Regulation system of server temperature

Abstract

The embodiment of the application provides a kind of server temperature's regulating system, wherein the system includes: server and cooling liquid distribution unit, server is connected with cooling liquid distribution unit, server includes: chip, cold plate and first energy storage component, wherein, cold plate is connected with chip and first energy storage component respectively;Cooling liquid distribution unit is used to send first cooling liquid to cold plate and first energy storage component respectively in the case where the temperature of chip is higher than first preset temperature, cold plate is used to reduce the temperature of chip by first cooling liquid, and first energy storage component is used to store the cold content in first cooling liquid, release cold content in the case where the temperature of chip is higher than second preset temperature, until the temperature of chip is not higher than second preset temperature.By the application, the problem that the temperature fluctuation of cooling liquid is larger when the temperature of chip in the server is adjusted is solved, and the effect of ensuring the temperature of chip in the normal range and prolonging the running time of chip is achieved.

Description

TECHNICAL FIELD

[0001] Embodiments of the present application relate to the field of servers, in particular, to a server temperature adjustment system. BACKGROUND

[0002] With the rapid development of informatization, the integration of communication equipment is becoming higher and higher, the application of servers is becoming more and more widespread, and the power density of the graphics processor of the artificial intelligence server is continuously increasing. A large amount of energy consumption makes the heat dissipation problem of the server particularly important.

[0003] In the related art, the server uses cooling liquid with heat dissipation function for heat dissipation. The liquid in the server absorbs heat from the key components such as the graphics processor, and then enters the cooling liquid distribution unit for heat exchange. The cooling liquid distribution unit only exchanges heat, and the water supply temperature is adjusted by adjusting the valve opening degree of the primary side. There is a technical problem of limited control range. Because, when the secondary side water temperature is low, relying on small valve opening degree to improve the temperature will cause large system fluctuations and other problems; moreover, the heat of the secondary side is completely wasted and cannot be used by the system; in addition, the related art also has the technical problem that the cold plate structure is single, and when the cooling liquid distribution unit stops running, the chip will overheat quickly. SUMMARY

[0004] Embodiments of the present application provide a server temperature adjustment system to at least solve the problem of large temperature fluctuation of the cooling liquid when adjusting the temperature of the chip in the server in the related art.

[0005] According to an embodiment of the present application, a server temperature adjustment system includes a server and a cooling liquid distribution unit, wherein the server is connected with the cooling liquid distribution unit, the server includes a chip, a cold plate and a first energy storage component, wherein the cold plate is connected with the chip and the first energy storage component respectively; the cooling liquid distribution unit is used to deliver first cooling liquid to the cold plate and the first energy storage component when the temperature of the chip is higher than a first preset temperature, wherein the cold plate is used to reduce the temperature of the chip by the first cooling liquid, the first energy storage component is used to store the cold energy in the first cooling liquid, and the first energy storage component is further used to release the cold energy when the temperature of the chip is higher than a second preset temperature until the temperature of the chip is not higher than the second preset temperature.

[0006] Optionally, the cooling liquid distribution unit is further used to receive second cooling liquid flowing through the cold plate, wherein the second cooling liquid is formed after the first cooling liquid absorbs the heat of the chip.

[0007] Optionally, the cooling liquid distribution unit further comprises a second energy storage component, wherein the second energy storage component is configured to store heat in the second cooling liquid; and the second energy storage component is further configured to release heat until the temperature of the cooling liquid is not lower than the third preset temperature, in the case that the temperature of the cooling liquid in the cooling liquid distribution unit is lower than the third preset temperature.

[0008] Optionally, the system further comprises a primary side heat exchange device, wherein the primary side heat exchange device is connected to the cooling liquid distribution unit, and is configured to receive target heat in the second cooling liquid and release the target heat to the atmosphere, in the case that the second energy storage component fails to store all the heat in the second cooling liquid, wherein the target heat is the heat in the second cooling liquid that cannot be stored by the second energy storage component.

[0009] Optionally, the cooling liquid distribution unit further comprises a power device, a first heat exchange device, and an electric valve, wherein the power device is connected to the first heat exchange device, and is configured to deliver the first cooling liquid to the cold plate and the first energy storage component, respectively; the first heat exchange device is connected to the power device, the second energy storage component, and the electric valve, respectively, and is configured to transfer the target heat in the second cooling liquid to the primary side heat exchange device; and the electric valve is connected to the first heat exchange device and the primary side heat exchange device, respectively, and is configured to control the heat entering the cooling liquid while the second energy storage component releases heat, in the case that the temperature of the cooling liquid in the cooling liquid distribution unit is lower than the third preset temperature.

[0010] Optionally, the server further comprises a fan, wherein the fan is configured to increase the rotation speed to dissipate heat from the chip, in the case that the temperature of the chip is higher than the second preset temperature.

[0011] Optionally, the number of the cooling liquid distribution units is multiple, and in the case that a first cooling liquid distribution unit stops running, a second cooling liquid distribution unit is switched within a preset switching duration, wherein the first cooling liquid distribution unit and the second cooling liquid distribution unit are any two different cooling liquid distribution units in the multiple cooling liquid distribution units.

[0012] Optionally, a first value corresponding to the first energy storage component is greater than a second value corresponding to the multiple cooling liquid distribution units, wherein the first value is a minimum value of the first energy storage component storing the heat in the first cooling liquid, and the second value is a maximum value of the heat generated by the chip within the preset switching duration.

[0013] Optionally, the server further comprises a fin, wherein the fin is connected to the first energy storage component, and is configured to dissipate the heat generated by the chip.

[0014] Optionally, the first energy storage component and the second energy storage component are composed of a phase change material.

[0015] According to the application, since the server comprises the energy storage component, the energy storage component can store the cooling liquid for reducing the temperature of the chip, and release the cold energy in the cooling liquid stored in the energy storage component when the temperature of the chip is higher than the normal temperature. Therefore, the problem that the temperature fluctuation of the cooling liquid is large when the temperature of the chip in the server is adjusted can be solved, the effect that the temperature of the chip is ensured to be in the normal range, and the running time of the chip is prolonged is achieved. BRIEF DESCRIPTION OF DRAWINGS

[0016] Figure 1 is a structural diagram of a server temperature adjustment system according to an embodiment of the application;

[0017] Figure 2 is a structural diagram of another server temperature adjustment system according to an embodiment of the application;

[0018] Figure 3 is a structural diagram of another server temperature adjustment system according to an embodiment of the application;

[0019] Figure 4 is a structural diagram of another server temperature adjustment system according to an embodiment of the application;

[0020] Figure 5 is a structural diagram of another server temperature adjustment system according to an embodiment of the application. DETAILED DESCRIPTION

[0021] Hereinafter, the embodiments of the application will be described in detail with reference to the accompanying drawings and in combination with the embodiments.

[0022] It should be noted that the terms "first", "second" and the like in the specification and claims of the application and the above-mentioned drawings are used to distinguish similar objects, and do not necessarily have to describe a specific order or sequence.

[0023] Figure 1 is a structural diagram of a server temperature adjustment system according to an embodiment of the application, as Figure 1 shown, the system comprises a server 102 and a cooling liquid distribution unit 104, wherein,

[0024] The server 102 is connected with the cooling liquid distribution unit 104, and the server 102 comprises a chip 10202, a cold plate 10204 and a first energy storage component 10206, wherein the cold plate 10204 is connected with the chip 10202 and the first energy storage component 10206 respectively.

[0025] According to an optional embodiment of the present application, the chip 10202 is a central processing unit (CPU) and / or a graphics processing unit (GPU). The chip is a main power consumption device in the server, and the power consumption ratio reaches more than one-half. The silicon-based chip (GPU and CPU) and the carbon-based human body have similar temperature adaptability. In general, the tolerance temperature of the chip is 70 DEG C. If the temperature exceeds the above-mentioned temperature, the performance of the chip is reduced, and the service life is shortened.

[0026] The central processing unit is one of the main devices of the electronic computer and the server. The server is an important device in the network, and can accept access of tens of people to thousands of people, so that the server has strict requirements of large data quantity, fast throughput, super stability, long time running and the like. Therefore, the central processing unit is the "brain" of the server, and is the primary index for measuring the performance of the server. The function of the central processing unit is mainly to interpret the instructions of the computer and the server and process the data in the server. The central processing unit is the core component responsible for reading instructions, decoding instructions and executing instructions in the server. The central processing unit mainly includes two parts, namely a controller and an arithmetic unit, and further includes a cache memory and a bus for realizing the connection between them. The functions of the central processing unit mainly include processing instructions, executing operations, controlling time and processing data.

[0027] The graphics processing unit, also known as a display core, a visual processor or a display chip, is a microprocessor specially used for image and graphics related operation in a personal computer, a workstation, a game machine and some mobile devices (such as a tablet computer and a smart phone). The graphics processing unit makes the display card reduce the dependence on the central processing unit, and performs part of the work of the central processing unit, especially when the graphics processing unit adopts core technologies such as hardware T&L (geometric conversion and illumination processing), cubic environment material mapping and vertex mixing, texture compression and concave-convex mapping, double-texture four-pixel 256-bit rendering engine, etc. in 3D graphics processing. The hardware T&L technology can be said to be the symbol of the graphics processing unit.

[0028] The work flow and physical structure of the central processing unit and the graphics processing unit are roughly similar, and the work of the graphics processing unit is more single compared with the central processing unit. In most personal computers, the graphics processing unit is only used to draw images. If the central processing unit wants to draw a two-dimensional figure, it only needs to give an instruction to the graphics processing unit, and the graphics processing unit can quickly calculate all the pixels of the figure and draw the corresponding figure at the specified position on the display. Since the graphics processing unit generates a large amount of heat, the display card usually has an independent cooling device.

[0029] The central processing unit has a powerful arithmetic operation unit, which can complete arithmetic calculation in a few clock cycles. At the same time, there is a large cache to save a lot of data in it. In addition, there is a complex logic control unit, when the program has multiple branches, the ability of branch prediction is provided to reduce the delay. The graphics processing unit is designed based on large throughput, with a large number of arithmetic operation units and a small cache. At the same time, the graphics processing unit supports a large number of threads running simultaneously, if the central processing unit and the graphics processing unit need to access the same data, the cache will merge these accesses, which will naturally cause delay problems. Although there is a delay, because of the large number of arithmetic operation units, a very large throughput effect can be achieved.

[0030] Because the central processing unit has a large cache and a complex logic control unit, it is very good at logical control and serial operation. In comparison, the graphics processing unit can execute a large number of calculation work at the same time because of a large number of arithmetic operation units. The graphics processing unit is good at large-scale concurrent computing, which requires a large amount of calculation and is repeated many times. Therefore, using the central processing unit for complex logical control and the graphics processing unit for simple but large arithmetic operation can greatly improve the running speed of the program.

[0031] It can be understood that both the central processing unit and the graphics processing unit will generate heat, and then the temperature may be too high. The temperature of the central processing unit and the graphics processing unit rises because the heat generated is greater than the heat dissipated.

[0032] In an optional embodiment, the first energy storage component 10206 is composed of a phase change material, wherein the phase change material (PCM) refers to a substance that changes the state of matter at a constant temperature and can provide latent heat. The process of changing physical properties is called a phase change process, at which time the phase change material will absorb or release a large amount of latent heat. The process of a substance changing from one phase to another. A uniform part of a substance system that has the same physical and chemical properties as other parts and has a clear interface with other parts is called a phase. Corresponding to the three states of solid, liquid and gas, the substance has solid, liquid and gas phases. The most common phase change material is water, which changes from liquid to solid (freezing) when the temperature is as low as 0℃. When the temperature is higher than 0℃, the water changes from solid to liquid (dissolution). A large amount of cold energy is absorbed and stored in the freezing process, and a large amount of heat energy is absorbed in the dissolution process. The larger the amount (volume) of ice, the longer the dissolution process takes. This is the most typical example of a phase change material.

[0033] The phase change material has the ability to change its physical state within a certain temperature range. Taking the solid-liquid phase change as an example, when heated to the melting temperature, a phase change from solid to liquid occurs, and during the melting process, the phase change material absorbs and stores a large amount of latent heat; when the phase change material cools down, the stored heat is dissipated to the environment within a certain temperature range, and the reverse phase change from liquid to solid occurs. In the two phase change processes, the ability to store or release is called phase change latent heat. When the physical state changes, the temperature of the material itself hardly changes before the phase change is completed.

[0034] In the case where the temperature of the central processor reaches the critical value of the phase change material in the energy storage component, the phase change material changes from one physical property to another, for example, from solid to liquid, and in the above transition process, the phase change material releases the energy it stores, for example, in the case of storing heat using the phase change material, when the ambient temperature reaches the critical value of the phase change material, the phase change material changes from one physical property to another and releases the heat it stores; in the case of storing cold using the phase change material, when the chip temperature reaches the critical value of the phase change material, the phase change material changes from one physical property to another and releases the cold it stores.

[0035] The cooling liquid distribution unit 104 is configured to, in the case where the temperature of the chip 10202 is higher than a first preset temperature, deliver the first cooling liquid to the cold plate 10204 and the first energy storage component 10206, respectively, wherein the cold plate 10204 is configured to reduce the temperature of the chip 10202 by the first cooling liquid, and the first energy storage component 10206 is configured to store cold in the first cooling liquid, and the first energy storage component 10206 is further configured to, in the case where the temperature of the chip 10202 is higher than a second preset temperature, release the cold until the temperature of the chip 10202 is not higher than the second preset temperature.

[0036] In the field of data centers, especially in the field of high-performance computing, there are core advantages in using liquid cooling instead of air cooling: liquid cooling can maintain the chip at a safe and stable temperature of 10-20℃ below the critical safety temperature, support high-power chips and chip overclocking, improve chip performance, and increase chip service life. Liquid cooling is a cooling technology that uses cooling liquid to carry away the heat of heat-generating devices (such as central processing units and graphics processing units), and is suitable for application scenarios that require improved computing power, energy efficiency, deployment density, etc. Liquid cooling can be divided into contact and non-contact liquid cooling. Contact liquid cooling refers to a form of liquid cooling in which the cooling liquid is in direct contact with the heat-generating device, including immersion and spray liquid cooling. Non-contact liquid cooling refers to a form of liquid cooling in which the cooling liquid is not in direct contact with the heat-generating device, including cold plate type and other schemes, wherein the cold plate type liquid cooling is a form of implementation in which the heat of the heat-generating device is indirectly transferred to the cooling liquid in the circulating pipeline through the cold plate (usually a closed cavity made of heat-conducting metals such as copper and aluminum), and the heat is carried away by the cooling liquid.

[0037] The cold plate is a heat exchanger or radiator with internal fluid channels and allows the cooling working fluid to flow through it. The cold plate is installed on the hot surface of the electronic components (central processing units and graphics processing units) that need to be cooled, and the heat generated by the components is transferred to the plate heat exchanger of the cooling liquid distribution unit through the cooling liquid. The design of the cold plate is diverse, and it can be structurally optimized according to different needs. The internal flow channel can be a groove, a snap-in fin, a spade tooth, a folded fin, etc. For some heat dissipation design of high-power or high-heat-density components, the flow channel is usually designed into a more complex micro-channel structure to increase the contact area and improve the heat dissipation performance.

[0038] The basic structure of the cold plate includes a heat dissipation module and a fixed module. The fixed module design should maximize the meet the snap-in force positive pressure cold plate. According to the connection method between the heat dissipation module and the fixed module, it can be divided into split type liquid cooling cold plate (the cold plate heat dissipation module and the fixed module are connected by screws or other methods, and can be disassembled and assembled according to needs), and integrated type liquid cooling cold plate (the cold plate heat dissipation module and the fixed module cannot be disassembled and assembled). According to the sealing form, it can be divided into sealing ring assembly type or welding sealing, etc. The selection of the cold plate depends on factors such as heat dissipation requirements, cost requirements, operating parameters, and the use of infiltration materials. The materials used for all components in contact with the cooling working fluid in the secondary side cooling circuit must be consistent with the cooling working fluid's infiltration material list (WML). According to the different temperature requirements of the components to be cooled, the cooling working fluid parameters such as flow rate, temperature, and heat transfer characteristics, the complexity and difficulty of the cold plate design are also different. For example, the commonly used micro-channel cold plate structure is relatively complex, and the micro-channel is mainly to increase the contact area with the liquid to improve the cooling performance. A simpler cold plate design is a module with a simple internal fluid channel. In this embodiment, a cold plate with a simple internal fluid channel module is used.

[0039] According to another optional embodiment of the present application, the cooling liquid distribution unit (CDU) is mainly composed of a circulating pump, a frequency converter, a filter, a heat exchanger, an electric control unit, a monitoring sensor unit, connecting pipelines and valves, etc. The CDU conducts the heat of the equipment to the secondary cooling loop water circulation system through the built-in heat exchanger, is the core of the liquid cooling scheme, is responsible for the utilization of the cold source and the heat dissipation and distribution of the heat source, and serves as a connecting bridge between the primary side and the secondary side. The CDU can be configured as needed to maximize the power consumption of heat dissipation. The main function of the cooling liquid distribution unit is to isolate the primary side and the secondary side loop and provide heat exchange capacity between the primary side and the secondary side inside the unit. The cooling liquid distribution unit is mainly divided into rack type (embedded cabinet type), cabinet type and platform type, etc. The cabinet type cooling liquid distribution unit usually provides cooling for one or more IT servers or even the entire computer room, has greater cooling capacity and liquid supply capacity than the rack type cooling liquid distribution unit. The platform type cooling liquid distribution unit is usually a type of cooling liquid distribution unit with greater cooling capacity and liquid supply capacity, and the maximum cooling capacity can reach more than 10 MW, which can provide cooling for the entire data center. In the server temperature regulation system provided by the embodiment, the cooling liquid distribution unit is of the rack type or the cabinet type. In order to avoid insufficient cooling capacity caused by failure of the cooling liquid distribution unit, the N+1, N+2 redundant backup of the cooling liquid distribution unit needs to be fully considered according to the actual situation, or the pump drive module inside the cooling liquid distribution unit needs to be designed with N+1 redundancy, so as to ensure sufficient cooling capacity for cooling the IT equipment (server), or to form a redundant design to realize online maintenance of the cooling liquid distribution unit.

[0040] Through the above system, the energy storage assembly included in the server 102 is utilized, wherein the energy storage assembly can store cooling liquid for reducing the temperature of the chip 10202, and release the cold energy in the stored cooling liquid in the energy storage assembly in the case that the temperature of the chip 10202 is higher than the normal temperature, so as to solve the problem that the temperature fluctuation of the cooling liquid is large when the temperature of the chip in the server is adjusted, ensure that the temperature of the chip is in the normal range, and prolong the running time of the chip.

[0041] According to an optional embodiment of the present application, the cooling liquid distribution unit 104 is further configured to receive the second cooling liquid flowing through the cold plate 10204, wherein the second cooling liquid is formed after the first cooling liquid absorbs the heat of the chip 10202.

[0042] According to another optional embodiment of the present application, the cooling liquid entering the chip 10202 end is a first cooling liquid, the first cooling liquid flows through the cold plate 10204 connected with the chip 10202, and the heat released by the chip 10202 is taken away by the cold plate 10204. The cooling liquid containing the heat released by the chip 10202 after flowing through the cold plate 10204 is defined as a second cooling liquid, that is, the second cooling liquid is formed after the first cooling liquid absorbs the heat of the chip 10202. After the first cooling liquid flows through the cold plate 10204 to form the second cooling liquid, it returns to the cooling liquid distribution unit 104, is heat-exchanged by the plate heat exchanger in the cooling liquid distribution unit 104, and transmits heat to the outdoor or atmosphere through the primary side outdoor system.

[0043] Figure 2 is a structural diagram of another server temperature adjustment system according to an embodiment of the present application, as shown in Figure 2 The cooling liquid distribution unit 104 further includes a second energy storage assembly 10402,

[0044] The second energy storage assembly 10402 is configured to store the heat in the second cooling liquid. The second energy storage assembly 10402 is further configured to release heat when the temperature of the cooling liquid in the cooling liquid distribution unit 104 is lower than a third preset temperature, until the temperature of the cooling liquid is not lower than the third preset temperature.

[0045] In some optional embodiments of the present application, the cooling liquid distribution unit 104 also includes an energy storage assembly, which is configured to store the heat in the cooling liquid flowing through the chip 10202 and the cold plate 10204 to complete the absorption of the heat of the chip 10202. When the supply liquid temperature in the cooling liquid distribution unit 104 is low, that is, lower than the third preset temperature, the energy storage assembly releases the stored heat of the chip 10202, which can effectively improve the supply liquid temperature.

[0046] The energy storage assembly in the cooling liquid distribution unit 104 achieves the technical effect of fully utilizing the heat generated by the chip 10202, and realizes the recycling of energy.

[0047] Figure 3 is a structural diagram of another server temperature adjustment system according to an embodiment of the present application, as shown in Figure 3 The system further includes a primary side heat exchange device 106, wherein

[0048] The primary side heat exchange device 106 is connected with the cooling liquid distribution unit 104, and is configured to receive target heat in the second cooling liquid and release the target heat to the atmosphere when the second energy storage assembly 10402 cannot store all the heat in the second cooling liquid. The target heat is the heat in the second cooling liquid that cannot be stored by the second energy storage assembly 10402.

[0049] As an optional embodiment of the present application, the energy storage efficiency of the energy storage component in the cooling liquid distribution unit 104 cannot store all the heat in the cooling liquid with the heat of the chip 10202, or the energy storage capacity of the energy storage component in the cooling liquid distribution unit 104 cannot meet the storage of all the heat in the cooling liquid with the heat of the chip 10202. Therefore, the first side heat exchange device 106 is connected with the cooling liquid distribution unit 104, and in the case that the energy storage component in the cooling liquid distribution unit 104 cannot store all the heat in the cooling liquid, the heat in the cooling liquid that cannot be stored by the energy storage component in the cooling liquid distribution unit 104 is received, and the heat is released to the outdoor or atmosphere.

[0050] Figure 4 is a structure diagram of another server temperature adjustment system according to an embodiment of the present application, as shown in Figure 4 The cooling liquid distribution unit 104 further comprises a power device 10404, a first heat exchange device 10406 and an electric valve 10408, wherein,

[0051] The power device 10404 is connected with the first heat exchange device 10406, and is used to deliver the first cooling liquid to the cold plate 10204 and the first energy storage component 10206 respectively.

[0052] The first heat exchange device 10406 is connected with the power device 10404, the second energy storage component 10402 and the electric valve 10408 respectively, and is used to transfer the target heat in the second cooling liquid to the first side heat exchange device 106.

[0053] The electric valve 10408 is connected with the first heat exchange device 10406 and the first side heat exchange device 106 respectively, and is used to control the heat entering the cooling liquid while releasing the heat by the second energy storage component 10402 in the case that the temperature of the cooling liquid in the cooling liquid distribution unit 104 is lower than the third preset temperature.

[0054] In some optional embodiments of the present application, the power device 10404 includes but is not limited to a pump, and the pump in the cooling liquid distribution unit 104 is used to deliver the cooling liquid to the cold plate 10204 and the energy storage assembly, respectively. The first heat exchange device 10406 includes but is not limited to a plate heat exchanger, wherein the plate heat exchanger is a high-efficiency heat exchanger formed by a series of metal sheets with a certain corrugated shape. Thin rectangular channels are formed between various plates, and heat exchange is carried out through the plates. The plate heat exchanger is an ideal device for liquid-liquid and vapor heat exchange. The plate heat exchanger has the characteristics of high heat exchange efficiency, small heat loss, compact and light structure, small floor area, wide application, long service life, etc. Under the same pressure loss condition, the heat transfer coefficient is 3-5 times higher than that of the tube heat exchanger, the floor area is one third of that of the tube heat exchanger, and the heat recovery rate can be as high as 90% or more. The target heat in the second cooling liquid can be transferred to the primary side heat exchange device 106 through the plate heat exchanger.

[0055] In an optional embodiment, the heated cooling liquid (second cooling liquid) flows through the energy storage assembly of the cooling liquid distribution unit, and the energy storage assembly stores heat. When the supply liquid temperature is lower than the preset temperature, the valve opening degree is adjusted, and at the same time, the energy storage assembly of the cooling liquid distribution unit releases heat to increase the supply liquid temperature, thereby ensuring the system supply liquid temperature and reducing the system internal liquid temperature fluctuation.

[0056] As another optional embodiment of the present application, the server further includes a fan, wherein the fan is used to increase the rotation speed to dissipate heat from the chip when the temperature of the chip is higher than the second preset temperature.

[0057] The energy storage assembly at the chip end is in contact with the cold plate. When the system provided by the present application is running, the energy storage assembly at the chip end stores cold energy. When the temperature of the chip is too high, the energy storage assembly at the chip end releases the cold energy stored therein to reduce the temperature of the chip. When the cooling liquid distribution unit stops running, the energy storage assembly at the chip end also provides cold energy, thereby prolonging the working time of the chip. At the same time, the fan in the server also increases the rotation speed to accelerate the heat dissipation of the chip and ensure the normal operation of the chip.

[0058] In an optional embodiment, the number of cooling liquid distribution units is multiple, and in the case that the first cooling liquid distribution unit stops running, the second cooling liquid distribution unit is switched within a preset switching time, wherein the first cooling liquid distribution unit and the second cooling liquid distribution unit are any two different cooling liquid distribution units in the multiple cooling liquid distribution units.

[0059] According to some preferred embodiments of the present application, the first value corresponding to the first energy storage component is greater than the second value corresponding to the plurality of cooling liquid distribution units, wherein the first value is the minimum value of the cold energy stored by the first energy storage component in the first cooling liquid, and the second value is the maximum value of the heat generated by the chip within the preset switching duration.

[0060] To avoid insufficient cold energy caused by failure of the cooling liquid distribution unit, N+1 or N+2 redundant backup of the cooling liquid distribution unit needs to be considered in combination with the actual situation, or the pump drive module inside the cooling liquid distribution unit needs to be designed with N+1 redundancy to ensure sufficient cold energy for cooling the IT equipment, or a redundant design is formed to realize online maintenance of the cooling liquid distribution unit.

[0061] According to some preferred embodiments of the present application, the cooling liquid distribution unit is multiple and divided into primary and standby, for example, three cooling liquid distribution units, one primary and two standby. When the primary cooling liquid distribution unit stops running due to various conditions, the standby cooling liquid distribution unit is switched, but there is a switching duration in the switching process. During the switching process of the cooling liquid distribution unit, the energy storage component at the chip end also provides cold energy, thereby prolonging the working time of the chip.

[0062] In a preferred embodiment, to avoid the cold energy stored at the chip end being unable to balance the heat generated by the chip during the switching process of the cooling liquid distribution unit, the preset switching duration between the cooling liquid distribution units needs to be determined, for example, one minute, and the maximum value of the heat generated by the chip within one minute is determined. At the same time, the minimum value of the cold energy stored by the energy storage component at the chip end in the cooling liquid is determined, and it is ensured that the minimum value of the cold energy stored by the energy storage component at the chip end in the cooling liquid is greater than the maximum value of the heat generated by the chip within the preset switching duration between the plurality of cooling liquid distribution units, that is, the situation that the cold energy stored at the chip end is less than the heat generated by the chip during the switching process of the cooling liquid distribution unit can be avoided.

[0063] Optionally, the energy storage component is composed of a phase change material. The process of changing the physical properties of the phase change material is called a phase change process, at which time the phase change material will absorb or release a large amount of latent heat. Generally, the energy (heat or cold energy) stored by the phase change material is dissipated to the environment within a certain temperature range. In this embodiment, the cold energy of the cooling liquid stored by the phase change material is released when the chip heats up to a certain temperature range, and the phase change material releases the cold energy stored therein, and the heat dissipation to the chip is realized through the cold plate connected to the chip. The energy absorbed or released by the phase change material is generally related to its volume and natural properties. Under the premise that the minimum value of the cold energy stored by the energy storage component at the chip end in the cooling liquid is greater than the maximum value of the heat generated by the chip within the preset switching duration between the plurality of cooling liquid distribution units, selecting a phase change material with small volume and strong energy storage capacity helps to reduce the volume of the server.

[0064] According to an optional embodiment of the present application, the server further comprises fins, wherein the fins are connected with the first energy storage component, and are used to dissipate the heat generated by the chip.

[0065] In the heat dissipation system, it is usually required to install fins on the surface of the heat source, i.e. to install fins on the surface of the chip. Since the chip is connected with the energy storage component through the cold plate, the fins are connected with the first energy storage component. The fins are used to increase the heat exchange area, and can dissipate the heat generated by the chip more efficiently.

[0066] According to another optional embodiment of the present application, the first energy storage component and the second energy storage component are composed of phase change materials.

[0067] Phase change material (PCM) refers to a substance that changes its state without changing its temperature and can provide latent heat. The process of changing physical properties is called phase change process, at which time the phase change material will absorb or release a large amount of latent heat. The process of changing from one phase to another phase of a substance. A uniform part in a substance system that has the same physical and chemical properties as other parts and has a clear interface is called a phase. Corresponding to the three states of solid, liquid and gas, the substance has solid phase, liquid phase and gas phase. The most common phase change material is water, which changes from liquid to solid (freezing) when the temperature is as low as 0℃. When the temperature is higher than 0℃, water changes from solid to liquid (melting). A large amount of cold energy is absorbed and stored in the freezing process, and a large amount of heat energy is absorbed in the melting process.

[0068] Figure 5 is another structure diagram of the server temperature regulation system according to the embodiment of the present application, as shown in Figure 5 The system is composed of a primary side outdoor heat exchange system, a secondary side cooling liquid distribution unit, a chip, a chip cold plate, an electric valve, a fan and an energy storage component.

[0069] The operation principle of the server temperature regulation system will be described below with a specific embodiment:

[0070] The pump inside the cooling liquid distribution unit delivers liquid (cooling liquid) to the GPU chip and CPU chip for heat exchange, the required cooling capacity for heat exchange is provided by the cooling liquid distribution unit, the liquid takes away the heat of the chip and returns to the cooling liquid distribution unit for heat exchange through the plate heat exchanger, and the heat is transmitted through the primary outdoor system. At the same time, the heated liquid flows through the phase change material, and the phase change material stores heat. When the supply liquid temperature is low, the energy storage assembly releases heat while adjusting the opening degree of the electric valve, thereby improving the supply liquid temperature and reducing the temperature fluctuation of the liquid inside the system. Moreover, the chip cooling cold plate also contacts the energy storage assembly. When the system is running, the energy storage assembly stores cold energy, and releases the stored cold energy when the temperature of the chip is too high, thereby reducing the temperature of the chip. After the cooling liquid distribution unit is stopped, the energy storage assembly also provides cold energy, thereby prolonging the working time of the chip. At the same time, the fan speed of the system is also increased to dissipate heat from the chip, thereby ensuring the working of the chip and prolonging the working time of the chip.

[0071] Optionally, the cooling liquid distribution unit inside includes a pump drive module, which controls the start and stop of the pump. Generally, the pump drive module is designed with N+1 redundancy to ensure sufficient cooling capacity for cooling the server, or the cooling liquid distribution unit is designed with redundancy to realize online maintenance.

[0072] Optionally, the cooling liquid distribution unit inside further includes a filter assembly. The filtering capacity of the filter assembly must match the requirements of the most sensitive components in the cooling circuit, such as self-sealing quick connectors and micro-channel cold plates, to ensure that potential particles in the cooling medium do not cause blockage in the fluid circuit and prevent the flow of the cooling medium, or the components lose the self-sealing ability when being plugged in or out. It is recommended that the filtering precision be 50 microns.

[0073] The system adds a phase change energy storage assembly in the cooling liquid distribution unit system and the cold plate heat exchange system. The liquid is cooled by the cooling liquid distribution unit and enters the cold plate and the phase change energy storage assembly at the same time. The material stores cold energy and does not participate in chip heat exchange. The liquid entering the cold plate cools the chip and takes away the heat. The heated liquid returns to the cooling liquid distribution unit system, and the liquid flows through the phase change material to store heat. The heat generated by the chip is used to release heat to improve the supply liquid temperature when the supply liquid temperature is low. The stored cold energy at the chip end is released when the temperature of the chip is too high or the cooling liquid distribution unit system is stopped, which is used to reduce the temperature of the chip and prolong the working time of the chip. The phase change material has fins, which can be used for air cooling. After the liquid cooling system is stopped, the fan speed is increased to cool the chip together with the phase change material and the fins thereon, thereby ensuring the temperature of the chip and prolonging the working time of the chip.

[0074] Through the system, the server includes the energy storage component, wherein the energy storage component can store the cooling liquid for reducing the temperature of the chip, and release the cold energy in the cooling liquid stored in the energy storage component when the temperature of the chip is higher than the normal temperature, so as to solve the problem that the temperature fluctuation of the cooling liquid is large when the chip in the server is temperature-regulated, ensure that the temperature of the chip is in the normal range, and prolong the running time of the chip.

[0075] PUE is a comprehensive index for measuring the energy efficiency of data center infrastructure universally accepted and adopted at home and abroad, and its calculation formula is: PUE=PTotal / PIT. Wherein, PTotal is the total power consumption of the data center, and PIT is the power consumption of the IT equipment in the data center. The actual meaning of PUE refers to how much of the total power provided to the data center is actually applied to the IT equipment.

[0076] By using the above system, the overall energy efficiency of the data center can be improved, the PUE of the data center can be effectively reduced, the design requirements of the data center intensification, scale and greenness can be met, and a large amount of operating expenses of the enterprise can be saved.

[0077] Through the description of the above embodiments, those skilled in the art can clearly understand that the method according to the above embodiments can be realized by means of software and the necessary general hardware platform, of course, it can also be realized by hardware, but in many cases, the former is a better embodiment. Based on such understanding, the technical solutions of the present application can be embodied in the form of a software product, which is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk) and includes a plurality of instructions for causing a terminal device (which can be a mobile phone, computer, server, or network device) to execute the method described in each embodiment of the present application.

[0078] The specific examples in the present embodiment can refer to the examples described in the above embodiments and exemplary embodiments, which will not be described here again.

[0079] Obviously, those skilled in the art should understand that the above-mentioned modules or steps of the present application can be realized by a general computing device, which can be concentrated on a single computing device, or distributed on a network composed of multiple computing devices, which can be realized by program code executable by the computing device, so that they can be stored in the storage device and executed by the computing device, and in some cases, the steps shown or described can be executed in different order, or they can be manufactured into individual integrated circuit modules, or multiple modules or steps can be manufactured into a single integrated circuit module. Thus, the present application is not limited to any specific hardware and software combination.

[0080] The above descriptions are only the preferred embodiments of the present application, and are not intended to limit the present application. The present application can have various modifications and changes for those skilled in the art. Any modifications, equivalent replacements, improvements, etc. within the principles of the present application shall be included in the protection scope of the present application.

Claims

1. A system for regulating the temperature of a server, comprising: A server and a cooling liquid distribution unit, wherein The server is connected with the cooling liquid distribution unit, and the server comprises a chip, a cold plate and a first energy storage component, wherein the cold plate is connected with the chip and the first energy storage component respectively; The cooling liquid distribution unit is configured to deliver first cooling liquid to the cold plate and the first energy storage component when the temperature of the chip is higher than a first preset temperature, wherein the cold plate is configured to reduce the temperature of the chip by the first cooling liquid, and the first energy storage component is configured to store cold energy in the first cooling liquid, and the first energy storage component is further configured to release cold energy when the temperature of the chip is higher than a second preset temperature until the temperature of the chip is not higher than the second preset temperature. The cooling liquid distribution unit further comprises a pump driving module, and the pump driving module is configured to control the start and stop of a pump in the cooling liquid distribution unit; and the number of the pump driving modules is plural. The server further comprises a fin connected with the first energy storage component and configured to dissipate heat generated by the chip. The number of the cooling liquid distribution units is plural, and a second cooling liquid distribution unit is switched within a preset switching time length when a first cooling liquid distribution unit stops running, wherein the first cooling liquid distribution unit and the second cooling liquid distribution unit are any two different cooling liquid distribution units in the plural cooling liquid distribution units. A first value corresponding to the first energy storage component is greater than a second value corresponding to the plural cooling liquid distribution units, wherein the first value is a minimum value of the first energy storage component storing cold energy in the first cooling liquid, and the second value is a maximum value of heat generated by the chip within the preset switching time length.

2. The server temperature regulation system of claim 1, wherein, The cooling liquid distribution unit is further configured to receive second cooling liquid flowing through the cold plate, wherein the second cooling liquid is formed after the first cooling liquid absorbs heat of the chip.

3. The server temperature regulation system of claim 2, wherein, The cooling liquid distribution unit further comprises a second energy storage component, wherein The second energy storage component is configured to store heat in the second cooling liquid. The second energy storage component is further configured to release heat when the temperature of the cooling liquid in the cooling liquid distribution unit is lower than a third preset temperature until the temperature of the cooling liquid is not lower than the third preset temperature.

4. The server temperature regulation system of claim 3, wherein, The system further comprises a primary side heat exchange device, wherein The primary side heat exchange device is connected with the cooling liquid distribution unit and configured to receive target heat in the second cooling liquid and release the target heat to the atmosphere when the second energy storage component cannot store all the heat in the second cooling liquid, wherein the target heat is heat in the second cooling liquid that cannot be stored by the second energy storage component.

5. The server temperature regulation system of claim 4, wherein, The cooling liquid distribution unit further comprises a power device, a first heat exchange device and an electric valve, wherein The power device is connected with the first heat exchange device and configured to deliver the first cooling liquid to the cold plate and the first energy storage component respectively; and The electric valve is connected with the first heat exchange device and configured to deliver the first cooling liquid to the cold plate and the first energy storage component respectively. The first heat exchange device is connected with the power device, the second energy storage component and the electric valve respectively, and is used for transferring target heat in the second cooling liquid to the primary side heat exchange device; The electric valve is connected with the first heat exchange device and the primary side heat exchange device respectively, and is used for controlling the heat entering the cooling liquid by releasing heat of the second energy storage component when the temperature of the cooling liquid in the cooling liquid distribution unit is lower than the third preset temperature.

6. The server temperature regulation system of claim 1, wherein, The server further comprises a fan, wherein The fan is used for increasing the rotating speed to dissipate heat for the chip when the temperature of the chip is higher than the second preset temperature.

7. The server temperature regulation system of claim 3, wherein, The first energy storage component and the second energy storage component are composed of phase change materials.

Images