Liquid cooling system

Abstract

Embodiments of the present disclosure provide a liquid cooling system, comprising: a first set of supply and return liquid loops, each connected to at least one cabinet for supplying a first cooling liquid to the corresponding cabinet; a first set of cold distribution units configured to supply the first cooling liquid to the first set of supply and return liquid loops, each cold distribution unit comprising a first circulation path and a second circulation path in thermal exchange with each other, the first cooling liquid being disposed in the first circulation path, and a second cooling liquid being disposed in the second circulation path, the first set of cold distribution units comprising at least one redundant cold distribution unit, the first circulation path of the at least one redundant cold distribution unit being connected to each of the first set of supply and return liquid loops through a first redundant distribution pipeline; and a primary side loop connected to the second circulation path of each cold distribution unit and configured to receive the second cooling liquid from an external cold source and provide the second cooling liquid to the second circulation path.

Description

Technical Field

[0001] The embodiments of this disclosure generally relate to the field of data center cooling technology, and more specifically, to a liquid cooling system. Background Technology

[0002] Currently, the mainstream liquid cooling solutions for data centers mainly include cold plate liquid cooling and immersion liquid cooling. Regardless of the type of liquid cooling used, the architectural design of the liquid cooling system, especially the redundancy of the cooling distribution units (CDUs), has a significant impact on the reliability, cooling performance, availability, cost, and data center space configuration of the liquid cooling system, thus affecting the design and deployment of liquid-cooled data centers. Conventional liquid cooling solutions can only make simple trade-offs between system design complexity and CDU redundancy, making it difficult to achieve good design results in both. For example, cooling solutions with lower design complexity and cost usually have lower CDU redundancy, while solutions with better redundancy tend to have more complex designs. Summary of the Invention

[0003] Embodiments of this disclosure provide a liquid cooling system that improves CDU redundancy while reducing system design complexity.

[0004] In one aspect of this disclosure, a liquid cooling system is provided, comprising: a first set of supply and return liquid loops, each of the first set of supply and return liquid loops being connected to at least one cabinet for supplying a first coolant to the respective cabinet; a first set of cooling capacity distribution units configured to supply the first coolant to the first set of supply and return liquid loops, each of the first set of cooling capacity distribution units including a first circulation path and a second circulation path for heat exchange with each other, the first circulation path containing the first coolant and the second circulation path containing a second coolant, wherein the first set of cooling capacity distribution units includes at least one redundant cooling capacity distribution unit, the first circulation path of the at least one redundant cooling capacity distribution unit being connected to each of the first set of supply and return liquid loops via a first redundant distribution conduit; and a primary side loop connected to the second circulation path of each of the first set of cooling capacity distribution units and configured to receive the second coolant from an external cold source and supply the second coolant to the second circulation path.

[0005] In another aspect of this disclosure, a data center is provided, including any of the liquid cooling systems described above.

[0006] It should be understood that the content described in this section is not intended to limit the key or essential features of the embodiments of this disclosure, nor is it intended to restrict the scope of this disclosure. Other features of this disclosure will become readily apparent from the following description. Attached Figure Description

[0007] The above and other features, advantages, and aspects of the embodiments of this disclosure will become more apparent from the accompanying drawings and the following detailed description. In the drawings, the same or similar reference numerals denote the same or similar elements, wherein:

[0008] Figure 1 and Figure 2 A schematic diagram of a liquid cooling system employing an N+1 redundancy scheme according to some embodiments of the present disclosure is shown;

[0009] Figures 3 to 5 A schematic diagram of a liquid cooling system employing an N+1 redundancy scheme according to other embodiments of the present disclosure is shown;

[0010] Figure 6 A schematic diagram of a liquid cooling system employing a distributed redundancy scheme according to an embodiment of the present disclosure is shown.

[0011] Figure 7 A schematic diagram of a liquid cooling system employing a distributed redundancy scheme according to another embodiment of the present disclosure is shown; and

[0012] Figure 8 and Figure 9 A schematic diagram of a liquid cooling system employing a distributed redundancy scheme according to some other embodiments of the present disclosure is shown. Detailed Implementation

[0013] Preferred embodiments of the present disclosure will now be described in more detail with reference to the accompanying drawings. While preferred embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure may be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided so that the present disclosure will be thorough and complete, and will fully convey the scope of the present disclosure to those skilled in the art.

[0014] The term "comprising" and its variations as used herein signify open inclusion, i.e., "including but not limited to". Unless otherwise stated, the term "or" means "and / or". The term "based on" means "at least partially based on". The terms "one example embodiment" and "one embodiment" mean "at least one example embodiment". The term "another embodiment" means "at least one additional embodiment". The terms "first", "second", etc., may refer to different or the same objects.

[0015] As mentioned above, conventional liquid cooling solutions can only make simple trade-offs between system design complexity and CDU redundancy, making it difficult to achieve good design results in both aspects. For example, cooling solutions with lower design complexity and cost usually have lower CDU redundancy, while solutions with better redundancy tend to have more complex designs. Next, we will first introduce some CDU redundancy solutions used in conventional liquid cooling systems in data centers, which can be selected based on the data center's requirements for CDU redundancy.

[0016] A conventional liquid cooling system can implement 1+1 redundancy for the circulating pumps built into the CDU, meaning redundant circulating pumps are installed inside the CDU instead of CDU-level redundancy. This redundancy scheme is low-cost; however, if a component within the CDU other than the circulating pump fails, the lack of redundant backup for that component will cause the corresponding CDU to malfunction. Therefore, this redundancy scheme is only suitable for scenarios where short-term cooling interruptions to the liquid cooling system are acceptable and where data center availability requirements are not very high.

[0017] Another conventional liquid cooling system can employ 1+1 redundancy at the CDU level. Due to the redundant CDUs, even if a single CDU fails, the redundant CDUs can still operate reliably, thus ensuring data center availability. Furthermore, the design of the secondary-side piping network in this redundancy scheme is relatively simple. However, the CDUs in this redundancy scheme typically need to operate in hot standby mode, which is costly. If the CDUs in this redundancy scheme operate in cold standby mode, problems with CDU switching can affect the reliability of the liquid cooling system. In addition, when the data center is large, this redundancy scheme usually requires very large CDU granularity design, such as at the 1-2MW level. Therefore, when the secondary-side piping network fails, the failure radius is large. Moreover, the large number of cabinets carried by the secondary-side piping network leads to poor flow balance. Therefore, manual or electric flow balancing valves are usually required for adjustment, resulting in longer deployment times and reduced reliability of the liquid cooling system.

[0018] Other conventional liquid cooling systems can employ CDU-level N+1 redundancy, where N > 1. The following will combine... Figures 1 to 5 An example of such a redundancy scheme is described. Figures 1 to 5 The principle of this disclosure is illustrated by showing an example of 3+1 redundancy as an N+1 redundancy scheme. It should be understood that N can take other values, such as 2+1 redundancy or 4+1 redundancy, and the embodiments of this disclosure are not limited thereto.

[0019] like Figure 1As shown, the liquid cooling system includes a set of CDUs 20, each CDU including a first circulation path and a second circulation path (not shown) for heat exchange with each other. A first coolant is disposed in the first circulation path. A second coolant is disposed in the second circulation path. The first and second coolants can be the same or different. The first and second circulation paths can exchange heat via plate heat exchangers and other conventional or future available heat exchange methods. The first circulation path of each CDU 20 is connected to a secondary-side loop 70 located on the secondary side of the liquid cooling system to supply the first coolant to the secondary-side loop 70. The secondary side of the liquid cooling system refers to the side of the liquid cooling system that is in communication with the electronic equipment to be cooled. Correspondingly, the primary side of the liquid cooling system refers to the side of the liquid cooling system that is in communication with an external cold source. The secondary-side loop 70 can be used as a fluid input and output static pressure pool for a set of supply and return loops 10. Each coolant supply / return loop 10 can be connected to at least one rack 40 to supply the first coolant received from the secondary-side loop 70 to the rack 40 and return the heated first coolant flowing out of the rack 40 to the secondary-side loop 70. The secondary-side loop 70 described herein can also be referred to as a secondary-side coolant supply / return mini-loop network. The coolant supply / return loop 10 described herein can also be referred to as a server mini-loop network. Accordingly, Figure 1 The liquid cooling scheme shown can also be called a small-ring network redundancy scheme. The second circulation path of each CDU 20 is connected to the primary-side loop 30 located on the primary side of the liquid cooling system. The primary-side loop 30 is connected to an external cold source to supply the second coolant provided by the external cold source to the second circulation path of each CDU 20 and to return the heated second coolant flowing out of the second circulation path to the external cold source.

[0020] Using the above arrangement, the supply and return loop 10 can uniformly obtain low-temperature first coolant from the static pressure pool formed by the secondary loop 70, carry away the heat from the cabinet 40 through the first coolant, and then return the heated first coolant to the secondary loop 70. Subsequently, the heat can be transferred to the primary loop 30 through the heat exchange inside the CDU 20, and finally dissipated to the external cold source.

[0021] exist Figure 1 In the N+1 redundancy scheme of the liquid cooling system shown, CDU 20 has a relatively low cost. Furthermore, when a single CDU 20 fails, the remaining CDUs 20 can collectively ensure the availability of the liquid cooling system. Additionally, N>1 ensures high reliability of the liquid cooling system during switchover or normal operation of each CDU 20. However, the secondary-side loop 70 increases the overall system cost. Moreover, the secondary-side loop 70 requires significant space, therefore CDU 20 can only be placed in a separate equipment room and cannot be arranged side-by-side with rack 40, meaning it cannot be integrated into the system.

[0022] When there are a large number of server racks (40) in the same server room of a data center, Figure 1 The redundancy scheme shown may encounter a surge in the required pipe diameter for the secondary loop 70. To avoid this surge, the CDU 20, secondary loop 70, and the connected supply and return fluid loop 10 need to be split into two or more groups, each achieving N+1 redundancy. Figure 2 Such an arrangement is shown in the image. Figure 2 As shown, the liquid cooling system includes two sets of CDU 20, two secondary side loops 70, and two sets of supply and return liquid loops 10. Figure 2 The N+1 redundancy scheme shown has a relatively appropriate CDU granularity. However, if the number of racks 40 in the data center does not match this granularity, good scalability cannot be achieved. Therefore, sufficient redundancy must be achieved by sacrificing economy, or economy must be maintained by sacrificing redundancy. For example, when the number of racks 40 in the data center is far lower than the optimal design number for the liquid cooling system, if each group still uses 3+1 redundancy, it will result in a huge waste of economy. If the number of CDUs is reduced to maintain economic performance, such as from 3+1 redundancy to 2+1 redundancy, or even 1+1 redundancy, the reliability of the liquid cooling system will be affected.

[0023] Figure 3 An N+1 redundancy scheme is shown, forming a large supply and return liquid loop network on the secondary side of the liquid cooling system. For example... Figure 3 As shown, the liquid cooling system includes a set of CDUs 20. The first circulation path of each CDU 20 is connected to a secondary-side loop 70 located on the secondary side of the liquid cooling system, forming a large supply and return loop network. Multiple internal pipes 71 are formed within the secondary-side loop 70, each internal pipe 71 connecting to multiple cabinets 40 to supply the first coolant from the secondary-side loop 70 to the cabinets 40 and return the heated first coolant flowing from the cabinets 40 to the secondary-side loop 70. The second circulation path of each CDU 20 is connected to a primary-side loop 30 located on the primary side of the liquid cooling system. The primary-side loop 30 is connected to an external cold source to supply the second coolant provided by the external cold source to the second circulation path of each CDU 20 and return the heated second coolant flowing from the second circulation path to the external cold source.

[0024] Using the above arrangement, CDU 20 can directly supply the first coolant to the secondary side loop 70 forming a large loop network. The secondary side loop 70 delivers the first coolant to the internal piping 71 at appropriate locations. Cabinet 40 directly receives the first coolant from the internal piping 71 to cool the electronic equipment within. The heat dissipated by the electronic equipment in cabinet 40 is carried away by the first coolant. The heated first coolant returns to the secondary side loop 70 via the return pipe of the internal piping 71. Subsequently, the heat can be transferred to the second coolant in the primary side loop 30 via the heat exchanger inside CDU 20, and finally dissipated to an external cold source.

[0025] Figure 3 The N+1 redundancy scheme shown has a simple design logic. Furthermore, when a single CDU 20 fails, the remaining CDUs 20 can work together to ensure the availability of the liquid cooling system. Additionally, N>1 ensures high reliability of the liquid cooling system during switching or normal operation of each CDU 20. Moreover, in this redundancy scheme, CDUs 20 can be placed side-by-side with cabinets 40, i.e., in parallel cabinet configuration, to save space. However, because the secondary loop 70 adopts a large loop network design, the pipe diameter is generally large, resulting in higher costs. Furthermore, the large loop network design easily leads to uneven flow distribution. When there are many cabinets 40, cabinet-level static flow balancing valves are generally required to achieve flow balance, which can lead to valve over-provisioning. Furthermore, when deploying and adjusting the static flow balancing valves, cabinet-level flow balance will affect each other. Since the flow balance requirements under normal operating conditions and worst-case operating conditions (when a single point of failure occurs in the pipeline or a CDU fails) are different, a large deviation in flow balance will occur when unfavorable operating conditions occur. Furthermore, large ring network designs make it difficult to achieve segmented isolation, resulting in a large fault domain.

[0026] Figure 4 and Figure 5 The redundancy scheme shown is Figure 3 An extended scheme of the redundancy scheme shown. For example... Figure 4 As shown, the number of CDU 20s has been increased and the diameter of the secondary loop 70 has been increased, while the diameter of the internal pipe 71 can remain essentially unchanged. This allows for cooling of more cabinets 40. Figure 5As shown, the number of CDU 20s remains unchanged; instead, the granularity of the CDU 20s is increased, i.e., the drive capacity of their internal circulation pumps is increased, for example, from 300kW to 600kW. However, when the number of racks 40 increases to a certain scale, both of the above solutions have insurmountable bottlenecks. First, as the number of racks 40 increases, the pipe diameter of the secondary side loop 70 will surge. When the number of racks 40 increases to a certain scale, it is almost impossible to expand further. In addition, the larger the large loop network formed by the secondary side loop 70, the more serious the problem of flow imbalance becomes. Therefore, in addition to over-provisioning static or dynamic balancing valves, adjusting the balance when deploying balancing valves also becomes very difficult. Especially when a single point of failure occurs, the flow of each rack 40 is very likely to become unbalanced, causing overheating of electronic equipment such as servers in the racks 40.

[0027] It is evident that balancing the design difficulty, cost, and reliability of liquid cooling systems is crucial for the design of large or ultra-large-scale data centers. A highly adaptable data center infrastructure solution is essential for large liquid-cooled data centers. To address the aforementioned issues with conventional liquid cooling solutions, embodiments of this disclosure provide a liquid cooling system employing a distributed redundancy scheme, achieving a highly adaptable data center liquid cooling solution. It achieves excellent results in terms of cost, redundancy, and design difficulty, rather than simply balancing these factors. This solution significantly reduces the design complexity of the N+1 redundancy architecture in data centers and improves redundancy. Furthermore, this solution achieves flow balancing requirements through a simple design and without the need for excessive valve configuration. Moreover, this solution is flexibly compatible with both parallel (co-located) and separate (non-co-located) CDU configurations with racks, achieving an N+1 redundancy architecture for CDUs without the need for complex CDU interconnection rings. Furthermore, this distributed redundancy architecture effectively pools CDU redundancy, improving redundancy in the event of a single point of failure while maintaining a simple architecture design.

[0028] The following will combine Figures 6 to 9 This disclosure describes a distributed redundancy scheme according to embodiments of the present disclosure. The embodiments of this disclosure will use cold plate liquid cooling as an example to describe the architecture and principle of a liquid cooling system; however, it should be understood that the architecture and principle of this disclosure are equally applicable to immersion liquid cooling. Furthermore, Figures 6 to 9 The principle of this disclosure is illustrated by showing an example of 3+1 redundancy as an N+1 redundancy scheme. It should be understood that N can take other values, such as 2+1 redundancy or 4+1 redundancy, and the embodiments of this disclosure are not limited thereto.

[0029] Figure 6A schematic diagram of a liquid cooling system employing a distributed redundancy scheme according to an embodiment of the present disclosure is shown. Figure 6 As shown, the liquid cooling system described herein generally includes a first set of liquid supply and return loops 11, a first set of cooling capacity distribution units (CDUs) 21, and a primary side loop 30.

[0030] Each of the first set of supply and return loops 11 is connected to at least one rack 40 for supplying a first coolant to the corresponding rack 40. The rack 40 houses electronic equipment, such as servers and switches. The first set of supply and return loops 11 can supply the first coolant to the rack 40 to remove heat generated by the electronic equipment during operation. The first coolant can remove the heat generated by the electronic equipment in the rack 40 through cold plate liquid cooling, immersion liquid cooling, or any other suitable form of liquid cooling, which is not limited by the embodiments of this disclosure. Each supply and return loop 11 includes a supply line 111 and a return line 112 connected to the rack. The supply line 111 can receive the first coolant from the first set of cooling capacity distribution units 21 and supply the first coolant to the rack 40. The first coolant, after being heated in the rack 40, can flow into the return line 112 and further return to the first set of cooling capacity distribution units 21 for further cooling.

[0031] Each cooling capacity distribution unit in the first set of cooling capacity distribution units 21 includes a first circulation path and a second circulation path (not shown) for heat exchange with each other. A first coolant is disposed in the first circulation path. A second coolant is disposed in the second circulation path. The first and second coolants may be the same or different. The first and second circulation paths may exchange heat via plate heat exchangers and other conventional or future available heat exchange methods. The first circulation path is connected to a first set of supply and return liquid loops 11 to supply the first coolant at a low temperature and to receive the heated first coolant returned from the first set of supply and return liquid loops 11. The second circulation path is connected to a primary side loop 30 to receive the heated second coolant from the primary side loop 30 and to return the heated second coolant to the primary side loop 30.

[0032] In one embodiment, such as Figure 6As shown, the first group of cooling capacity distribution units 21 includes multiple dedicated cooling capacity distribution units 212 corresponding one-to-one with the first group of supply and return liquid loops 11, and a first redundant cooling capacity distribution unit 210 shared by the first group of supply and return liquid loops 11. The first circulation path of each dedicated cooling capacity distribution unit 212 is connected to a corresponding supply and return liquid loop in the first group of supply and return liquid loops 11 to supply first coolant to the corresponding supply and return liquid loop and to receive heated first coolant from the corresponding supply and return liquid loop for further cooling. The first circulation path of the first redundant cooling capacity distribution unit 210 is connected to each supply and return liquid loop in the first group of supply and return liquid loops 11 via a first redundant distribution pipe 61 to supply first coolant to each supply and return liquid loop and to receive heated first coolant from each supply and return liquid loop for further cooling.

[0033] The primary loop 30 is connected to the second circulation path of each cooling capacity distribution unit in the first group of cooling capacity distribution units 21 for receiving low-temperature second coolant from an external cold source and supplying the low-temperature second coolant to the second circulation path. The primary loop 30 can supply the low-temperature second coolant received from the external cold source to the second circulation path of each cooling capacity distribution unit in the first group of cooling capacity distribution units 21, and receive heated second coolant from the second circulation path of each cooling capacity distribution unit in the first group of cooling capacity distribution units 21 and further return the heated second coolant to the external cold source to dissipate heat.

[0034] The liquid cooling system according to the embodiments of this disclosure has a simple architecture, and N+1 CDUs 20 do not require the same... Figure 1 and Figure 2 Sufficient redundancy can be achieved by forming a small loop network for secondary side liquid supply and return, as shown in the liquid cooling system. Therefore, each cooling capacity distribution unit in the first group of cooling capacity distribution units 21 can be set up independently with the cabinet 40, i.e., not in the same cabinet, or it can be set up side by side with the cabinet 40, i.e., in the same cabinet, thus facilitating the layout arrangement of the data center computer room.

[0035] Figure 6 The diagram shows a scheme in which each cooling capacity distribution unit in the first group of cooling capacity distribution units 21 is set up separately from the cabinet 40, i.e., a non-cabinet-connected scheme. Figure 7 This illustrates a configuration where each cooling capacity distribution unit in the first group of cooling capacity distribution units 21 is arranged side-by-side with the cabinet 40, i.e., a parallel cabinet configuration. For example... Figure 7 As shown, each cooling capacity distribution unit in the first group of cooling capacity distribution units 21 can be configured to have a similar size to the cabinet 40 so that it can be arranged side by side with the cabinet 40. Figure 7 Other structures of the liquid cooling system shown are similar to Figure 6 The liquid cooling system shown is similar and will not be described in detail here.

[0036] Alternatively or alternatively, in some embodiments, Figure 6 and Figure 7 At least one of the multiple dedicated cooling capacity distribution units 212 shown can be changed to a redundant cooling capacity distribution unit similar to the first redundant cooling capacity distribution unit 210, so that it can be shared by the first set of supply and return liquid loops 11. In some embodiments, Figure 6 and Figure 7 All or even all of the dedicated cooling capacity distribution units 212 shown can be converted into redundant cooling capacity distribution units. In such an embodiment, the redundancy performance of the liquid cooling system can be further improved, thereby enhancing the system's reliability.

[0037] When there are a large number of server racks (40) in the same server room of a data center, the liquid cooling system can include multiple supply and return liquid loops and multiple cooling capacity distribution units. Figure 8 Such an embodiment is shown. For example... Figure 8 As shown, in addition to the first set of supply and return liquid loops 11, the first set of cooling capacity distribution units 21, and the primary side loop 30, the liquid cooling system also includes a second set of supply and return liquid loops 12 and a second set of cooling capacity distribution units 22. The structure of the second set of supply and return liquid loops 12 can be similar to that of the first set of supply and return liquid loops 11. Each supply and return liquid loop in the second set of supply and return liquid loops 12 is connected to at least one cabinet 40 for supplying first coolant to the corresponding cabinet 40. The structure of the second set of cooling capacity distribution units 22 can be similar to that of the first set of cooling capacity distribution units 21. The second set of cooling capacity distribution units 22 is used to supply first coolant to the second set of supply and return liquid loops 12. Each cooling capacity distribution unit in the second set of cooling capacity distribution units 22 includes a first circulation path and a second circulation path. The second set of cooling capacity distribution units 22 includes a plurality of dedicated cooling capacity distribution units 212 corresponding one-to-one with the second set of supply and return liquid loops 12, and a second redundant cooling capacity distribution unit 220 shared by the second set of supply and return liquid loops 12. Each dedicated cooling capacity distribution unit 212 has a first circulation path connected to a corresponding supply and return liquid loop in the second set of supply and return liquid loops 12 to supply first coolant to the corresponding supply and return liquid loop and receive heated first coolant from the corresponding supply and return liquid loop for further cooling. The second redundant cooling capacity distribution unit 220 is connected to each supply and return liquid loop in the second set of supply and return liquid loops 12 via a second redundant distribution pipe 62 to supply first coolant to each supply and return liquid loop in the second set of supply and return liquid loops 12 and receive heated first coolant from each supply and return liquid loop for further cooling. The second circulation path of each cooling capacity distribution unit in the second set of cooling capacity distribution units 22 is connected to the primary side loop 30 to receive low-temperature second coolant from the primary side loop 30 and return heated second coolant to the primary side loop 30.

[0038] In some embodiments, such as Figure 8As shown, the second redundant distribution pipe 62 is connected to the first redundant distribution pipe 61. In this way, the first redundant cooling capacity distribution unit 210 and the second redundant cooling capacity distribution unit 220 can be shared by all supply and return liquid loops in the first set of supply and return liquid loops 11 and the second set of supply and return liquid loops 12, further improving the redundancy performance of the liquid cooling system. Therefore, when the number of server racks 40 in the same computer room of a data center is large, the redundancy of the system can be further improved simply by interconnecting the redundant distribution pipes, without increasing the pipe diameter. This allows the distributed liquid cooling system to be upgraded to a distributed dual-redundancy scheme with higher redundancy performance without a sudden increase in pipe diameter or sacrificing flow balance in the system design.

[0039] Figure 8 The redundancy expansion scheme shown is based on the combination of Figure 6 This is an extension based on the described non-consolidation scheme. It should be understood that such a redundant extension scheme is also applicable to combinations... Figure 7 The described cabinet-by-cabinet scheme. This redundant expansion scheme has strong scalability, does not need to consider the granularity and pipe diameter limitations of the cooling capacity distribution unit, and the redundancy can be appropriately increased as needed during expansion, thereby improving system reliability.

[0040] In some embodiments, similar to the first group of cooling capacity distribution units 21, one, several, or even all of the multiple dedicated cooling capacity distribution units 212 of the second group of cooling capacity distribution units 22 can be changed into redundant cooling capacity distribution units similar to the second redundant cooling capacity distribution unit 220, so that they can be shared by the second group of supply and return liquid loops 12, thereby further improving the redundancy performance of the liquid cooling system and improving the reliability of the system.

[0041] Figure 9 A schematic diagram of a liquid cooling system employing a distributed redundancy scheme according to some other embodiments of the present disclosure is shown. Figure 9 The liquid cooling system shown is Figure 8 The liquid cooling systems shown have similar structures. The differences between the two will be described in detail below, while the same parts will not be repeated.

[0042] In some embodiments, such as Figure 9 As shown, a fault isolation valve 50 is installed in the first redundant distribution pipeline 61 and the second redundant distribution pipeline 62, as well as between the first redundant distribution pipeline 61 and the second redundant distribution pipeline 62. Using this arrangement, when a single point of failure occurs in the pipeline, by controlling the opening and closing state of the fault isolation valve 50, fault isolation can be achieved while ensuring the normal operation of the system, thus limiting the maximum fault domain to only a single supply and return fluid loop, i.e., a single server small loop network.

[0043] The fault isolation valve 50 can be either a manual or an electric valve. When an electric valve is used, automatic fault isolation can be achieved in the event of a pipeline fault. Alternatively, the fault isolation valve 50 can be a seated valve, a ball valve, or any other suitable type of valve. When using a seated valve, not only can rapid fault isolation be achieved in the event of a pipeline fault, but the flow of redundant cooling capacity distribution units into each server's small ring network can also be adjusted during normal operation or fault mode, enhancing flexibility.

[0044] The advantages of the liquid cooling system employing a distributed redundancy scheme according to the embodiments of this disclosure are summarized as follows: The distributed design significantly reduces piping complexity and overall cost; the flow equalization of the server small ring network is guaranteed, achieving natural hydraulic balance without the need for over-configuration of static or dynamic balancing valves; the flow balance characteristics of the server small ring network are similar under normal and worst-case operating conditions, thus eliminating the need for repeated adjustments to system flow balancing during installation and deployment; the liquid cooling system can be flexibly expanded as needed, without requiring a dramatic increase in the diameter of redundant pipes, maintaining only the normal pipe diameter for interconnection, significantly reducing the space requirements for data center piping; it can be flexibly expanded at the server small ring network level as needed, and as the number of server small ring networks increases, the redundant CDUs, interconnected through redundant piping, achieve a redundant CDU pooling effect, resulting in a natural and synchronous increase in redundancy and enhanced reliability; due to the absence of a secondary-side supply and return small ring network design, CDUs can be flexibly integrated with or separated from the server racks, offering greater adaptability to data center layouts; in the event of a single point of failure, the fault can be quickly isolated, preventing large-scale fault domains and ensuring operational reliability.

[0045] An embodiment of this disclosure also provides a data center including any of the liquid cooling systems described above.

[0046] Embodiments of this disclosure are also embodied in the following examples.

[0047] Example 1. A liquid cooling system, comprising:

[0048] The first set of supply and return loops, each of the first set of supply and return loops being connected to at least one cabinet for supplying the first coolant to the corresponding cabinet;

[0049] A first set of cooling capacity distribution units is configured to supply the first coolant to a first set of supply and return fluid loops. Each cooling capacity distribution unit in the first set of cooling capacity distribution units includes a first circulation path and a second circulation path for heat exchange with each other. The first circulation path contains the first coolant, and the second circulation path contains the second coolant. The first set of cooling capacity distribution units includes at least one redundant cooling capacity distribution unit. The first circulation path of the at least one redundant cooling capacity distribution unit is connected to each supply and return fluid loop in the first set of supply and return fluid loops via a first redundant distribution pipeline.

[0050] A primary side loop is connected to the second circulation path of each of the first group of cooling capacity distribution units and is configured to receive the second coolant from an external cooling source and supply the second coolant to the second circulation path.

[0051] Example 2. The liquid cooling system according to Example 1, wherein each of the first group of cooling capacity distribution units is arranged side by side with or separately from the at least one cabinet.

[0052] Example 3. The liquid cooling system according to Example 1, wherein the first group of cooling capacity distribution units includes a plurality of dedicated cooling capacity distribution units corresponding one-to-one with the first group of supply and return liquid loops, and the first circulation path of each dedicated cooling capacity distribution unit is connected to the corresponding supply and return liquid loop.

[0053] Example 4. The liquid cooling system according to Example 3, wherein the at least one redundant cooling capacity distribution unit includes a first redundant cooling capacity distribution unit, the first redundant cooling capacity distribution unit being connected to each of the first set of supply and return liquid loops via the first redundant distribution pipeline.

[0054] Example 5. The liquid cooling system according to Example 4 further includes:

[0055] A second set of supply and return loops, each of which is connected to at least one cabinet, for supplying the first coolant to the corresponding cabinet; and

[0056] The second set of cooling capacity distribution units is configured to supply the first coolant to the second set of supply and return liquid loops. Each cooling capacity distribution unit in the second set of cooling capacity distribution units includes the first circulation path and the second circulation path. The second set of cooling capacity distribution units includes a plurality of dedicated cooling capacity distribution units corresponding one-to-one with the second set of supply and return liquid loops and a second redundant cooling capacity distribution unit. The second redundant cooling capacity distribution unit is connected to each supply and return liquid loop in the second set of supply and return liquid loops through a second redundant distribution pipeline.

[0057] Example 6. The liquid cooling system according to Example 5, wherein the second redundant distribution line is connected to the first redundant distribution line.

[0058] Example 7. The liquid cooling system according to Example 6, wherein a fault isolation valve is provided in the first redundant distribution line and the second redundant distribution line, and between the first redundant distribution line and the second redundant distribution line.

[0059] Example 8. The liquid cooling system according to Example 7, wherein the fault isolation valve includes at least one of a manual valve and an electric valve.

[0060] Example 9. The liquid cooling system according to Example 7, wherein the fault isolation valve includes at least one of a seat valve and a ball valve.

[0061] Example 10. A liquid cooling system according to any one of Examples 1 to 9, wherein the at least one cabinet is cooled by at least one of cold plate liquid cooling and immersion liquid cooling.

[0062] Example 11. A data center including a liquid cooling system according to any one of Examples 1 to 10.

[0063] The various embodiments of this disclosure have been described above. These descriptions are exemplary and not exhaustive, and are not limited to the disclosed embodiments. Many modifications and variations will be apparent to those skilled in the art without departing from the scope and spirit of the described embodiments. The terminology used herein is chosen to best explain the principles, practical application, or technical improvements to the embodiments in the market, or to enable others skilled in the art to understand the embodiments disclosed herein.

Claims

1. A liquid cooling system, comprising: A first set of supply and return loops (11), each of the first set of supply and return loops (11) is connected to at least one cabinet (40) for supplying a first coolant to the corresponding cabinet (40); A first set of cooling capacity distribution units (21) is configured to supply the first coolant to a first set of supply and return fluid loops (11). Each cooling capacity distribution unit in the first set of cooling capacity distribution units (21) includes a first circulation path and a second circulation path for heat exchange with each other. The first circulation path contains the first coolant, and the second circulation path contains the second coolant. The first set of cooling capacity distribution units (21) includes at least one redundant cooling capacity distribution unit. The first circulation path of the at least one redundant cooling capacity distribution unit is connected to each supply and return fluid loop in the first set of supply and return fluid loops (11) via a first redundant distribution pipe (61). A primary side loop (30) is connected to the second circulation path of each of the first set of cooling capacity distribution units (21) and is configured to receive the second coolant from an external cooling source and supply the second coolant to the second circulation path.

2. The liquid cooling system according to claim 1, wherein each of the first group of cooling capacity distribution units (21) is arranged side by side with or separately from the at least one cabinet (40).

3. The liquid cooling system according to claim 1, wherein the first group of cooling capacity distribution units (21) includes a plurality of dedicated cooling capacity distribution units (212) corresponding one-to-one with the first group of supply and return liquid loops (11), and the first circulation path of each dedicated cooling capacity distribution unit (212) is connected to the corresponding supply and return liquid loop.

4. The liquid cooling system according to claim 3, wherein the at least one redundant cooling capacity distribution unit includes a first redundant cooling capacity distribution unit (210), the first redundant cooling capacity distribution unit (210) being connected to each of the first set of supply and return liquid loops (11) via the first redundant distribution pipeline (61).

5. The liquid cooling system according to claim 4, further comprising: The second set of supply and return loops (12), each of the second set of supply and return loops (12) is connected to at least one cabinet (40) for supplying the first coolant to the corresponding cabinet (40); as well as The second set of cooling capacity distribution units (22) is configured to supply the first coolant to the second set of supply and return liquid loops (12). Each cooling capacity distribution unit in the second set of cooling capacity distribution units (22) includes the first circulation path and the second circulation path. The second set of cooling capacity distribution units (22) includes a plurality of dedicated cooling capacity distribution units (212) corresponding one-to-one with the second set of supply and return liquid loops (12) and a second redundant cooling capacity distribution unit (220). The second redundant cooling capacity distribution unit (220) is connected to each supply and return liquid loop in the second set of supply and return liquid loops (12) through a second redundant distribution pipeline (62).

6. The liquid cooling system according to claim 5, wherein the second redundant distribution line (62) is connected to the first redundant distribution line (61).

7. The liquid cooling system according to claim 6, wherein a fault isolation valve (50) is provided in the first redundant distribution pipeline (61) and the second redundant distribution pipeline (62) and between the first redundant distribution pipeline (61) and the second redundant distribution pipeline (62).

8. The liquid cooling system according to claim 7, wherein the fault isolation valve (50) comprises at least one of a manual valve and an electric valve.

9. The liquid cooling system according to claim 7, wherein the fault isolation valve (50) comprises at least one of a seat valve and a ball valve.

10. The liquid cooling system according to any one of claims 1 to 9, wherein the at least one cabinet (40) is cooled by at least one of cold plate liquid cooling and immersion liquid cooling.

11. A data center comprising a liquid cooling system according to any one of claims 1 to 10.

Images