a large number of servers are overloaded, the server power consumption may be too high, approaching or even exceeding the rated power consumption of the data center design, increasing the power burden of the data center server. For example, the business demand for some servers in the data center is not high, and it is necessary to reduce the server power consumption to reduce power consumption, save data center operating costs, and reduce data center carbon emissions. After the server runs a single cycle according to the current processor power consumption, the server power consumption can be obtained, and the processor power consumption is proportional to the processor frequency. Therefore, controlling the processor frequency can effectively control the server power consumption, and then control the server power consumption.

[0053] Currently, the optimal processor frequency curve is fitted according to the external temperature of the server, and the processor frequency is adaptively adjusted according to the curve to solve the problem of processor heating. This cannot take into account performance and cannot guarantee the normal execution of business. In addition, frequency adjustment based on temperature is not completely suitable for power consumption control. In other words, the current optimal frequency can reduce the processor temperature, but the server power consumption is not necessarily the minimum power consumption.

[0054] The terms "including" and "having" in the specification of the present invention and the above-mentioned drawings, as well as any variations thereof, are intended to cover non-exclusive inclusions. For example, a process, method, system, product or device including a series of steps or units is not limited to the listed steps or units, but may include steps or units that are not listed.

[0055] In order to enable those skilled in the art to better understand the solution of the present invention, the present invention is further described in detail below in conjunction with the accompanying drawings and specific implementation methods.

[0056] Next, a device power consumption control solution provided by an embodiment of the present invention is introduced in detail. Figure 1 A method for controlling power consumption of a device provided in an embodiment of the present invention includes:

[0057] Step S11: determining whether the task currently running on the target device is a periodic task; if the task currently running on the target device is a periodic task, determining a corresponding running time period.

[0058] For example Figure 2 A specific power consumption control structure layout diagram is shown. To realize the power consumption control of the device, a collection device, a computing device, and a control device are required. The collection device collects relevant indicators of the target device CPU (Central Processing Unit, i.e., central processing unit) through the in-band network, including CPU utilization, CPU power consumption and other indicators, and filters the collected indicators to filter out abnormal indicators; the computing device is used to calculate the optimal CPU frequency for the next cycle according to the filtered indicators; the control device is used to send instructions to the CPU to limit the CPU frequency, etc., wherein the target device can specifically be a server.

[0059] Determine whether the task running on the current target device is a periodic task. If so, the CPU utilization can be used to reflect the periodic business changes of the device, and the corresponding running time period can be determined by analyzing the main frequency component or the autocorrelation function.

[0060] S12: predicting the amount of computation of the processor in the target device during the running time period to obtain a corresponding target computation amount prediction value.

[0061] CPU utilization refers to the proportion of the CPU's busy time to the total time, and CPU frequency refers to the number of calculations that the CPU can perform per second, usually in Hertz (Hz). Therefore, if you want to predict the CPU's computing volume in a certain period of time, you can multiply the three factors of CPU utilization, CPU frequency and time to form a simplified calculation model to predict the CPU computing volume. The specific model formula is as follows:

[0062] N = f * T * U;

[0063] Wherein, N represents the amount of processor computation, which in this embodiment is the predicted value of the target amount of computation, f represents the CPU frequency, T represents the operating time period, and U represents the average CPU utilization rate during the operating time period.

[0064] Step S13: determining a target processor frequency required for the processor to complete an actual amount of computation consistent with a predicted value of the target amount of computation based on minimum power consumption.

[0065] In this embodiment, determining the target processor frequency required by the processor when completing an actual amount of calculation consistent with a predicted target amount of calculation based on minimum power consumption includes: obtaining a pre-constructed initial processor energy consumption model with processor frequency as an independent variable and processor power consumption as a dependent variable; inputting the predicted target amount of calculation into the initial processor energy consumption model as the processor calculation amount in the initial processor energy consumption model to obtain a target processor energy consumption model; and deriving the target processor energy consumption model, and determining the target processor frequency required by the processor when completing an actual amount of calculation consistent with the predicted target amount of calculation based on minimum power consumption based on the derived model.

[0066] In this embodiment, the initial processor energy consumption model includes a first component and a second component, wherein the first component is determined based on the product of the processor computing amount and the processor frequency, and the second component is determined based on the ratio of the processor computing amount to the processor frequency.

[0067] In this embodiment, the first component and the second component in the initial processor energy consumption model are respectively configured with corresponding constant coefficients, and the constant coefficients are coefficients determined based on the pre-acquired reference processor frequency and the corresponding power consumption of each reference processor.

[0068] For example Figure 3 A specific optimal frequency determination diagram is shown. Determining the optimal frequency is mainly divided into calculating the change cycle, that is, determining the running time cycle, followed by obtaining the power consumption model, that is, obtaining the target processor energy consumption model, and finally calculating the optimal CPU frequency, that is, determining the target processor frequency required for the processor to complete the actual computing amount consistent with the target computing amount prediction value based on the minimum power consumption.

[0069] Furthermore, in the process of obtaining the target processor energy consumption model, a pre-built initial processor energy consumption model is first obtained, the initial processor energy consumption model uses the processor frequency as an independent variable and the processor power consumption as a dependent variable, and the initial processor energy consumption model includes a first component and a second component, wherein the first component is determined based on the product of the processor calculation amount and the processor frequency, and the second component is determined based on the ratio between the processor calculation amount and the processor frequency, and the first component and the second component are respectively configured with corresponding constant coefficients, which are coefficients determined based on the pre-acquired benchmark processor frequency and the corresponding power consumption of each benchmark processor. The initial processor energy consumption model is specifically as follows:

[0070] W = aNf + bN / f;

[0071] Where W represents the power consumption of the processor, a and b represent the constant coefficients of the first component and the second component respectively, N represents the amount of processor calculation, f represents the processor frequency, aNf is the first component, and bN / f is the second component;

[0072] Secondly, the target computing amount prediction value is used as the processor computing amount in the initial processor energy consumption model and inputted into the initial processor energy consumption model to obtain the target processor energy consumption model. That is to say, at this time, the constant coefficient and the processor computing amount are both known quantities, and then the processor power consumption W changes with the processor frequency f;

[0073] It is understandable that by deriving the target processor energy consumption model, the target processor frequency required for the processor to complete the actual computing amount consistent with the target computing amount prediction value based on the minimum power consumption can be determined based on the derived model.

[0074] Furthermore, because only the processor power consumption W and the processor frequency f are variables in the target processor energy consumption model, an energy consumption curve graph of the target processor energy consumption model is drawn, with the processor frequency f as the horizontal axis and the processor power consumption W as the vertical axis. Therefore, the minimum power consumption and the target processor frequency corresponding to the minimum power consumption can be intuitively seen, thereby improving the efficiency of determining the target processor frequency and facilitating user understanding. Among them, when the processor frequency f is the horizontal axis, the smaller the step size of the horizontal axis, the higher the accuracy of the target processor frequency determined subsequently.

[0075] In this embodiment, obtaining the reference processor frequency and the corresponding power consumption of each reference processor includes: obtaining each reference processor frequency, and controlling the target device to run smoothly according to each reference processor frequency, so as to collect the reference processor power consumption of the target device at different reference processor frequencies. Obtaining each reference processor frequency, performing different frequency restrictions on the processor through the operating system, and waiting until the device runs smoothly, collecting the current device power consumption data, that is, the reference processor power consumption corresponding to the reference processor frequency.

[0076] In this embodiment, the constant coefficient is determined based on the reference processor frequency and the corresponding power consumption of each reference processor, including: fitting the reference processor frequency and the reference processor power consumption with a power function to obtain a target power consumption frequency relationship function; determining the constant coefficients corresponding to each of the first component and the second component in the initial processor energy consumption model based on each coefficient in the target power consumption frequency relationship function. Because the CPU power consumption is exponentially related to the frequency, after collecting each reference processor frequency and the reference processor power consumption corresponding to each reference processor frequency, a power function is fitted to each reference processor frequency and the reference processor power consumption to obtain a target power consumption frequency relationship function. The target power consumption frequency relationship function is as follows:

[0077] P=af 2 +b;

[0078] In the formula, P represents processor power consumption, f represents processor frequency, a and b are coefficients, and a is the constant coefficient corresponding to the first component in the initial processor energy consumption model, and b is the constant coefficient corresponding to the second component in the initial processor energy consumption model. In this way, after fitting the target power consumption frequency relationship function, the constant coefficients corresponding to the first component and the second component in the initial processor energy consumption model can be determined.

[0079] In this embodiment, the step of controlling the target device to run smoothly according to each of the reference processor frequencies to collect the reference processor power consumption of the target device at different reference processor frequencies includes: controlling the target device to run smoothly according to the current reference processor frequency to collect the reference processor power consumption of the target device at the current reference processor frequency; determining the sum of the preset frequency increase and the current reference processor frequency as the next reference processor frequency, and updating the next reference processor frequency to the current reference processor frequency, and jumping back to the step of controlling the target device to run smoothly according to the current reference processor frequency. For example, the preset frequency increase is set to the current reference processor frequency f n 10%, then the next benchmark processor frequency f n+1 That is f n+1 =f n +10%*f n In this way, multiple benchmark processor frequencies and the benchmark processor power consumption of the target device at different benchmark processor frequencies can be collected.

[0080] It should be noted that the processor power consumption and processor utilization can be considered to be linearly related, so the model for estimating processor power consumption can be simplified to multiply the three factors of CPU full load power consumption, running time and CPU utilization, as shown below:

[0081] W = P * T * U;

[0082] In the formula, W represents the power consumption of the processor, P represents the power consumption of the processor, T represents the running time cycle, and U represents the processor utilization;

[0083] Therefore, the initial processor energy consumption model can be obtained by combining the target power consumption frequency relationship function, the CPU computing amount prediction model, and the model for estimating the processor power consumption.

[0084] Step S14: Controlling the operation of the target device based on the target processor frequency, so that the target device uses the target processor frequency as the actual operating frequency of the processor in the next operation time period, thereby achieving power consumption control of the target device.

[0085] In this embodiment, the operation control of the target device based on the target processor frequency so that the target device adopts the target processor frequency as the actual operating frequency of the processor in the next operation time cycle to achieve power consumption control of the target device includes: determining the target processor power consumption corresponding to the operation of the target device in a single operation time cycle based on the current processor power consumption of the target device, and judging whether the target processor power consumption is greater than a preset threshold; if the target processor power consumption is greater than the preset threshold, the operation control of the target device based on the target processor frequency so that the target device adopts the target processor frequency as the actual operating frequency of the processor in the next operation time cycle to achieve power consumption control of the target device. Based on the current processor power consumption of the target device, the target processor power consumption corresponding to the target device running in a single operating time cycle can be determined. The maintenance personnel determine the preset threshold value according to the application scenario. If the target processor power consumption is greater than the preset threshold value, it means that the current power consumption is large and power consumption control is required. In this case, the current processor frequency of the target device can be directly updated to the target processor frequency. After the frequency is updated, the corresponding power consumption and power consumption are significantly reduced. If the target processor power consumption is not greater than the preset threshold value, it means that the current power consumption is small and power consumption control is not required. In other words, even if the target processor power consumption is not the minimum power consumption, it does not need to be controlled.

[0086] In this embodiment, after determining whether the target processor power consumption is greater than the preset threshold, it also includes: if the target processor power consumption is not greater than the preset threshold, determining the priority of the current task being processed by the target device, and if the priority of the current task is lower than the preset priority, then controlling the operation of the target device based on the target processor frequency, so that the target device uses the target processor frequency as the actual operating frequency of the processor in the next operating time cycle to achieve power consumption control of the target device. It can be understood that although the target processor power consumption is not greater than the preset threshold, the current processor frequency is not the target processor frequency, then it can be determined whether to adjust the processor frequency of the target device according to the priority of the current task being processed by the target device, that is, whether it is necessary to perform power consumption control at present. If the priority of the current task is lower than the preset priority, that is, the current task is not important, then the power consumption can be reduced, but if the current task priority is higher, it is necessary to ensure that the current task is stably processed, so the power consumption can be reduced. In this way, power consumption control can be performed more reasonably.

[0087] It can be seen that the present invention determines whether the task running on the current target device is a periodic task. If the task running on the current target device is a periodic task, the corresponding running time period is determined; the computing amount of the processor in the target device within the running time period is predicted to obtain the corresponding target computing amount prediction value; the target processor frequency required for the processor to complete the actual computing amount consistent with the target computing amount prediction value based on the minimum power consumption is determined; and the operation of the target device is controlled based on the target processor frequency, so that the target device adopts the target processor frequency as the actual operating frequency of the processor in the next running time period to achieve power consumption control of the target device.

[0088] The beneficial effect is as follows: when the task running on the current target device of the present application is a periodic task, the corresponding running time cycle is determined, the computing amount of the processor in the target device within the running time cycle is predicted to obtain the corresponding target computing amount prediction value, and the target processor frequency required for the processor to complete the actual computing amount consistent with the target computing amount prediction value based on the minimum power consumption is determined. That is to say, the determined target processor frequency can ensure that the power consumption of the processor is minimized when completing the actual workload matching the target computing amount prediction value. In this way, the target device is controlled to run according to the target processor frequency, and the target device uses the target processor frequency as the actual operating frequency of the processor in the next running time cycle. Then, the power consumption control of the target device within the running time cycle can be achieved, and there is no need to reduce the processor computing amount of the target device within the running time cycle. The processor computing amount can reflect the business execution status. That is to say, keeping the processor computing amount stable and unchanged means that the target device can execute the business normally, that is, the present application can not only execute the business normally, but also reasonably control the power consumption of the device.

[0089] See also Figure 4 The embodiment of the present invention discloses a specific method for controlling power consumption of a device. Compared with the previous embodiment, this embodiment further illustrates and optimizes the technical solution. It includes:

[0090] Step S21: Determine whether the task currently running on the target device is a periodic task. If the task currently running on the target device is a periodic task, collect the utilization of each target processor of the target device at different times within a preset time period, and determine the utilization of each target processor as the original time series data.

[0091] Collect the raw processor utilization of the target device at different times within a preset time period, and filter the utilization of each target processor to determine the utilization of each target processor obtained after filtering as the original timing data, so as to subsequently determine the operating time period based on the original timing data; wherein, the filtered data can be invalid data, that is, during the data transmission process, it may contain some data that is useless for periodic analysis or processing. The filtering operation can remove these invalid data and improve the efficiency of data processing; the filtered data can also be erroneous data, that is, during the data collection process, erroneous data may be generated due to various reasons, such as sensor failure, data transmission error, etc. The filtering operation can identify and remove these erroneous data to ensure the accuracy of the data.

[0092] Step S22: performing periodic analysis on the original time series data to determine the corresponding operating time period.

[0093] In a specific cycle determination embodiment, the periodic analysis of the original time series data to determine the corresponding operation time period includes: performing Fourier transform on the original time series data to obtain a frequency spectrum, and determining the corresponding operation time period based on the main frequency component of the frequency spectrum. Performing Fourier transform on the original time series data to convert the time series data from the time domain to the frequency domain to obtain a frequency spectrum, and analyzing the frequency spectrum to determine the main frequency component, and the inverse of the main frequency component can be used as an estimated value of the cycle.

[0094] In another specific cycle determination embodiment, the periodic analysis of the original time series data to determine the corresponding operation time period includes: analyzing the correlation between the time series data in the original time series data, and determining the corresponding operation time period according to the correlation. Using the autocorrelation function method, by calculating the correlation between each point in the time series and other points, the periodic pattern in the data is revealed, and if the data shows a significant correlation at a certain lag time point, the estimated value of its period is recorded.

[0095] Step S23: predicting the computational effort of the processor in the target device during the operating time period to obtain a corresponding target computational effort prediction value.

[0096] In this embodiment, the formula for predicting the amount of calculation of the processor in the running time period is as follows:

[0097] N = f * T * U;

[0098] Among them, N represents the processor computing amount, which is the target computing amount prediction value in this embodiment, f represents the CPU frequency, T represents the running time period, and U represents the average CPU utilization rate during the running time period. That is, the calculation formula of the computing amount is introduced to quantify the CPU computing task into a specific value.

[0099] Step S24: determining a target processor frequency required for the processor to complete an actual amount of computation consistent with a predicted value of the target amount of computation based on minimum power consumption.

[0100] An initial processor energy consumption model is obtained with processor frequency as an independent variable and processor power consumption as a dependent variable, wherein the initial processor energy consumption model also includes processor calculation amount, and the processor calculation amount at this time is an unknown quantity, then the target calculation amount prediction value is substituted into the initial processor energy consumption model to obtain the target processor energy consumption model. It can be understood that the target processor energy consumption model only has processor frequency and processor power consumption as variables. If the target processor frequency corresponding to the minimum power consumption, that is, the optimal processor frequency, is to be determined, then the target processor energy consumption model is derived, and based on the derived model, the target processor frequency required for the processor to complete the actual calculation amount consistent with the target calculation amount prediction value based on the minimum power consumption is determined.

[0101] The target processor energy consumption model in this embodiment can accurately compare the energy consumption required for the processor to complete the same amount of calculation at different frequencies. By replacing the processor frequency and power consumption model and combining it with the calculation formula of the calculation amount, it is possible to deduce the processor frequency with the lowest power consumption while ensuring that the total calculation amount of periodic tasks remains unchanged. The determination of this frequency not only helps to reduce the energy consumption of the processor, but also reduces heat and extends the service life of the equipment.

[0102] Step S25: Controlling the operation of the target device based on the target processor frequency, so that the target device uses the target processor frequency as the actual operating frequency of the processor in the next operation time period, thereby achieving power consumption control of the target device.

[0103] It can be seen that the present invention calculates the processor fluctuation cycle by collecting the processor utilization rate, and ensures that the total amount of processor calculation remains unchanged within the same operating time cycle, that is, it does not affect the normal execution of the business. By adjusting the processor frequency, the purpose of reducing the power consumption of the processor and reducing the total power consumption of the server equipment is achieved. This is very important for the operation of modern data centers and server farms, because in these environments, a large number of servers and devices need to run continuously and consume a lot of electricity. By reducing the power consumption of the CPU, it can not only save electricity costs, but also reduce environmental impact and contribute to sustainable development. The present invention is to ensure that the total amount of processor calculation remains unchanged within a cycle, thereby ensuring that the stability and efficiency of business operation are not affected in any way. At the same time, the present invention can select the optimal processor frequency without sacrificing performance by accurately controlling and optimizing the processor frequency. This intelligent frequency adjustment strategy can effectively reduce the power consumption of the processor, thereby reducing the energy consumption of the overall system.

[0104] Figure 5 A schematic diagram of the structure of a device power consumption control device provided by an embodiment of the present invention includes:

[0105] The period determination module 11 is used to determine whether the task currently running on the target device is a periodic task, and if the task currently running on the target device is a periodic task, determine the corresponding running time period;

[0106] A computing amount prediction module 12 is used to predict the computing amount of the processor in the target device during the running time period to obtain a corresponding target computing amount prediction value;

[0107] A frequency determination module 13, used to determine a target processor frequency required for the processor to complete an actual computation amount consistent with a predicted target computation amount based on minimum power consumption;

[0108] The power consumption control module 14 is used to control the operation of the target device based on the target processor frequency so that the target device uses the target processor frequency as the actual operating frequency of the processor in the next operation time period to achieve power consumption control of the target device.

[0109] It can be seen that the present invention determines whether the task running on the current target device is a periodic task. If the task running on the current target device is a periodic task, the corresponding running time period is determined; the computing amount of the processor in the target device within the running time period is predicted to obtain the corresponding target computing amount prediction value; the target processor frequency required for the processor to complete the actual computing amount consistent with the target computing amount prediction value based on the minimum power consumption is determined; and the operation of the target device is controlled based on the target processor frequency, so that the target device adopts the target processor frequency as the actual operating frequency of the processor in the next running time period to achieve power consumption control of the target device.

[0110] The beneficial effect is as follows: when the task running on the current target device of the present application is a periodic task, the corresponding running time cycle is determined, the computing amount of the processor in the target device within the running time cycle is predicted to obtain the corresponding target computing amount prediction value, and the target processor frequency required for the processor to complete the actual computing amount consistent with the target computing amount prediction value based on the minimum power consumption is determined. That is to say, the determined target processor frequency can ensure that the power consumption of the processor is minimized when completing the actual workload matching the target computing amount prediction value. In this way, the target device is controlled to run according to the target processor frequency, and the target device uses the target processor frequency as the actual operating frequency of the processor in the next running time cycle. Then, the power consumption control of the target device within the running time cycle can be achieved, and there is no need to reduce the processor computing amount of the target device within the running time cycle. The processor computing amount can reflect the business execution status. That is to say, keeping the processor computing amount stable and unchanged means that the target device can execute the business normally, that is, the present application can not only execute the business normally, but also reasonably control the power consumption of the device.

[0111] In some specific embodiments, the period determination module 11 includes:

[0112] A time series data acquisition unit, used to collect the utilization rate of each target processor of the target device at different times within a preset time period, and determine the utilization rate of each target processor as original time series data;

[0113] The periodicity analysis submodule is used to perform periodicity analysis on the original time series data to determine the corresponding operating time period.

[0114] In some specific embodiments, the periodicity analysis submodule includes:

[0115] A first period determination unit, configured to perform Fourier transform on the original time series data to obtain a frequency spectrum, and determine a corresponding operating time period based on a main frequency component of the frequency spectrum;

[0116] In some specific embodiments, the periodicity analysis submodule includes:

[0117] The second period determination unit is used to analyze the correlation between each time series data in the original time series data, and determine the corresponding operation time period according to the correlation.

[0118] In some specific embodiments, the frequency determination module 13 includes:

[0119] An initial model acquisition unit, used to acquire a pre-built initial processor energy consumption model with processor frequency as an independent variable and processor power consumption as a dependent variable;

[0120] a target model acquisition unit, configured to input the target computation amount prediction value into the initial processor energy consumption model as the processor computation amount in the initial processor energy consumption model, so as to obtain a target processor energy consumption model;

[0121] The processor frequency determination unit is used to derive the target processor energy consumption model and determine the target processor frequency required for the processor to complete the actual calculation amount consistent with the target calculation amount prediction value based on the minimum power consumption based on the derived model.

[0122] In some specific embodiments, the initial processor energy consumption model includes a first component and a second component, wherein the first component is determined based on the product of the processor computing amount and the processor frequency, and the second component is determined based on the ratio between the processor computing amount and the processor frequency.

[0123] In some specific embodiments, the first component and the second component in the initial processor energy consumption model are respectively configured with corresponding constant coefficients, and the constant coefficients are coefficients determined based on the pre-acquired reference processor frequency and the corresponding power consumption of each reference processor.

[0124] In some specific embodiments, the device power consumption control apparatus includes:

[0125] A benchmark information collection submodule, used to obtain each benchmark processor frequency, and control the target device to run smoothly according to each benchmark processor frequency, so as to collect the benchmark processor power consumption of the target device at different benchmark processor frequencies;

[0126] In some specific embodiments, the device power consumption control apparatus includes:

[0127] A relationship function fitting unit, used for performing power function fitting on each of the reference processor frequencies and the reference processor power consumption to obtain a target power consumption frequency relationship function;

[0128] A constant coefficient determination unit is used to determine the constant coefficients corresponding to the first component and the second component in the initial processor energy consumption model based on the coefficients in the target power consumption frequency relationship function.

[0129] In some specific embodiments, the reference information collection submodule includes:

[0130] A current benchmark information acquisition unit, used to control the target device to run smoothly according to the current benchmark processor frequency, so as to acquire the benchmark processor power consumption of the target device at the current benchmark processor frequency;

[0131] The jump unit is used to determine the sum of the preset frequency increase and the current reference processor frequency as the next reference processor frequency, update the next reference processor frequency to the current reference processor frequency, and jump again to the step of controlling the target device to run smoothly according to the current reference processor frequency.

[0132] In some specific embodiments, the power consumption control module 14 includes:

[0133] A judging unit, configured to determine the target processor power consumption corresponding to the target device when the target device is running in a single running time period based on the current processor power consumption of the target device, and to judge whether the target processor power consumption is greater than a preset threshold;

[0134] A power consumption control unit is used to control the operation of the target device based on the target processor frequency if the power consumption of the target processor is greater than the preset threshold, so that the target device adopts the target processor frequency as the actual operating frequency of the processor in the next operating time cycle to achieve power consumption control of the target device.

[0135] In some specific embodiments, the power consumption control module 14 includes:

[0136] A priority judgment unit is used to determine the priority of the current task being processed by the target device if the power consumption of the target processor is not greater than the preset threshold value; if the priority of the current task is lower than the preset priority, the target device is controlled to operate based on the target processor frequency, so that the target device adopts the target processor frequency as the actual operating frequency of the processor in the next operating time period to achieve power consumption control of the target device.

[0137] Furthermore, the present application also discloses an electronic device. Figure 6 This is a block diagram of an electronic device according to an exemplary embodiment, and the content of the diagram cannot be considered as any limitation on the scope of use of this application. The electronic device may specifically include: at least one processor 21, at least one memory 22, a power supply 23, a communication interface 24, an input / output interface 25, and a communication bus 26. The memory 22 is used to store a computer program, which is loaded and executed by the processor 21 to implement the following steps:

[0138] Determine whether the task currently running on the target device is a periodic task, and if the task currently running on the target device is a periodic task, determine the corresponding running time period;

[0139] Predicting the amount of computation of the processor in the target device during the running time period to obtain a corresponding target amount of computation prediction value;

[0140] Determining a target processor frequency required for the processor to complete an actual amount of computation consistent with a predicted value of a target amount of computation based on minimum power consumption;

[0141] The target device is controlled to operate based on the target processor frequency, so that the target device uses the target processor frequency as the actual operating frequency of the processor in the next operating time cycle, thereby achieving power consumption control of the target device.

[0142] In some specific implementations, the processor can implement the following steps by executing the computer program stored in the memory:

[0143] Collecting the utilization rate of each target processor of the target device at different times within a preset time period, and determining the utilization rate of each target processor as original time series data;

[0144] The original time series data is periodically analyzed to determine the corresponding operating time period.

[0145] In some specific implementations, the processor can implement the following steps by executing the computer program stored in the memory:

[0146] Performing Fourier transformation on the original time series data to obtain a frequency spectrum, and determining a corresponding operating time period based on a main frequency component of the frequency spectrum;

[0147] In some specific implementations, the processor can implement the following steps by executing the computer program stored in the memory:

[0148] The correlation between each time series data in the original time series data is analyzed, and the corresponding operation time period is determined according to the correlation.

[0149] In some specific implementations, the processor can implement the following steps by executing the computer program stored in the memory:

[0150] Obtain a pre-built initial processor energy consumption model with processor frequency as an independent variable and processor power consumption as a dependent variable;

[0151] Inputting the target computing amount prediction value into the initial processor energy consumption model as the processor computing amount in the initial processor energy consumption model to obtain a target processor energy consumption model;

[0152] The target processor energy consumption model is derived, and based on the derived model, the target processor frequency required for the processor to complete the actual computing amount consistent with the target computing amount prediction value based on the minimum power consumption is determined.

[0153] In some specific implementations, the processor can implement the following steps by executing the computer program stored in the memory:

[0154] The initial processor energy consumption model includes a first component and a second component, wherein the first component is determined based on the product of the processor computing amount and the processor frequency, and the second component is determined based on the ratio of the processor computing amount to the processor frequency.

[0155] In some specific implementations, the processor can implement the following steps by executing the computer program stored in the memory:

[0156] The first component and the second component in the initial processor energy consumption model are respectively configured with corresponding constant coefficients, and the constant coefficients are coefficients determined based on the pre-acquired reference processor frequency and the corresponding power consumption of each reference processor.

[0157] In some specific implementations, the processor can implement the following steps by executing the computer program stored in the memory:

[0158] Acquire each benchmark processor frequency, and control the target device to run smoothly according to each benchmark processor frequency, so as to collect the benchmark processor power consumption of the target device at different benchmark processor frequencies;

[0159] Accordingly, determining the constant coefficient based on the reference processor frequency and the corresponding power consumption of each reference processor includes:

[0160] Performing power function fitting on each of the reference processor frequencies and the reference processor power consumption to obtain a target power consumption frequency relationship function;

[0161] Based on the coefficients in the target power consumption frequency relationship function, constant coefficients corresponding to the first component and the second component in the initial processor energy consumption model are determined.

[0162] In some specific implementations, the processor can implement the following steps by executing the computer program stored in the memory:

[0163] Controlling the target device to run smoothly according to the current reference processor frequency to collect the reference processor power consumption of the target device at the current reference processor frequency;

[0164] The sum of the preset frequency increase and the current reference processor frequency is determined as the next reference processor frequency, and the next reference processor frequency is updated to the current reference processor frequency, and the process is then jumped again to the step of controlling the target device to run smoothly according to the current reference processor frequency.

[0165] In some specific implementations, the processor may further include the following steps by executing the computer program stored in the memory:

[0166] Determine the target processor power consumption corresponding to the target device when the target device is running in a single running time period based on the current processor power consumption of the target device, and determine whether the target processor power consumption is greater than a preset threshold;

[0167] If the power consumption of the target processor is greater than the preset threshold, the target device is controlled to operate based on the target processor frequency so that the target device uses the target processor frequency as the actual operating frequency of the processor in the next operating time cycle to achieve power consumption control of the target device.

[0168] In some specific implementations, the processor may further include the following steps by executing the computer program stored in the memory:

[0169] If the power consumption of the target processor is not greater than the preset threshold, the priority of the current task being processed by the target device is determined; if the priority of the current task is lower than the preset priority, the operation of the target device is controlled based on the target processor frequency, so that the target device uses the target processor frequency as the actual operating frequency of the processor in the next operating time cycle to achieve power consumption control of the target device.

[0170] In addition, the electronic device in this embodiment may specifically be an electronic computer.

[0171] In this embodiment, the power supply 23 is used to provide working voltage for each hardware device on the electronic device; the communication interface 24 can create a data transmission channel between the electronic device and the external device, and the communication protocol it follows is any communication protocol that can be applied to the technical solution of the present application, and is not specifically limited here; the input and output interface 25 is used to obtain external input data or output data to the outside world, and its specific interface type can be selected according to specific application needs and is not specifically limited here.

[0172] In addition, the memory 22, as a carrier for storing resources, can be a read-only memory, a random access memory, a disk or an optical disk, etc. The resources stored thereon can include an operating system 221, a computer program 222, etc., and the storage method can be temporary storage or permanent storage.

[0173] The operating system 221 is used to manage and control various hardware devices on the electronic device and the computer program 222, which can be Windows Server, Netware, Unix, Linux, etc. In addition to including a computer program that can be used to complete the device power consumption control method performed by the electronic device disclosed in any of the aforementioned embodiments, the computer program 222 can further include a computer program that can be used to complete other specific tasks.

[0174] Furthermore, the present application also discloses a computer-readable storage medium for storing a computer program; wherein the computer program, when executed by a processor, implements the aforementioned disclosed device power consumption control method. For the specific steps of the method, reference may be made to the corresponding contents disclosed in the aforementioned embodiments, and no further description will be given here.

[0175] An embodiment of the present application further discloses a computer program product, including a computer program / instruction, which, when executed by a processor, implements the steps of the device power consumption control method disclosed in any of the aforementioned embodiments.

[0176] In this specification, each embodiment is described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the embodiments can be referred to each other. For the device disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant parts can be referred to the method part.

[0177] Professionals may further appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of the two. In order to clearly illustrate the interchangeability of hardware and software, the composition and steps of each example have been generally described in the above description according to function. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professionals and technicians may use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0178] The steps of the method or algorithm described in conjunction with the embodiments disclosed herein may be implemented directly using hardware, a software module executed by a processor, or a combination of the two. The software module may be placed in a random access memory (RAM), a memory, a read-only memory (ROM), an electrically programmable ROM, an electrically erasable programmable ROM, a register, a hard disk, a removable disk, a CD-ROM, or any other form of storage medium known in the art.

[0179] Finally, it should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, the elements defined by the sentence "comprise a ..." do not exclude the presence of other identical elements in the process, method, article or device including the elements.

[0180] The technical solution provided by the present application is introduced in detail above. Specific examples are used in this article to illustrate the principles and implementation methods of the present application. The description of the above embodiments is only used to help understand the method of the present application and its core idea. At the same time, for general technicians in this field, according to the idea of ​​the present application, there will be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as a limitation on the present application.

Claims

1. A method for controlling power consumption of a device, It is characterized in that include: Determine whether the task currently running on the target device is a periodic task, and if the task currently running on the target device is a periodic task, determine the corresponding running time period; Predicting the amount of computation of the processor in the target device during the running time period to obtain a corresponding target amount of computation prediction value; Determining a target processor frequency required for the processor to complete an actual amount of computation consistent with a predicted value of a target amount of computation based on minimum power consumption; The target device is controlled to operate based on the target processor frequency, so that the target device uses the target processor frequency as the actual operating frequency of the processor in the next operating time cycle, thereby achieving power consumption control of the target device.

2. The device power consumption control method according to claim 1, It is characterized in that Determining the corresponding operating time period includes: Collecting the utilization rate of each target processor of the target device at different times within a preset time period, and determining the utilization rate of each target processor as original time series data; The original time series data is periodically analyzed to determine the corresponding operating time period.

3. The device power consumption control method according to claim 2, It is characterized in that The periodic analysis of the original time series data to determine the corresponding operation time period includes: The original time series data is subjected to Fourier transformation to obtain a frequency spectrum, and the corresponding operating time period is determined based on the main frequency component of the frequency spectrum.

4. The device power consumption control method according to claim 2, It is characterized in that The periodic analysis of the original time series data to determine the corresponding operation time period includes: The correlation between each time series data in the original time series data is analyzed, and the corresponding operation time period is determined according to the correlation.

5. The device power consumption control method according to claim 1, It is characterized in that The determining of the target processor frequency required by the processor to complete the actual amount of computation consistent with the predicted value of the target amount of computation based on the minimum power consumption includes: Obtain a pre-built initial processor energy consumption model with processor frequency as an independent variable and processor power consumption as a dependent variable; Inputting the target computing amount prediction value into the initial processor energy consumption model as the processor computing amount in the initial processor energy consumption model to obtain a target processor energy consumption model; The target processor energy consumption model is derived, and based on the derived model, the target processor frequency required for the processor to complete the actual computing amount consistent with the target computing amount prediction value based on the minimum power consumption is determined.

6. The device power consumption control method according to claim 5, It is characterized in that The initial processor energy consumption model includes a first component and a second component, wherein the first component is determined based on the product of the processor computing amount and the processor frequency, and the second component is determined based on the ratio of the processor computing amount to the processor frequency.

7. The device power consumption control method according to claim 6, It is characterized in that The first component and the second component in the initial processor energy consumption model are respectively configured with corresponding constant coefficients, and the constant coefficients are coefficients determined based on the pre-acquired reference processor frequency and the corresponding power consumption of each reference processor.

8. The device power consumption control method according to claim 7, It is characterized in that Get the benchmark processor frequency and the corresponding power consumption of each benchmark processor, including: Each benchmark processor frequency is obtained, and the target device is controlled to run smoothly according to each benchmark processor frequency, so as to collect the benchmark processor power consumption of the target device at different benchmark processor frequencies.

9. The device power consumption control method according to claim 8, It is characterized in that Determining the constant coefficient based on the reference processor frequency and the corresponding power consumption of each reference processor includes: Performing power function fitting on each of the reference processor frequencies and the reference processor power consumption to obtain a target power consumption frequency relationship function; Based on the coefficients in the target power consumption frequency relationship function, constant coefficients corresponding to the first component and the second component in the initial processor energy consumption model are determined.

10. The device power consumption control method according to claim 8, It is characterized in that The controlling the target device to run smoothly according to each of the reference processor frequencies to collect the reference processor power consumption of the target device at different reference processor frequencies includes: Controlling the target device to run smoothly according to the current reference processor frequency to collect the reference processor power consumption of the target device at the current reference processor frequency; The sum of the preset frequency increase and the current reference processor frequency is determined as the next reference processor frequency, and the next reference processor frequency is updated to the current reference processor frequency, and the process is then jumped again to the step of controlling the target device to run smoothly according to the current reference processor frequency.

11. The device power consumption control method according to claim 1, It is characterized in that The operating control of the target device based on the target processor frequency, so that the target device adopts the target processor frequency as the actual operating frequency of the processor in the next operating time cycle, so as to achieve power consumption control of the target device, includes: Determine the target processor power consumption corresponding to the target device when the target device is running in a single running time period based on the current processor power consumption of the target device, and determine whether the target processor power consumption is greater than a preset threshold; If the power consumption of the target processor is greater than the preset threshold, the operation of the target device is controlled based on the target processor frequency so that the target device uses the target processor frequency as the actual operating frequency of the processor in the next operating time cycle to achieve power consumption control of the target device.

12. The device power consumption control method according to claim 11, It is characterized in that After determining whether the power consumption of the target processor is greater than a preset threshold, the method further includes: If the power consumption of the target processor is not greater than the preset threshold, the priority of the current task being processed by the target device is determined; if the priority of the current task is lower than the preset priority, the operation of the target device is controlled based on the target processor frequency, so that the target device uses the target processor frequency as the actual operating frequency of the processor in the next operating time cycle to achieve power consumption control of the target device.

13. A computer program product comprising a computer program / instructions, It is characterized in that When the computer program / instruction is executed by a processor, the steps of the device power consumption control method according to any one of claims 1 to 12 are implemented.

14. An electronic device, It is characterized in that include: Memory for storing computer programs; A processor, configured to execute the computer program to implement the steps of the device power consumption control method according to any one of claims 1 to 12.

15. A computer-readable storage medium, It is characterized in that The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the device power consumption control method according to any one of claims 1 to 12 are implemented.

Images