GPU heat dissipation speed regulation offset determination method, device, equipment and storage medium

Abstract

The present invention discloses a method, device, equipment and storage medium for determining a heat dissipation speed adjustment offset of a GPU. The method comprises: selecting a target reference GPU from at least one candidate reference graphics processor GPU; wherein the target reference GPU is used to assist in adjusting the heat dissipation speed adjustment offset of a GPU to be adapted; obtaining a reference speed adjustment offset of the target reference GPU; determining a first temperature value of the GPU to be adapted and a second temperature value of the target reference GPU at the same fan speed; determining a candidate speed adjustment offset of the GPU to be adapted in the current iteration cycle according to the first temperature value, the second temperature value and the reference speed adjustment offset; determining a target speed adjustment offset of the GPU to be adapted according to the candidate speed adjustment offset of the GPU to be adapted in the current iteration cycle. The embodiment of the present invention realizes dynamic adaptation of the heat dissipation speed adjustment offset of an unknown model GPU.

Description

Technical Field

[0001] The present invention relates to the technical field of servers, and in particular to a method, device, equipment and storage medium for determining a GPU heat dissipation speed regulation offset. Background Art

[0002] Computing power, as a key productivity factor in the development of the digital economy, has become the core support and driving force for the development of the data economy. Correspondingly, the demand for servers and GPU (graphics processing unit) servers is increasing day by day. In order to meet the complex needs of the market, GPU servers have higher and higher requirements for the adaptation and iteration speed of GPUs.

[0003] Currently, the adaptation of GPU thermal speed control offset is usually performed by BMC (Baseboard Management Controller) through relevant protocols or querying relevant GPU configuration tables to obtain GPU related thermal parameters, and increase or decrease the speed control offset according to the characteristics of the GPU model to obtain the final GPU thermal speed control offset. However, when a new unknown model of GPU is introduced, it is necessary to additionally adapt the unknown model GPU, recalculate and maintain the relationship between the unknown model GPU and the thermal speed control offset, and it is impossible for BMC to dynamically adapt the thermal speed control offset of the unknown model GPU. Summary of the invention

[0004] The present invention provides a method, device, equipment and storage medium for determining a heat dissipation speed regulation offset of a GPU, so as to realize dynamic adaptation of the heat dissipation speed regulation offset of an unknown model GPU.

[0005] According to one aspect of the present invention, a method for determining a GPU heat dissipation speed regulation offset is provided, the method comprising:

[0006] Selecting a target reference GPU from at least one candidate reference graphics processor GPU; wherein the target reference GPU is used to assist in adjusting the heat dissipation speed adjustment offset of the GPU to be adapted;

[0007] Obtaining a reference speed adjustment offset of the target reference GPU;

[0008] Determine a first temperature value of the GPU to be adapted and a second temperature value of the target reference GPU at the same fan speed;

[0009] Determine a candidate speed adjustment offset of the GPU to be adapted in a current iteration cycle according to the first temperature value, the second temperature value and the reference speed adjustment offset;

[0010] A target speed adjustment offset of the GPU to be adapted is determined according to the candidate speed adjustment offset of the GPU to be adapted in the current iteration cycle.

[0011] Optionally, selecting a target reference GPU from at least one candidate reference graphics processor GPU includes:

[0012] Acquire a first operating temperature parameter of the GPU to be adapted, and acquire a second operating temperature parameter corresponding to each of the candidate reference GPUs;

[0013] A target reference GPU is selected from the candidate reference GPUs according to the first operating temperature parameter and the second operating temperature parameters.

[0014] The above technical solution achieves automatic selection of the target reference GPU by obtaining the operating temperature parameters corresponding to the GPU to be adapted and each candidate reference GPU, and selecting the target reference GPU from each candidate reference GPU based on the operating temperature parameters, thereby improving the accuracy and efficiency of selecting the target reference GPU.

[0015] Optionally, determining, according to the first temperature value, the second temperature value and the reference speed adjustment offset, a candidate speed adjustment offset of the GPU to be adapted in a current iteration cycle includes:

[0016] Determining a first temperature difference of the GPU to be adapted according to the first temperature value and a first alarm threshold of the GPU to be adapted; and

[0017] Determining a second temperature difference of the target reference GPU according to the second temperature value and a second alarm threshold of the target reference GPU;

[0018] The reference speed adjustment offset is adjusted according to the first temperature difference and the second temperature difference to obtain a candidate speed adjustment offset of the GPU to be adapted in a current iteration cycle.

[0019] The above technical solution determines the first temperature difference through the first temperature value and the first alarm threshold, thereby improving the accuracy of determining the first temperature difference; and determines the second temperature difference through the second temperature value and the second alarm threshold, thereby improving the accuracy of determining the second temperature difference. By adjusting the reference speed adjustment offset through the first temperature difference and the second temperature difference, the candidate speed adjustment offset of the GPU to be adapted in the current iteration cycle is obtained, thereby achieving accurate determination of the candidate speed adjustment offset of the GPU to be adapted, thereby achieving accurate determination of the target speed adjustment offset of the GPU to be adapted.

[0020] Optionally, the method further includes:

[0021] Determining a first alarm threshold of the GPU to be adapted according to a first operating temperature parameter of the GPU to be adapted;

[0022] A second alarm threshold of the target reference GPU is determined according to a second operating temperature parameter of the target reference GPU.

[0023] The above technical solution achieves accurate determination of the first alarm threshold and the second alarm threshold by determining the first alarm threshold of the GPU to be adapted according to the first operating temperature parameter of the GPU to be adapted, and determining the second alarm threshold of the target reference GPU according to the second operating temperature parameter of the target reference GPU, thereby further improving the accuracy of subsequent determination of the candidate speed regulation offset of the GPU to be adapted.

[0024] Optionally, adjusting the reference speed adjustment offset according to the first temperature difference and the second temperature difference to obtain a candidate speed adjustment offset of the GPU to be adapted in the current iteration cycle includes:

[0025] If the first temperature difference is less than the second temperature difference, determining the sum of the reference speed adjustment offset and a preset offset coefficient threshold as the candidate speed adjustment offset of the GPU to be adapted in the current iteration cycle;

[0026] If the first temperature difference is equal to the second temperature difference, determining the reference speed adjustment offset as the candidate speed adjustment offset of the GPU to be adapted in the current iteration cycle;

[0027] If the first temperature difference is greater than the second temperature difference, the difference between the reference speed regulation offset and a preset offset coefficient threshold is determined as a candidate speed regulation offset of the GPU to be adapted in the current iteration cycle.

[0028] The above technical solution adaptively adjusts the reference speed control offset according to the correlation between the difference between the first temperature difference and the second temperature difference, thereby achieving accurate determination of the candidate speed control offset of the GPU to be adapted in the current iteration cycle, thereby further improving the accuracy of subsequent determination of the target speed control offset of the GPU to be adapted.

[0029] Optionally, determining the target speed adjustment offset of the GPU to be adapted according to the candidate speed adjustment offset of the GPU to be adapted in the current iteration cycle includes:

[0030] If the correlation relationship between the first temperature difference and the second temperature difference in the current iteration cycle satisfies a preset iteration termination condition, the iteration is terminated, and the candidate speed adjustment offset in the current iteration cycle is determined as the target speed adjustment offset of the GPU to be adapted;

[0031] If the current iteration cycle meets the preset iteration cycle number threshold, the iteration is terminated, and a candidate speed adjustment offset that is closer to the reference speed adjustment offset is selected from the candidate speed adjustment offsets corresponding to each iteration cycle, and is determined as the target speed adjustment offset of the GPU to be adapted.

[0032] The above technical solution determines the termination condition of the iteration by determining whether the correlation between the first temperature difference and the second temperature difference satisfies the iteration termination condition; and can also determine the termination condition of the iteration by determining whether the current iteration cycle satisfies a preset iteration cycle number threshold, thereby achieving accurate determination of the termination condition of the iteration, thereby improving the efficiency of determining the target speed regulation offset of the GPU to be adapted, avoiding infinite occupation of resources caused by infinite loop iterations, and thus avoiding excessive waste of resources.

[0033] Optionally, after determining the target speed adjustment offset of the GPU to be adapted, the method further includes:

[0034] If the GPU to be adapted supports out-of-band read power consumption, obtaining the current power consumption of the GPU to be adapted in the current read cycle;

[0035] If the speed correlation relationship between the fan speed in the current reading cycle and the fan speed in the previous reading cycle meets the preset speed judgment condition, then determine whether the current power consumption and the historical power consumption in the previous reading cycle meet the preset power consumption judgment condition;

[0036] If so, the target speed regulation offset is adjusted based on a preset offset adjustment strategy until a preset offset adjustment end condition is met.

[0037] The above technical solution obtains the power consumption value of the GPU to be adapted that supports out-of-band read power consumption in each read cycle, and uses the power consumption parameters to further optimize and adjust the target speed regulation offset of the GPU to be adapted, thereby further improving the accuracy of the target speed regulation offset of the GPU to be adapted.

[0038] According to another aspect of the present invention, a device for determining a GPU heat dissipation speed regulation offset is provided, the device comprising:

[0039] A target reference GPU selection module is used to select a target reference GPU from at least one candidate reference graphics processor GPU; wherein the target reference GPU is used to assist in adjusting the heat dissipation speed adjustment offset of the GPU to be adapted;

[0040] A reference offset acquisition module, used to acquire a reference speed adjustment offset of the target reference GPU;

[0041] A temperature value determination module, used to determine a first temperature value of the GPU to be adapted and a second temperature value of the target reference GPU at the same fan speed;

[0042] a candidate offset determination module, configured to determine a candidate speed adjustment offset of the GPU to be adapted in a current iteration cycle according to the first temperature value, the second temperature value and the reference speed adjustment offset;

[0043] The target offset determination module is used to determine the target speed adjustment offset of the GPU to be adapted according to the candidate speed adjustment offset of the GPU to be adapted in the current iteration cycle.

[0044] According to another aspect of the present invention, an electronic device is provided, the electronic device comprising:

[0045] at least one processor; and

[0046] a memory communicatively connected to the at least one processor; wherein,

[0047] The memory stores a computer program executable by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can execute the GPU heat dissipation speed regulation offset determination method described in any embodiment of the present invention.

[0048] According to another aspect of the present invention, a computer-readable storage medium is provided, wherein the computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a processor to implement the method for determining the GPU heat dissipation speed regulation offset described in any embodiment of the present invention when executed.

[0049] The embodiment of the present invention determines the first temperature value of the GPU to be adapted and the second temperature value of the target reference GPU at the same fan speed, and determines the candidate speed adjustment offset of the GPU to be adapted in the current iteration cycle according to the first temperature value, the second temperature value and the reference speed adjustment offset of the target reference GPU, thereby determining the target speed adjustment offset of the GPU to be adapted according to the candidate speed adjustment offset in the current iteration cycle, thereby realizing the automatic determination of the heat dissipation speed adjustment offset of the GPU to be adapted of unknown model, thereby realizing the dynamic adaptation of the heat dissipation speed adjustment offset of the GPU to be adapted. By adopting the relevant heat dissipation parameters of the target reference GPU close to the GPU to be adapted to determine the heat dissipation speed adjustment offset of the GPU to be adapted, rapid compatibility in GPU heat dissipation is realized, and it is widely used and can meet a large number of GPU temporary adaptation scenarios.

[0050] It should be understood that the contents described in this section are not intended to identify the key or important features of the embodiments of the present invention, nor are they intended to limit the scope of the present invention. Other features of the present invention will become easily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of 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 creative work.

[0052] Figure 1 is a flow chart of a method for determining a GPU heat dissipation speed regulation offset provided according to Embodiment 1 of the present invention;

[0053] Figure 2 is a flow chart of a method for determining a GPU heat dissipation speed regulation offset provided in accordance with Embodiment 2 of the present invention;

[0054] Figure 3 is a flow chart of a method for determining a GPU heat dissipation speed regulation offset provided in accordance with Embodiment 3 of the present invention;

[0055] Figure 4 2 is a schematic diagram of a device for determining a GPU heat dissipation speed regulation offset provided in accordance with a fourth embodiment of the present invention;

[0056] Figure 5 It is a structural schematic diagram of an electronic device for implementing the method for determining the GPU heat dissipation speed regulation offset according to an embodiment of the present invention. DETAILED DESCRIPTION

[0057] In order to enable those skilled in the art to better understand the scheme of the present invention, the technical scheme in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present invention.

[0058] It should be noted that the terms "first", "second", etc. in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments of the present invention described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0059] Embodiment 1

[0060] Figure 1 This is a flowchart of a method for determining a GPU heat dissipation speed regulation offset provided by the first embodiment of the present invention. This embodiment is applicable to the case of determining a heat dissipation speed regulation offset of an unknown model GPU. The method can be executed by a GPU heat dissipation speed regulation offset determination device. The GPU heat dissipation speed regulation offset determination device can be implemented in the form of hardware and / or software. The GPU heat dissipation speed regulation offset determination device can be configured in an electronic device. Figure 1 As shown, the method includes:

[0061] S110. Select a target reference GPU from at least one candidate reference graphics processor GPU; wherein the target reference GPU is used to assist in adjusting the heat dissipation speed regulation offset of the GPU to be adapted.

[0062] The candidate reference GPU may be a GPU that has been adapted in the BMC, that is, the heat dissipation speed adjustment offset and other heat dissipation parameters of the candidate reference GPU have been adapted in the BMC, and the heat dissipation parameters of the candidate reference GPU can be obtained in real time. The heat dissipation parameters may include the heat dissipation speed adjustment offset and the maximum operating temperature supported by the GPU.

[0063] The target reference GPU may be a candidate reference GPU for serving to adjust the heat dissipation speed adjustment offset of the GPU to be adapted. It should be noted that there may be at least one candidate reference GPU that has been adapted in the BMC. The GPU to be adapted may be an unknown model GPU to be adapted for heat dissipation speed adjustment offset in the BMC.

[0064] The heat dissipation speed adjustment offset can be the offset of the GPU temperature when adjusting the speed of the heat dissipation fan; the heat dissipation fan is used to cool the GPU in operation. The speed of the heat dissipation fan can be adaptively adjusted when the GPU temperature reaches a certain operating temperature, which is related to the maximum operating temperature supported by the GPU and the heat dissipation speed adjustment offset.

[0065] Specifically, the heat dissipation speed control strategies corresponding to the same model GPU or similar model GPU are the same. The heat dissipation speed control strategy can be to control and adjust the fan speed when the GPU temperature value in the working state reaches a certain temperature range, so that the GPU temperature value is reduced under high speed wind force, and when it is reduced to a certain temperature, the fan speed is controlled to be reduced. In the case where the heat dissipation fan speed needs to be adjusted, the corresponding temperature range is specifically related to the heat dissipation speed control offset of the GPU.

[0066] For example, a relevant technician may manually select a target reference GPU from at least one candidate GPU based on actual needs and combined with actual experience or experimental values. It should be noted that, in order to further improve the accuracy and efficiency of selecting the target reference GPU, the target reference GPU may also be automatically selected based on the heat dissipation parameters corresponding to the target reference GPU and the GPU to be adapted.

[0067] In an optional embodiment, selecting a target reference GPU from at least one candidate reference graphics processor GPU includes: obtaining a first operating temperature parameter of the GPU to be adapted, and obtaining a second operating temperature parameter corresponding to each candidate reference GPU; and selecting a target reference GPU from each candidate reference GPU according to the first operating temperature parameter and each second operating temperature parameter.

[0068] The first operating temperature parameter may be the maximum operating temperature that the GPU to be adapted can support; the second operating temperature parameter may be the maximum operating temperature that the GPUs corresponding to each candidate reference GPU can support.

[0069] The first operating temperature parameter and the second operating temperature parameter can be directly obtained from the BMC.

[0070] It should be noted that the first operating temperature parameter of the GPU to be adapted can be pre-imported in the BMC by relevant technical personnel. The specific import method can be to obtain the GPU information to be imported by consulting the product specification of the GPU to be adapted. For example, the GPU information may include the GPU's Vendor ID (Vendor Identity document, manufacturer's unique identification code), Device ID (Device Identity document, device unique identification code) and operating temperature parameters. Among them, the operating temperature parameter is the maximum operating temperature that the GPU can support.

[0071] Exemplarily, the GPU information that needs to be imported into the BMC is filled into an EXCEL table according to a preset template format to obtain a GPU information table; the GPU information table is uploaded to the BMC through the BMC Web UI (Website User Interface) interface. Specifically, the GPU information of the same model of GPUs can be filled into the same row in the EXCEL table, and at least one GPU information to be adapted can be filled row by row to obtain a GPU information table. For example, the first column in the GPU information table can be the Vendor ID of the GPU, in hexadecimal format; the second column can be the Device ID of the GPU, in hexadecimal format; the third column can be the operating temperature parameter of the GPU, in decimal format. The data format of different GPU data in the GPU information table can be set by relevant technical personnel according to actual needs, and this embodiment does not limit this.

[0072] Log in to the BMC management page through the BMC IP (Internet Protocol) address, and change the file suffix of the GPU information table to be uploaded to the xlsx format; if the file format of the GPU information table is not xlsx, the file upload may fail during the file upload process. Upload the GPU information table in the xlsx file format to the BMC through the BMC management page. Optionally, the BMC can change the file name of the uploaded GPU information table to the GPUInfo file and copy it to the var path to facilitate the subsequent acquisition and use of the data in the GPU information table.

[0073] It can be understood that, between the GPUs, the closer the corresponding maximum supported operating temperatures are, the more similar or identical the models of the two GPUs are, that is, the corresponding GPU information or heat dissipation parameters are also closer.

[0074] Exemplarily, a second operating temperature parameter that is closer to the first operating temperature parameter (for example, the closest) may be selected from the second operating temperature parameters, and the candidate reference GPU corresponding to the second operating temperature parameter may be used as the target reference GPU.

[0075] This optional embodiment achieves automatic selection of the target reference GPU by obtaining the operating temperature parameters corresponding to the GPU to be adapted and each candidate reference GPU, and selecting the target reference GPU from each candidate reference GPU based on the operating temperature parameters, thereby improving the accuracy and efficiency of selecting the target reference GPU.

[0076] S120: Obtain a reference speed adjustment offset of a target reference GPU.

[0077] The reference speed adjustment offset may be the heat dissipation speed adjustment offset of the target reference GPU. The reference speed adjustment offset of the GPU may be recorded in the BMC when the adaptation is completed in the BMC. Therefore, the reference speed adjustment offset of the target reference GPU may be directly obtained from the BMC.

[0078] S130: Determine a first temperature value of the GPU to be adapted and a second temperature value of the target reference GPU at the same fan speed.

[0079] The first temperature value may be a GPU temperature value obtained by the GPU monitoring process of the BMC at the same fan speed. The second temperature value may be a GPU temperature value at the same fan speed obtained by the BMC from the stored historical monitoring data of the target reference GPU. The historical monitoring data of the target reference GPU may include the GPU temperature values ​​of the target reference GPU at different fan speeds in different time periods.

[0080] The GPU monitoring process of the BMC can cyclically monitor the temperature value of the GPU to be adapted based on a preset iteration cycle. The iteration cycle can be preset by relevant technical personnel according to actual needs. For example, the iteration cycle can be 5 seconds.

[0081] Exemplarily, the GPU monitoring process of the BMC cyclically monitors the temperature value of the GPU to be adapted based on a preset iteration cycle. If in the current iteration cycle, the first temperature value monitored is 55°C, and the corresponding fan speed is 40 rpm, where 40 rpm means that the fan blades of the cooling fan rotate 40 times per minute, then the second temperature value of the target reference GPU corresponding to the target reference GPU at the fan speed of 40 rpm is obtained from the historical monitoring data of the target reference GPU.

[0082] S140: Determine a candidate speed adjustment offset of the GPU to be adapted in the current iteration cycle according to the first temperature value, the second temperature value, and the reference speed adjustment offset.

[0083] The candidate speed regulation offset may be a better (for example, an optimal) heat dissipation speed regulation offset after adjusting the heat dissipation speed regulation offset of the GPU to be adapted in the current iteration cycle.

[0084] Exemplarily, if the temperature difference between the first temperature value and the second temperature value is not greater than a preset temperature difference threshold, the reference speed adjustment offset is determined as the candidate speed adjustment offset of the GPU to be adapted in the current iteration cycle. The temperature difference threshold can be preset by relevant technical personnel, for example, the temperature difference threshold can be 3°C. If the temperature difference between the first temperature value and the second temperature value is greater than the preset temperature difference threshold, the reference speed adjustment offset is adjusted proportionally according to the ratio between the first temperature value and the second temperature value, and the adjusted reference speed adjustment offset is determined as the candidate speed adjustment offset of the GPU to be adapted in the current iteration cycle.

[0085] Specifically, if the temperature difference threshold is 3°C, the first temperature value is 50°C, and the second temperature value is 55°C, it can be determined that the temperature difference between the first temperature value and the second temperature value is greater than the preset temperature difference threshold. According to the first temperature value and the second temperature value, it can be determined that the ratio between the first temperature value and the second temperature value is 10:11. If the reference speed adjustment offset is 5°C, then according to the ratio between the first temperature value and the second temperature value, it can be determined that the candidate speed adjustment offset of the GPU to be adapted in the current iteration cycle is approximately 4°C. Optionally, if the value of the candidate speed adjustment offset determined according to the ratio is a decimal, the candidate speed adjustment offset can be rounded up or down.

[0086] S150: Determine a target speed adjustment offset of the GPU to be adapted according to the candidate speed adjustment offsets of the GPU to be adapted in the current iteration cycle.

[0087] It should be noted that after determining the candidate speed regulation offset of the GPU to be adapted in the current iteration cycle, the temperature range of the GPU to be adapted can be corrected according to the candidate speed regulation offset in the current iteration cycle. The temperature range may be the range in which the speed of the cooling fan needs to be controlled and adjusted. For example, if the initial temperature under the cooling speed regulation strategy corresponding to the GPU to be adapted is 65°C, and the candidate speed regulation offset is 5°C, then the temperature range of the GPU to be adapted after the initial temperature is offset up and down is 60°C to 70°C. In the next iteration cycle, when the temperature of the GPU to be adapted is between 60°C and 70°C, the speed of the cooling fan is controlled and adjusted to ensure that the GPU to be adapted is within a safe temperature range.

[0088] Exemplarily, in the process of determining the candidate speed adjustment offset of the GPU to be adapted in the next iteration cycle, the first temperature value of the GPU to be adapted at the same fan speed is determined, and the second temperature value of the target reference GPU at the same fan speed is obtained, and the candidate speed adjustment offset of the GPU to be adapted in the next iteration cycle is adjusted according to the first temperature value and the second temperature value in the next iteration cycle, and the candidate reference offset in the current iteration cycle, until the iteration end condition is met. The iteration end condition may be, for example, terminating the iteration after a preset number of iterations is met, and determining the candidate speed adjustment offset obtained in the last iteration cycle as the target speed adjustment offset of the GPU to be adapted. The preset number of iterations may be 10 times.

[0089] It should be noted that the influence of the candidate speed adjustment offset determined in each iteration cycle on the next iteration cycle may specifically be to change the temperature range of the GPU to be adapted in the next iteration cycle. Due to the change in the temperature range of the GPU to be adapted, the wind speed adjustment timing of the cooling fan changes, so that at the same wind speed of the cooling fan, the cooling effect of the GPU to be adapted and the target reference GPU is infinitely close, thereby achieving the purpose of accurately configuring the cooling speed adjustment offset of the GPU to be adapted.

[0090] The embodiment of the present invention determines the first temperature value of the GPU to be adapted and the second temperature value of the target reference GPU at the same fan speed, and determines the candidate speed adjustment offset of the GPU to be adapted in the current iteration cycle according to the first temperature value, the second temperature value and the reference speed adjustment offset of the target reference GPU, thereby determining the target speed adjustment offset of the GPU to be adapted according to the candidate speed adjustment offset in the current iteration cycle, thereby realizing the automatic determination of the heat dissipation speed adjustment offset of the GPU to be adapted of unknown model, thereby realizing the dynamic adaptation of the heat dissipation speed adjustment offset of the GPU to be adapted. By adopting the relevant heat dissipation parameters of the target reference GPU close to the GPU to be adapted to determine the heat dissipation speed adjustment offset of the GPU to be adapted, rapid compatibility in GPU heat dissipation is realized, and it is widely used and can meet a large number of GPU temporary adaptation scenarios.

[0091] It should be noted that, in order to further improve the accuracy of determining the target speed adjustment offset of the GPU to be adapted, the target speed adjustment offset of the GPU to be adapted may be further adjusted by obtaining the GPU power consumption of the GPU to be adapted.

[0092] In an optional embodiment, after determining the target speed regulation offset of the GPU to be adapted, it also includes: if the GPU to be adapted supports out-of-band read power consumption, obtaining the current power consumption of the GPU to be adapted in the current read cycle; if the speed correlation relationship between the fan speed in the current read cycle and the fan speed in the previous read cycle meets the preset speed judgment condition, then judging whether the current power consumption and the historical power consumption in the previous read cycle meet the preset power consumption judgment condition; if so, adjusting the target speed regulation offset based on the preset offset adjustment strategy until the preset offset adjustment end condition is met.

[0093] It is understandable that the GPU has the characteristics of out-of-band protocol. Only when the business is running and the system driver is used, the GPU power consumption can be read out-of-band. Therefore, when the GPU is not used in the system, it may be impossible to read the GPU power consumption out-of-band.

[0094] If the GPU to be adapted supports out-of-band reading of power consumption, the target speed regulation offset is optimized and adjusted based on the read power consumption value; if the GPU to be adapted does not support out-of-band reading of power consumption, there is no need to further optimize and adjust the target speed regulation offset of the GPU to be adapted.

[0095] Exemplarily, if the GPU to be adapted supports out-of-band reading of power consumption, the current power consumption of the GPU to be adapted in the current read cycle is obtained. The power consumption value of the GPU to be adapted in each read cycle can be read cyclically through the network transmission protocol. While obtaining the current power consumption of the GPU to be adapted in the current read cycle, the fan speed corresponding to the current power consumption is obtained. It is determined whether the speed correlation relationship between the fan speed in the current read cycle and the fan speed in the previous read cycle meets the preset speed judgment condition. The speed judgment condition can be predetermined by relevant technical personnel. For example, the speed judgment condition can be that the speed difference between the fan speed in the current read cycle and the fan speed in the previous read cycle is not greater than a preset speed difference threshold. The speed difference threshold can be preset by relevant technical personnel. For example, the speed difference threshold can be 5rmp.

[0096] If the speed correlation relationship between the fan speed in the current read cycle and the fan speed in the previous read cycle meets the preset speed judgment condition, then further determine whether the current power consumption and the historical power consumption in the previous read cycle meet the preset power consumption judgment condition; if the speed correlation relationship between the fan speed in the current read cycle and the fan speed in the previous read cycle does not meet the preset speed judgment condition, then continue to read the fan speed in the next read cycle and the power consumption value of the GPU to be adapted until the preset speed judgment condition is met. Among them, the power consumption judgment condition can be preset by relevant technical personnel. For example, the power consumption judgment condition can be that the power consumption difference between the current power consumption and the historical power consumption is not less than a preset power consumption difference threshold. Among them, the power consumption difference threshold can be preset by relevant technical personnel. For example, the power consumption difference threshold can be 20 watts.

[0097] If the current power consumption and the historical power consumption in the previous reading cycle meet the preset power consumption judgment conditions, the target speed regulation offset is adjusted based on the preset offset adjustment strategy until the preset offset adjustment end conditions are met; if the current power consumption and the historical power consumption in the previous reading cycle do not meet the preset power consumption judgment conditions, there is no need to adjust the target speed regulation offset, and continue to read the GPU power consumption value and fan speed in the next reading cycle.

[0098] The offset adjustment strategy may be preset by relevant technical personnel. For example, the offset adjustment strategy may be that when the current power consumption and the historical power consumption meet the preset power consumption judgment condition, if the current power consumption is greater than the historical power consumption, then the preset offset coefficient is increased on the basis of the current target speed regulation offset; if the current power consumption is less than the historical power consumption, then the preset offset coefficient is reduced on the basis of the current target speed regulation offset. The offset coefficient may be preset by relevant technical personnel. For example, the offset coefficient may be set to 1°C.

[0099] The end condition of the offset adjustment can be preset by relevant technical personnel. For example, the end condition of the offset adjustment can be that when the reading cycle reaches a preset number of readings, the adjustment of the target speed regulation offset is ended. For example, the number of readings can be 100 times. Optionally, the end condition of the offset adjustment can also be that in two adjacent reading cycles, the fan speeds are close to each other, and the corresponding power consumption values ​​are also close to each other, the adjustment of the target speed regulation offset is ended.

[0100] This optional embodiment obtains the power consumption value of the GPU to be adapted that supports out-of-band read power consumption in each read cycle, and uses power consumption parameters to further optimize and adjust the target speed regulation offset of the GPU to be adapted, thereby further improving the accuracy of the target speed regulation offset of the GPU to be adapted.

[0101] Embodiment 2

[0102] Figure 2 This is a flow chart of a method for determining a GPU heat dissipation speed regulation offset provided in Embodiment 2 of the present invention. This embodiment is optimized and improved on the basis of the above-mentioned technical solutions.

[0103] Furthermore, the step of "determining the candidate speed regulation offset of the GPU to be adapted in the current iteration cycle according to the first temperature value, the second temperature value and the reference speed regulation offset" is refined into "determining the first temperature difference of the GPU to be adapted according to the first temperature value and the first alarm threshold of the GPU to be adapted; and determining the second temperature difference of the target reference GPU according to the second temperature value and the second alarm threshold of the target reference GPU; adjusting the reference speed regulation offset according to the first temperature difference and the second temperature difference to obtain the candidate speed regulation offset of the GPU to be adapted in the current iteration cycle." so as to improve the method for determining the candidate speed regulation offset.

[0104] like Figure 2 As shown, the method comprises the following specific steps:

[0105] S210. Select a target reference GPU from at least one candidate reference graphics processor GPU; wherein the target reference GPU is used to assist in adjusting the heat dissipation speed regulation offset of the GPU to be adapted.

[0106] S220: Obtain a reference speed adjustment offset of a target reference GPU.

[0107] S230: Determine a first temperature value of the GPU to be adapted and a second temperature value of the target reference GPU at the same fan speed.

[0108] S240, determining a first temperature difference of the GPU to be adapted according to the first temperature value and a first alarm threshold of the GPU to be adapted; and determining a second temperature difference of the target reference GPU according to the second temperature value and a second alarm threshold of the target reference GPU.

[0109] The first alarm threshold may be a temperature threshold at which a temperature sensor alarms when the temperature of the GPU to be adapted is too high; the second alarm threshold may be a temperature threshold at which a temperature sensor alarms when the temperature of the target reference GPU is too high.

[0110] The first alarm threshold and the second alarm threshold can be preset by relevant technical personnel according to actual needs, combined with actual experience values ​​or experimental values. The first alarm threshold and the second alarm threshold can be the same or different; for example, the first alarm threshold and the second alarm threshold can be 60°C.

[0111] Exemplarily, the difference between the first temperature value and the first alarm threshold of the GPU to be adapted may be determined based on the first temperature value and the first alarm threshold, and the difference may be determined as the first temperature difference of the GPU to be adapted. The difference between the second temperature value and the second alarm threshold of the target reference GPU may be determined based on the second temperature value and the second alarm threshold, and the difference may be determined as the second temperature difference of the target reference GPU.

[0112] It should be noted that in order to further improve the accuracy of determining the first alarm threshold and the second alarm threshold, and thus further improve the accuracy of determining the candidate speed adjustment offset of the GPU to be adapted, the alarm threshold may also be determined based on the operating temperature parameters of the GPU.

[0113] In an optional embodiment, a first alarm threshold of the GPU to be adapted is determined according to a first operating temperature parameter of the GPU to be adapted; and a second alarm threshold of the target reference GPU is determined according to a second operating temperature parameter of the target reference GPU.

[0114] The first operating temperature parameter may be the maximum operating temperature that the GPU to be adapted can support; the second operating temperature parameter may be the maximum operating temperature that the GPUs corresponding to each candidate reference GPU can support.

[0115] For example, a preset increment threshold may be added based on the first operating temperature parameter of the GPU to be adapted, and the result after the increase is used as the first alarm threshold; a preset increment threshold may be added based on the second operating temperature parameter of the target reference GPU, and the result after the increase is used as the second alarm threshold. The preset increment threshold may be pre-set by relevant technical personnel, for example, the preset increment threshold may be 1°C.

[0116] Optionally, different levels of alarm thresholds can be obtained based on actual sensor alarm requirements. For example, 1°C is added to the operating temperature parameters as a normal alarm threshold; 2°C is added to the operating temperature parameters as a severe alarm threshold; 3°C is added to the operating temperature parameters as an irrecoverable alarm threshold. It can be set specifically according to actual needs. It should be noted that if the first alarm threshold corresponds to the severe alarm threshold of the GPU to be adapted, the second alarm threshold must also correspond to the severe alarm threshold of the target reference GPU.

[0117] This optional embodiment determines the first alarm threshold of the GPU to be adapted according to the first operating temperature parameter of the GPU to be adapted, and determines the second alarm threshold of the target reference GPU according to the second operating temperature parameter of the target reference GPU, thereby achieving accurate determination of the first alarm threshold and the second alarm threshold, and further improving the accuracy of subsequent determination of candidate speed regulation offsets for the GPU to be adapted.

[0118] S250. Adjust the reference speed adjustment offset according to the first temperature difference and the second temperature difference to obtain a candidate speed adjustment offset of the GPU to be adapted in the current iteration cycle.

[0119] Exemplarily, the reference offset may be adaptively adjusted according to the difference correlation between the first temperature difference value and the second temperature difference value, thereby obtaining a candidate speed adjustment offset of the GPU to be adapted in the current iteration cycle. For example, the difference correlation may be a difference size relationship between the first temperature difference value and the second temperature difference value.

[0120] In an optional embodiment, the reference speed regulation offset is adjusted according to the first temperature difference and the second temperature difference to obtain a candidate speed regulation offset of the GPU to be adapted in the current iteration cycle, including: if the first temperature difference is less than the second temperature difference, the sum of the reference speed regulation offset and a preset offset coefficient threshold is determined as the candidate speed regulation offset of the GPU to be adapted in the current iteration cycle; if the first temperature difference is equal to the second temperature difference, the reference speed regulation offset is determined as the candidate speed regulation offset of the GPU to be adapted in the current iteration cycle; if the first temperature difference is greater than the second temperature difference, the difference between the reference speed regulation offset and the preset offset coefficient threshold is determined as the candidate speed regulation offset of the GPU to be adapted in the current iteration cycle.

[0121] The offset coefficient threshold may be preset by relevant technicians. For example, the offset coefficient threshold may be 0.5°C.

[0122] For example, if the first temperature difference is smaller than the second temperature difference, it may indicate that the heat dissipation requirement of the GPU to be adapted is more stringent than the heat dissipation requirement of the target reference GPU. Therefore, the reference speed regulation offset may be adjusted upward. Specifically, the sum of the reference speed regulation offset and a preset offset coefficient threshold may be determined as the candidate speed regulation offset of the GPU to be adapted in the current iteration cycle.

[0123] If the first temperature difference is equal to the second temperature difference, it can indicate that the heat dissipation requirement of the GPU to be adapted is very close to the heat dissipation requirement of the target reference GPU. Therefore, the reference speed adjustment offset can be determined as the candidate speed adjustment offset of the GPU to be adapted in the current iteration cycle.

[0124] If the first temperature difference is greater than the second temperature difference, it may indicate that the heat dissipation requirement of the GPU to be adapted is more relaxed than the heat dissipation requirement of the target reference GPU. Therefore, the reference speed regulation offset may be adjusted downward. Specifically, the difference between the reference speed regulation offset and a preset offset coefficient threshold may be determined as the candidate speed regulation offset of the GPU to be adapted in the current iteration cycle.

[0125] This optional embodiment adaptively adjusts the reference speed regulation offset according to the correlation relationship between the difference between the first temperature difference and the second temperature difference, thereby achieving accurate determination of the candidate speed regulation offset of the GPU to be adapted in the current iteration cycle, thereby further improving the accuracy of subsequent determination of the target speed regulation offset of the GPU to be adapted.

[0126] S260: Determine a target speed adjustment offset of the GPU to be adapted according to the candidate speed adjustment offsets of the GPU to be adapted in the current iteration cycle.

[0127] The present embodiment determines the first temperature difference through the first temperature value and the first alarm threshold, thereby improving the accuracy of determining the first temperature difference; and determines the second temperature difference through the second temperature value and the second alarm threshold, thereby improving the accuracy of determining the second temperature difference. By adjusting the reference speed adjustment offset through the first temperature difference and the second temperature difference, the candidate speed adjustment offset of the GPU to be adapted in the current iteration cycle is obtained, thereby achieving accurate determination of the candidate speed adjustment offset of the GPU to be adapted, thereby achieving accurate determination of the target speed adjustment offset of the GPU to be adapted.

[0128] It is understandable that in order to further improve the iteration efficiency of the candidate speed adjustment offsets for the GPU to be adapted, the timing for terminating the iteration may be determined by the correlation between the first temperature difference and the second temperature difference, thereby improving the efficiency of determining the target speed adjustment offset.

[0129] In an optional embodiment, a target speed offset of the GPU to be adapted is determined based on the candidate speed offset of the GPU to be adapted in the current iteration cycle, including: if the correlation between the first temperature difference and the second temperature difference in the current iteration cycle satisfies a preset iteration termination condition, the iteration is terminated, and the candidate speed offset in the current iteration cycle is determined as the target speed offset of the GPU to be adapted; if the current iteration cycle satisfies a preset iteration cycle number threshold, the iteration is terminated, and from the candidate speed offsets corresponding to each iteration cycle, a candidate speed offset that is closer to the reference speed offset is selected and determined as the target speed offset of the GPU to be adapted.

[0130] The iteration termination condition may be preset by relevant technicians. For example, the iteration termination condition may be that the first temperature difference and the second temperature difference are infinitely close. That is, the difference between the first temperature difference and the second temperature difference is not greater than a preset difference threshold. The difference threshold may be preset by relevant technicians. For example, the difference threshold may be 0.3°C.

[0131] In a specific example, the difference threshold is 0.3°C. If the difference between the first temperature difference and the second temperature difference in the current iteration cycle is not greater than 0.3°C, the iteration is terminated, and the candidate speed adjustment offset in the current iteration cycle is determined as the target speed adjustment offset of the GPU to be adapted. If the difference between the first temperature difference and the second temperature difference in the current iteration cycle is less than 0.3°C, the iteration continues until the iteration termination condition is met.

[0132] The threshold of the number of iteration cycles may be preset by relevant technical personnel. For example, the threshold of the number of iteration cycles may be 10 times.

[0133] Exemplarily, if the current iteration cycle meets a preset iteration number threshold, the iteration is terminated. For example, the iteration number threshold is 10 times, and if the number of cycles corresponding to the current iteration cycle is the 10th, the iteration is terminated. From each candidate speed adjustment offset corresponding to each iteration cycle, a candidate speed adjustment offset that is closer to the reference speed adjustment offset is selected and determined as the target speed adjustment offset of the GPU to be adapted. For example, from each candidate speed adjustment offset determined by 10 iteration cycles, a candidate speed adjustment offset that is closest to the reference speed adjustment offset can be selected as the target speed adjustment offset of the GPU to be adapted.

[0134] This optional embodiment determines the termination condition of the iteration by determining whether the correlation between the first temperature difference and the second temperature difference satisfies the iteration termination condition; and can also determine the termination condition of the iteration by determining whether the current iteration cycle satisfies a preset iteration cycle number threshold, thereby achieving accurate determination of the termination condition of the iteration, thereby improving the efficiency of determining the target speed regulation offset of the GPU to be adapted, avoiding infinite occupation of resources caused by infinite loop iterations, and thus avoiding excessive waste of resources.

[0135] Embodiment 3

[0136] Figure 3 This is a flow chart of a method for determining a GPU heat dissipation speed regulation offset provided in Embodiment 3 of the present invention. This embodiment provides a preferred example based on the above embodiments.

[0137] like Figure 3 As shown, the method comprises the following specific steps:

[0138] S310: Obtain a first operating temperature parameter of the GPU to be adapted, and obtain a second operating temperature parameter corresponding to each candidate reference GPU.

[0139] S320 . Select a target reference GPU from each candidate reference GPU according to the first operating temperature parameter and each second operating temperature parameter.

[0140] S330: Obtain a reference speed adjustment offset of a target reference GPU.

[0141] S340: Determine a first temperature value of the GPU to be adapted and a second temperature value of the target reference GPU at the same fan speed.

[0142] S350: Determine a first alarm threshold of the GPU to be adapted according to a first operating temperature parameter of the GPU to be adapted; and determine a second alarm threshold of the target reference GPU according to a second operating temperature parameter of the target reference GPU.

[0143] S360: Determine a first temperature difference of the GPU to be adapted according to the first temperature value and a first alarm threshold of the GPU to be adapted; and determine a second temperature difference of the target reference GPU according to the second temperature value and a second alarm threshold of the target reference GPU.

[0144] S370A: If the first temperature difference is less than the second temperature difference, a sum of a reference speed adjustment offset and a preset offset coefficient threshold is determined as a candidate speed adjustment offset of the GPU to be adapted in the current iteration cycle.

[0145] S370B: If the first temperature difference is equal to the second temperature difference, determine the reference speed adjustment offset as the candidate speed adjustment offset of the GPU to be adapted in the current iteration cycle.

[0146] S370C: If the first temperature difference is greater than the second temperature difference, determine the difference between the reference speed adjustment offset and the preset offset coefficient threshold as the candidate speed adjustment offset of the GPU to be adapted in the current iteration cycle.

[0147] S380A: If the correlation between the first temperature difference and the second temperature difference in the current iteration cycle satisfies the preset iteration termination condition, the iteration is terminated, and the candidate speed adjustment offset in the current iteration cycle is determined as the target speed adjustment offset of the GPU to be adapted.

[0148] S380B. If the current iteration cycle meets the preset iteration cycle number threshold, the iteration is terminated, and a candidate speed adjustment offset that is closer to the reference speed adjustment offset is selected from the candidate speed adjustment offsets corresponding to each iteration cycle, and is determined as the target speed adjustment offset of the GPU to be adapted.

[0149] S390, determine whether the GPU to be adapted supports out-of-band reading of power consumption; if so, execute S3100A; if not, execute S3100B.

[0150] S3100A. Obtain the current power consumption of the GPU to be adapted in the current read cycle.

[0151] S3100B: No further optimization adjustment is required for the target speed adjustment offset of the GPU to be adapted.

[0152] S3110, determine whether the speed correlation relationship between the fan speed in the current reading cycle and the fan speed in the previous reading cycle meets the preset speed judgment condition, if so, execute S3120A; if not, execute S3120B.

[0153] S3120A, determine whether the current power consumption and the historical power consumption in the previous reading cycle meet the preset power consumption judgment condition; if so, execute S3130A; if not, execute S3130B.

[0154] S3120B, continue to read the fan speed and the power consumption value of the GPU to be adapted in the next reading cycle until the preset speed judgment condition is met.

[0155] S3130A. Adjust the target speed regulation offset based on a preset offset adjustment strategy until a preset offset adjustment end condition is met.

[0156] S3130B, no need to adjust the target speed adjustment offset, continue to read the GPU power consumption value and fan speed in the next reading cycle.

[0157] Embodiment 4

[0158] Figure 4 A schematic diagram of the structure of a GPU heat dissipation speed regulation offset determination device provided in the fourth embodiment of the present invention. A GPU heat dissipation speed regulation offset determination device provided in the embodiment of the present invention can be applied to the case of determining the heat dissipation speed regulation offset of an unknown model GPU. The GPU heat dissipation speed regulation offset determination device can be implemented in the form of hardware and / or software, such as Figure 4 As shown, the device specifically includes: a target reference GPU selection module 401, a reference offset acquisition module 402, a temperature value determination module 403, a candidate offset determination module 404 and a target offset determination module 405.

[0159] A target reference GPU selection module 401 is used to select a target reference GPU from at least one candidate reference graphics processor GPU; wherein the target reference GPU is used to assist in adjusting the heat dissipation speed adjustment offset of the GPU to be adapted;

[0160] A reference offset acquisition module 402 is used to acquire a reference speed adjustment offset of the target reference GPU;

[0161] The temperature value determination module 403 is used to determine a first temperature value of the GPU to be adapted and a second temperature value of the target reference GPU at the same fan speed;

[0162] A candidate offset determination module 404 is used to determine a candidate speed adjustment offset of the GPU to be adapted in a current iteration cycle according to the first temperature value, the second temperature value and the reference speed adjustment offset;

[0163] The target offset determination module 405 is used to determine the target speed adjustment offset of the GPU to be adapted according to the candidate speed adjustment offsets of the GPU to be adapted in the current iteration cycle.

[0164] The embodiment of the present invention determines the first temperature value of the GPU to be adapted and the second temperature value of the target reference GPU at the same fan speed, and determines the candidate speed adjustment offset of the GPU to be adapted in the current iteration cycle according to the first temperature value, the second temperature value and the reference speed adjustment offset of the target reference GPU, thereby determining the target speed adjustment offset of the GPU to be adapted according to the candidate speed adjustment offset in the current iteration cycle, thereby realizing the automatic determination of the heat dissipation speed adjustment offset of the GPU to be adapted of unknown model, thereby realizing the dynamic adaptation of the heat dissipation speed adjustment offset of the GPU to be adapted. By adopting the relevant heat dissipation parameters of the target reference GPU close to the GPU to be adapted to determine the heat dissipation speed adjustment offset of the GPU to be adapted, rapid compatibility in GPU heat dissipation is realized, and it is widely used and can meet a large number of GPU temporary adaptation scenarios.

[0165] Optionally, the target reference GPU selection module 401 includes:

[0166] a temperature parameter acquisition unit, configured to acquire a first operating temperature parameter of the GPU to be adapted, and to acquire a second operating temperature parameter corresponding to each of the candidate reference GPUs;

[0167] The target reference GPU selection unit is used to select a target reference GPU from each of the candidate reference GPUs according to the first operating temperature parameter and each of the second operating temperature parameters.

[0168] Optionally, the candidate offset determination module 404 includes:

[0169] a first temperature difference determining unit, configured to determine a first temperature difference of the GPU to be adapted according to the first temperature value and a first alarm threshold of the GPU to be adapted; and

[0170] a second temperature difference determining unit, configured to determine a second temperature difference of the target reference GPU according to the second temperature value and a second alarm threshold of the target reference GPU;

[0171] The candidate offset determination unit is used to adjust the reference speed adjustment offset according to the first temperature difference and the second temperature difference to obtain the candidate speed adjustment offset of the GPU to be adapted in the current iteration cycle.

[0172] Optionally, the device further comprises:

[0173] A first alarm threshold determination module, configured to determine a first alarm threshold of the GPU to be adapted according to a first operating temperature parameter of the GPU to be adapted;

[0174] The second alarm threshold determination module is used to determine the second alarm threshold of the target reference GPU according to the second operating temperature parameter of the target reference GPU.

[0175] Optionally, the candidate offset determination unit includes:

[0176] a first offset determination subunit, configured to determine, if the first temperature difference is less than the second temperature difference, a sum of the reference speed adjustment offset and a preset offset coefficient threshold as a candidate speed adjustment offset of the GPU to be adapted in a current iteration cycle;

[0177] a second offset determination subunit, configured to determine the reference speed adjustment offset as a candidate speed adjustment offset of the GPU to be adapted in a current iteration cycle if the first temperature difference is equal to the second temperature difference;

[0178] The third offset determination subunit is used to determine the difference between the reference speed adjustment offset and a preset offset coefficient threshold as the candidate speed adjustment offset of the GPU to be adapted in the current iteration cycle if the first temperature difference is greater than the second temperature difference.

[0179] Optionally, the target offset determination module 405 includes:

[0180] A first target offset determination unit, configured to terminate the iteration if the correlation between the first temperature difference and the second temperature difference in the current iteration cycle satisfies a preset iteration termination condition, and determine the candidate speed adjustment offset in the current iteration cycle as the target speed adjustment offset of the GPU to be adapted;

[0181] The second target offset determination unit is used to terminate the iteration if the current iteration cycle meets a preset iteration cycle number threshold, and select a candidate speed adjustment offset that is closer to the reference speed adjustment offset from each candidate speed adjustment offset corresponding to each iteration cycle, and determine it as the target speed adjustment offset of the GPU to be adapted.

[0182] Optionally, the device further comprises:

[0183] A current power consumption acquisition module, configured to acquire the current power consumption of the GPU to be adapted in a current read cycle if the GPU to be adapted supports out-of-band reading of power consumption after determining the target speed adjustment offset of the GPU to be adapted;

[0184] A power consumption judgment module, configured to judge whether the current power consumption and the historical power consumption in the previous reading cycle meet the preset power consumption judgment condition if the speed correlation relationship between the fan speed in the current reading cycle and the fan speed in the previous reading cycle meets the preset speed judgment condition;

[0185] The offset adjustment judgment module is used to adjust the target speed regulation offset based on a preset offset adjustment strategy if the current power consumption and the historical power consumption in the previous reading cycle meet the preset power consumption judgment condition, until the preset offset adjustment end condition is met.

[0186] The GPU heat dissipation speed regulation offset determination device provided in the embodiment of the present invention can execute the GPU heat dissipation speed regulation offset determination method provided in any embodiment of the present invention, and has the corresponding functional modules and beneficial effects of the execution method.

[0187] Embodiment 5

[0188] Figure 5 A schematic diagram of an electronic device 50 that can be used to implement an embodiment of the present invention is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital processing, cellular phones, smart phones, wearable devices (such as helmets, glasses, watches, etc.) and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the present invention described and / or required herein.

[0189] like Figure 5As shown, the electronic device 50 includes at least one processor 51, and a memory connected to the at least one processor 51 in communication, such as a read-only memory (ROM) 52, a random access memory (RAM) 53, etc., wherein the memory stores a computer program that can be executed by at least one processor, and the processor 51 can perform various appropriate actions and processes according to the computer program stored in the read-only memory (ROM) 52 or the computer program loaded from the storage unit 58 to the random access memory (RAM) 53. In the RAM 53, various programs and data required for the operation of the electronic device 50 can also be stored. The processor 51, the ROM 52, and the RAM 53 are connected to each other via a bus 54. An input / output (I / O) interface 55 is also connected to the bus 54.

[0190] A number of components in the electronic device 50 are connected to the I / O interface 55, including: an input unit 56, such as a keyboard, a mouse, etc.; an output unit 57, such as various types of displays, speakers, etc.; a storage unit 58, such as a disk, an optical disk, etc.; and a communication unit 59, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 59 allows the electronic device 50 to exchange information / data with other devices through a computer network such as the Internet and / or various telecommunication networks.

[0191] The processor 51 may be a variety of general and / or special processing components with processing and computing capabilities. Some examples of the processor 51 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various dedicated artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, a digital signal processor (DSP), and any appropriate processor, controller, microcontroller, etc. The processor 51 executes the various methods and processes described above, such as a method for determining a GPU heat dissipation speed regulation offset.

[0192] In some embodiments, the GPU thermal speed regulation offset determination method may be implemented as a computer program, which is tangibly contained in a computer-readable storage medium, such as a storage unit 58. In some embodiments, part or all of the computer program may be loaded and / or installed on the electronic device 50 via the ROM 52 and / or the communication unit 59. When the computer program is loaded into the RAM 53 and executed by the processor 51, one or more steps of the GPU thermal speed regulation offset determination method described above may be performed. Alternatively, in other embodiments, the processor 51 may be configured to perform the GPU thermal speed regulation offset determination method in any other appropriate manner (e.g., by means of firmware).

[0193] Various implementations of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on chips (SOCs), load programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various implementations can include: being implemented in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which can be a special purpose or general purpose programmable processor that can receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit data and instructions to the storage system, the at least one input device, and the at least one output device.

[0194] Computer programs for implementing the methods of the present invention may be written in any combination of one or more programming languages. These computer programs may be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, so that when the computer program is executed by the processor, the functions / operations specified in the flow chart and / or block diagram are implemented. The computer program may be executed entirely on the machine, partially on the machine, partially on the machine and partially on a remote machine as a stand-alone software package, or entirely on a remote machine or server.

[0195] In the context of the present invention, a computer-readable storage medium may be a tangible medium that may contain or store a computer program for use by or in combination with an instruction execution system, device or equipment. A computer-readable storage medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices or equipment, or any suitable combination of the foregoing. Alternatively, a computer-readable storage medium may be a machine-readable signal medium. A more specific example of a machine-readable storage medium may include an electrical connection based on one or more lines, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0196] To provide interaction with a user, the systems and techniques described herein may be implemented on an electronic device having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and a pointing device (e.g., a mouse or trackball) through which the user can provide input to the electronic device. Other types of devices may also be used to provide interaction with the user; for example, the feedback provided to the user may be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user may be received in any form (including acoustic input, voice input, or tactile input).

[0197] The systems and techniques described herein may be implemented in a computing system that includes backend components (e.g., as a data server), or a computing system that includes middleware components (e.g., an application server), or a computing system that includes frontend components (e.g., a user computer with a graphical user interface or a web browser through which a user can interact with implementations of the systems and techniques described herein), or a computing system that includes any combination of such backend components, middleware components, or frontend components. The components of the system may be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: a local area network (LAN), a wide area network (WAN), a blockchain network, and the Internet.

[0198] A computing system may include a client and a server. The client and the server are generally remote from each other and usually interact through a communication network. The client and server relationship is generated by computer programs running on the corresponding computers and having a client-server relationship with each other. The server may be a cloud server, also known as a cloud computing server or cloud host, which is a host product in the cloud computing service system to solve the defects of difficult management and weak business scalability in traditional physical hosts and VPS services.

[0199] It should be understood that the various forms of processes shown above can be used to reorder, add or delete steps. For example, the steps described in the present invention can be executed in parallel, sequentially or in different orders, as long as the desired results of the technical solution of the present invention can be achieved, and this document does not limit this.

[0200] The above specific implementations do not constitute a limitation on the protection scope of the present invention. It should be understood by those skilled in the art that various modifications, combinations, sub-combinations and substitutions can be made according to design requirements and other factors. Any modification, equivalent substitution and improvement made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for determining a GPU heat dissipation speed regulation offset, characterized in that: include: Selecting a target reference GPU from at least one candidate reference graphics processor GPU; wherein the target reference GPU is used to assist in adjusting the heat dissipation speed adjustment offset of the GPU to be adapted; the heat dissipation speed adjustment offset is the degree of GPU temperature offset when controlling and adjusting the speed of the heat dissipation fan; Obtaining a reference speed adjustment offset of the target reference GPU; Determine a first temperature value of the GPU to be adapted and a second temperature value of the target reference GPU at the same fan speed; Determine a candidate speed adjustment offset of the GPU to be adapted in a current iteration cycle according to the first temperature value, the second temperature value and the reference speed adjustment offset; A target speed adjustment offset of the GPU to be adapted is determined according to the candidate speed adjustment offset of the GPU to be adapted in the current iteration cycle.

2. The method according to claim 1, characterized in that: The step of selecting a target reference GPU from at least one candidate reference graphics processor GPU comprises: Obtaining a first operating temperature parameter of the GPU to be adapted, and obtaining a second operating temperature parameter corresponding to each candidate reference GPU; A target reference GPU is selected from the candidate reference GPUs according to the first operating temperature parameter and the second operating temperature parameters.

3. The method according to claim 1, characterized in that The step of determining the candidate speed adjustment offset of the GPU to be adapted in the current iteration cycle according to the first temperature value, the second temperature value and the reference speed adjustment offset includes: Determining a first temperature difference of the GPU to be adapted according to the first temperature value and a first alarm threshold of the GPU to be adapted; and Determining a second temperature difference of the target reference GPU according to the second temperature value and a second alarm threshold of the target reference GPU; The reference speed adjustment offset is adjusted according to the first temperature difference and the second temperature difference to obtain a candidate speed adjustment offset of the GPU to be adapted in a current iteration cycle.

4. The method according to claim 3, characterized in that The method further comprises: Determining a first alarm threshold of the GPU to be adapted according to a first operating temperature parameter of the GPU to be adapted; A second alarm threshold of the target reference GPU is determined according to a second operating temperature parameter of the target reference GPU.

5. The method according to claim 3, characterized in that: The step of adjusting the reference speed adjustment offset according to the first temperature difference and the second temperature difference to obtain a candidate speed adjustment offset of the GPU to be adapted in the current iteration cycle includes: If the first temperature difference is less than the second temperature difference, determining the sum of the reference speed adjustment offset and a preset offset coefficient threshold as the candidate speed adjustment offset of the GPU to be adapted in the current iteration cycle; If the first temperature difference is equal to the second temperature difference, determining the reference speed adjustment offset as the candidate speed adjustment offset of the GPU to be adapted in the current iteration cycle; If the first temperature difference is greater than the second temperature difference, the difference between the reference speed regulation offset and a preset offset coefficient threshold is determined as a candidate speed regulation offset of the GPU to be adapted in the current iteration cycle.

6. The method according to claim 3, characterized in that The step of determining a target speed adjustment offset of the GPU to be adapted according to the candidate speed adjustment offset of the GPU to be adapted in the current iteration cycle includes: If the correlation relationship between the first temperature difference and the second temperature difference in the current iteration cycle satisfies a preset iteration termination condition, the iteration is terminated, and the candidate speed adjustment offset in the current iteration cycle is determined as the target speed adjustment offset of the GPU to be adapted; If the current iteration cycle meets the preset iteration cycle number threshold, the iteration is terminated, and the candidate speed offset with the smallest difference with the reference speed offset is selected from the candidate speed offsets corresponding to each iteration cycle, and determined as the target speed offset of the GPU to be adapted.

7. The method according to any one of claims 1 to 5, characterized in that: After determining the target speed adjustment offset of the GPU to be adapted, the method further includes: If the GPU to be adapted supports out-of-band read power consumption, obtaining the current power consumption of the GPU to be adapted in the current read cycle; If the speed correlation relationship between the fan speed in the current reading cycle and the fan speed in the previous reading cycle meets the preset speed judgment condition, then determine whether the current power consumption and the historical power consumption in the previous reading cycle meet the preset power consumption judgment condition; If so, the target speed regulation offset is adjusted based on a preset offset adjustment strategy until a preset offset adjustment end condition is met.

8. A device for determining a GPU heat dissipation speed regulation offset, characterized in that: include: A target reference GPU selection module is used to select a target reference GPU from at least one candidate reference graphics processor GPU; wherein the target reference GPU is used to assist in adjusting the heat dissipation speed adjustment offset of the GPU to be adapted; the heat dissipation speed adjustment offset is the degree of GPU temperature offset when controlling and adjusting the speed of the heat dissipation fan; A reference offset acquisition module, used to acquire a reference speed adjustment offset of the target reference GPU; A temperature value determination module, used to determine a first temperature value of the GPU to be adapted and a second temperature value of the target reference GPU at the same fan speed; a candidate offset determination module, configured to determine a candidate speed adjustment offset of the GPU to be adapted in a current iteration cycle according to the first temperature value, the second temperature value and the reference speed adjustment offset; The target offset determination module is used to determine the target speed adjustment offset of the GPU to be adapted according to the candidate speed adjustment offset of the GPU to be adapted in the current iteration cycle.

9. An electronic device, characterized in that: The electronic device comprises: at least one processor; and a memory communicatively connected to the at least one processor; wherein, The memory stores a computer program executable by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can execute the GPU heat dissipation speed regulation offset determination method described in any one of claims 1-7.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a processor to implement the method for determining a GPU heat dissipation speed regulation offset according to any one of claims 1 to 7 when executed.

Images