Methods, devices, equipment and storage media for heat dissipation control of storage servers

Abstract

This invention relates to a heat dissipation control method, apparatus, device, and storage medium for a storage server, aiming to optimize the server's heat dissipation performance by real-time monitoring of chip operating characteristic parameters and power consumption data. First, the workload and power consumption information of each chip within the server are acquired, and their temperature distribution is calculated to generate an initial temperature data set. Next, the heat generation characteristics of this data set are analyzed to obtain a heat generation model under the current load. Subsequently, a heat transfer path model is established based on this model and the server structure to analyze heat transfer between chips. According to the heat transfer path, the configuration parameters of the server's air-cooling and liquid-cooling systems are optimized to generate optimized cooling parameter settings. Finally, the optimized parameters are applied to the server's heat dissipation control system and dynamically adjusted based on real-time load and temperature feedback to form an adaptive heat dissipation strategy, thereby improving heat dissipation efficiency and ensuring stable server operation.

Description

Technical Field

[0001] This invention relates to the field of data processing technology, and in particular to a heat dissipation control method, apparatus, computer equipment, and storage medium for a storage server. Background Technology

[0002] In modern data centers, with the continuous improvement of storage server performance, thermal control has become a crucial factor affecting system stability and energy consumption. In particular, advancements in semiconductor and chip technology have significantly increased the power density of computing units within servers, leading to increasingly severe heat dissipation problems. During the efficient processing of large amounts of data, server chips often operate at high frequencies, inevitably increasing heat generation. Existing cooling solutions are inadequate in handling frequent thermal cycles and dense heat sources. Furthermore, traditional air-cooling and liquid-cooling technologies suffer from performance bottlenecks in some high-performance storage scenarios, failing to meet the requirements for uniform heat dissipation and response speed.

[0003] As semiconductor manufacturing processes continue to shrink, more functional units are integrated into chips, leading to greater localized heat concentration in server chips. Increased thermal stress not only threatens chip reliability but can also cause performance degradation or even system failure. Current thermal control strategies are mostly based on fixed parameter settings, lacking the ability to dynamically adapt to the actual workload of the server and the characteristics of the chip, thus failing to respond effectively to different application scenarios and temperature conditions. This limitation in thermal control makes it difficult for the system to achieve efficient energy management between light and high loads, resulting in a low overall system energy efficiency ratio.

[0004] Furthermore, with the increasing demand for large-scale data processing in storage servers, the interplay of heat dissipation between chips has become a new challenge. Within a storage server, when multiple chips perform high-frequency calculations simultaneously, the cumulative heat generation makes it difficult for existing cooling systems to effectively identify and regulate the heat dissipation needs of each chip. This can lead to some chips overheating and even triggering thermal protection mechanisms, impacting the stability of the entire system. Therefore, an innovative heat dissipation control method based on semiconductor and chip characteristics is needed to dynamically adapt to changes in workload, optimize heat dissipation, and improve the overall energy efficiency of the server. Summary of the Invention

[0005] The main objective of this invention is to provide a heat dissipation control method, apparatus, device, and storage medium for storage servers, in order to solve the problems of uneven heat dissipation and low energy efficiency caused by increased chip power density and heat accumulation in modern storage servers.

[0006] To achieve the above objectives, the present invention provides a heat dissipation control method for a storage server, comprising the following steps: acquiring the operating characteristic parameters and real-time power consumption data of each chip in the storage server, and determining the temperature distribution of each chip in the server based on the operating characteristic parameters and the real-time power consumption data to obtain an initial temperature data set; performing heat generation characteristic analysis on the initial temperature data set to obtain a heat generation model of each chip in the storage server under the current load; establishing a heat transfer path model of each chip in the server based on the heat generation model and the server's preset structure to obtain a heat transfer path set; optimizing the configuration parameters of the heat dissipation control system inside the server based on the heat transfer path set to obtain optimized cooling parameter settings; and applying the optimized cooling parameter settings in the storage server to control the heat dissipation control system of the storage server, wherein the heat dissipation control system includes at least an air cooling system and a liquid cooling system.

[0007] The present invention also provides a heat dissipation control device for a storage server, comprising: a first acquisition unit, configured to acquire the operating characteristic parameters and real-time power consumption data of each chip in the storage server, and determine the temperature distribution of each chip in the server based on the operating characteristic parameters and the real-time power consumption data, thereby obtaining an initial temperature data set; a second acquisition unit, configured to perform heat generation characteristic analysis on the initial temperature data set, thereby obtaining a heat generation model of each chip in the storage server under the current load; a third acquisition unit, configured to establish a heat transfer path model of each chip in the server based on the heat generation model and a preset structure of the server, thereby obtaining a heat transfer path set; a fourth acquisition unit, configured to optimize the configuration parameters of the heat dissipation control system inside the server based on the heat transfer path set, thereby obtaining optimized cooling parameter settings; and a control unit, configured to apply the optimized cooling parameter settings in the storage server to control the heat dissipation control system of the storage server, wherein the heat dissipation control system includes at least an air cooling system and a liquid cooling system.

[0008] The present invention also provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps of any of the methods described above.

[0009] The present invention also provides a computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the steps of any of the methods described above.

[0010] The present invention provides a heat dissipation control method, apparatus, device, and storage medium for storage servers. By dynamically acquiring the operating characteristic parameters and real-time power consumption data of each chip within the storage server, it can analyze the heat generation of the chips in real time, construct a heat generation model, and establish a heat transfer path model for each chip based on the server's internal structure. By optimizing the configuration parameters of the air-cooling and liquid-cooling systems based on these models, precise control of the server's heat dissipation system is achieved, thereby effectively solving the problem of uneven heat dissipation caused by increased chip power density and heat accumulation.

[0011] Furthermore, the method can dynamically adjust cooling parameters based on the server's actual workload and temperature feedback, providing an adaptive heat dissipation control strategy, thereby improving the server's overall heat dissipation and energy efficiency. Through this adaptive dynamic control approach, the present invention can also ensure server stability and heat dissipation efficiency under different load conditions, avoiding chip overheating or system crashes caused by localized heat accumulation.

[0012] Compared to existing fixed-parameter heat dissipation solutions, the heat dissipation control method of this invention is more flexible and responsive, and can adjust the heat dissipation system according to actual working conditions to meet the high-efficiency heat dissipation requirements of modern data centers for high-performance storage servers.

[0013] It should be noted that this invention is applicable to a variety of high-density computing and storage scenarios, especially in chip power-intensive servers. Attached Figure Description

[0014] Figure 1 This is a schematic diagram of the heat dissipation control method steps for a storage server in one embodiment of the present invention;

[0015] Figure 2 This is a structural block diagram of the heat dissipation control device for a storage server in one embodiment of the present invention;

[0016] Figure 3 This is a schematic block diagram of the structure of a computer device according to an embodiment of the present invention.

[0017] The realization of the objective, functional features and advantages of the present invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0018] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0019] Reference Figure 1 This invention provides a heat dissipation control method for a storage server, comprising the following steps:

[0020] S1, obtain the operating characteristic parameters and real-time power consumption data of each chip in the storage server, and determine the temperature distribution of each chip in the server based on the operating characteristic parameters and the real-time power consumption data to obtain an initial temperature data set.

[0021] S2, perform heat generation characteristic analysis on the initial temperature data set to obtain the heat generation model of each chip in the storage server under the current load.

[0022] S3. Based on the heat generation model and combined with the server's preset structure, establish the heat transfer path model for each chip in the server to obtain a set of heat transfer paths.

[0023] S4. Based on the set of heat transfer paths, the configuration parameters of the heat dissipation control system inside the server are optimized to obtain optimized cooling parameter settings.

[0024] S5, apply the optimized cooling parameter settings to the storage server to control the heat dissipation control system of the storage server, wherein the heat dissipation control system includes at least: an air cooling system and a liquid cooling system.

[0025] In other words, acquiring the operating characteristic parameters and real-time power consumption data of each chip within the storage server is fundamental to thermal control. Monitoring the chip's operating characteristic parameters allows us to understand its status and performance, while collecting real-time power consumption data enables us to monitor the chip's power consumption under different load conditions. This data comes not only from the chip's own measurement sensors but may also involve environmental monitoring within the server, such as temperature sensors capturing the overall air temperature inside the server. This information provides the foundation for subsequent thermal strategies. After acquiring this data, based on these parameters and data, we can further determine the temperature distribution of each chip within the server, obtaining an initial temperature dataset. This initial temperature dataset reflects the thermal distribution state of each chip under the current load.

[0026] Next, the initial temperature dataset is analyzed for heat generation characteristics to generate a heat generation model for each chip within the server under the current load. The construction of this heat generation model aims to analyze how each chip generates heat under different operating loads. This process typically involves performing complex heat change trend analysis based on temperature data to generate a model reflecting the heat generation patterns of the chips under various loads. Through this analysis, not only can the temperature rise rate of each chip be understood, but also chips with abnormally rapid temperature changes or those generating large amounts of heat under high loads can be identified, allowing for targeted heat dissipation measures.

[0027] After obtaining the heat generation model, and combining it with the server's pre-defined structure, a heat transfer path model for each chip within the server is established, resulting in a set of heat transfer paths. To do this, it's necessary to obtain the physical parameters of the storage server, including not only the chip locations but also the overall geometric dimensions and material properties of the server structure. Combining these physical parameters with the thermal property parameters and the heat generation model, heat transfer equations between the chips can be established, thus mathematically describing the heat flow between them. Using the finite element method, heat transfer paths within the server can be simulated. These paths not only reflect the direction of heat flow between chips but also show the density distribution of heat flow. Through this simulation, the main heat transfer paths can be extracted, filtered, and simplified, ultimately yielding a network diagram containing the main heat transfer paths between chips. In this process, topology optimization is also used to analyze thermal resistance and thermal coupling effects, aiming to determine the optimal heat transfer path, thus providing an important basis for subsequent heat dissipation optimization.

[0028] Based on the set of heat transfer paths, the configuration parameters of the server's internal thermal management system are optimized to obtain optimized cooling parameter settings. Specifically, the set of heat transfer paths provides a basis for analyzing the heat transfer efficiency and heat dissipation requirements of each chip. Based on these heat transfer efficiencies and requirements, key cooling areas can be identified, determining the areas that the air-cooling and liquid-cooling systems should focus on cooling. The fan layout in the air-cooling system can be reconfigured according to these key cooling areas, including adjusting fan speed and airflow direction to optimize heat dissipation. The piping in the liquid-cooling system is optimized for areas with high heat flux density within the server, including replanning the path and flow rate of the liquid cooling pipes to provide more effective cooling in these high heat flux density areas. These optimized cooling solutions not only require coordinated configuration of the air-cooling and liquid-cooling system parameters but also necessitate real-time monitoring of the temperature of each chip within the server. The operating parameters of the air-cooling and liquid-cooling systems are dynamically adjusted based on temperature changes to ultimately obtain optimized cooling parameter settings.

[0029] Finally, these optimized cooling parameter settings are applied to the storage server to control its thermal management system. This system includes at least an air-cooling system and a liquid-cooling system. By rationally allocating and dynamically adjusting the operating parameters of air and liquid cooling, precise control of the server's internal temperature can be achieved. Furthermore, continuous temperature monitoring ensures that the thermal management system can adaptively respond to load changes, thereby guaranteeing high energy efficiency and stable heat dissipation performance under various operating conditions. Overall, this thermal control method, from real-time chip data acquisition, heat generation model construction, and heat transfer path analysis to cooling system parameter optimization, forms a complete and dynamic thermal management mechanism. This significantly improves the heat dissipation efficiency of the chips within the server, thereby ensuring system stability and sustained performance.

[0030] In one example, each chip is equipped with a miniature temperature sensor and a voltage current sensor. The miniature temperature sensor measures the surface temperature parameter of the chip, and the voltage current sensor measures the actual power consumption parameter of the chip. An ambient temperature sensor is also installed at multiple preset locations inside the server to capture the air temperature parameter inside the server. The acquisition of the operating characteristic parameters and real-time power consumption data of each chip within the storage server includes: monitoring the operating characteristic parameters of the storage server chips using the miniature temperature sensor and the ambient temperature sensor, and monitoring the real-time power consumption data of the storage server chips using the voltage current sensor.

[0031] Furthermore, based on the operating characteristic parameters and the real-time power consumption data, the temperature distribution of each chip within the server is determined to obtain an initial temperature data set. This includes: estimating the real-time temperature of each chip within the server using a preset heat dissipation model based on the real-time power consumption data and the operating characteristic parameters, obtaining estimated temperature values ​​for each chip and module; performing outlier removal and data interpolation on the temperature of each chip based on the estimated temperature values, obtaining a corrected temperature data set; performing feature mapping processing based on topological network relationships on the corrected temperature data set, mapping the temperature data to a multi-dimensional thermal coupling feature space, and constructing a thermal coupling relationship graph between each chip; and extracting the heat conduction characteristics and local heat accumulation effects between each chip based on the thermal coupling relationship graph, generating feature vectors with complex thermal relationships to generate the initial temperature data set for each chip within the server.

[0032] In this example, miniature temperature and voltage current sensors are placed on each chip to obtain more accurate real-time operating data. The miniature temperature sensors primarily monitor the chip's surface temperature, while the voltage current sensors measure the chip's actual power consumption. This meticulous sensor configuration ensures that the status of each chip can be accurately monitored in real time. Furthermore, ambient temperature sensors are placed in multiple pre-designed locations within the server to capture the internal air temperature data. These multi-faceted temperature and power consumption monitoring methods provide a solid data foundation for subsequent temperature distribution determination and heat dissipation control.

[0033] Specifically, the process of acquiring the operating characteristic parameters and real-time power consumption data of each chip is accomplished collaboratively by these sensors. Miniature temperature sensors and ambient temperature sensors work together to monitor temperature changes in the chip and its surrounding environment, thus constructing a comprehensive temperature dataset, while current and voltage sensors continuously monitor the chip's real-time power consumption. This multi-sensor monitoring not only allows temperature and power consumption data to complement each other but also enables the accurate recording of chip behavior under various operating load conditions, thereby providing more comprehensive information support for heat generation analysis and the optimization of heat dissipation strategies.

[0034] Based on these collected operating characteristic parameters and real-time power consumption data, the temperature distribution of each chip inside the server can be further determined, resulting in an initial temperature data set. To accurately determine these temperature distributions, the real-time temperature of each chip is first estimated using a pre-defined heat dissipation model based on the real-time power consumption data and the chip's operating characteristic parameters. This generates a temperature estimate for each chip, allowing us to understand its thermal state under the current load. However, since there may be outliers in the actual monitoring data, these estimates need to undergo outlier removal and data interpolation to improve the accuracy of the temperature data, ultimately resulting in a corrected temperature data set.

[0035] After obtaining the corrected temperature data, it is necessary to perform feature mapping processing based on topological network relationships to map the temperature data into a multi-dimensional thermal coupling feature space. This processing constructs a thermal coupling relationship diagram between the chips, clarifying the interactions and thermal coupling effects between them in heat transfer. Based on this thermal coupling relationship diagram, the heat conduction characteristics and local heat accumulation effects between the chips are further extracted, generating feature vectors with complex thermal relationships to generate the initial temperature data set for each chip within the server. This process, by transforming temperature data into thermal coupling features, not only helps identify which chips have strong heat transfer relationships but also reveals local heat accumulation phenomena within the chips, which is crucial for further optimizing the heat dissipation path and cooling strategy within the server.

[0036] By monitoring and analyzing the temperature of each chip in such detail, we can better understand the thermal behavior and heat distribution of the chip during operation, especially the temperature changes and heat accumulation under high load conditions. This understanding provides a key reference for the establishment of subsequent heat dissipation models and the optimization of heat dissipation control.

[0037] In one example, heat generation characteristic analysis is performed on the initial temperature data set to obtain a heat generation model for each chip in the storage server under the current load. This includes: analyzing the temperature change trend of each chip in the storage server based on the initial temperature data set to obtain the temperature rise rate of each chip and generate a set of temperature change rates; performing region clustering analysis on the set of temperature change rates to identify the key heat generation regions of each chip and obtain a set of key heat generation regions; applying energy conservation analysis based on the first law of thermodynamics to the set of key heat generation regions to quantitatively calculate the heat generation of each chip in the key regions and obtain a set of heat generation rates for each chip under different loads; constructing a dynamic thermal model based on the set of heat generation rates to predict the time series changes in heat generation of each chip under the current load and obtain a set of dynamic features of the heat generation model; performing multidimensional correlation analysis on the set of dynamic features to determine the thermal coupling relationship and mutual influence between chips and obtain a set of global characteristics of heat generation within the server; and generating a heat generation model for each chip in the server under the current load based on the set of global characteristics.

[0038] This example primarily describes how to determine the heat generation of each chip under different load conditions, thereby effectively guiding server heat dissipation optimization. Specifically, based on an initial temperature data set, the temperature change trends of each chip within the storage server are analyzed. Through in-depth study of this temperature data, the temperature rise rate of each chip can be calculated, ultimately generating a set of temperature change rates. The temperature rise rate is a crucial indicator of the chip's heat generation process, reflecting the chip's heat dissipation under the current load, particularly the rate of heat generation. Based on this rate set, it is possible to further determine which chips generate the most heat under the current environmental and load conditions, thereby identifying key areas of high heat generation.

[0039] Next, region clustering analysis is performed on the temperature change rate set to identify the key heat generation regions of each chip, resulting in a set of key heat generation regions. The goal of region clustering analysis is to group chips that generate similar heat changes together to identify key heat dissipation areas within the server. These key heat generation regions represent the parts that require the most cooling and heat dissipation under specific loads. Accurately identifying these regions provides more targeted information for the design and optimization of the cooling system.

[0040] Then, for these key heat-generating regions, energy conservation analysis based on the first law of thermodynamics is further applied to quantitatively calculate the heat generation of each chip in these key regions, obtaining a set of heat generation rates for each chip under different loads. The application of the first law of thermodynamics means that precise thermal analysis can be performed based on energy conservation, thereby accurately quantifying the heat generation rate of each chip. This quantitative analysis makes the heat generation behavior of each chip in the server clearer, especially under high load conditions, revealing which chips generate the most heat and their heat generation patterns under different load conditions.

[0041] Based on a set of heat generation rates, a dynamic thermal model was constructed to predict the time-series changes in heat generation of each chip under the current load, resulting in a dynamic feature set of the heat generation model. Through time-series prediction, the future heat generation of each chip can be estimated, allowing for the implementation of necessary heat dissipation measures in advance to ensure the system maintains good heat dissipation performance and temperature control under varying loads. The dynamic feature set reflects the evolution of heat generation over time and is a crucial basis for predicting and managing chip heat.

[0042] To better understand the thermal relationships between chips, a multidimensional correlation analysis was performed on the dynamic feature set to determine the thermal coupling relationships and the degree of mutual influence between chips, thereby obtaining a global feature set of heat generation within the server. The purpose of the multidimensional correlation analysis is to study the interrelationships between the heat generation behaviors of each chip. These coupling relationships may include direct heat conduction or indirect effects caused by heat accumulation in adjacent chips. Through this analysis, we can understand how the heat between chips influences each other and how these thermal couplings affect the overall temperature state of the server. This global feature set is crucial for understanding the overall behavior of heat generation within the server and can provide guidance for system-level thermal optimization.

[0043] Finally, based on the global feature set, a thermal generation model of each chip within the server under the current load is generated. This thermal generation model provides a comprehensive description of the thermal generation behavior within the server, reflecting not only the temperature rise of each chip but also revealing the heat transfer and coupling relationships between chips.

[0044] In one example, based on the heat generation model and the server's preset structure, a heat transfer path model for each chip within the server is established to obtain a set of heat transfer paths. This includes: acquiring the physical parameters of the storage server and determining a set of spatial geometric relationships within the storage server based on these parameters, wherein the physical parameters include at least chip locations, server structural dimensions, and material properties; extracting thermal property parameters from the set of spatial geometric relationships, and establishing heat transfer equations between chips based on the heat generation model to obtain a set of mathematical descriptions of heat transfer between chips; simulating the heat transfer paths between chips within the storage server using finite element analysis based on the set of mathematical descriptions to obtain the heat transfer paths and heat flux density distribution between chips, generating a set of heat flux paths; extracting the main heat transfer paths from the set of heat flux paths, filtering and simplifying the paths, and establishing a heat transfer path network diagram containing each chip to obtain a set of heat transfer paths; performing topology optimization on the set of heat transfer paths, and determining the optimal heat transfer path through thermal resistance and thermal coupling effect analysis to generate a set of heat transfer paths for subsequent heat dissipation optimization.

[0045] In this example, the heat transfer path model of each chip inside the server is established based on the heat generation model and the server's preset structure. This process of obtaining the set of heat transfer paths clarifies the way heat is transferred inside the server, thus providing a basis for optimizing the heat dissipation system.

[0046] Specifically, the first step is to obtain the physical parameters of the storage server. These parameters include the chip locations, the server's structural geometry, and its material properties. These parameters form the basis for constructing a heat transfer path model, as the relative positions of the chips, the overall server structure, and the thermal conductivity of the materials used all directly affect heat transfer efficiency. Based on these physical parameters, the set of spatial geometric relationships within the server can be determined. These relationships reflect the layout of the chips within the server, thus providing a geometric foundation for subsequent heat transfer analysis.

[0047] After understanding the spatial geometric relationships, the next step is to extract the thermal property parameters from these sets. Combined with the heat generation model established in the previous step, heat transfer equations between the chips are constructed, thus obtaining a mathematical description of heat transfer between chips. These heat transfer equations are established to mathematically describe the heat flow between chips. The heat transfer equations typically consider parameters such as thermal conductivity, distance, and temperature gradient between chips. Through these mathematical descriptions, we can gain a preliminary understanding of how heat is transferred between chips and which chip relationships are most critical for heat transfer.

[0048] After obtaining the mathematical description of heat transfer, the heat transfer paths between the chips inside the storage server are simulated using the finite element method (FEM). FEM is a sophisticated numerical simulation method that divides the complex internal structure of the server into many small units to accurately simulate the heat transfer between these units. Through FEM, the heat transfer paths and heat flux density distributions between the chips can be obtained, thus generating a set of heat flux paths. These heat flux paths reflect how heat flows inside the server, particularly indicating which paths have higher heat concentrations. This information is crucial for understanding the distribution of heat within the system.

[0049] Based on the set of heat flow paths, these paths are then extracted and simplified to ultimately construct a heat transfer path network diagram encompassing each chip. The purpose of constructing this network diagram is to identify and retain the most important heat transfer paths, thereby simplifying the complex heat transfer network into a single key heat transfer model. During this process, path filtering removes less important or less contributing heat transfer paths, making the model more concise and easier to analyze and optimize.

[0050] To further optimize heat transfer paths, topology optimization was performed on the heat transfer path set. This step aimed to determine the optimal heat transfer path by analyzing thermal resistance and thermal coupling effects. The magnitude of thermal resistance determines the ease of heat transfer, while thermal coupling reflects the thermal interactions between different chips. By comprehensively analyzing these two aspects, the optimal heat transfer path within the server can be determined, providing guidance for subsequent cooling system design. The optimized heat transfer path set allows for a clearer understanding of how heat should flow within the server under different load conditions, enabling priority consideration of these paths in the design and configuration of the cooling system, thereby achieving more effective heat dissipation control.

[0051] In one example, based on the set of heat transfer paths, the configuration parameters of the server's internal heat dissipation control system are optimized to obtain optimized cooling parameter settings. This includes: analyzing the heat transfer efficiency and heat dissipation requirements of each chip based on the set of heat transfer paths to obtain a set of heat dissipation requirements for each chip; identifying key cooling areas in the set of heat dissipation requirements to determine the key cooling areas of the air-cooled system and the liquid-cooled system to obtain a set of key cooling areas; and reconfiguring the fan layout in the air-cooled system based on the set of key cooling areas, adjusting the speed and airflow direction of each fan to obtain fan configuration parameters. The system optimizes the liquid cooling system's piping based on the fan configuration parameter set. It then redesigns the path and flow rate of the liquid cooling piping, considering high heat flux density areas in the heat transfer path, resulting in an optimized liquid cooling piping scheme. This scheme is combined with the fan configuration parameter set, and adaptive cooling control is implemented by real-time monitoring of temperature changes in each chip. This dynamically adjusts the operating parameters of the air-cooled and liquid-cooled systems, resulting in dynamic cooling parameter settings. Based on these dynamic cooling parameter settings, the overall cooling system's energy efficiency ratio is calculated, and the cooling parameters are further adjusted based on this ratio to obtain optimized cooling parameter settings.

[0052] In this example, based on the set of heat transfer paths, a detailed analysis of the heat transfer efficiency and cooling requirements of each chip within the server is conducted, resulting in a set of cooling requirements for each chip. Heat transfer efficiency is derived using the previously mentioned heat transfer path model; combining these models allows for the quantification of heat flow and heat accumulation for each chip. Analysis of this data identifies which chips require more cooling and which areas have more pressing cooling needs, thus obtaining the set of cooling requirements for each chip. These sets of requirements provide direction for optimizing the cooling system, determining which areas require focused cooling design.

[0053] Next, the set of heat dissipation requirements is analyzed to identify key cooling areas for both air-cooling and liquid-cooling systems, resulting in a set of key cooling areas. This involves analyzing the thermal requirements of each chip within the server and identifying critical areas requiring additional cooling. Based on these key cooling areas, resource allocation for the air-cooling and liquid-cooling systems can be determined; for example, which areas need more fans and which areas require increased density of liquid cooling pipes. The goal of this process is to ensure that cooling resources are distributed where they are most needed, avoiding resource waste and uneven heat dissipation.

[0054] After identifying the key cooling areas, the fan layout in the air-cooling system needs to be reconfigured based on this information. Fan layout is a crucial factor in the heat dissipation efficiency of an air-cooling system; adjusting fan position, speed, and airflow direction is necessary to ensure that cooling airflow covers the key cooling areas to the maximum extent. To implement this, simulation software can be used to simulate the fan airflow paths and determine the air-cooling effect under different configurations. This yields the optimal set of fan configuration parameters, including the position, power setting, speed, and airflow angle of each fan, ensuring that cooling airflow effectively covers the key heat dissipation areas of the chip.

[0055] Simultaneously, the piping design of the liquid cooling system is optimized based on the fan configuration parameters. Liquid cooling is a highly efficient heat dissipation method, particularly suitable for handling the heat dissipation needs of high heat density areas in servers. To optimize the design of the liquid cooling piping, the path of the liquid cooling piping and the flow rate of the coolant need to be replanned, taking into account the high heat flux density areas in the heat transfer path. In implementing this part, the analysis results of the heat transfer path can be used to determine which areas have the highest heat flux density, and the density of liquid cooling pipes can be increased in these areas. Optimizing the flow rate is also crucial; hydraulic calculations are needed to ensure that the coolant has a sufficient flow rate in critical areas to quickly remove heat from the chip surface. These optimization designs can be simulated using computer-aided design (CAD) software to determine parameters such as pipe diameter, bending radius, and flow rate, ultimately resulting in an optimized liquid cooling piping scheme.

[0056] Next, the fan configuration parameters and liquid cooling pipeline optimization scheme are combined. By monitoring the temperature changes of each chip in real time, adaptive cooling control is implemented to dynamically adjust the operating parameters of the air-cooling and liquid-cooling systems, thereby obtaining dynamic cooling parameter settings. This means that during actual server operation, temperature data of each chip needs to be collected in real time by sensors to determine the current heat dissipation effect. If the temperature in a certain area rises above a set threshold, the fan speed in that area can be increased immediately, or the flow rate of the liquid cooling pipeline can be increased. This adaptive cooling control requires an intelligent control system to manage all cooling equipment. It can dynamically adjust the operating status of each fan and liquid cooling pipeline through integrated data analysis algorithms to ensure that the server can maintain an appropriate temperature under any operating conditions.

[0057] Finally, based on the dynamic cooling parameter settings, the overall cooling system's energy efficiency ratio (EER) is calculated, and the cooling parameters are adjusted a second time based on the EER to obtain the optimized cooling parameter settings. The EER is a crucial indicator for evaluating the performance of a heat dissipation system, reflecting the cooling effect per unit of energy consumption. In this stage, by analyzing real-time data on the energy consumption of the entire system and the temperature of each chip, the EER of the cooling system under the current configuration is calculated, and the calculation results are used to determine if there is any waste of cooling resources or insufficient heat dissipation. If the EER is low, it indicates poor cooling efficiency, requiring secondary optimization of the operating parameters for both air cooling and liquid cooling to improve the overall heat dissipation effect and reduce energy consumption. In practice, short-term performance verification can be conducted to compare the heat dissipation effects under different parameter settings, thereby selecting the optimal cooling parameters.

[0058] By applying optimized cooling parameters to the server's thermal management system, precise control of the server's internal temperature and maximization of overall energy efficiency can be achieved. The entire thermal management process, through continuous data acquisition, analysis, and optimization, forms a dynamic, adaptive closed-loop control mechanism. This not only improves the efficiency of the thermal management system but also reduces energy consumption, enabling stable and efficient server operation.

[0059] In one example, a set of fan configuration parameters is combined with an optimized liquid cooling pipeline scheme, and adaptive cooling control is performed by real-time monitoring of the temperature changes of each chip to dynamically adjust the operating parameters of the air-cooling and liquid-cooling systems, resulting in dynamic cooling parameter settings. This includes: combining the fan configuration parameter set with the liquid cooling pipeline optimization scheme to obtain a preliminary cooling parameter set, and real-time monitoring of the temperature data of each chip inside the server to obtain a real-time temperature data set; performing adaptive cooling control based on the preliminary cooling parameter set and the real-time temperature data set to dynamically adjust the fan speed of the air-cooling system and the coolant flow rate of the liquid-cooling system to obtain the first round of dynamic cooling parameters; verifying the first round of dynamic cooling parameters by analyzing the temperature response of each chip to evaluate the cooling effect and energy efficiency ratio, obtaining a cooling effect evaluation result; performing a second optimization adjustment on the first round of dynamic cooling parameters based on the cooling effect evaluation result to obtain the second-optimized cooling parameter settings; applying the second-optimized cooling parameter settings to the server's internal cooling system, and obtaining the final dynamic cooling parameter settings through continuous temperature monitoring and cooling parameter adjustment.

[0060] In other words, the fan configuration parameters and liquid cooling pipeline optimization scheme are combined to form a preliminary cooling parameter set. This means that the previous optimization process has yielded the configuration schemes for fan position, speed, and airflow direction, as well as the optimization schemes for liquid cooling pipeline path and flow rate. Combining these two parts forms a comprehensive preliminary cooling parameter set, used for initial control of the entire server's heat dissipation system. Simultaneously, the temperature of each chip inside the server needs to be monitored in real time to obtain a real-time temperature data set. Through miniature temperature sensors installed on each chip and ambient temperature sensors within the server, the chip surface temperature and the temperature of the air inside the server can be captured in real time. This temperature data collection is the basis for dynamically adjusting the cooling system, as it reflects the heat dissipation effect of each chip under the current cooling conditions.

[0061] After obtaining the initial set of cooling parameters and real-time temperature data, the system begins adaptive cooling control. The core of this control process is adjusting the operating states of the air-cooling and liquid-cooling systems based on changes in temperature data. For example, when the temperature of a chip rises to a set upper limit, the system automatically increases the fan speed in that area or increases the coolant flow rate in the liquid-cooling system to ensure that the chip's heat is quickly dissipated. Simultaneously, for areas with lower temperatures or lower heat demand, the fan speed or coolant flow rate can be appropriately reduced to save energy. The key to this adaptive cooling lies in the real-time nature of the temperature monitoring data and the flexibility of adjustment. This requires an intelligent control system to dynamically adjust the parameters of the fan and liquid-cooling systems through complex data processing algorithms to obtain the first round of dynamic cooling parameters.

[0062] After implementing the first round of dynamic cooling parameters, they need to be validated to ensure that these parameters can effectively reduce temperature and improve energy efficiency in actual operation. Specifically, the cooling effect and energy efficiency can be evaluated by analyzing the temperature response of each chip, resulting in a cooling effect assessment. The cooling effect assessment aims to determine whether the current cooling configuration is reasonable and can effectively cope with the heat generated by the server. If the temperature of some chips drops significantly, it indicates a significant cooling effect; however, if the temperature in some areas drops slowly or remains above safe levels, further optimization of the cooling configuration may be necessary. The energy efficiency assessment aims to ensure the energy utilization efficiency of the heat dissipation system, that is, to achieve the best cooling effect with the lowest possible energy consumption.

[0063] Based on the cooling effect evaluation results, the first round of dynamic cooling parameters were optimized a second time, resulting in the optimized cooling parameter settings. This second optimization process mainly focuses on improving areas that performed poorly in the first round of cooling. This is achieved by further adjusting the fan layout, speed, or the flow rate and path of the liquid cooling system to improve the heat dissipation in these areas. The goal of this second optimization is to refine the cooling system configuration, ensuring more balanced heat dissipation across all areas and minimizing the possibility of overall server performance degradation due to excessively high temperatures in any one area. During this step, temperature data under different load conditions may be compared and analyzed to identify the optimal cooling solution.

[0064] The optimized cooling parameters are applied to the server's internal cooling system, and through continuous temperature monitoring and dynamic adjustment of the cooling parameters, the optimal dynamic cooling parameter settings are ultimately obtained. This means that during actual server operation, the system continuously collects temperature data and makes corresponding adjustments based on this data. Even when the workload changes, such as during peak load periods or low load conditions, the cooling system can automatically adjust the operating status of air cooling and liquid cooling according to real-time temperature changes to ensure that the internal temperature of the server remains within a safe and efficient range.

[0065] This dynamic cooling control enables the entire heat dissipation system to adaptively address heat generation under varying load conditions. It not only ensures precise chip temperature control but also achieves efficient energy utilization by dynamically adjusting the cooling system's operating parameters.

[0066] In one example, based on the dynamic cooling parameter settings, the overall cooling system's energy efficiency ratio (EER) is calculated, and the cooling parameters are adjusted a second time based on the EER to obtain optimized cooling parameter settings. This includes: based on the dynamic cooling parameter settings, real-time acquisition of energy consumption data from the air-cooling and liquid-cooling systems, as well as temperature data from each chip, to calculate the overall cooling system's EER and obtain an EER evaluation result; based on the EER evaluation result, multi-objective optimization of the dynamic cooling parameter settings is performed to determine the final adjustment direction of the cooling parameters, resulting in a final adjusted cooling parameter set; the final adjusted cooling parameter set undergoes short-term performance verification to ensure the stability and high energy efficiency of the cooling system under different load conditions; the final adjusted cooling parameter set is applied to the server's thermal management system to obtain optimized cooling parameter settings, thereby achieving precise control of the server's internal temperature and maximizing overall energy efficiency.

[0067] In other words, based on dynamic cooling parameter settings, real-time energy consumption data from both air-cooling and liquid-cooling systems, as well as temperature data from each chip, are collected to calculate the overall cooling system's energy efficiency ratio (EER) and obtain an EER evaluation result. The dynamic cooling parameter settings are the cooling scheme obtained through preliminary optimization and adaptive adjustment in the preceding steps. Therefore, these parameters need to be practically applied to the server, and energy consumption data and temperature changes of each chip must be collected during operation. Specifically, energy consumption data for the air-cooling system is typically obtained by monitoring fan power consumption, while energy consumption data for the liquid-cooling system is obtained by measuring pump power consumption and coolant flow. Simultaneously, temperature data for each chip is acquired in real-time using temperature sensors installed on the chip surface and inside the server. Based on this data, the EER of the entire cooling system can be calculated. The EER is a key indicator for measuring the cooling effect of a heat dissipation system per unit of energy consumption.

[0068] After obtaining the energy efficiency ratio (EER) assessment results, multi-objective optimization of the dynamic cooling parameter settings is needed to determine the final adjustment direction of the cooling parameters and obtain the final adjusted set of cooling parameters. The core of multi-objective optimization lies in comprehensively considering multiple factors, including cooling effect, energy consumption, and temperature stability. If the EER is low, it indicates that the system consumes too much energy but the cooling effect is not ideal, requiring reconfiguration of the cooling system. Specific implementation of multi-objective optimization may include further adjusting the speed of each fan in the air-cooled system, and even adjusting the fan's operating period, increasing its operating frequency when the temperature rises rapidly and reducing energy consumption when the temperature is low. Simultaneously, for liquid-cooled systems, the coolant flow rate can be optimized to ensure that the flow rate reaches an appropriate level in critical hot areas, while appropriately reducing the flow rate in areas with lower heat demand to reduce pump power consumption. Through the multi-objective optimization process, the combination of cooling parameters with the highest EER can be found, thereby maximizing the cooling effect while minimizing energy consumption.

[0069] The final adjusted cooling parameter set undergoes short-term performance verification to ensure the stability and high energy efficiency of the cooling system under different load conditions. This means that after multi-objective optimization, the newly obtained cooling parameter settings need to be practically verified. The focus of performance verification is to check the system's performance under different load conditions, including heat dissipation capacity under high load and energy consumption under low load. By placing the server under simulated high and low load conditions, the cooling effect and energy efficiency of these parameter settings are verified respectively, ensuring that the cooling system can maintain effective heat dissipation performance while maintaining optimal energy efficiency regardless of the system's state. During performance verification, environmental simulation tools can be used to simulate the server's load conditions in a real operating environment, and by monitoring the temperature of each chip and the energy consumption of the cooling system in real time, it can be ensured that the final adjusted cooling parameter set is optimal.

[0070] The final adjusted cooling parameter set is applied to the server's thermal management system to obtain optimized cooling parameter settings, achieving precise control of the server's internal temperature and maximizing overall energy efficiency. This means that the optimized cooling parameters will become the system's regular operating parameters, continuously used to control the server's internal air-cooling and liquid-cooling systems. After applying these parameters to the actual cooling system, the server can maintain stable chip temperatures under different workloads and environmental conditions through intelligent control of the cooling system, thereby reducing performance degradation or hardware damage caused by overheating. Simultaneously, through these optimized cooling parameter settings, the cooling system can minimize energy consumption while ensuring effective heat dissipation, maximizing the overall energy efficiency of the server, which is crucial for the server's economic and environmental friendliness.

[0071] The entire process is an iterative optimization of cooling parameters through continuous energy efficiency assessment. After each parameter adjustment, the optimal settings for the cooling system are gradually identified through steps such as energy efficiency assessment, performance verification, and multi-objective optimization, thereby ensuring that both the server's heat dissipation performance and energy efficiency reach their best state. Through this design and implementation, the entire server's cooling system can not only effectively cope with heat generation under high load conditions, but also achieve high energy efficiency by continuously adjusting cooling parameters. This is of great significance for the stable operation and energy management of server clusters in modern data centers.

[0072] It is worth emphasizing that maintenance costs for large-scale chip servers are extremely high, especially when the cooling system malfunctions. The impact extends beyond equipment repair costs; it also includes significant economic losses and operational risks due to prolonged downtime. These servers generate enormous amounts of heat during continuous operation, requiring extremely stringent cooling measures. Any poor cooling can lead to system overheating, resulting in a sharp decline in performance or even complete shutdown. Therefore, maintaining the server temperature within a safe range is crucial for equipment stability and extending its lifespan. For this reason, the application for such advanced adaptive thermal control methods is particularly necessary to ensure the efficient and safe operation of servers!

[0073] To facilitate understanding of the present invention, an embodiment is now provided for illustrative purposes:

[0074] The experimental subject is a large-scale storage server at the data center level. The server is equipped with 48 XYZ-1000 high-performance processing chips. The maximum power consumption of a single chip is 120W. Under full load, the surface temperature of the chip can exceed 90℃. The temperature must be controlled below 75℃. The server is equipped with both air cooling and liquid cooling systems to cope with the stringent heat dissipation requirements.

[0075] In the initial stage of the experiment, miniature temperature and current / voltage sensors were used to monitor the real-time temperature and power consumption of the chips. Complete operating characteristic parameters and power consumption data were obtained. Based on this data, the temperature distribution was estimated using a pre-defined heat dissipation model. The results showed that the temperature of the 16 chips in the middle of the server rapidly rose to 85°C under high load (within the first 5 minutes), becoming the critical heat generation area.

[0076] After identifying the key heat generation regions, energy conservation analysis was performed on these chips based on the first law of thermodynamics. The heat generation rate set under full load was calculated. Approximately 720 joules of heat are generated per minute. Finite element simulation was performed using a heat transfer path model to identify the concentrated heat flux regions. The liquid cooling system piping was optimized in conjunction with high heat flux density. The piping density in the key hot regions was increased, and the coolant flow rate was increased to 3 liters per minute.

[0077] An adaptive control strategy combining air cooling and liquid cooling is employed to dynamically optimize and adjust the cooling system. Based on initial cooling parameters, the temperature of each chip is monitored in real time, and dynamic cooling control is implemented. The fan speed and coolant flow rate are adjusted through an intelligent control system. After 3 hours of operation, monitoring data shows that the chip temperature in the central high-temperature area dropped from 85℃ to 70℃. The overall system energy efficiency ratio is improved by 15%.

[0078] Performance verification after multi-objective optimization showed that the air-cooled system consumed too much energy under low load conditions. Fan speed was further optimized under low load by reducing it by 20%, without affecting the efficient cooling effect of the liquid cooling system. Short-term performance verification results showed that the finally adjusted cooling parameters had good adaptability under different load conditions.

[0079] The optimized cooling parameters are then applied to the thermal management system. Through real-time monitoring and adaptive adjustments, the cooling system is managed with precision. Chip temperatures are maintained within a safe range. Cooling system energy consumption is significantly reduced. Energy efficiency is further improved. The risk of system downtime due to overheating is avoided. Maintenance costs are significantly reduced. Overall energy efficiency and operational reliability are enhanced.

[0080] The heat dissipation control method in this project has significantly improved server heat dissipation efficiency through experiments, reducing system energy consumption and maintenance costs. For large servers with high maintenance costs and significant downtime consequences, the solution of this invention is indispensable. Its effectiveness and necessity have been fully verified.

[0081] like Figure 2 As shown, this embodiment of the invention also provides a heat dissipation control device for a storage server, comprising:

[0082] The first acquisition unit 1 is used to acquire the operating characteristic parameters and real-time power consumption data of each chip in the storage server, and based on the operating characteristic parameters and the real-time power consumption data, determine the temperature distribution of each chip in the server to obtain an initial temperature data set.

[0083] The second acquisition unit 2 is used to perform heat generation characteristic analysis on the initial temperature data set to obtain the heat generation model of each chip in the storage server under the current load.

[0084] The third acquisition unit 3 is used to establish a heat transfer path model for each chip in the server based on the heat generation model and the preset structure of the server, so as to obtain a set of heat transfer paths.

[0085] The fourth acquisition unit 4 is used to optimize the configuration parameters of the heat dissipation control system inside the server based on the set of heat transfer paths, so as to obtain the optimized cooling parameter settings.

[0086] Control unit 5 is used to apply the optimized cooling parameter settings in the storage server to control the heat dissipation control system of the storage server, wherein the heat dissipation control system includes at least an air cooling system and a liquid cooling system.

[0087] In this embodiment, the specific implementation of each unit in the above device embodiment is described in the above method embodiment, and will not be repeated here.

[0088] Reference Figure 3 This invention also provides a computer device, which can be a server, and its internal structure can be as follows: Figure 3 As shown, the computer device includes a processor, memory, display screen, input device, network interface, and database connected via a system bus. The processor provides computing and control capabilities. The memory includes a non-volatile storage medium and internal memory. The non-volatile storage medium stores the operating system, computer programs, and database. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage medium. The database stores the data corresponding to this embodiment. The network interface is used to communicate with external terminals via a network connection. When the computer program is executed by the processor, it implements the above-described method.

[0089] Those skilled in the art will understand that Figure 3 The structures shown are merely block diagrams of some structures related to the present invention and do not constitute a limitation on the computer devices on which the present invention is applied.

[0090] An embodiment of the present invention also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the above-described method. It is understood that the computer-readable storage medium in this embodiment can be a volatile readable storage medium or a non-volatile readable storage medium.

[0091] Those skilled in the art will understand that all or part of the processes in the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes of the embodiments of the above methods. Any references to memory, storage, databases, or other media used in the present invention and embodiments can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual-rate SDRAM (SSRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), Rambus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM, etc.

[0092] It should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, apparatus, article, or method that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, apparatus, article, or method. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, apparatus, article, or method that includes that element.

[0093] The above description is merely a preferred embodiment of the present invention and does not limit the patent scope of the present invention. Any equivalent structural or procedural transformations made based on the content of the present invention's specification and drawings, or direct or indirect applications in other related technical fields, are similarly included within the patent protection scope of the present invention.

Claims

1. A heat dissipation control method for a storage server, characterized in that, Includes the following steps: The operating characteristic parameters and real-time power consumption data of each chip in the storage server are obtained, and the temperature distribution of each chip in the storage server is determined based on the operating characteristic parameters and the real-time power consumption data to obtain an initial temperature data set. Based on the initial temperature data set, the temperature change trend analysis of each chip in the storage server is performed to obtain the temperature rise rate of each chip and generate a set of temperature change rates. A region clustering analysis is performed on the temperature change rate set to identify the key heat generation regions of each chip, resulting in a key heat generation region set. Based on the first law of thermodynamics, energy conservation analysis is applied to the key heat generation region set to quantitatively calculate the heat generation of each chip within the key regions, resulting in a heat generation rate set for each chip under different loads. A dynamic thermal model is constructed based on this heat generation rate set to predict the time series changes in heat generation of each chip under the current load, resulting in a dynamic feature set of the thermal generation model. Multidimensional correlation analysis is performed on the dynamic feature set to determine the thermal coupling relationship and mutual influence between chips, resulting in a global characteristic set of the internal thermal generation model of the storage server. Based on the global characteristic set, a thermal generation model for each chip within the storage server under the current load is generated. The process involves: acquiring the physical parameters of the storage server; determining the set of spatial geometric relationships within the storage server based on these parameters, wherein the physical parameters include at least chip locations, storage server structural dimensions, and material properties; extracting thermal property parameters from the set of spatial geometric relationships; establishing heat transfer equations between chips using the heat generation model to obtain a mathematical description set of heat transfer between chips; simulating the heat transfer paths between chips within the storage server using finite element analysis to obtain the heat transfer paths and heat flux density distribution between chips, generating a set of heat flux paths; extracting the main heat transfer paths from the set of heat flux paths; filtering and simplifying the paths to establish a heat transfer path network diagram containing each chip, obtaining a set of heat transfer paths; performing topology optimization on the set of heat transfer paths; determining the optimal heat transfer path through thermal resistance and thermal coupling effect analysis, generating a set of heat transfer paths for subsequent heat dissipation optimization; Based on the set of heat transfer paths, the configuration parameters of the heat dissipation control system inside the storage server are optimized to obtain optimized cooling parameter settings. The optimized cooling parameter settings are applied in the storage server to control the heat dissipation control system of the storage server, wherein the heat dissipation control system includes at least an air cooling system and a liquid cooling system.

2. The heat dissipation control method according to claim 1, characterized in that, Each chip is equipped with a miniature temperature sensor and a current voltage sensor. The miniature temperature sensor is used to measure the surface temperature parameter of the chip, and the current voltage sensor is used to measure the actual power consumption parameter of the chip. In addition, multiple preset locations inside the storage server are equipped with ambient temperature sensors, which are used to capture the air temperature parameter inside the storage server. Acquiring the operating characteristic parameters and real-time power consumption data of each chip in the storage server includes: monitoring the operating characteristic parameters of the storage server chips through the micro temperature sensor and the ambient temperature sensor, and monitoring the real-time power consumption data of the storage server chips through the current voltage sensor. And / or, Based on the aforementioned operating characteristic parameters and the real-time power consumption data, the temperature distribution of each chip within the storage server is determined, resulting in an initial temperature data set, including: Based on the real-time power consumption data and the operating characteristic parameters, the temperature of each chip in the storage server is estimated in real time using a preset heat dissipation model, and the estimated temperature values ​​of each chip and module are obtained. Based on the temperature estimates, outlier values ​​are removed and data interpolation is performed on the temperatures of each chip to obtain a corrected temperature data set. The corrected temperature data set is subjected to feature mapping processing based on topological network relationships to map the temperature data to a multidimensional thermal coupling feature space, thereby constructing a thermal coupling relationship diagram between each chip. Based on the thermal coupling relationship diagram, the thermal conduction characteristics and local heat accumulation effects between each chip are extracted to generate a feature vector with complex thermal relationships, so as to generate the initial temperature data set of each chip in the storage server.

3. The heat dissipation control method according to claim 1, characterized in that, Based on the set of heat transfer paths, the configuration parameters of the heat dissipation control system inside the storage server are optimized to obtain optimized cooling parameter settings, including: Based on the set of heat transfer paths, the heat transfer efficiency and heat dissipation requirements of each chip are analyzed to obtain the set of heat dissipation requirements for each chip. The set of heat dissipation requirements is used to identify key cooling areas, and the key cooling areas of the air-cooled system and the liquid-cooled system are determined to obtain a set of key cooling areas. Based on the set of key cooling areas, the fan layout in the air-cooling system is reconfigured, and the speed and airflow direction of each fan are adjusted to obtain a set of fan configuration parameters. Based on the set of fan configuration parameters, the piping of the liquid cooling system is optimized. Combining the high heat flux density region in the heat transfer path, the path and flow rate of the liquid cooling piping are replanned to obtain an optimized liquid cooling piping scheme. By combining the fan configuration parameter set with the liquid cooling pipeline optimization scheme, and by monitoring the temperature changes of each chip in real time for adaptive cooling control, the operating parameters of the air cooling and liquid cooling systems are dynamically adjusted to obtain dynamic cooling parameter settings. Based on the dynamic cooling parameter settings, the energy efficiency ratio of the overall cooling system is calculated, and the cooling parameters are adjusted a second time based on the energy efficiency ratio to obtain the optimized cooling parameter settings.

4. The heat dissipation control method according to claim 3, characterized in that, By combining the fan configuration parameter set with the liquid cooling pipeline optimization scheme, and performing adaptive cooling control by monitoring the temperature changes of each chip in real time, the operating parameters of the air cooling and liquid cooling systems are dynamically adjusted to obtain dynamic cooling parameter settings, including: By combining the fan configuration parameter set with the liquid cooling pipeline optimization scheme, a preliminary cooling parameter set is obtained, and the temperature data of each chip inside the storage server is monitored in real time to obtain a real-time temperature data set. Adaptive cooling control is performed based on the preliminary cooling parameter set and the real-time temperature data set to dynamically adjust the fan speed of the air-cooled system and the coolant flow rate of the liquid-cooled system, thereby obtaining the first round of dynamic cooling parameters. The first round of dynamic cooling parameters were verified. The cooling effect and energy efficiency ratio were evaluated by analyzing the temperature response of each chip, and the cooling effect evaluation results were obtained. Based on the cooling effect evaluation results, the first round of dynamic cooling parameters were optimized and adjusted a second time to obtain the optimized cooling parameter settings. The optimized cooling parameters are applied to the internal cooling system of the storage server, and the final dynamic cooling parameters are obtained through continuous temperature monitoring and adjustment.

5. The heat dissipation control method according to claim 3, characterized in that, Based on the dynamic cooling parameter settings, the overall cooling system's energy efficiency ratio is calculated, and the cooling parameters are adjusted a second time based on the energy efficiency ratio to obtain optimized cooling parameter settings, including: Based on dynamic cooling parameter settings, real-time energy consumption data of air cooling and liquid cooling systems and temperature data of each chip are collected to calculate the overall cooling system energy efficiency ratio and obtain energy efficiency ratio evaluation results. Based on the energy efficiency ratio assessment results, the dynamic cooling parameter settings are optimized through multiple objectives to determine the final adjustment direction of the cooling parameters and obtain the final set of adjusted cooling parameters. The final adjusted cooling parameter set was tested for performance within a short period of time to ensure the stability and high energy efficiency of the cooling system under different load conditions. The final adjusted cooling parameter set is applied to the heat dissipation control system of the storage server to obtain optimized cooling parameter settings, so as to achieve precise control of the internal temperature of the storage server and maximize overall energy efficiency.

6. A heat dissipation control device for a storage server, characterized in that, include: The first acquisition unit is used to acquire the operating characteristic parameters and real-time power consumption data of each chip in the storage server, and based on the operating characteristic parameters and the real-time power consumption data, determine the temperature distribution of each chip in the storage server to obtain an initial temperature data set. The second acquisition unit is used to perform temperature change trend analysis on each chip in the storage server based on the initial temperature data set, obtain the temperature rise rate of each chip, and generate a temperature change rate set. A region clustering analysis is performed on the temperature change rate set to identify the key heat generation regions of each chip, resulting in a key heat generation region set. Based on the first law of thermodynamics, energy conservation analysis is applied to the key heat generation region set to quantitatively calculate the heat generation of each chip within the key regions, resulting in a heat generation rate set for each chip under different loads. A dynamic thermal model is constructed based on this heat generation rate set to predict the time series changes in heat generation of each chip under the current load, resulting in a dynamic feature set of the thermal generation model. Multidimensional correlation analysis is performed on the dynamic feature set to determine the thermal coupling relationship and mutual influence between chips, resulting in a global characteristic set of the internal thermal generation model of the storage server. Based on the global characteristic set, a thermal generation model for each chip within the storage server under the current load is generated. The third acquisition unit is used to acquire the physical parameters of the storage server and determine the set of spatial geometric relationships within the storage server based on the physical parameters of the storage server. The physical parameters include at least chip locations, storage server structural dimensions, and material properties. Thermal property parameters are extracted from the set of spatial geometric relationships. Combined with the thermal generation model, heat transfer equations between chips are established to obtain a mathematical description set of heat transfer between chips. Based on the mathematical description set, the heat transfer paths between chips within the storage server are simulated using finite element analysis to obtain the heat transfer paths and heat flux density distribution between chips, generating a set of heat flux paths. According to the set of heat flux paths, the main heat transfer paths are extracted, and path filtering and simplification are performed to establish a heat transfer path network diagram containing each chip, obtaining a set of heat transfer paths. Topology optimization is performed on the set of heat transfer paths. Through thermal resistance and thermal coupling effect analysis, the optimal heat transfer path is determined, generating a set of heat transfer paths for subsequent heat dissipation optimization. The fourth acquisition unit is used to optimize the configuration parameters of the heat dissipation control system inside the storage server based on the set of heat transfer paths, so as to obtain the optimized cooling parameter settings. A control unit is used to apply the optimized cooling parameter settings in the storage server to control the heat dissipation control system of the storage server, wherein the heat dissipation control system includes at least an air cooling system and a liquid cooling system.

7. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the method according to any one of claims 1 to 5.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 5.

Images