Liquid cooling apparatus for prioritized processing of data center server racks

By employing a series-parallel cooling loop configuration in the data center server rack and optimizing the distribution of cooling liquid, the high cost of traditional liquid cooling systems is solved, achieving efficient and economical cooling and ensuring effective cooling of high-performance electronic processing components.

CN122269632APending Publication Date: 2026-06-23OVH
View PDF 0 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
OVH
Filing Date
2025-12-19
Publication Date
2026-06-23

Smart Images

  • Figure CN122269632A_ABST
    Figure CN122269632A_ABST
Patent Text Reader

Abstract

A liquid cooling arrangement system for cooling a plurality of server clusters is disclosed. The plurality of server clusters each contain heat generating devices with liquid cooling units, the liquid cooling arrangement system comprising: a cooling loop configured to circulate cooling liquid and circulate heated liquid from the plurality of server clusters back to a dry cooler unit; the plurality of server clusters comprising: a first cluster and a second cluster connected in series by the cooling loop, the first cluster comprising a first set of liquid cooling units connected in parallel by the cooling loop, the first set having a first number of liquid cooling units, the second cluster comprising a second set of liquid cooling units connected in parallel by the cooling loop, the second set having a second number of liquid cooling units, wherein the first number is greater than the second number.
Need to check novelty before this filing date? Find Prior Art

Description

Cross-reference to related applications

[0001] This patent application claims priority to European Patent Application No. 24307221.2, filed on 20 December 2024, the disclosure of which is incorporated herein by reference in its entirety. Technical Field

[0002] This technology relates to liquid cooling configurations for data center server racks. Background Technology

[0003] Data centers are configured to house numerous server racks containing electronic equipment, such as computer systems (e.g., server components), repositories, etc., to process massive amounts of data in near real-time. During operation, the electronic equipment in the server racks generates a significant amount of heat, which must be dissipated to ensure the continuous and efficient operation of the electronic equipment. To address this heat generation issue, many cooling solutions have been implemented, including liquid cooling of heat-generating components by directly mounting liquid cooling blocks (often referred to as liquid block units or water block units) to certain heat-generating parts.

[0004] While water-cooled units are effective at cooling heat-generating components, their implementation in server racks typically requires a liquid distribution infrastructure to power numerous server racks and the large amount of electronics they support. Furthermore, some server racks contain servers with high-performance electronic processing units that generate higher temperature profiles, unlike servers containing electronic components that do not generate such high temperature profiles.

[0005] Understandably, liquid cooling distribution infrastructure traditionally requires extensive piping across the data center and numerous pumps to maintain the necessary coolant flow to supply large volumes of water, especially for high-performance electronic processing components. For data centers, the use of extensive piping and numerous pumps can be prohibitively expensive in terms of initial investment and operating costs.

[0006] Therefore, it seems desirable to provide a liquid cooling arrangement for data center server racks that can alleviate at least some of the high costs associated with conventional pipework and pumps by providing a configuration that effectively addresses the cooling needs of both servers with processing components that generate high temperature distributions due to electronic components and servers with processing components that generate standard temperature distributions.

[0007] It is important to note that the topics discussed in the background section should not be considered prior art simply because they are mentioned therein. Similarly, the problems mentioned in the background section should not be interpreted as having been recognized in the prior art. Summary of the Invention

[0008] The purpose of this technology is to alleviate at least some of the problems of excessive cost that are common in existing technologies.

[0009] According to one aspect of the present technology, a liquid cooling arrangement system is provided for cooling multiple server clusters, each of the multiple server clusters including a heat-generating device with a liquid cooling unit. The liquid cooling arrangement system includes: a cooling loop configured to circulate cooling liquid to the multiple server clusters and to circulate heated liquid from the multiple server clusters back to a dry cooler unit; the multiple server clusters include: a first cluster and a second cluster, the first cluster and the second cluster being connected in series via the cooling loop; the first cluster including a first group of liquid cooling units connected in parallel via the cooling loop, the first group having a first number of liquid cooling units; the second cluster including a second group of liquid cooling units connected in parallel via the cooling loop, the second group having a second number of liquid cooling units; the second cluster being located further downstream along the cooling loop; and the second cluster being configured to receive the cooling liquid flowing out from the first cluster, wherein the first number is greater than the second number.

[0010] The implementation of this technology has at least one of the above-mentioned objectives and / or aspects, but not necessarily all of them. It should be understood that some aspects of this technology resulting from attempts to achieve the above objectives may not satisfy those objectives and / or may satisfy other objectives not specifically described herein.

[0011] Additional and / or alternative features, aspects and advantages of the implementation of this technology will become apparent from the following description, drawings and appended claims. Attached Figure Description

[0012] To better understand this technology and its other aspects and additional features, reference will be made to the following description used in conjunction with the accompanying drawings, in which: Figure 1 A functional block diagram depicts an internal server cluster configuration of a data center server rack according to a non-limiting embodiment of the present disclosure; Figure 2 A functional block diagram of a liquid cooling system that optimizes cooling efficiency by prioritizing the processing of server clusters within a data center rack, according to some non-limiting embodiments of the present disclosure, is depicted. Figure 3A graph depicting the upper operating limit of the housing temperature and the cooling liquid temperature of a heat-generating electronic processing component according to some non-limiting embodiments of the present disclosure; and Figure 4 A graph depicting the relationship between flow rate and pressure drop according to some non-limiting embodiments of the present disclosure is provided. Detailed Implementation

[0013] This disclosure aims to address at least some of the problems associated with the conventional use of various piping configurations and numerous pumps to supply liquid flows to water blocks serving the cooling needs of a large number of heat-generating components. In particular, this disclosure proposes embodiments of liquid cooling devices that prioritize the cooling of higher-heat-generating servers to optimize their cooling while simplifying the arrangement of liquid cooling piping.

[0014] The examples and conditions described herein are primarily intended to aid the reader's understanding of the principles of this technology, rather than limiting its scope to the specific examples and conditions described herein. It will be understood that those skilled in the art can devise various arrangements that, while not explicitly described or shown herein, still embody the principles of this technology.

[0015] Furthermore, to aid understanding, the following description may depict a relatively simplified implementation of this technology. Those skilled in the art will understand that various implementations of this technology may involve greater complexity.

[0016] In some cases, examples that are considered helpful to modifications of the present technology may also be illustrated. This is done merely to aid understanding and is not intended to limit the scope of the present technology or to define its boundaries. These modifications are not exhaustive, and those skilled in the art can make other modifications while still remaining within the scope of the present technology. Furthermore, the absence of examples illustrating modifications should not be construed as impossibility of modification or as the description being the only way to implement that element of the present technology.

[0017] Furthermore, all statements herein describing the principles, aspects, and implementations of the technology, and specific examples thereof, are intended to cover both their structural and functional equivalents, whether they are currently known or will be developed in the future. Therefore, for example, those skilled in the art will understand that any block diagram herein represents a conceptual view of an exemplary system embodying the principles of the technology.

[0018] Having grasped these fundamental principles, we will now consider some non-limiting examples to illustrate how various inventive aspects of this disclosure can be realized.

[0019] Figure 1A functional block diagram depicts a representative configuration of multiple internal server clusters (such as server cluster 10) within a plurality of data center server racks (such as rack assembly 20) according to a non-limiting embodiment of the present disclosure. Generally, each server cluster includes heat-generating electronic processing components.

[0020] In the context of this application, the term "heat-generating component" refers broadly to any electronic or mechanical device, system, or subcomponent that generates heat as a byproduct during its operation. These components may include, but are not limited to, central processing units (CPUs), graphics processing units (GPUs), application-specific integrated circuits (ASICs), memory modules (e.g., RAM), power supply units (PSUs), data storage devices (e.g., hard disk drives and solid-state drives), network hardware (e.g., routers and switches), and other similar components commonly found in computing systems, server architectures, or industrial control systems. Heat may be generated due to resistance within these components, high processing loads, or rapid switching operations. In some embodiments, the heat-generating component may also include motors, actuators, or other electromechanical devices integrated into a hybrid system. The cooling systems described herein are designed to effectively dissipate this heat, regardless of the specific type or configuration of the heat-generating component.

[0021] As shown in the figure, each data processing component (such as data processing component 100) incorporates at least one corresponding liquid cooling block unit (such as block unit 102) configured to be in direct thermal contact with the heat-generating electronic processing components. Block unit 102 is configured with internal conduits (not shown) to accommodate a circulating flow of channeled cooling liquid through these internal conduits. The channeled cooling liquid is supplied by a cooling liquid supplier that delivers the liquid to each of the liquid cooling block units in series via an internal server cluster liquid circulation channel 30 to absorb heat energy from the heat-generating electronic components and to discharge the heated liquid from the liquid cooling block units.

[0022] While each server cluster is associated with a corresponding rack component in the illustrated implementation, this is not a limitation. It is possible to install multiple server clusters on a single rack component. Alternatively, it is also possible to install multiple server clusters on multiple server rack components.

[0023] Given the configuration of the internal data center server cluster 10 described above, Figure 2 A functional block diagram of a liquid cooling device 200, according to a non-limiting embodiment of this disclosure, is depicted to optimize overall cooling efficiency by prioritizing the processing of server clusters in a data center rack. While the embodiments and implementations described herein are directed to data center racks, it should be understood that extending such concepts to encompass multiple data center racks is clearly within the scope of this technology.

[0024] As shown in the figure, the liquid cooling device 200 includes a single liquid distribution loop 208, which is configured with a supply side and a return side. The supply side is used to supply cooling liquid from the cooling facility 207 to multiple server clusters 201, including a first server cluster 202, a second server cluster 204, and a third server cluster 206. The return side is used to return the heated liquid from the multiple server clusters 201 to the cooling facility 207 for recooling and recirculation back to the multiple server clusters 201. As described above relative to... Figure 1 Each of the first server cluster 202, the second server cluster 204, and the third server cluster 206 includes a data processing component, which incorporates at least one corresponding liquid cooling block unit for direct thermal contact with the heat-generating electronic components of the data processing component to dissipate heat from the heat-generating electronic components. Although only three server clusters are shown in the illustrated embodiment, this is for ease of illustration, and it is conceivable that the liquid cooling device 200 can be implemented with two or more server clusters.

[0025] A liquid distribution loop 208 is configured with a liquid distribution inlet 209 along the supply side for supplying cooling liquid to multiple fluidly connected server clusters 201; and a liquid distribution outlet 211 is configured along the return side for receiving heated liquid from the multiple server clusters 201 and returning the heated liquid to the cooling facility 207 for recooling and recirculating it back to the multiple server clusters 201. The liquid distribution loop 208 can be constructed of flexible materials (e.g., rubber, plastic, etc.), rigid materials (e.g., metal, PVC pipes, etc.), or any combination thereof. It should be understood that the liquid supplied may include water, alcohol, or any suitable liquid capable of maintaining a sufficient cooling temperature. It should be noted that the first server cluster 202, the second server cluster 204, and the third server cluster 206 are arranged in series (i.e., one after another) via the liquid distribution loop 208.

[0026] Cooling facility 207 may include a dry cooler unit 210 configured to process and re-regulate liquid received from multiple server clusters 201 to provide cooling liquid, which is then recirculated back to the multiple server clusters 201 via liquid distribution loop 208. Cooling facility 207 may also include a pump 216 configured to provide the required pressure increase and volumetric flow rate of the cooling liquid from dry cooler unit 210 through liquid distribution loop 208. Alternatively, the cooling facility may include a cooling tower and / or chiller unit instead of dry cooler unit 210.

[0027] The liquid cooling device 200 also includes a plurality of heat exchangers (HEXs) 212 to 214. In the illustrated embodiment, HEXs 212 to 214 are fluidly connected in parallel via a liquid distribution loop 208. However, it is understood that HEXs 212 to 214 may be fluidly interconnected in other configurations, such as, for example, in series via the liquid distribution loop 208 without departing from the concept of the disclosed art. Additionally, it is conceivable that HEXs 212 to 214 be implemented in parallel with a plurality of server clusters 201, in series with a plurality of server clusters 201 (i.e., before or after a plurality of server clusters 201), between a plurality of server clusters 201 (e.g., between a first server cluster 202 and a second server cluster 204), or in parallel with one or more of a first server cluster 202, a second server cluster 204, and a third server cluster 206. In some embodiments, the HEX may be implemented as an air-liquid heat exchanger (ALHEX), a copper heat exchanger, or a plate heat exchanger.

[0028] The function of HEX 212 to 214 is to adequately cool the ambient air surrounding the first server cluster 202, the second server cluster 204, and the third server cluster 206. HEX 212 to 214 can embody any suitable configuration that reduces liquid temperature through supplied airflow (e.g., via compact fans), such as internal cooling coils, heat-absorbing airflow fins, etc. For example, HEX 212 to 214 can be positioned on the rear door of a rack housing multiple server clusters 201 to directly cool the air exhausted from the multiple server clusters 201 and heated by the air-cooling components therein.

[0029] As mentioned above, some servers containing high-performance electronic processing components experience higher generation temperature distributions than other servers. Furthermore, as a general rule, liquid cooling components arranged in parallel reduce the fluid pressure drop experienced by the liquid distribution loop 208, thereby enabling the pump 216 to operate efficiently and provide sufficient flow to circulate throughout the liquid distribution loop 208.

[0030] In view of these operational factors, the liquid cooling device 200 provides a configuration in which one or more server clusters incorporating high-performance electronic processing units are designated as "high-priority" server clusters, while server clusters not incorporating high-performance electronic processing units are designated as "standard-priority" server clusters.

[0031] In other words, such as Figure 2As shown, the liquid cooling system 200 configures one or more high-priority server clusters to be arranged in parallel and fluidly connected to HEX 212 to 214 via liquid distribution loop 208. This configuration also positions one or more high-priority server clusters very close to HEX 212 to 214. In this configuration, one or more high-priority server clusters (such as the first server cluster 202) are configured to receive the “coldest” cooling liquid supplied by liquid distribution loop 208 first, specifically to meet the specific cooling needs of the high-priority server clusters (such as the first server cluster 202).

[0032] After the "coldest" cooling liquid circulates within each data processing component of the "high priority" server cluster, the outflowing liquid is then supplied to the "standard priority" server cluster via liquid distribution loop 208 for cooling.

[0033] Reference Figure 3 The diagram 300 illustrates the gradual changes in the upper limit of the casing temperature and the cooling liquid temperature for multiple server clusters 201. It is assumed that the first server cluster 202 corresponds to the "high priority" server cluster as described above, while each of the second server cluster 204 and the third server cluster 206 corresponds to a "standard priority" server cluster.

[0034] In some implementations, the upper limit housing temperatures for the first server cluster 202, the second server cluster 204, and the third server cluster 206 are configured to increase sequentially along the coolant flow path. This arrangement ensures that the coolant, having absorbed heat as it flows through each server cluster, remains sufficiently effective in cooling downstream clusters. By allowing downstream clusters to withstand progressively higher housing temperatures, this technique minimizes the risk of the coolant becoming overheated and unable to maintain proper thermal management. This configuration optimizes the thermal efficiency of the liquid distribution loop 208 while ensuring that all server clusters operate within their respective thermal limits.

[0035] In some implementations, the server’s electronic processing components are enclosed in a housing that is thermally connected to the water block; a high-priority server cluster (i.e., the first server cluster 202) may require an operating upper limit housing temperature of less than about 50°C, while a standard-priority server cluster (such as the second server cluster 204 and the third server cluster 206) may require an operating upper limit housing temperature of about 50°C to 80°C.

[0036] For the purposes of explanation, assume that the first server cluster 202 has a first operating upper limit housing temperature 302 of 40°C, the second server cluster 204 has a second operating upper limit housing temperature 304 of 55°C, and the third server cluster 206 has a third operating upper limit housing temperature 306 of 70°C.

[0037] Although the liquid flowing from the high-priority server cluster (i.e., the first server cluster 202) via the liquid distribution loop 208 is typically warmer than the "coldest" cooling liquid due to its respective heat-generating components, the temperature of this "warm" liquid is generally maintained at a level sufficient to cool the subsequent standard-priority server cluster (i.e., the second server cluster 204). This is because the standard-priority server cluster does not employ high-performance electronic processing components with a higher generation temperature distribution, thus requiring a lower-temperature (i.e., "coldest") cooling liquid. In some implementations, the coldest liquid temperature can be between approximately 25°C and 35°C, while the warm liquid temperature can be between approximately 35°C and 45°C.

[0038] After the "warm" liquid circulates within each data processing component and cooling unit in the third server cluster 206, the outflowing liquid is heated. The "heated" liquid is then transported to the return side of the liquid distribution loop 208 to return to the cooling facility 207 for recooling and recirculation back to the multiple server clusters 201. In some implementations, the heated liquid temperature can be between approximately 45°C and 65°C.

[0039] It should be understood that the temperature of the coolant entering the respective server cluster must be lower than the operating temperature of the respective server cluster. In other words, the temperature of the coolant entering the first server cluster 202 must be lower than the first operating limit housing temperature 302, and entering the second server cluster 204 must be lower than the second operating limit housing temperature 304, and entering the third server cluster 206 must be lower than the third operating limit housing temperature 306.

[0040] Although this embodiment describes only a combination of one “high priority” server cluster (i.e., the first server cluster 202) and two “standard priority” server clusters (the second server cluster 204 and the third server cluster 206), it should be understood that other combinations of more or fewer “high priority” / “standard priority” server clusters may be considered.

[0041] Back Figure 2We will now focus on the first server cluster 202, the second server cluster 204, and the third server cluster 206, which are sequentially connected within the liquid distribution loop 208. Each of the first server cluster 202, the second server cluster 204, and the third server cluster 206 includes a set of electronic processing components and a corresponding set of liquid cooling units, which are connected in parallel within the given server cluster.

[0042] The first server cluster 202 includes a first group of electronic processing units 202b1 to 202b5 and a first group of liquid cooling units 202a1 to 202a5. Each liquid cooling unit (e.g., 202a1) is in direct thermal contact with a corresponding electronic processing unit (e.g., 202b1). The first group of liquid cooling units 202a1 to 202a5 receive cooling liquid from a distribution loop 208 and return the cooled liquid to the distribution loop 208. Although five liquid cooling units are illustrated in the first server cluster 202, it is not limited to this. The first server cluster 202 may be considered to include N cooling units.

[0043] The second server cluster 204 includes a second set of electronic processing units 204b1 to 204b4 and a second set of liquid cooling units 204a1 to 204a4. Each liquid cooling unit (e.g., 204a1) is in direct thermal contact with the corresponding electronic processing unit (e.g., 204b1). The second set of liquid cooling units 204a1 to 204a4 receives cooling liquid flowing from the first server cluster 202 from the liquid distribution circuit 208 and returns the cooling liquid to the liquid distribution circuit 208. Although four liquid cooling units are illustrated in the second server cluster 204, it is not limited to this. It is conceivable that the second server cluster 204 includes M cooling units.

[0044] The third server cluster 206 includes a third group of electronic processing units 206b1 to 206b3 and a third group of liquid cooling units 206a1 to 206a3. Each liquid cooling unit (e.g., 206a1) is in direct thermal contact with the corresponding electronic processing unit (e.g., 206b1). The third group of liquid cooling units 206a1 to 206a3 receive cooling liquid flowing from the second server cluster 204 from the liquid distribution circuit 208 and return the cooled liquid to the liquid distribution circuit 208. Although three liquid cooling units are illustrated in the third server cluster 206, it is not limited to this. It is conceivable that the third server cluster 206 includes P cooling units.

[0045] As described above, the number of liquid cooling units included in each of the first server cluster 202(N), the second server cluster 204(M), and the third server cluster 206(P) is different, where M > N > P. In other words, as the liquid distribution loop 208 circulates the cooling liquid through the plurality of server clusters 201, each subsequent downstream server cluster includes fewer liquid cooling units than its upstream server cluster. In some non-limiting embodiments, a configuration can be considered in which the number of liquid cooling units does not decrease at each subsequent server cluster, depending on the type of electronic processing components included in the server cluster. For example, a configuration can be considered in which M < N < P, or M > N > P.

[0046] In some non-limiting embodiments of the present technology, it can also be considered that the types of liquid cooling units included in the first server cluster 202, the second server cluster 204, and the third server cluster 206 are different from each other. In other words, it can be considered that the docking portions of the corresponding first set of liquid cooling units 202a1 to 202a5, the second set of liquid cooling units 204a1 to 204a4, and the third set of liquid cooling units 206a1 to 206a3 have different geometries for circulating the cooling liquid. This variation allows the liquid cooling device to adapt to different cooling requirements at each cluster. Needless to say, it can also be considered that a given set of liquid cooling units (such as the first set of liquid cooling units 202a1 to 202a5) can include liquid cooling units with docking portions of different geometries.

[0047] As shown in the figure, each data processing component in the data processing components (such as the data processing component 100) incorporates at least one corresponding liquid cooling block unit (such as the block unit 102) arranged to be in direct thermal contact with the heat-generating electronic processing component. The block unit 102 is configured with internal conduits (not shown) to accommodate a circulating flow of channelized cooling liquid flowing through the internal conduits. The channelized cooling liquid is provided by a cooling liquid supplier, which is serially conveyed via an internal server cluster liquid circulation channel 30 to each of the liquid cooling block units in the liquid cooling block units to absorb the thermal energy from the heat-generating electronic components and discharge the heated liquid from the liquid cooling block units.

[0048] By reducing the number of liquid cooling units, the present technology allows the flow rate and pressure difference (ΔP) of the cooling liquid of the liquid cooling units to increase at each subsequent server cluster because the same volume of cooling liquid has to flow through fewer liquid cooling units.

[0049] Refer to Figure 4 , in which a schematic diagram 400 is depicted, which illustrates the flow rate and ΔP of the cooling liquid of the liquid distribution loop 208.

[0050] Pump 216 is configured to control the flow rate of coolant entering the first server cluster 202. The flow rate of coolant gradually increases downstream, such that the flow rate of coolant entering the first group of liquid cooling units 202a1 to 202a5 is lower than the flow rate of coolant entering the second group of liquid cooling units 204a1 to 204a4, and the flow rate of coolant entering the second group of liquid cooling units 204a1 to 204a4 is lower than the flow rate of coolant entering the third group of liquid cooling units 206a1 to 206a3.

[0051] In some non-limiting embodiments, the flow state of the cooling liquid can vary between the first group of liquid cooling units 202a1 to 202a5, the second group of liquid cooling units 204a1 to 204a4, and the third group of liquid cooling units 206a1 to 206a3. For example, the cooling liquid can circulate in a laminar flow state 402 only in the first group of liquid cooling units 202a1 to 202a5, while the cooling liquid circulates in a turbulent flow state 404 in the second group of liquid cooling units 204a1 to 204a4 and the third group of liquid cooling units 206a1 to 206a3. In other non-limiting embodiments, it is conceivable that the cooling liquid circulates in a turbulent flow state 404 in the first group of liquid cooling units 202a1 to 202a5, the second group of liquid cooling units 204a1 to 204a4, and the third group of liquid cooling units 206a1 to 206a3.

[0052] The transition from laminar flow state 402 to turbulent flow state 404 in the liquid cooling device 200 is influenced by the physical properties of the cooling liquid, which change as the cooling liquid absorbs heat from the electronic processing components (as described above). As the cooling liquid warms downstream, its viscosity decreases, thereby reducing flow resistance and increasing the Reynolds number (Re), a key factor determining the flow state. Furthermore, the reduced number of liquid cooling units in the downstream server cluster leads to higher flow velocities and ΔP, further contributing to the increase in Re. This interaction of increased flow velocity and decreased viscosity causes the cooling liquid to tend towards turbulence within the liquid cooling units of the liquid distribution loop 208 located in the downstream server cluster (i.e., the second group of liquid cooling units 204a1 to 204a4 and the third group of liquid cooling units 206a1 to 206a3).

[0053] The advantage of circulating the coolant in turbulent flow state 404 lies in the improved heat capture of the coolant. Indeed, in turbulent flow state 404, irregular motion minimizes the thermal boundary layer near the heated surface, allowing the coolant to contact the heated surface more frequently, thereby enhancing the heat transfer rate compared to laminar flow. Therefore, this technology allows for varying the coolant flow rate to selectively place it in either laminar or turbulent flow states at the liquid cooling unit without requiring additional pumps between server clusters.

[0054] Although in the illustrated embodiment, the cooling liquid exhibits a laminar flow state 402 only in the first group of liquid cooling units 202a1 to 202a5, and a turbulent flow state 404 in the second group of liquid cooling units 204a1 to 204a4 and the third group of liquid cooling units 206a1 to 206a3, the embodiment is not limited to this. For example, the cooling liquid may achieve a turbulent flow state 404 only in the third group of liquid cooling units 206a1 to 206a3.

[0055] In this way, the liquid cooling unit 200 provides a configuration that optimizes overall cooling efficiency by separately handling the liquid cooling needs of high-priority server clusters and standard-priority server clusters.

[0056] Modifications and improvements to the above implementation of this technology will be apparent to those skilled in the art. The above description is intended to be exemplary and not restrictive. Therefore, the scope of this technology is limited only by the scope of the appended claims.

Claims

1. A liquid cooling system (200) for cooling multiple server clusters (202, 204, 206), each of the multiple server clusters including heat-generating devices (202b1-202b5, 204b1-204b4, 206b1-206b3) having liquid cooling units (202a1-202a5, 204a1-204a4, 206a1-202a3), the liquid cooling system comprising: Cooling circuit (208) is configured to circulate coolant to the plurality of server clusters (202, 204, 206) and to circulate heat-transforming liquid from the plurality of server clusters (202, 204, 206) back to the dry cooler unit (210). The multiple server clusters (202, 204, 206) include: The first cluster (202) and the second cluster (204) are connected in series through the cooling circuit (208). The first cluster (202) includes a first group of liquid cooling units (202a1-202a5), which are connected in parallel through the cooling circuit, and the first group has a first number of liquid cooling units (202a1-202a5). The second cluster (204) includes a second group of liquid cooling units (204a1-204a5) connected in parallel via the cooling circuit. The second group has a second number of liquid cooling units (204a1-204a5). The second cluster is located further downstream along the cooling circuit (208) and is configured to receive cooling liquid flowing from the first cluster (202). The first quantity is greater than the second quantity.

2. The liquid cooling system according to claim 1, wherein, The first cluster (202) further includes a first group of heating components (202b1-202b5), which has a first number of heating components (202b1-202b5), and the second cluster (204) further includes a second group of heating components (204b1-204b5), which has a second number of heating components (204b1-204b5).

3. The liquid cooling system according to claim 2, wherein, At least one of the first group of heating elements (202b1-202b5) is configured to operate at a temperature below a first operating upper limit housing temperature (302), and at least one of the second group of heating elements (204b1-204b5) is configured to operate at a temperature below a second operating upper limit housing temperature (304). The first operating upper limit housing temperature (302) is lower than the second operating upper limit housing temperature (304).

4. The liquid cooling system according to claim 3, wherein, The coolant entering the first cluster (202) is below the first operating limit housing temperature (302), and the coolant flowing out of the first cluster (202) is below the second operating limit housing temperature (304).

5. The liquid cooling system according to any one of claims 1 to 4, wherein, The cooling circuit (208) is configured to circulate the cooling liquid in the first set of liquid cooling units (202a1-202a5) at a first flow rate (402), and to circulate the cooling liquid in the second set of liquid cooling units (204a1-204a5) at a second flow rate (404). The first flow rate (402) is lower than the second flow rate (404).

6. The liquid cooling system according to claim 5, wherein, The cooling liquid circulates at the first flow rate (402) in either a laminar or turbulent flow state, and the cooling liquid circulates at the second flow rate (404) in a turbulent flow state.

7. The liquid cooling system according to any one of claims 1 to 6, wherein, The cooling circuit (208) has a first docking portion that docks with the first group of heating components and a second docking portion that docks with the second group of heating components, wherein the first geometry of the first docking portion is different from the second geometry of the second docking portion.

8. The liquid cooling system according to claim 1, further comprising one or more heat exchangers (212, 214), said one or more heat exchangers being fluidly connected to the cooling circuit (208), wherein, The one or more heat exchangers (212-214) include one of an air-liquid heat exchanger, a copper heat exchanger, or a plate heat exchanger.

9. The liquid cooling system according to claim 1, wherein, The dry cooler unit (210) is configured to cool the heated coolant received from the plurality of server clusters (202, 204, 206).

10. The liquid cooling system of claim 1, further comprising a pump (216) configured to control the flow of cooling liquid into the first cluster (202).

11. The liquid cooling system according to claim 1, wherein, The plurality of server clusters (202, 204, 206) further includes a third cluster (206), which is connected in series with the first cluster (202) and the second cluster (204) via the cooling circuit (208). The third cluster (206) includes a third group of liquid cooling units (206a1-206a3), which are connected in parallel via the cooling circuit. The third group (206a1-206a3) has a third number of liquid cooling units. The third cluster is located further downstream along the cooling circuit and is configured to receive cooling liquid flowing from the second cluster (204), wherein the second number is greater than the third number.

12. The liquid cooling system according to claim 6, wherein, The third cluster also includes a third set of heating elements (206b1-206b3) configured to operate at a temperature below a third operating upper limit housing temperature (306), the second operating upper limit housing temperature (304) being lower than the third operating upper limit housing temperature (306), and wherein the cooling liquid flowing out from the second cluster (204) is lower than the third operating upper limit housing temperature (306).

13. The liquid cooling system according to claim 1, wherein, The multiple server clusters are installed on server rack components.

14. The liquid cooling system according to claim 1, wherein, The multiple server clusters are installed on multiple server rack components.

15. A liquid cooling system (200) for cooling multiple server clusters (202, 204, 206), each of the multiple server clusters comprising heat-generating devices (202b1-202b5, 204b1-204b4, 206b1-206b3) having liquid cooling units (202a1-202a5, 204a1-204a4, 206a1-202a3), the liquid cooling system comprising: Cooling circuit (208) is configured to circulate coolant to the plurality of server clusters (202, 204, 206) and to circulate heat-transforming liquid from the plurality of server clusters (202, 204, 206) back to the dry cooler unit (210). One or more heat exchangers (212, 214), said one or more heat exchangers being fluidly connected to said cooling circuit (208), The multiple server clusters (202, 204, 206) include: The first cluster (202) and the second cluster (204) are connected in series through the cooling circuit (208). The first cluster (202) includes a first group of liquid cooling units (202a1-202a5), which are connected in parallel through the cooling circuit, and the first group has a first number of liquid cooling units (202a1-202a5). The second cluster (204) includes a second group of liquid cooling units (204a1-204a5) connected in parallel via the cooling circuit. The second group has a second number of liquid cooling units (204a1-204a5). The second cluster is located further downstream along the cooling circuit (208) and is configured to receive cooling liquid flowing from the first cluster (202). The first quantity is less than the second quantity.