Server heat dissipation intelligent regulation system

Abstract

The application discloses a server heat dissipation intelligent regulation and control system, which comprises: a plurality of modular heat conduction patch units, each unit comprising a flexible heat conduction base, an internal micro flow channel, a temperature sensor and a quick connector, and being used for being attached to different heat sources; a cold source and a distribution unit, which are connected with each patch unit through a branch pipe provided with an independent flow control valve; and an intelligent control system, wherein a central controller independently adjusts the opening degree of a corresponding flow valve according to the deviation of the real-time temperature of each heat source from a target value, so that precise and independent closed-loop temperature control is realized, and the use method is as follows: installing the patch and setting a temperature target, monitoring the temperature in real time, independently calculating and adjusting the flow of each branch, and coordinating the system energy efficiency. The application realizes flexible, precise and self-adaptive heat dissipation for diversified heat sources in the server, improves the heat dissipation efficiency and the system energy efficiency, and has the advantages of simple deployment and easy maintenance.

Description

Technical Field

[0001] This invention relates to the field of server heat dissipation technology, and in particular to an intelligent server heat dissipation control system. Background Technology

[0002] With the continuous increase in computing density in data centers, the power consumption and heat generation of key components such as CPUs, GPUs, and memory inside servers are increasing dramatically, and traditional air cooling methods are gradually approaching their heat dissipation limits. Liquid cooling technology has become the mainstream choice due to its higher heat dissipation efficiency. Existing liquid cooling solutions are mostly centralized cold plate designs for CPUs / GPUs, or immersion cooling of the entire server rack. These solutions have the following shortcomings: Poor flexibility: Cold plates are usually customized for specific hardware models, making it difficult to adapt to the diverse heat sources in the server (such as different models of memory, SSD, power supply modules, etc.). Hardware upgrades or replacements require redesigning the cold plate.

[0003] Lack of fine-grained control: Existing solutions mostly use a uniform flow supply, which cannot perform differentiated and precise heat dissipation control based on the real-time temperature of different heat source components in the server. This can easily lead to local overheating or overcooling, resulting in low energy efficiency.

[0004] Deployment and maintenance are complex: the installation of an integrated liquid cooling circuit is complicated, and a leak may affect the entire system, and troubleshooting is difficult.

[0005] Therefore, there is an urgent need for a server cooling solution that can flexibly adapt to various heat sources, achieve precise independent temperature control, and is easy to deploy and maintain. Summary of the Invention

[0006] The purpose of this application is to provide an intelligent server heat dissipation control system to solve the problems mentioned in the background art.

[0007] To achieve the above objectives, this application provides the following technical solution: a server heat dissipation intelligent control system, comprising several modular thermal conductive patch units, each unit including: Flexible thermally conductive substrate, whose shape and size are adapted to the surface of specific heat source components within the server; Microchannels embedded within the flexible thermally conductive substrate; A temperature sensor integrated on the thermally conductive patch unit for sensing the temperature of the heat source element to which it is attached; Quick connectors are provided on the thermally conductive patch unit; The cold source and distribution unit includes: Cold storage box; Connect the main liquid supply pipe and the main liquid return pipe of the cold source box; Multiple branch pipes branching off from the main liquid supply pipe are connected to the microchannel inlet of one of the modular thermal conductive patch units via the quick connector. A flow control valve installed on each of the branch pipes for regulating the flow rate; The intelligent control system includes a central controller, which is connected to all the temperature sensors and all the flow control valves. The central controller receives real-time temperature data from each temperature sensor and independently calculates and outputs control signals to the corresponding flow control valves based on the deviation between the real-time temperature of each heat source element and the preset temperature parameters, so as to dynamically adjust their opening degree.

[0008] Preferably, the flexible thermally conductive substrate is made of a highly thermally conductive flexible composite material, and its bottom surface is configured to fit tightly to the surface of the heat source element and can be torn off without damage.

[0009] Preferably, the quick connector is a self-sealing quick-connect connector.

[0010] Preferably, the flow control valve is a proportional regulating valve or a pulse width modulation valve.

[0011] Preferably, the system further includes a total flow sensor for monitoring the liquid output from the cold source box, the total flow sensor being connected to the central controller, and the cold source box being equipped with a refrigeration device for cooling its internal cold source.

[0012] Preferably, the control logic of the central controller is based on a PID control algorithm, fuzzy logic control, or a machine learning model.

[0013] Preferably, the system further includes a configuration management module for storing server hardware configurations and corresponding heat dissipation strategies, and for automatically applying the corresponding strategies when hardware replacement is detected.

[0014] Preferably, the modular thermal pad unit has different preset shapes and sizes depending on the type of heat source element it is adapted to, including large square pads for CPUs / GPUs, long strip pads for memory modules, and small pads for SSDs or power supply modules.

[0015] The intelligent control method for server heat dissipation includes the following steps: Install corresponding modular thermal pad units for each heat source component inside the server and connect them to the piping; Set temperature control targets for each heat source element; The central controller continuously reads data from all temperature sensors; For each heat source element, based on the deviation between its real-time temperature and the temperature control target, the opening of the flow control valve on the corresponding branch pipe is independently calculated and adjusted to achieve independent closed-loop temperature control for each heat source element. Coordinate the overall flow and pressure of the system to achieve overall energy efficiency optimization.

[0016] In summary, the technical effects and advantages of this invention are as follows: Standardized patch units are provided for heat sources of different shapes and sizes, enabling plug-and-play functionality and greatly simplifying the deployment of liquid cooling systems within complex servers. This also supports rapid hardware replacement and upgrades. Through integrated temperature sensors on each patch and independent flow valves on each branch, the system can perform independent closed-loop temperature control for each heat source, achieving on-demand cooling, avoiding insufficient or excessive cooling, and improving heat dissipation efficiency. The central controller employs advanced control algorithms, which not only respond quickly to temperature changes but also optimize control strategies by learning hardware load patterns. The configuration management module further enhances the system's automated management capabilities. The modular design localizes the impact of faults, ensuring that a fault in one patch unit does not affect other units. Quick-connect connectors facilitate rapid maintenance and replacement. By implementing refined and differentiated flow management and coordinating overall energy consumption, the system can significantly reduce the overall energy consumption of the cooling system while ensuring effective heat dissipation, thereby improving the data center's PUE value. Attached Figure Description

[0017] To more clearly illustrate the technical solutions in the embodiments of 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 only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0018] Figure 1 This is a schematic diagram of the allocation structure of the server heat dissipation intelligent control system in the embodiments of this application; Figure 2 This is a partial cross-sectional view of the flexible thermally conductive substrate of the server heat dissipation intelligent control system in this application embodiment; Figure 3 This is a schematic diagram of the control method of the server heat dissipation intelligent control system in the embodiments of this application.

[0019] In the diagram: 110-Flexible thermal conductive substrate, 111-Microchannel, 120-Temperature sensor, 130-Quick connector, 210-Cold source box, 220-Main liquid supply pipe, 230-Main liquid return pipe, 240-Branch pipe, 250-Flow control valve, 260-Total flow sensor, 310-Central controller. Detailed Implementation

[0020] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0021] In the description of this disclosure, it should be noted that the terms center, up, down, left, right, vertical, horizontal, inner, outer, top, bottom, etc., indicating orientation or positional relationships, are based on the orientation or positional relationships shown in the accompanying drawings and are only for the convenience of describing this disclosure and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of this disclosure. Furthermore, the terms first, second, and third are used for descriptive purposes only and should not be construed as indicating or implying relative importance.

[0022] It should also be noted that all standard parts used in this application are commercially available, and can be custom-made according to the description and drawings. Unless otherwise explicitly specified and limited, the terms installation, connection, and linking should be interpreted broadly, for example, as fixed connections, detachable connections, or integral connections; as mechanical connections or electrical connections; as direct connections or indirect connections through an intermediate medium; and as internal communication between two components. Those skilled in the art can understand the specific meaning of the above terms in this disclosure based on the specific circumstances, and unless explicitly limited, machinery, parts, and equipment can adopt conventional models in the prior art.

[0023] In this document, the term "include" is intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that includes a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitation, an element defined by the term "include" does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0024] Example 1: A server heat dissipation intelligent control system, comprising multiple modular thermally conductive patch units: each unit consists of a flexible thermally conductive substrate 110, the bottom surface of which is coated with a high-performance thermally conductive silicone grease layer to enhance contact with the heat source. The flexible thermally conductive substrate 110 has serpentine microchannels 111 etched inside. A high-precision miniature digital temperature sensor 120 is integrated on the side of the flexible thermally conductive substrate 110 near the heat source. A self-sealing quick-connect connector 130 is mounted on the edge of the flexible thermally conductive substrate 110, in fluid communication with the inlet and outlet of the microchannels 111.

[0025] Cooling source and distribution unit: Includes a cooling source tank 210 with an internal coolant storage compartment. The cooling source tank 210 integrates a compression refrigeration unit for continuous cooling of the circulating coolant. A main supply pipe 220 extends from the cooling source tank 210 and branches into multiple independent branch pipes 240 via multi-port connectors. Each branch pipe 240 is connected in series with an electronically controlled proportional regulating valve as a flow control valve 250. The ends of all branch pipes 240 are connected to quick-connect fittings 130 of the corresponding modular thermal conductive patch units. The coolant outlets of all patch units converge through piping to a main return pipe 230 and flow back to the cooling source tank 210. A total flow sensor 260 is installed on the main supply pipe 220 near the outlet of the cooling source tank 210 to monitor the total circulating flow of the system.

[0026] Intelligent control system: Its core is the central controller 310, which is connected to all temperature sensors 120, all flow control valves 250, and the total flow sensor 260 via signal lines. The central controller 310 has an embedded temperature control program based on the PID control algorithm and is equipped with a configuration management module. This module is used to store and manage the identity information, installation location, and corresponding preset temperature control targets of each hardware component in the server.

[0027] Operation Process: After the system powers on, the central controller 310 continuously polls and reads the real-time temperature data reported by each temperature sensor 120. For example, when the server GPU is under high load, and its corresponding temperature sensor 120 detects a temperature rise approaching a preset threshold, the central controller 310 calculates the deviation between the current temperature and the target value, and generates a control command accordingly. This command increases the opening of the flow control valve 250 on the branch pipe 240 leading to the GPU patch unit, thereby increasing the flow rate of coolant through its internal microchannels 111 and enhancing the GPU's cooling performance. Simultaneously, if the CPU temperature stabilizes near the target value, the opening of the flow control valve 250 on its corresponding branch pipe 240 remains stable; if the memory temperature is low, the opening of its corresponding valve may be reduced to minimize unnecessary cooling. Data from the total flow sensor 260 is used to monitor the overall system cooling load, allowing the central controller 310 to coordinate the power output of the cooling devices within the cold source box 210, achieving system-level energy efficiency optimization. When upgrading server hardware, you only need to replace or add the corresponding modular thermal pad unit and update the heat dissipation strategy of the hardware in the configuration management module.

[0028] Example 2 illustrates an intelligent control method based on the above system, such as... Figure 3 As shown, the specific steps are as follows: S1: Deployment and Initialization. Based on the model and location of each heat-generating component within the server, such as the CPU, GPU, and memory modules, select modular thermal pad units of corresponding shape and size. Attach their flexible thermally conductive substrates 110 to the component surfaces, and connect each unit to the corresponding branch pipes 240 of the cold source and distribution unit via quick-connect connectors 130. In the configuration management module of the central controller 310, set independent temperature control targets for each heat-generating component, for example: CPU core temperature 70℃, GPU core temperature 75℃, and memory temperature 65℃.

[0029] S2: Data Acquisition. After the system starts, the central controller 310 continuously reads the real-time temperature data of each heat source element collected by all temperature sensors 120 at a fixed sampling period.

[0030] S3: Independent Decision-Making and Control. For each heat source component, the central controller 310 compares its real-time temperature with the target temperature set in step S1 and calculates the temperature deviation. Subsequently, based on the built-in control algorithm, it independently calculates the required coolant flow rate adjustment for each component and generates a corresponding control signal, which is sent to the flow control valve 250 corresponding to that component to dynamically adjust its opening. For example, when a GPU's temperature rises to 78°C due to a surge in load, the central controller 310 will immediately increase the opening of the flow control valve 250 on its branch pipe 240 to improve cooling capacity.

[0031] S4: System Coordination and Energy Efficiency Optimization. While performing fine-grained control of the flow rate of each independent branch, the central controller 310 integrates the flow demand of all branch pipes 240, the total flow data fed back by the total flow sensor 260 on the main liquid supply pipe 220, and the liquid supply temperature of the cold source box 210 to perform global optimization control on the speed of the circulating pump and the operating power of the refrigeration unit in the cold source box 210, ensuring that the overall energy consumption of the entire cooling system is minimized while meeting all heat dissipation requirements.

[0032] Finally, it should be noted that the above are merely preferred embodiments of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A server heat dissipation intelligent control system, characterized in that: It includes several modular thermally conductive patch units, each unit comprising: Flexible thermally conductive substrate, whose shape and size are adapted to the surface of specific heat source components within the server; Microchannels embedded within the flexible thermally conductive substrate; A temperature sensor integrated on the thermally conductive patch unit for sensing the temperature of the heat source element to which it is attached; Quick connectors are provided on the thermally conductive patch unit; The cold source and distribution unit includes: Cold storage box; Connect the main liquid supply pipe and the main liquid return pipe of the cold source box; Multiple branch pipes branching off from the main liquid supply pipe are connected to the microchannel inlet of one of the modular thermally conductive patch units via the quick connector. A flow control valve installed on each of the branch pipes for regulating the flow rate; The intelligent control system includes a central controller, which is connected to all the temperature sensors and all the flow control valves. The central controller receives real-time temperature data from each temperature sensor and independently calculates and outputs control signals to the corresponding flow control valves based on the deviation between the real-time temperature of each heat source element and the preset temperature parameters, so as to dynamically adjust their opening degree.

2. The intelligent server heat dissipation control system according to claim 1, characterized in that, The flexible thermally conductive substrate is made of a highly thermally conductive flexible composite material, and its bottom surface is configured to fit tightly to the surface of the heat source element and can be torn off without damage.

3. The intelligent server heat dissipation control system according to claim 1, characterized in that, The quick connector is a self-sealing quick-connect connector.

4. The intelligent server heat dissipation control system according to claim 1, characterized in that, The flow control valve is a proportional regulating valve or a pulse width modulation valve.

5. The intelligent server heat dissipation control system according to claim 1, characterized in that, The system also includes a total flow sensor for monitoring the liquid output from the cold source box, the total flow sensor being connected to the central controller, and the cold source box being equipped with a refrigeration device for cooling its internal cold source.

6. The intelligent server heat dissipation control system according to claim 1, characterized in that, The control logic of the central controller is based on PID control algorithm, fuzzy logic control or machine learning model.

7. The intelligent server heat dissipation control system according to claim 1, characterized in that, The system also includes a configuration management module, which stores server hardware configurations and corresponding heat dissipation strategies, and can automatically apply the corresponding strategies when hardware replacement is detected.

8. The server heat dissipation intelligent control system according to any one of claims 1 to 7, characterized in that, The modular thermal pad unit has different preset shapes and sizes depending on the type of heat source component it is adapted to, including large square pads for CPUs / GPUs, long strip pads for memory modules, and small pads for SSDs or power supply modules.

9. A method for intelligent control of server heat dissipation based on the system according to any one of claims 1 to 8, characterized in that, Includes the following steps: Install corresponding modular thermal conductive patch units for each heat source component inside the server and connect them to the piping; Set temperature control targets for each heat source element; The central controller continuously reads data from all temperature sensors; For each heat source element, based on the deviation between its real-time temperature and the temperature control target, the opening of the flow control valve on the corresponding branch pipe is independently calculated and adjusted to achieve independent closed-loop temperature control for each heat source element. Coordinate the overall flow and pressure of the system to achieve overall energy efficiency optimization.

Images