Partitioned heat flow control method and system for server heat dissipation module
By using a zoned heat flow control method for server cooling modules, and leveraging functional equivalence analysis and dynamic regulation, the problem of frequent switching of cooling loops was solved, resulting in reduced energy consumption and improved system stability.
Patent Information
- Application Number
- CN202511503561.8
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-10-21
- Publication Date
- 2026-02-17
AI Technical Summary
The lack of detailed analysis of the functional equivalence and performance differences between different heat dissipation circuits in the existing technology leads to the frequent activation of new heat dissipation circuits, which affects the energy utilization efficiency and operational stability of the system.
By using a partitioned heat flow control method for server heat dissipation modules and a heat dissipation loop resource manager for functional equivalence analysis, a startup and rollback heat dissipation loop table is generated. Based on the equivalence score results, intelligent selection and dynamic control of heat dissipation loops are performed, including reusing rollback loops and starting new loops.
The frequency of heat dissipation circuit switching was reduced, energy consumption was decreased, and system stability was improved.
Smart Images
Figure CN121541753A_ABST
Abstract
Description
Technical Field
[0001] This application relates to the field of heat dissipation module technology, and in particular to a partitioned heat flow control method and system for a server heat dissipation module. Background Technology
[0002] In large server clusters, data centers, and high-performance computing platforms, thermal management technology is a key element in ensuring system stability and energy efficiency.
[0003] Currently, the common practice in existing technologies is to allocate multiple loops to servers, and switch loops to meet the cooling needs at different stages when the load changes or thermal risks occur. While this approach offers a degree of flexibility, it has several drawbacks in practical applications. For example, excessively high loop switching frequencies can lead to significant energy consumption, and the switching process can cause short-term delays in thermal response, potentially even causing localized temperature fluctuations and affecting hardware stability.
[0004] In summary, the existing technology suffers from a lack of detailed analysis on the functional equivalence and performance differences between different heat dissipation circuits. This leads to the frequent activation of new heat dissipation circuits even when existing fallback circuits can meet the requirements through parameter adjustments, further impacting the system's energy utilization efficiency and operational stability. Summary of the Invention
[0005] The purpose of this application is to provide a partitioned heat flow control method and system for a server heat dissipation module, in order to solve the technical problem in the prior art that, due to the lack of detailed analysis on the functional equivalence and performance differences between different heat dissipation loops, new heat dissipation loops are frequently activated even when the existing fallback loops can meet the requirements through parameter adjustment, which further affects the energy utilization efficiency and operational stability of the system.
[0006] In view of the above problems, this application provides a method and system for partitioned heat flow control of a server heat dissipation module.
[0007] In a first aspect, this application provides a partitioned heat flow control method for a server heat dissipation module, implemented through a partitioned heat flow control system for the server heat dissipation module, comprising: accessing a heat dissipation loop resource manager of the heat dissipation module and extracting a set of heat dissipation loops from the heat dissipation loop resource manager; obtaining multiple partitions of the server and generating a start-up heat dissipation loop table and a fallback heat dissipation loop table for each partition based on the real-time heat dissipation demand information of each partition and the set of heat dissipation loops; when the start-up heat dissipation loop table is triggered for update, obtaining a newly added first heat dissipation loop to be started, performing functional equivalence analysis on the first heat dissipation loop to be started and the heat dissipation loops in the fallback heat dissipation loop table, and outputting a functional equivalence score result; outputting a heat dissipation loop execution decision result based on the functional equivalence score result, and issuing control instructions for the corresponding partitions to the heat dissipation module according to the heat dissipation loop execution decision result, wherein the heat dissipation loop execution result includes reusing the fallback heat dissipation loop and starting the newly added heat dissipation loop.
[0008] Preferably, the partitioned heat flow control method for a server heat dissipation module further includes: extracting a set of heat dissipation loops from the heat dissipation loop resource manager, wherein the heat dissipation loop resource manager stores a resource description semantic vector for each heat dissipation loop; wherein the resource description semantic vector includes loop ID, loop type, loop connection relationship and capability parameters, and the loop type of the heat dissipation module includes air-cooled loop, cooling loop and refrigeration loop.
[0009] Preferably, the partitioned heat flow control method for a server heat dissipation module further includes: recording real-time operating condition monitoring data of each partition; extracting key features of heat dissipation impact from the real-time operating condition monitoring data, wherein the key features of heat dissipation impact include heat accumulation index, temperature gradient, and load power consumption; performing heat dissipation demand quantification analysis on the key features of heat dissipation impact using a thermal resistance network model, and outputting real-time heat dissipation demand information for each partition, including heat dissipation quantity demand information, heat dissipation rate demand information, and heat dissipation response delay demand information.
[0010] Preferably, the partitioned heat flow control method for a server heat dissipation module further includes: obtaining the historical heat dissipation startup status of each partition; determining whether the current partition is starting heat dissipation for the first time based on the historical heat dissipation startup status of each partition; if the current partition is starting heat dissipation for the first time, obtaining a set of optional heat dissipation loops for the current partition based on the usage tag of each heat dissipation loop in the heat dissipation loop set, performing a heat dissipation demand matching analysis with the real-time heat dissipation demand information of the current partition in the set of optional heat dissipation loops, and outputting a matching heat dissipation loop; constructing a startup heat dissipation loop table based on the matching heat dissipation loops, at which time the fallback heat dissipation loop table is empty.
[0011] Preferably, the partitioned heat flow control method for a server heat dissipation module further includes: if the current partition is not the first time it is started for heat dissipation, calculating the heat dissipation demand gap based on the real-time heat dissipation demand information of the current partition and the current start-up heat dissipation loop table; performing a heat dissipation demand gap matching analysis on the available heat dissipation loop set based on the heat dissipation demand gap, and outputting a matching gap heat dissipation loop; at this time, the start-up heat dissipation loop table is triggered to update, and the matching gap heat dissipation loop is added to the start-up heat dissipation loop table as a newly added first heat dissipation loop to be started.
[0012] Preferably, the partitioned heat flow control method for a server heat dissipation module further includes: performing redundancy analysis on the heat dissipation loops in the current activated heat dissipation loop table based on the real-time heat dissipation demand information of the current partition, and obtaining a redundancy index; extracting heat dissipation loops with a redundancy greater than a preset redundancy threshold as fallback heat dissipation loops, and constructing a fallback heat dissipation loop table.
[0013] Preferably, the partitioned heat flow control method for a server heat dissipation module further includes: extracting the heat dissipation characteristic index of the first heat dissipation loop to be started; performing functional equivalence analysis on the heat dissipation loops in the fallback heat dissipation loop table according to the heat dissipation characteristic index of the first heat dissipation loop to be started, and obtaining the loop type difference index, partition coverage overlap and response time difference; calculating the loop type difference index, partition coverage overlap and response time difference, and outputting the functional equivalence score result.
[0014] Preferably, the partitioned heat flow control method for a server heat dissipation module further includes: if the heat dissipation loops with a value greater than a preset equivalence threshold in the equivalence score results return empty, the newly added first heat dissipation loop to be started is activated, and the heat dissipation loop execution decision result for activating the newly added heat dissipation loop is output; if the heat dissipation loops with a value greater than the preset equivalence threshold in the equivalence score results return not empty, the heat dissipation loops with a value greater than the preset equivalence threshold are selected and marked for reuse, and the heat dissipation loop execution result for reusing and rolling back the heat dissipation loop is obtained.
[0015] Preferably, the partitioned heat flow control method for a server heat dissipation module further includes: if the heat dissipation loop execution result outputs the result of reusing the fallback heat dissipation loop, obtaining the reuse performance difference between the newly added first heat dissipation loop to be started and the fallback heat dissipation loop; adjusting the heat dissipation parameters of the fallback heat dissipation loop according to the reuse performance difference, including the loop cooling temperature and the loop flow rate.
[0016] Secondly, this application also provides a partitioned heat flow control system for a server heat dissipation module, used to execute a partitioned heat flow control method for a server heat dissipation module as described in the first aspect, comprising: a heat dissipation loop set extraction module, used to access the heat dissipation loop resource manager of the heat dissipation module and extract the heat dissipation loop set in the heat dissipation loop resource manager; a loop table generation module, used to obtain multiple partitions of the server and generate a start-up heat dissipation loop table and a fallback heat dissipation loop table for each partition based on the real-time heat dissipation demand information of each partition and the heat dissipation loop set; a functional equivalence analysis module, used to obtain a newly added first heat dissipation loop to be started when the start-up heat dissipation loop table is triggered to be updated, perform functional equivalence analysis on the first heat dissipation loop to be started and the heat dissipation loops in the fallback heat dissipation loop table, and output a functional equivalence score result; and a control command issuance module, used to output a heat dissipation loop execution decision result based on the functional equivalence score result, and issue control commands for the corresponding partitions to the heat dissipation module according to the heat dissipation loop execution decision result, wherein the heat dissipation loop execution result includes reusing the fallback heat dissipation loop and starting the newly added heat dissipation loop.
[0017] The technical solution provided in this application has at least the following technical effects or advantages: by achieving the technical goal of intelligent selection and dynamic control of heat dissipation circuits based on equivalence scoring, it achieves the technical effects of reducing the switching frequency of heat dissipation circuits, reducing energy consumption, and improving the stability of system operation.
[0018] The above description is merely an overview of the technical solution of this application. To enable a clearer understanding of the technical means of this application and to facilitate its implementation according to the description, and to make the above and other objects, features, and advantages of this application more apparent, specific embodiments of this application are described below. It should be understood that the content described in this section is not intended to identify key or important features of the embodiments of this application, nor is it intended to limit the scope of this application. Other features of this application will become readily apparent through the following description. Attached Figure Description
[0019] To more clearly illustrate the technical solutions in this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are merely exemplary. For those skilled in the art, other drawings can be obtained based on the provided drawings without creative effort.
[0020] Figure 1 This is a flowchart illustrating a partitioned heat flow control method for a server heat dissipation module according to this application.
[0021] Figure 2 This is a schematic diagram of the partitioned heat flow control system for a server heat dissipation module according to this application.
[0022] Figure labeling: 1. Heat dissipation loop set extraction module; 2. Loop table generation module; 3. Functional equivalence analysis module; 4. Control command issuance module. Detailed Implementation
[0023] This application provides a partitioned heat flow control method and system for a server heat dissipation module, solving the technical problem in existing technologies where the lack of detailed analysis of the functional equivalence and performance differences between different heat dissipation loops leads to the frequent activation of new heat dissipation loops even when existing fallback loops can meet the requirements through parameter adjustments, further affecting the system's energy utilization efficiency and operational stability. It achieves the technical goal of intelligent selection and dynamic control of heat dissipation loops based on equivalence scoring, thereby reducing the frequency of heat dissipation loop switching, reducing energy consumption, and improving system operational stability.
[0024] The technical solutions of this application will now be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of them. It should be understood that this application is not limited to the exemplary embodiments described herein. All other embodiments obtained by those skilled in the art based on the embodiments of this application without creative effort are within the scope of protection of this application. It should also be noted that, for ease of description, only the parts related to this application are shown in the accompanying drawings, not all of them.
[0025] Example 1, please refer to the appendix. Figure 1 This application provides a partitioned heat flow control method for a server heat dissipation module, applied to a partitioned heat flow control system for a server heat dissipation module, specifically including the following steps: S1: Access the heat dissipation circuit resource manager of the heat dissipation module and extract the heat dissipation circuit set in the heat dissipation circuit resource manager.
[0026] Specifically, the server's heat dissipation module is connected to the management system of the heat dissipation loop resource manager. The heat dissipation module is a hardware component responsible for the heat release of different partitions inside the server, while the heat dissipation loop resource manager is a controller used to manage and schedule these heat dissipation loops, and can record the status and characteristics of each heat dissipation loop.
[0027] Then, all available heat dissipation loops are retrieved from the heat dissipation loop resource manager to form a heat dissipation loop set, which includes the identifier, type, connection method and capability parameters of each loop, for subsequent selection of appropriate loops based on the heat load of different zones.
[0028] S2: Obtain multiple partitions of the server, and generate a startup cooling loop table and a fallback cooling loop table for each partition based on the real-time heat dissipation requirements of each partition and the set of cooling loops.
[0029] Specifically, the process identifies multiple partitions within the server. These partitions can be independent areas divided according to functional modules, rack locations, or hardware types. For example, a server might be divided into processor, storage, and power supply areas, each with different heat generation characteristics during operation. Next, real-time cooling requirements for each partition are used, including the current required cooling volume, cooling rate, and allowable response time, to measure the intensity of the partition's current cooling demand. Simultaneously, data from a set of available cooling loops is used to generate a startup cooling loop table and a fallback cooling loop table for each partition. The startup cooling loop table lists loops that should currently be used to meet cooling requirements. The fallback cooling loop table records loops that can be used as backups, replacements, or deactivated at this stage, allowing for flexible switching during load changes or system adjustments.
[0030] S3: When the start-up heat dissipation circuit table is triggered to be updated, the newly added first heat dissipation circuit to be started is obtained, and the first heat dissipation circuit to be started and the heat dissipation circuit in the rollback heat dissipation circuit table are subjected to functional equivalence analysis, and the functional equivalence score result is output.
[0031] Specifically, when the activation of the cooling circuit table triggers an update, the existing cooling circuit combination can no longer fully meet the cooling needs of the current partition, requiring adjustment or the addition of new cooling circuits. The update trigger condition may arise from increased load, temperature changes, or a failure of an existing circuit. Subsequently, the first newly added cooling circuit to be activated is identified; it may be an air-cooled circuit, a carrier-cooled circuit, or a refrigeration circuit, and its performance parameters need to be accurately recorded. Next, a functional equivalence analysis is performed between the first activated cooling circuit and the cooling circuits in the rollback cooling circuit table to compare the similarity of the two circuits in terms of cooling type, coverage area, and response speed, thereby determining their substitutability. Finally, the functional equivalence score is output.
[0032] S4: Based on the functional equivalence score, output the heat dissipation loop execution decision result, and issue the corresponding partition control command to the heat dissipation module according to the heat dissipation loop execution decision result. The heat dissipation loop execution result includes reusing the back-off heat dissipation loop and starting the new heat dissipation loop.
[0033] Specifically, the functional equivalence score outputs the execution decision results of the heat dissipation loop, thereby determining the specific operation plan. The functional equivalence score results integrate factors such as loop type difference indicators, zone coverage overlap, and response time differences, resulting in a clear action plan. Next, according to the heat dissipation loop execution decision results, control commands for the corresponding zones are issued to the heat dissipation module. The heat dissipation module is the hardware unit actually responsible for heat dissipation. Upon receiving the control commands, it adjusts the fan speed, coolant flow, or starts the compressor, thereby controlling the temperature of the target zone.
[0034] The execution results of the heat dissipation loop include reusing the backed-up heat dissipation loop and starting a new heat dissipation loop. Reusing the backed-up heat dissipation loop means directly using the existing backup loop without additional hardware startup, which is fast and energy-efficient. Starting a new heat dissipation loop means activating a loop that was not previously running, which is used when the existing loop cannot meet the demand. Although it can increase the heat dissipation capacity, the startup time and energy consumption will be higher.
[0035] Furthermore, this application also includes: extracting the heat dissipation circuit set in the heat dissipation circuit resource manager, wherein the heat dissipation circuit resource manager stores the resource description semantic vector of each heat dissipation circuit; wherein the resource description semantic vector includes circuit ID, circuit type, circuit connection relationship and capability parameters, and the circuit type of the heat dissipation module includes air-cooled circuit, cooling circuit and refrigeration circuit.
[0036] Specifically, information on all available heat dissipation loops is obtained from the heat dissipation loop resource manager. The heat dissipation loop resource manager not only records whether each heat dissipation loop is available, but also saves detailed characteristic information of each heat dissipation loop, providing a basis for subsequent partitioned heat flow control.
[0037] Then, each heat dissipation loop is represented in the resource manager using a resource description semantic vector. The resource description semantic vector encodes the structured information of the heat dissipation loop, including the loop ID, loop type, loop connection relationship, and capability parameters. The loop ID is a unique identifier for the heat dissipation loop, the loop type indicates whether the heat dissipation loop dissipates heat through air, liquid, or refrigerant, the loop connection relationship describes the topology of the heat dissipation loop in the heat dissipation module, and the capability parameters represent the heat dissipation capacity of the heat dissipation loop, such as maximum cooling capacity or flow rate.
[0038] Next, the main types of circuits in the heat dissipation module are air-cooled circuits, cooling circuits, and refrigerant circuits. Air-cooled circuits use fans or airflow to remove heat, cooling circuits transfer heat through circulating liquid, and refrigerant circuits rely on refrigerant to lower the temperature. Table 1 shows the semantic vector data table for heat dissipation circuit resource descriptions.
[0039] Table 1: Semantic Vector Data Table for Heat Dissipation Loop Resources Loop ID circuit type loop connection relationship Capability parameters CL001 air-cooled circuit Partition A → Heat Dissipation Module 1 Maximum air volume 3000 cubic meters per hour, temperature difference reduced by 8 degrees Celsius. CL002 air-cooled circuit Partition B → Heat dissipation module 1 Maximum air volume 2800 cubic meters per hour, temperature difference reduced by 7 degrees Celsius CL003 Cooling circuit Partition A → Partition C → Heat Dissipation Module 2 With a coolant flow rate of 60 liters per minute, the temperature difference decreases by 12 degrees Celsius. CL004 Refrigeration circuit Partition D → Heat dissipation module 3 Cooling power of 5 kW, temperature difference reduced by 18 degrees Celsius CL005 Cooling circuit Partition B → Partition E → Heat Dissipation Module 2 With a coolant flow rate of 55 liters per minute, the temperature difference decreases by 10 degrees Celsius. CL006 air-cooled circuit Partition E → Heat Dissipation Module 1 Maximum air volume 2500 cubic meters per hour, temperature difference reduced by 6 degrees Celsius CL007 Refrigeration circuit Partition C → Heat dissipation module 3 Cooling power 4.5 kW, temperature difference reduced by 16 degrees Celsius. Furthermore, this application also includes: recording real-time operating condition monitoring data of each partition; extracting key features of heat dissipation impact from the real-time operating condition monitoring data, wherein the key features of heat dissipation impact include heat accumulation index, temperature gradient and load power consumption; performing heat dissipation demand quantification analysis on the key features of heat dissipation impact through a thermal resistance network model, and outputting real-time heat dissipation demand information for each partition, including heat dissipation quantity demand information, heat dissipation rate demand information and heat dissipation response delay demand information.
[0040] Specifically, the operating status of each partition in the server is monitored in real time to obtain real-time operating condition monitoring data. Each partition may have different computing loads and heat generation conditions. The real-time operating condition monitoring data includes dynamic data that affects heat dissipation, such as temperature, current, fan speed, and power consumption, which can reflect the current thermal environment of the partition.
[0041] Next, key features that directly affect heat dissipation are extracted from real-time operating condition monitoring data, including the heat accumulation index, temperature gradient, and load power consumption. The heat accumulation index indicates the degree of heat accumulation in the partition; the higher the value, the less easily the heat dissipates. The temperature gradient refers to the temperature difference at different locations within the partition, reflecting the driving force of heat transfer. Load power consumption is the electrical power consumed by the equipment within the partition.
[0042] Then, a thermal resistance network model is used to quantify the heat dissipation demand by analyzing the key characteristics affecting heat dissipation. The thermal resistance network model is a method that uses a circuit model to analogize heat exchange processes such as heat conduction, convection, and radiation, deriving the heat flow distribution through the calculation of nodes and thermal resistance values. The analysis results in the transformation of heat-influencing factors into quantifiable heat dissipation demands. The output heat dissipation demand information includes heat dissipation quantity demand information, heat dissipation rate demand information, and heat dissipation response delay demand information. The heat dissipation quantity demand represents the total heat to be removed, the heat dissipation rate demand represents how much heat needs to be removed per second, and the heat dissipation response delay demand represents the time required to complete effective heat dissipation.
[0043] Furthermore, this application also includes: obtaining the historical heat dissipation startup status of each partition; determining whether the current partition is starting heat dissipation for the first time based on the historical heat dissipation startup status of each partition; if the current partition is starting heat dissipation for the first time, obtaining the set of optional heat dissipation circuits for the current partition based on the usage tag of each heat dissipation circuit in the heat dissipation circuit set, performing a heat dissipation demand matching analysis with the real-time heat dissipation demand information of the current partition in the set of optional heat dissipation circuits, and outputting the matching heat dissipation circuits; constructing a startup heat dissipation circuit table based on the matching heat dissipation circuits, at which time the fallback heat dissipation circuit table is empty.
[0044] Specifically, the historical cooling startup status of each partition in the server is obtained. The historical cooling startup status refers to whether the partition has started cooling in the past operation, the startup time, the duration, and the cooling loop used. It can be obtained from the system operation log or the cooling management module and is used to determine the cooling history of the partition at the current moment.
[0045] Next, the system determines whether a partition is starting its cooling system for the first time based on its historical cooling activation status. If a partition has never used cooling before, it is considered to be starting its cooling system for the first time. If a partition has previously started its cooling system, it is considered to be starting its cooling system for the first time. This determination affects the subsequent loop selection strategy. During a first-time startup, the system lacks experience with the cooling operation of the current partition and requires a completely new loop to be allocated. For non-first-time startups, optimizations and adjustments can be made based on existing loops.
[0046] Then, if this is the first time the partition is being cooled, a set of available cooling loops for that partition needs to be selected based on the usage flag of each loop in the cooling loop set. The usage flag records whether each loop is currently occupied or available, ensuring that loops that conflict with multiple partitions are not assigned. Within the set of available cooling loops, a cooling demand matching analysis needs to be performed against the partition's real-time cooling requirements. This analysis compares the cooling capacity and response speed of the loops with the partition's heat requirements to select loops that effectively meet those requirements as the matching cooling loops.
[0047] Finally, a startup cooling loop table is constructed based on the selected matching cooling loops. This startup cooling loop table is a list of cooling loops that a partition can immediately activate, used to issue startup commands. In the case of initial cooling startup, the rollback cooling loop table is empty because the partition previously had no running cooling loops, thus eliminating the possibility of rollback and reuse.
[0048] Furthermore, this application also includes: if the current partition is not the first time to start heat dissipation, calculate the heat dissipation demand gap based on the real-time heat dissipation demand information of the current partition and the current start heat dissipation loop table; perform heat dissipation demand gap matching analysis on the available heat dissipation loop set and output the matching gap heat dissipation loop; at this time, the start heat dissipation loop table is triggered to update, and the matching gap heat dissipation loop is added to the start heat dissipation loop table as the newly added first heat dissipation loop to be started.
[0049] Specifically, if the current partition is not starting cooling for the first time, it is necessary to calculate the cooling demand gap by combining the real-time cooling demand information of the current partition with the existing startup cooling loop table. The real-time cooling demand information includes indicators such as the heat that the partition needs to remove at this moment, the cooling speed, and the response time. The startup cooling loop table is a list of cooling loops that are currently running. The cooling demand gap refers to the difference between the cooling capacity that the existing loops can provide and the cooling capacity that the partition currently needs, reflecting the additional cooling capacity that needs to be supplemented.
[0050] Next, based on the heat dissipation demand gap, a matching analysis is performed on the set of available heat dissipation loops. The set of available heat dissipation loops refers to the set of loops that are currently unoccupied and can be allocated to the partition. The heat dissipation demand gap matching analysis compares whether the capability parameters, types and response characteristics of the available heat dissipation loops can make up for the gap, and finally selects the heat dissipation loops that match the gap.
[0051] Then, since a heat dissipation loop that can meet the gap has been found, the heat dissipation loop table needs to be updated. The heat dissipation loop that matches the gap is added to the existing heat dissipation loop table. The heat dissipation loop table then includes the heat dissipation loops that were originally running and the newly added heat dissipation loops. The newly added heat dissipation loops will be marked as the first heat dissipation loop to be started, indicating that they will be the new start objects to be executed in the next step.
[0052] Furthermore, this application also includes: performing redundancy analysis on the heat dissipation loops in the current startup heat dissipation loop table based on the real-time heat dissipation demand information of the current partition, and obtaining a redundancy index; extracting heat dissipation loops with a redundancy greater than a preset redundancy threshold as fallback heat dissipation loops, and constructing a fallback heat dissipation loop table.
[0053] Specifically, based on the real-time heat dissipation requirements of the current partition, a redundancy analysis is performed on the currently running heat dissipation loops in the currently activated heat dissipation loop table. The real-time heat dissipation requirements include the partition's current heat dissipation volume, heat dissipation rate, and response time. The currently activated heat dissipation loop table records all currently operational heat dissipation loops. Redundancy analysis is used to assess whether there is any overcapacity in the heat dissipation loops. The redundancy index is a numerical value used to quantify the degree to which a loop exceeds actual demand.
[0054] Next, cooling loops with redundancy levels exceeding a pre-set redundancy threshold are selected from the redundancy indicators. These loops can then be replaced or shut down during current operation to save resources and serve as fallback cooling loops. The pre-set redundancy threshold is a pre-defined standard value. Subsequently, the fallback cooling loops are recorded to form a fallback cooling loop table, which is used for subsequent functional comparisons with newly selected loops to be started.
[0055] Furthermore, this application also includes: extracting the heat dissipation characteristic index of the first heat dissipation circuit to be started; performing functional equivalence analysis on the heat dissipation circuits in the fallback heat dissipation circuit table according to the heat dissipation characteristic index of the first heat dissipation circuit to be started, and obtaining the circuit type difference index, the partition coverage overlap degree and the response time difference; calculating the circuit type difference index, the partition coverage overlap degree and the response time difference, and outputting the functional equivalence score result.
[0056] Specifically, the heat dissipation characteristic indicators of the first heat dissipation circuit to be started are extracted, and its performance-related parameters are obtained from the new heat dissipation circuit that is about to be put into operation. These parameters may include heat dissipation capacity, cooling medium type, heat transfer efficiency, and response speed, etc., to accurately describe the working characteristics of the circuit.
[0057] Next, based on the heat dissipation characteristic index of the first heat dissipation circuit to be started, a functional equivalence analysis is performed on each heat dissipation circuit in the fallback heat dissipation circuit table to compare the degree of functional similarity between the new circuit and the backup circuit. Among them, the circuit type difference index is used to measure the difference in cooling methods, for example, the difference between air cooling and liquid cooling is large; the zone coverage overlap is used to measure whether the two circuits can provide heat dissipation for the same zone. If the coverage area is completely consistent, the overlap is 100%; the response time difference indicates the difference in the time required for the two to reach an effective heat dissipation state from startup.
[0058] Finally, by calculating the loop type difference index, partition coverage overlap, and response time difference, multiple analysis results are combined into a functional equivalence score. The higher the functional equivalence score, the more the heat dissipation fallback loop can replace the heat dissipation loop to be started, and thus the functional equivalence score result is output.
[0059] Furthermore, this application also includes: if the heat dissipation circuit with a value greater than a preset equivalence threshold in the equivalence score results returns empty, the newly added first heat dissipation circuit to be started is started, and the heat dissipation circuit execution decision result of starting the newly added heat dissipation circuit is output; if the heat dissipation circuit with a value greater than a preset equivalence threshold in the equivalence score results returns not empty, the heat dissipation circuit with a value greater than the preset equivalence threshold is selected and marked for reuse, and the heat dissipation circuit execution result of reusing the fallback heat dissipation circuit is obtained.
[0060] Specifically, when the equivalence score for a heat dissipation loop exceeding a preset equivalence threshold returns empty, the functional equivalence analysis finds no fallback loop with a performance similarity higher than the threshold to the first heat dissipation loop to be activated. The preset equivalence threshold is a score line used to determine the substitutability of a loop, for example, set to 80 points. Only loops with a score higher than 80 points are considered capable of replacing the new loop. If no heat dissipation loop is found, the newly added first heat dissipation loop to be activated is directly activated, and the corresponding heat dissipation loop execution decision result is output, recording information such as the activated loop number, type, and target partition.
[0061] Next, when a heat dissipation loop with an equivalence score greater than the preset equivalence threshold returns a non-empty result, it indicates that a fallback loop with sufficiently high performance similarity exists to replace the new loop. At this point, all heat dissipation loops with scores exceeding the threshold are selected from the newly added heat dissipation loops and marked for reuse, indicating that this heat dissipation loop will be directly put back into operation instead of starting a completely new loop. Finally, the execution result of a reused fallback heat dissipation loop is obtained, recording the reused loop's number, operating parameters, and target partition, thereby achieving efficient resource reuse.
[0062] Furthermore, this application also includes: if the heat dissipation circuit execution result outputs the result of multiplexing the back-off heat dissipation circuit, obtaining the multiplexing performance difference between the newly added first heat dissipation circuit to be started and the back-off heat dissipation circuit; adjusting the heat dissipation parameters of the back-off heat dissipation circuit according to the multiplexing performance difference, including the circuit cooling temperature and the circuit flow rate.
[0063] Specifically, when the output of the heat dissipation loop execution result is a reuse of the fallback heat dissipation loop, after equivalence analysis, it is chosen not to start a new loop, but to let the existing fallback loop continue to run. At this time, it is necessary to obtain the reuse performance difference between the newly added first heat dissipation loop to be started and the selected fallback heat dissipation loop to quantify their performance differences in terms of heat dissipation capacity, response speed, and cooling medium efficiency.
[0064] Next, the heat dissipation parameters of the return heat dissipation loop are adjusted according to the difference in reuse performance. The loop cooling temperature refers to the temperature of the cooling medium when it enters the heat dissipation target. Lowering the temperature can improve the efficiency of heat removal. The loop flow rate refers to the flow rate of the cooling medium per unit time. Increasing the flow rate can enhance the heat exchange effect and ensure that the heat dissipation effect meets the expected standard.
[0065] In summary, the partitioned heat flow control method for a server heat dissipation module provided in this application has the following technical effects: by achieving the technical goal of intelligent selection and dynamic control of heat dissipation loops based on equivalence scoring, it achieves the technical effects of reducing the switching frequency of heat dissipation loops, reducing energy consumption, and improving the stability of system operation.
[0066] Example 2: Based on the same inventive concept as the partitioned heat flow control method for a server heat dissipation module in the foregoing examples, this application also provides a partitioned heat flow control system for a server heat dissipation module. Please refer to the appendix. Figure 2The system includes: a heat dissipation loop set extraction module 1, used to access the heat dissipation loop resource manager of the heat dissipation module and extract the heat dissipation loop set in the heat dissipation loop resource manager; a loop table generation module 2, used to obtain multiple partitions of the server and generate a start heat dissipation loop table and a fallback heat dissipation loop table for each partition based on the real-time heat dissipation demand information of each partition and the heat dissipation loop set; a functional equivalence analysis module 3, used to obtain a newly added first heat dissipation loop to be started when the start heat dissipation loop table is triggered to be updated, perform functional equivalence analysis on the first heat dissipation loop to be started and the heat dissipation loops in the fallback heat dissipation loop table, and output a functional equivalence score result; and a control command issuance module 4, used to output a heat dissipation loop execution decision result based on the functional equivalence score result, and issue control commands for the corresponding partitions to the heat dissipation module according to the heat dissipation loop execution decision result, wherein the heat dissipation loop execution result includes reusing the fallback heat dissipation loop and starting the newly added heat dissipation loop.
[0067] Furthermore, the partitioned heat flow control system of the server heat dissipation module is also used to: extract the heat dissipation loop set in the heat dissipation loop resource manager, wherein the heat dissipation loop resource manager stores the resource description semantic vector of each heat dissipation loop; wherein the resource description semantic vector includes loop ID, loop type, loop connection relationship and capability parameters, and the loop type of the heat dissipation module includes air-cooled loop, cooling loop and refrigeration loop.
[0068] Furthermore, the partitioned heat flow control system of the server heat dissipation module is also used to: record real-time operating condition monitoring data of each partition; extract key features of heat dissipation impact from the real-time operating condition monitoring data, including heat accumulation index, temperature gradient, and load power consumption; perform heat dissipation demand quantification analysis on the key features of heat dissipation impact through a thermal resistance network model, and output real-time heat dissipation demand information for each partition, including heat dissipation quantity demand information, heat dissipation rate demand information, and heat dissipation response delay demand information.
[0069] Furthermore, the partitioned heat flow control system of the server heat dissipation module is also used for: acquiring the historical heat dissipation start-up status of each partition; determining whether the current partition is starting heat dissipation for the first time based on the historical heat dissipation start-up status of each partition; if the current partition is starting heat dissipation for the first time, acquiring the set of optional heat dissipation loops for the current partition based on the usage tag of each heat dissipation loop in the heat dissipation loop set, performing heat dissipation demand matching analysis with the real-time heat dissipation demand information of the current partition in the set of optional heat dissipation loops, and outputting the matching heat dissipation loops; constructing a start-up heat dissipation loop table based on the matching heat dissipation loops, at which time the fallback heat dissipation loop table is empty.
[0070] Furthermore, the partitioned heat flow control system of the server heat dissipation module is also used for: if the current partition is not the first time to start heat dissipation, calculating the heat dissipation demand gap based on the real-time heat dissipation demand information of the current partition and the current start heat dissipation loop table; based on the heat dissipation demand gap, performing a heat dissipation demand gap matching analysis with the heat dissipation demand gap in the set of optional heat dissipation loops, and outputting a matching gap heat dissipation loop; at this time, the start heat dissipation loop table is triggered to update, and the matching gap heat dissipation loop is added to the start heat dissipation loop table as a newly added first heat dissipation loop to be started.
[0071] Furthermore, the partitioned heat flow control system of the server heat dissipation module is also used to: perform redundancy analysis on the heat dissipation loops in the current activated heat dissipation loop table based on the real-time heat dissipation demand information of the current partition, and obtain the redundancy index; extract heat dissipation loops with a redundancy greater than a preset redundancy threshold as fallback heat dissipation loops, and construct a fallback heat dissipation loop table.
[0072] Furthermore, the partitioned heat flow control system of the server heat dissipation module is also used to: extract the heat dissipation characteristic index of the first heat dissipation circuit to be started; perform functional equivalence analysis on the heat dissipation circuits in the fallback heat dissipation circuit table according to the heat dissipation characteristic index of the first heat dissipation circuit to be started, and obtain the circuit type difference index, partition coverage overlap and response time difference; calculate the circuit type difference index, partition coverage overlap and response time difference, and output the functional equivalence score result.
[0073] Furthermore, the partitioned heat flow control system of the server heat dissipation module is also used for: if the heat dissipation loops with a value greater than a preset equivalence threshold in the equivalence score results return empty, starting the newly added first heat dissipation loop to be started, and outputting the heat dissipation loop execution decision result for starting the newly added heat dissipation loop; if the heat dissipation loops with a value greater than the preset equivalence threshold in the equivalence score results return not empty, filtering the heat dissipation loops with a value greater than the preset equivalence threshold for reuse marking, and obtaining the heat dissipation loop execution result for reused and rolled-back heat dissipation loops.
[0074] Furthermore, the partitioned heat flow control system of the server heat dissipation module is also used for: if the heat dissipation loop execution result outputs the result of multiplexing the back-off heat dissipation loop, obtaining the multiplexing performance difference between the newly added first heat dissipation loop to be started and the back-off heat dissipation loop; adjusting the heat dissipation parameters of the back-off heat dissipation loop according to the multiplexing performance difference, including the loop cooling temperature and the loop flow rate.
[0075] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on the differences from other embodiments. The partitioned heat flow control method and specific examples of a server heat dissipation module in the foregoing embodiment one are also applicable to the partitioned heat flow control system of a server heat dissipation module in this embodiment. Through the foregoing detailed description of the partitioned heat flow control method of a server heat dissipation module, those skilled in the art can clearly understand the partitioned heat flow control system of a server heat dissipation module in this embodiment. Therefore, for the sake of brevity, it will not be described in detail here.
[0076] The above description of the disclosed embodiments enables those skilled in the art to make or use this application. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of this application. Therefore, this application is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.
[0077] Obviously, those skilled in the art can make various modifications and variations to this application without departing from the spirit and scope of this application. Therefore, if such modifications and variations fall within the scope of this application and its equivalents, this application also intends to include such modifications and variations.
Claims
1. A method for zoned heat flow control in a server heat dissipation module, characterized in that, The method comprises: Accessing a heat dissipation loop resource manager of a heat dissipation module, extracting a heat dissipation loop set in the heat dissipation loop resource manager; Obtaining multiple partitions of a server, generating a start heat dissipation loop table and a fallback heat dissipation loop table of each partition according to real-time heat dissipation demand information of each partition and the heat dissipation loop set; When the start heat dissipation loop table triggers an update, obtaining a first to-be-started heat dissipation loop added, performing functional equivalence analysis on the first to-be-started heat dissipation loop and heat dissipation loops in the fallback heat dissipation loop table, and outputting a functional equivalence score result; Based on the functional equivalence score result, outputting a heat dissipation loop execution decision result, and issuing a control instruction of a corresponding partition to the heat dissipation module according to the heat dissipation loop execution decision result, wherein the heat dissipation loop execution result comprises multiplexing a fallback heat dissipation loop and starting a newly added heat dissipation loop.
2. The method of claim 1, wherein the server heat dissipation module is divided into a plurality of zones, and the heat flow control is performed in each of the zones. Extracting a heat dissipation loop set in the heat dissipation loop resource manager, wherein the heat dissipation loop resource manager stores a resource description semantic vector of each heat dissipation loop; The resource description semantic vector comprises a loop ID, a loop type, a loop connection relationship, and a capability parameter, and the loop type of the heat dissipation module comprises an air cooling loop, a cold carrier loop, and a refrigeration loop.
3. The method of claim 1, wherein the server heat dissipation module is divided into a plurality of zones, and the heat flow control is performed in each of the zones. The real-time heat dissipation demand information of each partition is obtained by monitoring the operating condition data of each partition, and the method comprises: Recording real-time operating condition monitoring data of each partition; Extracting heat dissipation influence key features of the real-time operating condition monitoring data, wherein the heat dissipation influence key features comprise a heat accumulation index, a temperature gradient, and a load power consumption; Performing heat dissipation demand quantification analysis on the heat dissipation influence key features through a thermal resistance network model, and outputting real-time heat dissipation demand information of each partition, including heat dissipation amount demand information, heat dissipation rate demand information, and heat dissipation response delay demand information.
4. The method of claim 3, wherein the server heat dissipation module is divided into a plurality of zones, and the heat flow control is performed in each of the zones. Generating a start heat dissipation loop table and a fallback heat dissipation loop table of each partition according to real-time heat dissipation demand information of each partition and the heat dissipation loop set, and the method comprises: Obtaining a historical heat dissipation start state of each partition; Judging whether the current partition is starting heat dissipation for the first time according to the historical heat dissipation start state of each partition; If the current partition is starting heat dissipation for the first time, obtaining a selectable heat dissipation loop set of the current partition based on a use mark of each heat dissipation loop in the heat dissipation loop set, performing heat dissipation demand matching analysis on the real-time heat dissipation demand information of the current partition in the selectable heat dissipation loop set, and outputting a matching heat dissipation loop; Constructing a start heat dissipation loop table according to the matching heat dissipation loop, and the fallback heat dissipation loop table is empty at this time.
5. The partitioned heat flow control method for a server heat dissipation module as described in claim 4, characterized in that, Judging whether the current partition is starting heat dissipation for the first time according to the historical heat dissipation start state of each partition, and the method further comprises: If the current partition is not starting heat dissipation for the first time, calculating a heat dissipation demand gap according to the real-time heat dissipation demand information of the current partition and the current start heat dissipation loop table; Performing heat dissipation demand gap matching analysis on the heat dissipation demand gap in the selectable heat dissipation loop set, and outputting a matching gap heat dissipation loop; At this time, the starting heat dissipation loop table trigger updates, and the matching gap heat dissipation loop is added to the starting heat dissipation loop table as a newly added first to-be-started heat dissipation loop.
6. The partitioned heat flow control method for a server heat dissipation module as described in claim 5, characterized in that, If the current partition is not the first time to start heat dissipation, the method further comprises: If the redundancy of the heat dissipation loop in the current starting heat dissipation loop table is analyzed according to the real-time heat dissipation demand information of the current partition, a redundancy index is obtained. If the redundancy of the heat dissipation loop in the current starting heat dissipation loop table is analyzed according to the real-time heat dissipation demand information of the current partition, a redundancy index is obtained.
7. The partitioned heat flow control method for a server heat dissipation module as described in claim 1, characterized in that, Extract the heat dissipation loop with a greater than preset redundancy threshold as a fallback heat dissipation loop, and construct a fallback heat dissipation loop table. The first to-be-started heat dissipation loop and the heat dissipation loop in the fallback heat dissipation loop table are functionally equivalent, and a functionally equivalent score result is output, the method comprising: Extracting the heat dissipation characteristic index of the first to-be-started heat dissipation loop; According to the heat dissipation characteristic index of the first to-be-started heat dissipation loop, the functionally equivalent analysis of the heat dissipation loop in the fallback heat dissipation loop table is carried out, and the loop type difference index, the partition coverage coincidence degree and the response time difference are obtained.
8. The partitioned heat flow control method for a server heat dissipation module as described in claim 1, characterized in that, The loop type difference index, the partition coverage coincidence degree and the response time difference are calculated, and the functionally equivalent score result is output. Based on the equivalent score result, a heat dissipation loop execution decision result is output, the method comprising: If the heat dissipation loop greater than the preset equivalent threshold in the equivalent score result returns empty, start the newly added first to-be-started heat dissipation loop, and output the heat dissipation loop execution decision result of starting the newly added heat dissipation loop.
9. The partitioned heat flow control method for a server heat dissipation module as described in claim 1, characterized in that, If the heat dissipation loop greater than the preset equivalent threshold in the equivalent score result returns not empty, filter the heat dissipation loop greater than the preset equivalent threshold for reuse marking, and obtain the heat dissipation loop execution result of the reuse fallback heat dissipation loop. If the heat dissipation loop execution result outputs the result of the reuse fallback heat dissipation loop, the reuse performance difference degree of the newly added first to-be-started heat dissipation loop and the fallback heat dissipation loop is obtained.
10. A zoned heat flow control system for a server heat dissipation module, characterized in that, According to the reuse performance difference degree, the heat dissipation parameters of the fallback heat dissipation loop are adjusted, including the loop cooling temperature and the loop flow. The steps for implementing the partition heat flow control method of the server heat dissipation module according to any one of claims 1-9, comprising: A heat dissipation loop set extraction module is used to access the heat dissipation loop resource manager of the heat dissipation module, and extract the heat dissipation loop set in the heat dissipation loop resource manager; A loop table generation module is used to obtain a plurality of partitions of a server, and generate a starting heat dissipation loop table and a fallback heat dissipation loop table for each partition according to the real-time heat dissipation demand information of each partition and the heat dissipation loop set; A functionally equivalent analysis module is used to obtain a newly added first to-be-started heat dissipation loop when the starting heat dissipation loop table triggers updates, and perform functionally equivalent analysis on the first to-be-started heat dissipation loop and the heat dissipation loop in the fallback heat dissipation loop table, and output a functionally equivalent score result; A control instruction issuing module is used to output a heat dissipation loop execution decision result based on the functionally equivalent score result, and issue a control instruction corresponding to the partition to the heat dissipation module according to the heat dissipation loop execution decision result, wherein the heat dissipation loop execution result comprises a reuse fallback heat dissipation loop and a starting newly added heat dissipation loop.