Thermal management board for key processors

Abstract

The cooling plate module includes: a first cooling plate layer having a single-phase region inside; and a second cooling plate layer having a phase change region inside. The first cooling plate layer includes: a first liquid inlet port to receive a first cooling liquid into the single-phase region; and a first liquid outlet port to discharge the first cooling liquid from the single-phase region. The second cooling plate layer includes: a second liquid inlet port to receive a second cooling liquid into the phase change region; and a vapor outlet port to discharge the second cooling liquid in a vapor state from the phase change region, wherein the first cooling plate layer is in thermal contact with the second cooling plate layer, and the first cooling plate layer is in thermal contact with an IT component to be cooled.

Description

Technical Field

[0001] Embodiments of the present invention generally relate to server and data center cooling. More specifically, embodiments of the present invention relate to thermal management plates for critical processors. Background Art

[0002] Modern information technology (IT) equipment (such as servers, blade servers, routers, edge servers, etc.) generates a lot of heat during operation. The heat generated by individual components (especially high-power components such as processors) makes many of these individual components impossible or difficult to effectively cool with air cooling systems. Therefore, modern IT equipment requires liquid cooling or liquid-air hybrid cooling.

[0003] A cold plate / cold plate is a component used in many liquid cooling solutions. A cold plate is a component that is coupled to a heat generating component such as a critical processor. Once the cold plate is coupled to the heat generating component, liquid is circulated through the cold plate to remove the heat. Existing electronic device cooling and thermal management solutions for processors do not provide redundancy at the cold plate module level, which means there is a single point of failure at the cold plate module level. Summary of the invention

[0004] An embodiment of the present invention provides a cooling plate module, comprising: a first cooling plate layer having a single-phase zone inside, wherein the first cooling plate layer comprises: a first liquid inlet port, which is connected to the first cooling plate layer to receive a first cooling liquid into the single-phase zone; and a first liquid outlet port, which is connected to the first cooling plate layer to discharge the first cooling liquid from the single-phase zone; and a second cooling plate layer, which has a phase change zone inside, wherein the second cooling plate layer comprises: a second liquid inlet port, which is connected to the second cooling plate layer to receive a second cooling liquid into the phase change zone; and a vapor outlet port, which is connected to the second cooling plate layer to discharge the second cooling liquid in a vapor state from the phase change zone, wherein a first portion of the first cooling plate layer is in thermal contact with a portion of the second cooling plate layer, and a second portion of the first cooling plate layer is in thermal contact with an IT component to be cooled.

[0005] In some embodiments, the first cooling plate layer further includes: a plurality of first liquid cooling fins to form first liquid channels in the single-phase region.

[0006] In some embodiments, the second cooling plate layer further includes: a plurality of second liquid cooling fins to form second liquid channels in the phase change region.

[0007] In some embodiments, each of the plurality of first liquid-cooling fins is spaced apart at a first pitch, and each of the plurality of second liquid-cooling fins is spaced apart at a second pitch.

[0008] In some embodiments, the plurality of first liquid-cooling fins are parallel to the plurality of second liquid-cooling fins and staggered with the plurality of second liquid-cooling fins to form staggered liquid channels.

[0009] In some embodiments, the plurality of first liquid-cooling fins are perpendicular to the plurality of second liquid-cooling fins, thereby forming a vertical liquid channel.

[0010] In some embodiments, the first cooling plate is coupled to a liquid manifold, and the liquid manifold actively promotes liquid flow in the single-phase region.

[0011] In some embodiments, flow in the phase change zone is facilitated by gravity because when the temperature of the second cooling plate layer is above a predetermined temperature, the liquid flowing in the phase change zone undergoes a phase change to a vapor state, wherein the vapor is discharged from the vapor outlet port and the liquid enters the second inlet port of the phase change zone.

[0012] In some embodiments, flow in the phase change zone is facilitated by pressure because the liquid flowing in the phase change zone undergoes a phase change to a vapor state, and the volume change in the phase change zone causes vapor to be discharged from the vapor outlet port, wherein the discharged vapor causes liquid at the liquid supply container to enter the second inlet port of the phase change zone.

[0013] In some embodiments, the first cooling plate layer and the second cooling plate layer are independently machined, and the single phase region is separated from the phase change region.

[0014] In some embodiments, the first cooling plate layer and the second cooling plate layer together are an integrated cooling plate machined into a single unit, wherein the integrated cooling plate includes a plurality of first liquid cooling fins on a first surface of the integrated cooling plate, wherein the integrated cooling plate includes a plurality of second liquid cooling fins on an opposite surface of the integrated cooling plate.

[0015] In some embodiments, the integrated cooling plate further comprises: a top frame covering the plurality of first liquid cooling fins; and a bottom frame covering the plurality of second liquid cooling fins, wherein the bottom frame is thermally coupled to the IT components to be cooled.

[0016] An embodiment of the present invention also provides a cooling system, comprising: a first liquid supply part and a liquid reflux part; a second liquid supply part and a steam reflux part; and at least one cooling plate module according to the above embodiment, wherein the at least one cooling plate module is connected to the first liquid supply part and the liquid reflux part and the second liquid supply part and the steam reflux part.

[0017] An embodiment of the present invention also provides a cooling plate module, comprising: an intermediate cooling plate; a base frame fixed to the bottom surface of the intermediate cooling plate, with a single-phase zone between the intermediate cooling plate and the base frame; a top frame fixed to the top surface of the intermediate cooling plate, with a phase change zone between the intermediate cooling plate and the top frame; a first liquid inlet port and a first liquid outlet port, the first liquid inlet port and the first liquid outlet port being connected to the intermediate cooling plate to form a first liquid channel in the single-phase zone; and a second liquid inlet port and a steam outlet port, the second liquid inlet port and the steam outlet port being connected to the intermediate cooling plate to form a second liquid channel in the phase change zone, wherein a portion of the base frame is in thermal contact with the IT component to be cooled.

[0018] In some embodiments, the intermediate cooling plate includes a plurality of first fins on a bottom surface of the intermediate cooling plate and a plurality of second fins on a top surface of the intermediate cooling plate. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] Embodiments of the invention are illustrated by way of example and not by way of limitation in the figures of the accompanying drawings in which like references indicate similar elements.

[0020] Figure 1 is a block diagram of a thermal management system according to one embodiment.

[0021] Figure 2 is a block diagram of a thermal management plate according to one embodiment.

[0022] Figure 3 is a block diagram of a thermal management plate according to one embodiment.

[0023] Figure 4 is a block diagram of a thermal management plate according to one embodiment.

[0024] Figure 5 is a block diagram illustrating thermal management for single phase mode according to one embodiment.

[0025] Figure 6 is a block diagram illustrating thermal management for phase change mode according to one embodiment.

[0026] Figure 7is a block diagram illustrating thermal management for hybrid mode according to one embodiment.

[0027] Figure 8 is a block diagram of one embodiment of an IT container including an electronics rack having electronics and a cooling system housed therein. DETAILED DESCRIPTION

[0028] Various embodiments and aspects of the present invention will be described with reference to the details discussed below, and the accompanying drawings will illustrate various embodiments. The following description and drawings are illustrative of the present invention and should not be construed as limiting the present invention. Many specific details are described to provide a thorough understanding of various embodiments of the present invention. However, in some cases, in order to concisely discuss embodiments of the present invention, known or conventional details are not described.

[0029] References to "one embodiment" or "an embodiment" in the specification mean that a particular feature, structure, or characteristic described in conjunction with the embodiment may be included in at least one embodiment of the present invention. The phrase "in one embodiment" appearing in different places in the specification does not necessarily refer to the same embodiment.

[0030] Aspects of the present disclosure relate to a thermal management plate module that can operate in a single-phase mode, a phase change mode, or a hybrid mode to circulate a cooling medium. A single-phase plate layer is designed on the bottom and is directly assembled to the processing unit to be cooled. A phase change or two-phase plate layer is designed on top of the single-phase plate layer. In one embodiment, an intermediate plate is designed with a microchannel structure on opposite sides, and the intermediate plate serves as an integrated unit for the single-phase plate layer and the phase change plate layer. The single-phase plate layer can be used for normal thermal management operations. The phase change layer is self-sensing and self-regulating, and the phase change layer can be activated when the interface temperature between the single-phase layer and the phase change layer is higher than the boiling temperature of the circulating liquid.

[0031] According to a first aspect, a cooling plate module includes: a first cooling plate layer having a single-phase region inside; and a second cooling plate layer having a phase change region inside. The first cooling plate layer includes: a first liquid inlet port, which is connected to the first cooling plate layer to receive a first cooling liquid into the single-phase region; and a first liquid outlet port, which is connected to the first cooling plate layer to discharge the first cooling liquid from the single-phase region. The second cooling plate layer includes: a second liquid inlet port, which is connected to the second cooling plate layer to receive a second cooling liquid into the phase change region; and a vapor outlet port, which is connected to the second cooling plate layer to discharge the second cooling liquid in a vapor state from the phase change region. A first portion of the first cooling plate layer is in thermal contact with a portion of the second cooling plate layer, and a second portion of the first cooling plate layer is in thermal contact with an IT component to be cooled.

[0032] In one embodiment, the first cooling plate layer further comprises a first number of liquid cooling fins to form a first liquid channel in the single phase region. In one embodiment, the second cooling plate layer further comprises a second number of liquid cooling fins to form a second liquid channel in the phase change region.

[0033] In one embodiment, each of the first liquid cooling fins is spaced apart by a first spacing, and each of the second liquid cooling fins is spaced apart by a second spacing. In one embodiment, the first liquid cooling fins are parallel to the second liquid cooling fins and staggered with the second liquid cooling fins to form staggered liquid channels.

[0034] In one embodiment, the first liquid cooling fin is perpendicular to the second liquid cooling fin, thereby forming a vertical liquid channel. In one embodiment, the first cooling plate is coupled to a liquid manifold, and the liquid manifold actively promotes liquid flow in the single-phase region.

[0035] In one embodiment, flow in the phase change zone is facilitated by gravity because when the temperature of the second cooling plate layer is above a predetermined temperature, the liquid flowing in the phase change zone undergoes a phase change to a vapor state, wherein the vapor is discharged from the vapor outlet port, and the liquid enters the second inlet port of the phase change zone by gravity. In one embodiment, flow in the phase change zone is facilitated by pressure because the liquid flowing in the phase change zone undergoes a phase change to a vapor state, and a density change in the phase change zone causes the vapor to be discharged from the vapor outlet port, wherein the discharged vapor causes liquid at the liquid supply container to enter the second inlet port of the phase change zone.

[0036] In one embodiment, the first cooling plate layer and the second cooling plate layer are independently machined, and the single phase region is separated from the phase change region. In one embodiment, the first cooling plate layer and the second cooling plate layer are together an integrated cooling plate machined into a single unit, wherein the integrated cooling plate includes a first number of liquid cooling fins on a first surface of the integrated cooling plate, wherein the integrated cooling plate includes a second number of liquid cooling fins on an opposite surface of the integrated cooling plate. In one embodiment, the integrated cooling plate further includes: a top frame covering the first liquid cooling fins; and a bottom frame covering the second liquid cooling fins, wherein the bottom frame is thermally coupled to the IT components to be cooled.

[0037] According to a second aspect, a cooling system includes: a first liquid supply and a liquid return; a second liquid supply and a vapor return, and at least one cooling plate module. At least one cooling plate module is coupled to the first liquid supply and the liquid return and the second liquid supply and the vapor return. The at least one cooling plate module includes: a first cooling plate layer having a single-phase region inside and a second cooling plate layer having a phase change region inside. The first cooling plate layer includes: a first liquid inlet port coupled to the first cooling plate layer to receive a first cooling liquid into the single-phase region; and a first liquid outlet port coupled to the first cooling plate layer to discharge the first cooling liquid from the single-phase region. The second cooling plate layer includes: a second liquid inlet port coupled to the second cooling plate layer to receive a second cooling liquid into the phase change region; and a vapor outlet port coupled to the second cooling plate layer to discharge the second cooling liquid in a vapor state from the phase change region. A first portion of the first cooling plate layer is in thermal contact with a portion of the second cooling plate layer, and a second portion of the first cooling plate layer is in thermal contact with an IT component to be cooled.

[0038] According to a third aspect, a cooling plate module includes an intermediate cooling plate. The cooling plate module includes a base frame fixed to a bottom surface of the intermediate cooling plate, with a single-phase region between the intermediate cooling plate and the base frame. The cooling plate module includes a top frame fixed to a top surface of the intermediate cooling plate, with a phase change region between the intermediate cooling plate and the top frame. The cooling plate module includes a first liquid inlet port and a first liquid outlet port, which are coupled to the intermediate cooling plate to form a first liquid channel in the single-phase region. The cooling plate module includes a second liquid inlet port and a vapor outlet port, which are coupled to the intermediate cooling plate to form a second liquid channel in the phase change region, wherein a portion of the base frame is in thermal contact with an IT component to be cooled.

[0039] Figure 1 is a side view of a thermal management system 100 according to one embodiment. The thermal management system 100 may include one or more processing units 101 (e.g., a central processing unit or CPU, a graphics processing unit (GPU), a tensor processing unit (TPU), or any type of processor or application specific integrated circuit (ASIC)). The processing unit 101 may perform data processing tasks by executing executable instructions stored in a storage device (not shown) and loaded into a memory (not shown) to perform data processing tasks.

[0040] Processing unit 101 may be part of a host server or a compute server (e.g., a CPU server or a GPU server). A host server typically interfaces with a client over a network (e.g., the Internet) to receive requests for specific services, such as storage services (e.g., cloud-based storage services, such as backup and / or recovery), or services to execute applications to perform certain operations (e.g., image processing, deep data learning algorithms or modeling, etc., as part of a software as a service or SaaS platform). In response to the request, the host server assigns the task to a compute server (having one or more GPUs) managed by the host server. The host server or compute server performs specific tasks, which generates heat during operation.

[0041] like Figure 1 As shown, in one embodiment, the cooling plate module (or plate module) 103 has a double plate layer implementation, wherein the plate module 103 includes a thermal management plate layer 105 (or simply cooling plate 105) and a thermal management plate layer 107 (or simply cooling plate 107). The cooling plate 105 may include a single-phase region 106 (e.g., a spatial volume or pocket within the cooling plate 106 as a fluid region), and the cooling plate 107 may include a phase change region 108 (e.g., a spatial volume or pocket within the cooling plate 107 that stores liquid / vapor). The bottom surface of the cooling plate 105 of the plate module 103 is directly attached to the surface of the processor unit 101 to extract the heat generated by the processor unit 101.

[0042] In one embodiment, the cooling plate 105 includes an inlet port 115 and an outlet port 125. The inlet port 115 and the outlet port 125 are used to exchange liquid (e.g., receive and discharge) in the single-phase region 106, wherein the inlet port 115 and the outlet port 125 are connected to a liquid supply line 116 and a liquid return line 126, respectively. In one embodiment, the cooling plate 107 includes an inlet port 117 and an outlet port 127. The inlet port 117 and the outlet port 127 are used to exchange liquid / vapor (e.g., receive and discharge) in the phase change region 108, wherein the inlet port 117 and the outlet port 127 are connected to a liquid supply line 118 and a vapor return line 128, respectively. Note that each cooling plate 105, 107 uses a dual port pair (e.g., 115, 125 and 117, 127), and each dual port pair can be connected to a corresponding heat exchange unit 135-137.

[0043] In one embodiment, heat exchange unit 135 is a facility-level single-phase (liquid) heat exchange unit with centralized pumping or a server-level single-phase heat exchange unit with localized pumping. In one embodiment, heat exchange unit 137 is a closed container with no active moving parts. That is, heat exchange unit 137 is designed as a self-activated loop to circulate cooling medium in a liquid and / or vapor state, wherein heat exchange unit 137 has no active pump to supply liquid or receive vapor return. The loop between 135 and 106 can include a fluid pump on 116 or 126.

[0044] In one embodiment, heat exchange unit 135 and heat exchange unit 137 are separated by a predetermined distance 140 to ensure thermal decoupling of exchange units 135 - 137 .

[0045] Figure 2 1 is a block diagram of a thermal management plate module 103 according to one embodiment. The thermal management plate module 103 includes a cooling plate 107 stacked on a cooling plate 105. The processing unit 101 may represent one or more high power density processors. As previously described, the cooling plate 105 is coupled to a liquid supply 116 and a liquid return 126. The cooling plate 107 is coupled to a liquid supply 118 and a vapor return 128.

[0046] In one embodiment, multiple heat transfer mechanisms are involved within the plate module 103. For example, the first heat transfer mechanism is conduction I at the interface 201, which conducts heat from the processing unit 101 to the cooling plate 105. Within the cooling plate 105, the second heat transfer mechanism is forced convection I, where liquid is forced through one or more liquid channels (such as Figure 3 to Figure 4 106. Forced convection I involves an active pump (not shown) that circulates liquid between a liquid supply 116 and a liquid return 126 to regulate the temperature at the cooling plate 105. Here, forced convection I is the main cooling or primary cooling mechanism for the plate module 103.

[0047] The third heat transfer mechanism is conduction II at the interface 203, which conducts heat between the single-phase zone 106 and the phase change zone 108. The fourth heat transfer mechanism is natural convection II. Natural convection II uses a liquid supply 118 and a vapor return 128 to circulate the cooling medium. Natural convection II is activated when the temperature at the interface 203 is higher than the boiling point of the cooling medium circulated by the cooling plate 107. Here, the cooling medium undergoes a phase change in the phase change zone 108, where different cooling media can be selected for different boiling points based on the desired temperature maintained at the interface 203. For example, the cooling medium can be water, a refrigerant coolant, or any other cooling medium treated to a certain pressure to select the desired boiling point temperature. In one embodiment, the cooling medium circulating in the phase change zone 108 is separate from the cooling liquid circulating in the single-phase zone 106. That is, there are two thermal circuits in the cooling plate module 103, and the cooling media of the two circuits are separated.

[0048] From an implementation perspective, the primary function of the plate module 103 is to maintain the temperature at the interface 201 below the specified operating temperature required by the processing unit 101. During normal operation, the temperature at the interface 203 can be maintained below the boiling point of the phase change fluid. When the temperature at the interface 203 is higher than the boiling point of the selected cooling medium, due to a malfunction, lack of cooling capacity, or any other similar situation, the cooling medium undergoes a phase change within the phase change zone 108. For example, when heat cannot be extracted quickly or sufficiently by forced convection I, natural convection II will be automatically triggered.

[0049] In one embodiment, during the design phase, based on the desired temperature at the interface 201, T1 may be selected as the supply liquid temperature for the cooling plate 105, while T2 may be selected as the return liquid temperature for the cooling plate 105. T3 may be the temperature at the supply line 118 and vapor return line 128 for the cooling medium. Here, the cooling medium in the phase change region 108 undergoes a phase change from liquid to vapor while extracting heat from the interface 203, even though the cooling medium, liquid, and vapor at the ports 118, 128 remain relatively the same at temperature T3.

[0050] Figure 3 is a block diagram of a thermal management board module 300 according to one embodiment. The thermal management board module 300 may represent Figure 1 to Figure 2 As shown, the thermal management plate module 300 includes cooling plates 105 and 107, wherein cooling plate 107 is stacked on top of cooling plate 105. In this design, the two heat exchange areas (106 and 108) can be designed and machined separately. Once the machining is completed, cooling plate 105 can be permanently (welded) or temporarily (hot melt adhesive) connected to cooling plate 107. In one embodiment, cooling plates 105 and 107 can include liquid cooling fins 301-303, such as Figure 3 As shown. Liquid cooling fins 301-303 can form one or more fluid channels to promote convection (e.g., forced convection I or natural convection II) for cooling plates 105 and 107, respectively, to promote the flow of cooling medium between corresponding inlet ports and outlet ports. In one embodiment, fins 301 are parallel to and staggered with fins 303, so that the fluid channels of fins 301 are staggered with the fluid channels of fins 303. In one embodiment, fins 301 are perpendicular to fins 303, so that the inlet port / outlet port for cooling plate 105 can be placed at a different position than cooling plate 107. In this case, the fluid channels in zone 106 are perpendicular to the fluid channels in zone 108.

[0051] Figure 4 is a block diagram of a thermal management board module 400 according to one embodiment. The thermal management board module 400 may represent Figure 1 to Figure 2 1. As shown, the thermal management plate includes an intermediate cooling plate 401 (or simply an intermediate plate 401), a bottom frame 403, and a top frame 405. The bottom frame 403 and the intermediate plate 401 together form a single-phase region 106. The top frame 405 and the intermediate plate 401 together form a phase change region 108. In one embodiment, the intermediate plate 401 includes a sheet 407 near the bottom surface. In one embodiment, the intermediate plate 401 includes a sheet 409 near the top surface. The sheets 407-409 can increase the contact surface area to improve heat dissipation and can form channels to guide the cooling medium from the inlet port to the outlet port. In one embodiment, the intermediate plate 401 is machined into a single unit with sheets 407-409. In one embodiment, the sheets 407 are staggered with the sheets 409, wherein the sheets 407-409 are all machined together in the intermediate plate 401. In one embodiment, the channel formed by the sheet 407 is separated from the channel formed by the sheet 409. For example, the liquid flowing in the single-phase region 106 is different from the cooling medium flowing in the phase change region 108 .

[0052] Figure 5 is a block diagram illustrating the thermal management plate 103 operating in a single phase mode according to one embodiment. Figure 5 A normal operating situation is shown, in which only the single-phase region 106 is used for cooling. In this case, since heat is extracted sufficiently quickly by the single-phase liquid circulating in the cooling plate 105, the temperature at the interface 203 will be lower than the boiling temperature of the cooling medium maintained in the phase change region 108. Here, no phase change is activated at the phase change region 108, and the cooling medium in the phase change region 108 remains in a liquid state, for example, no steam is generated.

[0053] Figure 6 is a block diagram illustrating the operation of the thermal management plate 103 in a phase change mode according to one embodiment. Figure 6A fault scenario is shown in which an external pump fails to circulate cooling liquid in the single-phase region 106, or the corresponding port of the plate 105 is stuck. In this case, the temperature at the interface 203 is higher than the boiling temperature of the cooling medium circulated by the phase change region 108 because heat is not extracted from the cooling plate 105 quickly enough. The higher temperature at the interface 203 automatically activates the phase change region 108, for example, causing the liquid medium in the phase change region 108 to undergo a phase change from liquid to gas. In one case, the thermal management system may not support heat removal from the processing unit 101. Here, the plate 107 and the liquid medium supply required by the plate 107 need to be designed to be able to remove a predetermined heat load threshold to ensure that the operator has enough time to diagnose the fault condition. In another case, when the temperature of the interface 203 is higher than the boiling point of the cooling medium in the phase change region 108, the higher temperature may cause the processing unit 101 to throttle performance to limit its heat output, whereby the processing unit 101 may continue to operate with limited performance.

[0054] Figure 7 is a block diagram illustrating thermal management for hybrid mode according to one embodiment. Figure 7 A normal operating situation with a mixed mode is shown, where both the single-phase region 106 and the phase change region 108 are active. This situation occurs when the cooling capacity from the single-phase region 106 is less than the heat generated from the processing unit 101 due to various reasons. For example, when the power of the processing unit 101 is much higher under a high computing environment (e.g., video rendering, training, etc.), even if the liquid is circulating in the single-phase region 106, the temperature at the interface 203 will be higher than the temperature at the interface 201. Here, the phase change region 108 can assist in extracting the heat load overflow from the single-phase region 106. In one embodiment, the liquid medium in the phase change region 108 can be selected so that a certain proportion of the heat is removed by the plate 105, and the rest of the heat (as designed by the cooling plate module) is removed by the plate 107.

[0055] Figure 81 is a block diagram illustrating an electronic rack according to one embodiment. The electronic rack 1200 may represent any electronic rack described in the present application. According to one embodiment, the electronic rack 1200 includes, but is not limited to, a coolant distribution unit (CDU) 1201, a rack management unit (RMU) 1202, and one or more server chassis 1203A-1203E (collectively referred to as server chassis 1203). The server chassis 1203 may be inserted into a server slot array (e.g., a standard shelf) from the front end 1204 or the rear end 1205 of the electronic rack 1200, respectively. Note that although five server chassis 1203A-1203E are shown here, more or fewer server chassis may be maintained within the electronic rack 1200. It is also noted that the specific locations of the CDU 1201, RMU 1202, and / or server chassis 1203 are for illustrative purposes only; other arrangements or configurations of the CDU 1201, RMU 1202, and / or server chassis 1203 may also be implemented. In one embodiment, electronics rack 1200 can be either open to the environment or partially contained by a rack container, as long as the cooling fans are able to generate airflow from front to back.

[0056] In one embodiment, the server chassis 1203A-1203E include cold plates / cold plate modules 103A-103E. Each cold plate module 103A-103E may include a heat exchange unit 137 (not shown). In one embodiment, the cold plate modules 103A-103E may share a heat exchange unit 137 (not shown). The heat exchange unit 137 may be a passive reservoir to store a liquid / vapor medium.

[0057] In addition, for at least some of the server chassis 1203, an optional fan module (not shown) is associated with the server chassis. Each fan module includes one or more cooling fans. The fan module can be mounted at the rear end of the server chassis 1203 or on the electronic rack to generate an airflow that flows from the front end 1204, through the air space of the server chassis 1203, and is present at the rear end 1205 of the electronic rack 1200.

[0058] In one embodiment, the CDU 1201 mainly includes a heat exchange unit 1211, a liquid pump 1212 and a pump controller (not shown) and some other components, such as a liquid reservoir, a power supply, a monitoring sensor, etc. The heat exchange unit 1211 can be a liquid-liquid heat exchange unit. The heat exchange unit 1211 includes a first loop having an inlet port and an outlet port, and the inlet and outlet ports have a first pair of liquid connectors connected to external liquid supply / return lines 1231-1232 to form a primary loop. The connector connected to the external liquid supply / return lines 1231-1232 can be placed or installed on the rear end 1205 of the electronic rack 1200. The liquid supply / return lines 1231-1232 (also referred to as room liquid supply / return lines) can be connected to an external cooling system.

[0059] In addition, the heat exchange unit 1211 further includes a secondary loop having two ports of a second pair of liquid connectors coupled to a liquid manifold 1225 (also referred to as a rack manifold) to form a secondary loop, which may include: a supply manifold (also referred to as a rack liquid supply line or a rack supply manifold) to supply cooling liquid to the server chassis 1203; and a return manifold (also referred to as a rack liquid return line or a rack return manifold) to return warmer liquid to the CDU 1201. Please note that the CDU 1201 may be any commercially available or custom CDU. Therefore, the details of the CDU 1201 will not be described here.

[0060] Each server chassis 1203 may include one or more IT components (e.g., a central processing unit or CPU, a general / graphics processing unit (GPU), memory, and / or storage). Each IT component may perform data processing tasks, wherein the IT component may include software installed in a storage device, loaded into a memory, and executed by one or more processors to perform data processing tasks. The server chassis 203 may include a host server (referred to as a host node) connected to one or more computing servers (also referred to as computing nodes, such as CPU servers and GPU servers). The host server (having one or more CPUs) is typically connected to a client interface via a network (e.g., the Internet) to receive requests for specific services, such as storage services (e.g., cloud-based storage services, such as backup and / or recovery), execute applications to perform certain operations (e.g., image processing, deep data learning algorithms or modeling, etc., as part of a software as a service or SaaS platform). In response to the request, the host server assigns the task to one or more computing nodes or computing servers (with one or more GPUs) managed by the host server. The computing server performs the actual task, which may generate heat during operation.

[0061] The electronics rack 1200 further includes an optional RMU 1202 configured to provide and manage power supplied to the servers 1203. The RMU 1202 can be coupled to a power supply unit (not shown) to manage the power consumption of the power supply unit. The power supply unit can include necessary circuitry (e.g., an AC to DC or DC to DC power converter, a battery, a transformer or a voltage regulator, etc.) to provide power to the remaining components of the electronics rack 1200.

[0062] In one embodiment, the RMU 1202 includes an optimization module 1221 and a rack management controller (RMC) 1222. The RMC 1222 may include a monitor to monitor the operating status of various components (e.g., computing nodes 1203, CDU 1201, and fan modules) within the electronic rack 1200. Specifically, the monitor receives operating data representing the operating environment of the electronic rack 1200 from various sensors. For example, the monitor may receive operating data representing the temperature of the processor, cooling fluid, and airflow, which may be captured and collected via various temperature sensors. The monitor may also receive data representing the fan power and pump power generated by the fan module and the liquid pump 1212, which may be proportional to their respective speeds. These operating data are referred to as real-time operating data. Please note that the monitor may be implemented as a separate module within the RMU 1202.

[0063] Based on the operation data, the optimization module 1221 performs optimization using a predetermined optimization function or optimization model to obtain a set of optimal fan speeds for the fan module and an optimal pump speed for the liquid pump 1212, so that the total power consumption of the liquid pump 1212 and the fan module is minimized, and the operation data related to the cooling fan of the liquid pump 1212 and the fan module are within their respective design specifications. Once the optimal pump speed and the optimal fan speed are determined, the RMC 1222 configures the liquid pump 1212 and the cooling fan of the fan module according to the optimal pump speed and fan speed.

[0064] As an example, based on the optimal pump speed, the RMC 1222 communicates with the pump controller of the CDU 1201 to control the speed of the liquid pump 1212, which in turn controls the liquid flow rate of the cooling liquid supplied to the liquid manifold 1225 for distribution to at least some of the server chassis 1203. Similarly, based on the optimal fan speed, the RMC 1222 communicates with each fan module to control the speed of each cooling fan of the fan module, which in turn controls the airflow speed of the fan module. Note that each fan module can individually control its specific optimal fan speed, and different fan modules and / or different cooling fans within the same fan module can have different optimal fan speeds.

[0065] Note that the rack configuration as shown is for illustrative purposes only; other configurations or arrangements may also be applicable. For example, CDU 1201 may be an optional unit. The cooling plates of the server chassis 1203 may be coupled to the rack manifolds, which may be coupled directly to the room manifolds 1231-1232 without using a CDU. Although not shown, a power supply unit may be arranged within the electronic rack 1200. The power supply unit may be implemented as a standard chassis that is the same or similar to the server chassis, wherein the power supply chassis may be inserted into any standard shelf, replacing any server chassis 1203. In addition, the power supply chassis may further include a battery backup unit (BBU) to provide battery power to the server chassis 1203 when the main power source is unavailable. The BBU may include one or more battery packs, each battery pack including one or more battery cells and the necessary charging and discharging circuits for charging and discharging the battery cells.

[0066] In the foregoing description, embodiments of the present invention have been described with reference to specific exemplary embodiments thereof. It will be apparent that various modifications may be made to the present invention without departing from the broader spirit and scope of the invention as set forth in the following claims. Therefore, the present description and drawings should be regarded as illustrative rather than restrictive.

Claims

1. A cooling plate module, comprising: A first cooling plate layer having a single-phase region therein, wherein the first cooling plate layer comprises: a first liquid inlet port coupled to the first cooling plate layer to receive a first cooling liquid into the single-phase region; and a first liquid outlet port coupled to the first cooling plate layer to discharge the first cooling liquid from the single-phase region; and A second cooling plate layer having a phase change region therein, wherein the second cooling plate layer comprises: a second liquid inlet port coupled to the second cooling plate layer to receive a second cooling liquid into the phase change region; and a vapor outlet port coupled to the second cooling plate layer to discharge the second cooling liquid in a vapor state from the phase change region, wherein a first portion of the first cooling plate layer is in thermal contact with a portion of the second cooling plate layer, and a second portion of the first cooling plate layer is in thermal contact with an IT component to be cooled; a plurality of first liquid cooling fins to form first liquid channels in the single-phase region; A plurality of second liquid cooling fins are provided to form second liquid channels in the phase change region. 2 . The cooling plate module of claim 1 , wherein each of the plurality of first liquid cooling fins is spaced apart at a first pitch, and each of the plurality of second liquid cooling fins is spaced apart at a second pitch.

3. The cooling plate module of claim 1, wherein the plurality of first liquid cooling fins are parallel to the plurality of second liquid cooling fins and staggered with the plurality of second liquid cooling fins to form staggered liquid channels. 4 . The cooling plate module according to claim 2 , wherein the plurality of first liquid cooling fins are perpendicular to the plurality of second liquid cooling fins to form vertical liquid channels.

5. The cooling plate module according to any one of claims 1 to 4, wherein the first cooling plate is coupled to a liquid manifold, and the liquid manifold actively promotes liquid flow in the single-phase region.

6. A cooling plate module according to any one of claims 1 to 4, wherein the flow in the phase change zone is facilitated by gravity, because when the temperature of the second cooling plate layer is higher than a predetermined temperature, the liquid flowing in the phase change zone undergoes a phase change to a vapor state, wherein the vapor is discharged from the vapor outlet port and the liquid enters the second inlet port of the phase change zone.

7. A cooling plate module according to any one of claims 1 to 4, wherein the flow in the phase change zone is promoted by pressure because the liquid flowing in the phase change zone undergoes a phase change into a vapor state, and the volume change in the phase change zone causes the vapor to be discharged from the vapor outlet port, wherein the discharged vapor causes the liquid at the liquid supply container to enter the second inlet port of the phase change zone.

8. The cooling plate module according to any one of claims 1 to 4, wherein the first cooling plate layer and the second cooling plate layer are independently machined, and the single-phase region is separated from the phase change region.

9. The cooling plate module according to any one of claims 1 to 4, wherein the first cooling plate layer and the second cooling plate layer together are an integrated cooling plate machined into a single unit, wherein the integrated cooling plate comprises a plurality of first liquid cooling fins on a first surface of the integrated cooling plate, wherein the integrated cooling plate comprises a plurality of second liquid cooling fins on an opposite surface of the integrated cooling plate.

10. The cooling plate module of claim 9, wherein the integrated cooling plate further comprises: a top frame covering the plurality of first liquid cooling fins; and a bottom frame covering the plurality of second liquid cooling fins, wherein the bottom frame is thermally coupled to the IT components to be cooled.

11. A cooling system comprising: a first liquid supply portion and a liquid return portion; a second liquid supply portion and a vapor reflux portion; as well as According to any one of claims 1 to 10, the at least one cooling plate module is connected to the first liquid supply part and the liquid return part and the second liquid supply part and the steam return part.

12. A cooling plate module, comprising: Intermediate cooling plate; a bottom frame fixed to the bottom surface of the intermediate cooling plate, with a single-phase region between the intermediate cooling plate and the bottom frame; A top frame fixed to the top surface of the intermediate cooling plate, wherein a phase change zone is provided between the intermediate cooling plate and the top frame; a first liquid inlet port and a first liquid outlet port coupled to the intercooling plate to form a first liquid channel in the single-phase region; as well as A second liquid inlet port and a vapor outlet port are coupled with the intercooling plate to form a second liquid channel in the phase change zone, wherein a portion of the chassis is in thermal contact with an IT component to be cooled. 13 . The cooling plate module of claim 12 , wherein the intermediate cooling plate comprises a plurality of first fins on a bottom surface of the intermediate cooling plate and a plurality of second fins on a top surface of the intermediate cooling plate.

Images