Split heat sink cooling system for improved maintainability

Abstract

The present disclosure relates to split heat sink cooling systems for improved maintainability. A heat sink includes a first section for extracting heat generated by a high power component and at least a second section that is attachable to and detachable from the first section, the at least second section for extracting heat generated by a low power component. When the at least second section is attached to the first section, the at least second section is thermally coupled to the first section such that heat generated by the high power component is further dissipated through the at least second section. The heat sink can correspond to an air or liquid cooled heat sink. When liquid cooled, the heat sink can include a cold plate, fluid flow channels, etc.

Description

[0001] Cross-reference to related applications This application is a continuation-in-part of U.S. Application No. 19 / 314,081, filed August 29, 2025, entitled “Split Radiator Cooling System for Improved Maintainability,” which claims the benefit and priority of U.S. Provisional Application No. 63 / 790,302, filed April 17, 2025, entitled “Split Radiator,” pursuant to 35 USC §119. The entire disclosure of the above applications is incorporated herein by reference and is applicable to all purposes. Technical Field

[0002] The embodiments of this disclosure generally relate to heat dissipation solutions for computing hardware, and more specifically to cooling components used in data centers. Background Technology

[0003] As data centers continue to operate at higher speeds (e.g., 100Gb and above) and semiconductor die sizes increase with processing demands, the heat dissipation requirements for data center components also rise. The heat generated in a data center primarily comes from the electrical energy consumed by servers, storage devices, and other hardware components. Almost all the energy consumed is converted into heat as a byproduct of data processing; essentially, every watt of electrical energy consumed generates one watt of heat. This heat needs careful management; otherwise, data center components may fail under extreme thermal loads. For example, processing units generate heat during operation, which can lead to decreased processing power and, in some cases, failure of the processing unit itself and related system components. Summary of the Invention

[0004] As data center infrastructure evolves, improving the maintainability of servers within data centers has become increasingly important. Of particular concern is addressing the assembly and disassembly challenges required to service critical components on the substrate. Components beneath heatsinks on server substrates are often difficult to maintain because each operation often requires removing the entire heatsink. This problem is especially pronounced when using high-power components such as graphics processing units (GPUs) or central processing units (CPUs). (Similar issues exist with CPU packages, parallel processor unit (PPU) packages, quantum processing unit (QPU) packages, and other processing unit packages. It should be understood throughout this disclosure that references to CPU, CPU packages, GPUs, and GPU packages also apply to other types of processing units and processing unit packages, including any other types of processing units and processing unit packages described herein.) Specifically, cooling these high-power components may require heatsinks with considerable surface area. Therefore, achieving optimal cooling often necessitates large heatsinks, which can obscure a significant portion of the substrate, making it impossible to directly maintain pluggable components.

[0005] This problem is particularly pronounced when maintaining memory modules located around processing units (e.g., GPUs and / or CPUs). While memory modules (e.g., Small Outline Compression Attached Memory Modules (SOCAMMs), Dual In-Line Memory Modules (DIMMs), network modules, optical modules, retimers, and Mid-Board Optical Modules (MBOMs)) can be pluggable components, the processing unit's heatsink may cover these modules. This means that each time the memory modules need to be operated, the large heatsink of the processing unit must be removed. This component layout makes field maintenance of the memory modules extremely difficult.

[0006] Embodiments of this disclosure provide a solution to the aforementioned problems. One aspect of this disclosure is to provide a split-type heat sink. In some embodiments, the heat sink is divided into multiple mechanically independent but thermally interconnected parts, thereby providing a solution that enables uniform heat conduction between the parts. The proposed solution can create a substantially uniform air temperature downstream of the heat sink, avoiding large temperature gradients (e.g., minimizing the temperature gradient of the cooled area to at least 5 degrees Celsius below), ensuring the effective operation of downstream sensitive components. Through this conduction path, the parts are thermally interconnected and can be considered as a single unit from a heat transfer perspective, although they can be assembled and disassembled independently. Therefore, the proposed solution provides both high-efficiency heat transfer capabilities and relatively simple maintainability. In other words, it is possible to maintain pluggable components (e.g., storage modules) simply by removing one part of the heat sink from one or more other parts, without removing the other parts of the heat sink.

[0007] Various aspects of this disclosure include the ability to efficiently cool high-power components (e.g., GPUs and / or CPUs) even when the heatsink is broken down into multiple independently maintainable parts; to provide a way to equip each component with an independent heatsink and efficiently cool it; and to enable extremely simple maintenance of pluggable components (e.g., storage modules) without removing other parts of the heatsink by simply removing one part of the heatsink from the other parts.

[0008] This disclosure offers several advantages. First, in the embodiments of this disclosure, thermal connections are provided between the heat sink portions, thereby forming a heat transfer path between the heat sink portions that is not present in existing solutions. Second, the embodiments of this disclosure improve maintainability, including maintenance of the processing unit and / or pluggable components surrounding the processing unit package. Third, the embodiments of this disclosure can create a substantially uniform air temperature downstream of high-power components (e.g., processing units) and low-power components (e.g., SOCAMM, DIMM, network modules, optical modules, retimers, MBOM). Fourth, the embodiments of this disclosure have lower local pressure drop because efficient heat transfer can be achieved without designing dense fin spacing. As a non-limiting example, the fin spacing can be at least 1.4 mm. Fifth, the embodiments of this disclosure have lower system pressure drop and reduced complexity because the system does not require air ducts to direct airflow to the heat sinks of high-power components (e.g., processing units). As a non-limiting example, without departing from the scope of this disclosure, each thermally coupled heatsink segment may have a local pressure drop of no more than 1.1 psi at an airflow of 100 cfm. In some embodiments, each thermally coupled heatsink segment may have a substantially constant pressure drop over its entire surface.

[0009] A heat sink according to an embodiment of this disclosure may include: a first segment for extracting heat generated by a high-power component; and at least a second segment, which can be attached to and detached from the first segment, the at least second segment being used to extract heat generated by a low-power component. When the at least second segment is attached to the first segment, the at least second segment is thermally coupled to the first segment, such that the heat generated by the high-power component is further dissipated through the at least second segment.

[0010] According to other embodiments of this disclosure, a system may include: a processing unit; a first storage module; a first heat sink segment for extracting heat generated by the processing unit; and a second heat sink segment for extracting heat generated by the first storage module, wherein the second heat sink segment can be assembled to and detached from the first heat sink segment, and wherein the first heat sink segment and the second heat sink segment, in the assembled state, jointly transfer heat from the processing unit.

[0011] According to other embodiments of this disclosure, a split-type heat sink may be or include: a central heat sink configured to extract heat generated by a processing unit; and a first side wing heat sink configured to extract heat generated by a storage module, the first side wing heat sink including at least one heat pipe and at least one fin; wherein the first side wing heat sink can be detachably mounted to the central heat sink, and the first side wing heat sink is configured to extract a portion of the heat generated by the processing unit when mounted to the central heat sink. It should be understood that in some embodiments, such as liquid-cooled embodiments, the side wing heat sink may take the form of a finless thermal spreader.

[0012] In some embodiments, a heat sink is provided, comprising: a first section for extracting heat generated by a high-power component; and at least a second section that can be fitted to and removed from the first section, the at least second section for extracting heat generated by a low-power component, wherein, when the at least second section is fitted to the first section, the at least second section is thermally coupled to the first section such that the heat generated by the high-power component is further dissipated through the at least second section.

[0013] In some aspects, the high-power components include at least one of a graphics processing unit (GPU) and / or a central processing unit (CPU).

[0014] In some respects, the low-power components include pluggable modules. As an example, the storage module includes at least one of a Small Form Factor Compact Attached Module (SOCAMM), a Dual In-line Module (DIMM), a network module, an optical module, a retimer, and an Onboard Optical Module (MBOM).

[0015] In some respects, the low-power components are maintainable without removing the high-power components from the first segment.

[0016] In some respects, the first segment is assembled to the at least second segment by one or more screws.

[0017] In some respects, the at least second segment houses one or more heat pipes. As an example, the one or more heat pipes are thermally coupled to the first segment and conduct heat from the first segment to the at least second segment.

[0018] In some respects, both the first segment and the at least second segment include one or more fin structures.

[0019] In some aspects, the at least second segment includes: a second segment thermally coupled to a first low-power component, and a third segment thermally coupled to a second low-power component. In some aspects, the first segment is located between the second segment and the third segment.

[0020] In some respects, a flow channel for conveying cooling fluid is provided between the first segment and the high-power component.

[0021] In some aspects, the first segment includes: a cold plate assembly having a cold plate body, a cold plate cover, and a cold plate thermal contact, wherein the cold plate body is located between the cold plate cover and the cold plate thermal contact, and wherein the flow channel passes between the cold plate body and the cold plate thermal contact to allow heat generated by high-power components to be transferred to the cooling fluid via the cold plate thermal contact.

[0022] In some aspects, the cold plate thermal contact element includes a plurality of fins through which the cooling fluid flows.

[0023] In some embodiments, a system is provided, the system comprising: a processing unit; a first storage module; a first heat sink segment for extracting heat generated by the processing unit; and a second heat sink segment for extracting heat generated by the first storage module, wherein the second heat sink segment can be assembled to and detached from the first heat sink segment, and wherein the first heat sink segment and the second heat sink segment, in an assembled state, jointly transfer heat from the processing unit.

[0024] In some aspects, the system may further include a printed circuit board (PCB) connected to both the processing unit and the first storage module. In some aspects, the PCB transmits electrical signals between the processing unit and the first storage module.

[0025] In some aspects, the system may further include: a second storage module; and a third heat sink section for extracting heat generated by the second storage module, wherein the third heat sink section is detachable from and can be attached to the first heat sink section, and wherein the first heat sink section and the third heat sink section jointly transfer heat from the processing unit in the assembled state.

[0026] In some respects, the first storage module can be maintained by removing the first heat sink section from the second heat sink section.

[0027] In some respects, the first radiator segment is assembled to the second radiator segment by one or more screws.

[0028] In some aspects, the first heat sink section and the second heat sink section include a thermally conductive material, said thermally conductive material including at least one of copper, aluminum, and alloys thereof.

[0029] In some aspects, the system further includes one or more heat pipes. In some aspects, the one or more heat pipes are embedded in the base of the second radiator segment. In some aspects, the first radiator segment includes a vapor chamber, and the one or more heat pipes embedded in the base of the second radiator segment abut against the vapor chamber of the first radiator segment.

[0030] In some respects, the first heat sink section and the second heat sink section, in their assembled state, distribute the heat generated by the processing unit and the storage module substantially evenly.

[0031] In some aspects, the processing unit includes at least one of a graphics processing unit (GPU) and a central processing unit (CPU).

[0032] In some aspects, the first storage module includes at least one of a small-size compact laminated memory module (SOCAMM), a dual in-line memory module (DIMM), a network module, an optical module, a retimer, and an onboard optical module (MBOM).

[0033] In some aspects, the system may also include a reinforcing base that provides structural support for the first radiator segment and provides a connection interface between the first radiator segment and the second radiator segment.

[0034] In some aspects, the system may also include a thermal interface material (TIM). In some aspects, the TIM is disposed between the reinforcing base and the second heat sink section. In some aspects, the TIM is disposed between the second heat sink section and the first storage module. In some aspects, the TIM includes at least one of thermal grease, spacer pad, phase change material, and thermal putty, wherein the TIM provides substantially uniform cooling between the reinforcing base, the first heat sink section, and / or the second heat sink section. In some aspects, the TIM is adapted to tolerance differences between the reinforcing base, the first heat sink section, the second heat sink section, the processing unit, and / or the storage module. Non-limiting examples of materials that can be used for the TIM include substrates (e.g., silicone rubber, acrylic resin, epoxy resin, and / or thermoplastic), conductive fillers (e.g., alumina, boron nitride, zinc oxide, graphite, graphene, and / or other polymers for improving conductivity), combinations thereof, etc.

[0035] In some aspects, the system further includes: a fluid inlet; a fluid outlet; an inlet pipe that delivers cooling fluid between the fluid inlet and the first radiator section; and a flow channel contained within the first radiator section, through which the cooling fluid flows to extract the heat generated by the processing unit.

[0036] In some respects, the first heat sink section is liquid-cooled.

[0037] In some respects, the first heat sink section is air-cooled.

[0038] In some embodiments, a split-type heat sink assembly is provided, comprising: a central heat sink for extracting heat generated by a processing unit; and a first side fin heat sink for extracting heat generated by a storage module, the first side fin heat sink including at least one heat pipe and at least one fin; wherein the first side fin heat sink is detachably mountable to the central heat sink, and the first side fin heat sink is configured to extract a portion of the heat generated by the processing unit when mounted to the central heat sink.

[0039] In some aspects, the split radiator assembly further includes: a plurality of radiator segments, each of the plurality of radiator segments being mechanically connected to at least one other radiator segment, wherein at least one of the plurality of radiator segments is detachable from the split radiator assembly to allow operation of an adjacent radiator segment among the plurality of radiator segments.

[0040] In some respects, the multiple heat sink segments are thermally coupled to each other so as to function as a single thermal unit.

[0041] In some respects, the thermally coupled radiator section heats the surrounding air substantially uniformly, thereby giving the air downstream of the thermally coupled radiator section a substantially uniform temperature.

[0042] In some respects, the heat dissipated from the thermally coupled heat sink section will not impede the effective operation of sensitive components located downstream of the thermally coupled heat sink section.

[0043] In some respects, removing at least one of the plurality of radiator segments will not prevent the remaining radiator segments from functioning as a single thermal unit.

[0044] In some aspects, the thermally coupled heat sink segment includes multiple fin structures with a fin spacing of at least 1.4 mm.

[0045] In some respects, the thermally coupled radiator section has a local pressure drop of no more than 1.1 psi at an airflow of 100 cfm.

[0046] In some respects, the split-type radiator assembly allows sufficient airflow to pass through the multiple fin structures without the need for air ducts during operation.

[0047] In some respects, the thermally coupled heat sink section provides a substantially constant voltage drop.

[0048] In some respects, one of the plurality of radiator segments is a central radiator.

[0049] In some aspects, the central heat sink includes a heat spreader configured to be in thermal communication with the processing unit or processing unit package.

[0050] In some aspects, the central radiator includes a heat spreader that is configured to be in thermal communication with high-power components during use.

[0051] In some embodiments, a system is provided, the system comprising: a first split-type radiator assembly including: a first central radiator that is liquid-cooled, and at least one side radiator thermally coupled to the first central radiator; and a second split-type radiator assembly including: a second central radiator that is liquid-cooled, and at least one additional side radiator thermally coupled to the second central radiator.

[0052] In some aspects, the first central radiator is connected to the at least one side radiator by at least a first screw, wherein the second central radiator is connected to the at least one additional side radiator by at least a second screw, wherein the at least one side radiator is mountable to and removable from the first central radiator, and wherein the at least one additional side radiator is mountable to and removable from the second central radiator.

[0053] In some aspects, the first central radiator is coupled to the second central radiator via a coupling pipe that delivers cooling fluid between the first and second central radiators.

[0054] In some aspects, the at least one side fin radiator is thermally coupled to the first central radiator via at least a first heat pipe, and wherein the at least one additional side fin radiator is thermally coupled to the second central radiator via at least a second heat pipe.

[0055] In some aspects, the first central radiator extracts heat generated by the first high-power component, wherein the second central radiator extracts heat generated by the second high-power component, wherein at least one side radiator extracts heat generated by the first low-power component, wherein at least one additional side radiator extracts heat generated by the second low-power component, and wherein cooling fluid flows along a flow path such that the cooling fluid flows through the first low-power component and the second low-power component before flowing through the first high-power component and the second high-power component, to avoid damage to the first low-power component and the second low-power component by the heat generated by the first high-power component and the second high-power component.

[0056] In some respects, the cooling fluid flows along a fluid flow path having at least one set of parallel channels.

[0057] Embodiments of this disclosure also envision a method for maintaining a computing system, the method comprising: removing at least one fastener coupling a side fin heatsink to a central heatsink; sliding the side fin heatsink relative to the central heatsink in at least a first direction while the central heatsink remains in fluid communication with a source of cooling fluid; and operating components cooled by the side fin heatsink while maintaining the fluid communication.

[0058] Any aspect in combination with any or more other aspects.

[0059] Any one or more features disclosed herein.

[0060] This article discloses one or more of the features.

[0061] Combination of any one or more features substantially disclosed herein with any one or more other features substantially disclosed herein.

[0062] Any aspect / feature / implementation method combined with any or more other aspects / features / implementation methods.

[0063] Use any one or more aspects or features disclosed herein.

[0064] It should be understood that any feature described herein may be combined with any one or more other features described herein to form a claim, regardless of whether such features are derived from the same embodiment described.

[0065] Details of one or more aspects of this disclosure are set forth in the accompanying drawings and the following description. Other features, objects, and advantages of the technology described in this disclosure will be apparent from the description, the drawings, and the claims.

[0066] Those skilled in the art will appreciate the many additional features and advantages of this disclosure upon considering the following description of the implementation methods provided.

[0067] Additional features and advantages are described herein and will be apparent from the following description and figures. Attached Figure Description

[0068] Having provided a general description of this disclosure, the following description will now be given with reference to the accompanying drawings, which are not necessarily drawn to scale: Figure 1A This is a top perspective view of a split-type heat sink according to at least some embodiments of the present disclosure; Figure 1B This is a block diagram of a split heat sink according to at least some embodiments of the present disclosure; Figure 2A This is a side view of a split radiator according to at least some embodiments of the present disclosure, showing two side wing radiators separate from the central radiator; Figure 2B This is a bottom plan view of a split radiator according to at least some embodiments of the present disclosure, showing two side radiators separate from the central radiator; Figure 3A This is a bottom plan view of a split heat sink according to at least some embodiments of the present disclosure, wherein one or more components have been removed to show details of the split heat sink; Figure 3B This is a bottom plan view of a split heat sink according to at least some embodiments of the present disclosure, wherein one or more components have been removed to show details of the split heat sink; Figure 4 This is a bottom perspective view of a split heat sink according to at least some embodiments of the present disclosure, wherein one or more components have been removed to show details of the split heat sink; Figure 5A This is a top perspective view of two side wing heat sinks of a split heat sink according to at least some embodiments of the present disclosure; Figure 5B This is a top perspective view of the central heat sink of a split-type heat sink according to at least some embodiments of the present disclosure; Figure 6 This is a bottom perspective view of the heat pipe arrangement of a split-type radiator according to at least some embodiments of the present disclosure; Figure 7 This is a bottom perspective view of the components of a split-type heat sink according to at least some embodiments of the present disclosure; Figure 8 This is a side plan view of a split heatsink operatively connected to a CPU and a plurality of memories according to at least some embodiments of the present disclosure; Figure 9It is a series of thermal maps showing the heat transfer results achieved using two different configurations of a split heat sink according to at least some embodiments of the present disclosure; Figure 10 This is a block diagram illustrating a computer system according to at least some embodiments of the present disclosure; Figure 11 A data center according to at least some embodiments of the present disclosure is depicted; Figure 12 A block diagram of a system according to at least some embodiments of the present disclosure is depicted; Figure 13 A block diagram depicts a computing environment according to at least some embodiments of the present disclosure; Figure 14 This is a top perspective view of an alternative configuration of a system according to at least some embodiments of the present disclosure; Figure 15 yes Figure 14 Second top perspective view of the system shown; Figure 16 This is a side sectional view illustrating fluid flow in a liquid cooling system according to at least some embodiments of the present disclosure; Figure 17 This is a perspective view of a split heat sink assembly according to at least some embodiments of the present disclosure; Figure 18A This is a perspective view of a cold plate thermal contact component according to at least some embodiments of the present disclosure; Figure 18B This is a side view of a cold plate thermal contact element according to at least some embodiments of the present disclosure; Figure 19A This is a first top perspective view of the cold plate body according to at least some embodiments of the present disclosure; Figure 19B This is a second top perspective view of the cold plate body according to at least some embodiments of the present disclosure; Figure 20A This is a first bottom perspective view of the cold plate body and the cold plate thermal contact member according to at least some embodiments of the present disclosure; Figure 20B This is a second bottom perspective view of a cold plate body with the cold plate thermal contact element removed according to at least some embodiments of the present disclosure; Figure 21 This is a perspective view of a side wing heat sink according to at least some embodiments of the present disclosure; Figure 22 This is a perspective view of a first central heat sink section and a second central heat sink section according to at least some embodiments of the present disclosure; Figure 23A This is a top perspective view of the cold plate cover according to at least some embodiments of the present disclosure; Figure 23BThis is a bottom perspective view of the cold plate cover according to at least some embodiments of the present disclosure; Figure 23C This is a side view of the cold plate cover according to at least some embodiments of the present disclosure; Figure 23D This is a top view of the cold plate cover according to at least some embodiments of the present disclosure; Figure 24 This is an exploded view of a cold plate assembly according to at least some embodiments of the present disclosure; and Figure 25 This is a flowchart illustrating example methods according to at least some embodiments of the present disclosure.

[0069] The same reference numerals and markings in the accompanying drawings may represent the same elements. Detailed Implementation

[0070] The following description provides only examples and is not intended to limit the scope, applicability, or configuration of the claims. Rather, the following description will provide a sufficient illustration for those skilled in the art to implement the described embodiments. It should be understood that various changes may be made to the function and arrangement of the elements without departing from the spirit and scope of the appended claims.

[0071] As used herein, the phrases “at least one,” “one or more,” “or,” and “and / or” are open-ended expressions that combine both parallel and selective usage. For example, each of the expressions “at least one of A, B, and C,” “at least one of A, B, or C,” “one or more of A, B, and C,” “one or more of A, B, or C,” “A, B, and / or C,” and “A, B, or C” represents a single A, a single B, a single C, a combination of A and B, a combination of A and C, a combination of B and C, or a combination of A, B, and C.

[0072] Various aspects of this disclosure will be described herein with reference to the accompanying drawings, which are schematic illustrations of idealized configurations.

[0073] Unless otherwise defined, all terms used herein (including technical and scientific terms) shall have the same meaning as commonly understood by one of ordinary skill in the art to which this disclosure pertains. It should also be understood that terms, such as those defined in a general dictionary, shall be interpreted as having the same meaning as they have in the relevant technical context and within the context of this disclosure.

[0074] As used herein, the singular forms “a,” “an,” and “described” are intended to include the plural forms as well, unless the context clearly indicates otherwise. It should also be understood that, when used in this specification, the terms “comprising,” “including,” and / or “having” specify the presence of the stated features, integral units, steps, operations, elements, and / or components, but do not exclude the presence or addition of one or more other features, integral units, steps, operations, elements, components, and / or combinations thereof. The term “and / or” includes any and all combinations of one or more of the listed related items.

[0075] The present disclosure will be described more fully below with reference to the accompanying drawings, which illustrate some, but not all, embodiments of the present disclosure. In fact, the present disclosure may be embodied in many different forms and should not be construed as limited to the embodiments set forth herein.

[0076] Throughout this document, the same reference numerals refer to the same elements. As used herein, the terms “front,” “rear,” “top,” etc., are used in the examples provided below to describe the location of certain components or parts of components in their installed and operational state. As used herein, the term “module” encompasses hardware, software, and / or firmware configured to perform one or more specific functions, including but not limited to the conversion between electrical and optical signals and their transmission. As will be apparent to those skilled in the art based on this disclosure, the term “substantially” indicates that the mentioned elements or related descriptions are accurate within applicable engineering tolerances.

[0077] The embodiments of this disclosure are envisioned for deployment in data center environments. While embodiments will be described in conjunction with certain examples of data center environments, it should be understood that the embodiments of this disclosure are not limited thereto. In fact, embodiments of this disclosure can be deployed in any number of environments, including data center environments or any other suitable environment requiring management of the thermal load generated by computing components.

[0078] The following will refer to Figure 1 to... Figure 13 A split-type heat sink according to embodiments of the present disclosure is shown and described, as well as exemplary data center environments and components that help to fully understand the present disclosure.

[0079] Figure 1- Figure 8 and Figures 14-24 Various aspects of a split-type heat sink 100 according to embodiments of the present disclosure are depicted, and although some parts of the following description may refer to specific drawings in describing particular aspects of the split-type heat sink 100, the same or similar components can be found in FIG1- Figure 8 and / or Figures 14-24 Other figures are shown in the accompanying drawings.

[0080] Refer to Figure 1 for the overall layout. Figure 8 And refer to the specific details. Figure 1Aand Figure 1B The split-type heat sink 100 includes a central heat sink 104, a first side wing heat sink 108A, and a second side wing heat sink 108B. Throughout this disclosure, the central heat sink 104 may be referred to as, for example, a first segment or a first heat sink section, and each of the first side wing heat sink 108A and the second side wing heat sink 108B may be referred to as, for example, a second segment, a second heat sink section, a third segment, and / or a third heat sink section. Both the first side wing heat sink 108A and the second side wing heat sink 108B can be independently removed from the central heat sink 104, thereby facilitating maintenance of computing hardware (including, for example, storage modules) located beneath one or both of the side wing heat sinks 108A and / or 108B. However, as further shown and described herein, when the first side wing heat sink 108A and the second side wing heat sink 108B are assembled or connected to the central heat sink 104, each side wing heat sink 108A, 108B is thermally coupled to the central heat sink 104, and the split heat sink 100 provides integrated cooling for processing units, processing unit packages or other high-power components 168 operatively connected to the central heat sink 104, and also provides cooling for storage modules or other low-power components 172 operatively connected to each side wing heat sink 108A, 108B.

[0081] As used herein, the phrase "thermal coupling" refers to two or more objects connected in a manner that facilitates heat transfer between the two objects, such that the two or more objects operate as a single entity (or substantially as a single entity) from a thermal perspective. Within the meaning of this disclosure, "low-power component" includes any component on a motherboard, motherboard, blade server, node, or other computing device that consumes relatively less power (and therefore generates relatively less heat) compared to a CPU, GPU, or other processing unit or processing unit package. Thus, for example, a low-power component can be or may include random access (or similar) memory (including any memory units described herein), chipset or platform controller hubs, high-speed interconnects, substrate management controllers, storage controllers, voltage regulators, and / or power supply circuitry. In some embodiments, a low-power component can be or may include any component that consumes less than about 100W of power during operation.

[0082] Despite Figure 1- Figure 8The illustrated split-type heat sink includes a first section (e.g., a central heat sink 104), a second section (e.g., a first side wing heat sink 108A), and a third section (e.g., a second side wing heat sink 108B). However, a split-type heat sink according to embodiments of this disclosure may include only the first and second sections, one configured to cool a high-power component and the other configured to primarily cool a low-power component while simultaneously assisting in cooling the high-power component. In other embodiments, the split-type heat sink may include more than three sections. For example, a split-type heat sink according to embodiments of this disclosure may include four, five, six, or more sections. In such embodiments, the first section may be configured to extract heat generated by the high-power component, and each of the remaining sections may be configured to extract heat generated by a separate low-power component. Similarly, in such embodiments, each section configured to extract heat generated by a separate low-power component may also be configured to assist in dissipating heat generated by a high-power component operatively connected to the first section, in a manner substantially the same as described herein with respect to the second and third sections.

[0083] Embodiments of this disclosure employing more than three segments can utilize various arrangements for each segment. For example, in embodiments with four or five segments, each of the second, third, fourth, and (if any) fifth segments can be arranged adjacent to one side of the first segment, which is substantially square or rectangular. In embodiments with six segments, the first segment can have a substantially pentagonal shape, and each of the second to sixth segments can be arranged adjacent to one side of the first segment. In embodiments with seven segments, the first segment can have a substantially hexagonal shape, and each of the second to seventh segments can be arranged adjacent to one side of the first segment.

[0084] Alternatively, additional segments beyond the third segment can be fixed adjacent to the second and third segments (e.g., to increase the length of the split heatsink without increasing its width, or to increase the width of the split heatsink without increasing its length). In yet other embodiments, a split heatsink comprising more than three segments can be configured such that the second segment and any additional segment each include at least a portion extending into and thermally coupled to the first segment, even if that portion may not be substantially in direct contact with the entire edge of the first segment.

[0085] The central heatsink 104 includes a vapor chamber 112 configured to be in thermal communication (whether through direct contact or otherwise) with a processing unit (e.g., CPU or GPU), a processing unit package, or other high-power components during use. A reinforcing base 116 is mounted above and partially surrounds the vapor chamber 112 to provide structural support for the components of the central heatsink 104 and to ensure that the weight of the central heatsink 104 is not exerted on the processing unit or other high-power components mounted on the central heatsink 104 (e.g., ...). Figure 8 The CPU shown is 804.

[0086] Each heatsink segment of the split-type heatsink 100 includes one or more fin structures. More specifically, one or more vertical fins 120 may be mounted on a reinforcing base 116, and one or more horizontal fins 124 may be stacked above the reinforcing base 116 at a uniform or irregular vertical spacing. One or more heat pipes 128A may also be embedded and / or mounted on the reinforcing base 116 and extend upward through the horizontal fins 124 to transfer heat to these horizontal fins 124.

[0087] Each side wing heatsink 108A, 108B includes a side wing base 132, similar to the reinforcing base 116, which provides structural support for the components of the respective side wing heatsink 108A or 108B and helps ensure that the weight of the side wing heatsink 108A or 108B is not exerted on the storage module or other components thermally coupled to the side wing heatsink 108A or 108B (e.g., Figure 8 The storage modules 808A and 808B are shown. Each side wing base 132 may include one or more mounting tabs 140, each mounting tab 140 having a mounting hole 144 configured to receive a removable mounting fastener 148 (e.g., screw, bolt, pin) that secures the side wing base 132 to a printed circuit board (PCB) to which the cooled component is connected (e.g., Figure 8 (PCB 812 shown) and / or data center racks or other structural components. Mounting lug 140 may be integrally formed with side wing base 132, or manufactured separately from side wing base 132 and then attached to side wing base 132 using any known connection method (e.g., mechanical fasteners, welding, adhesive).

[0088] Each side wing base 132 may also have one or more mounting holes 144 within itself, each mounting hole 144 being configured to receive a removable mounting fastener 148 that secures the side wing base 132 to the reinforcing base 116 of the central heatsink 104. Using the removable mounting fastener 148 to secure the side wing base 132 (and consequently the side wing heatsinks 108A, 108B) to the central heatsink 104 and / or the underlying substrate (e.g., a PCB) allows the side wing heatsinks 108A, 108B to be mounted to and removed from the central heatsink 104 and / or the underlying substrate, thereby facilitating the individual removal of the side wing heatsinks 108A, 108B without removing the central heatsink 104.

[0089] The side wing base 132 can be made of copper, aluminum, copper alloys, aluminum alloys, other metals, and / or other metal alloys. The material and / or thickness of the side wing base 132 can be selected to ensure that the side wing heat sinks 108A, 108B have sufficient structural rigidity. The side wing base 132 includes a plurality of heat pipe cutouts 134 configured to fix heat pipes 128B such that when the side wing heat sinks 108A, 108B are mounted to the central heat sink 104, the flat bottom surface of the heat pipe 128B can directly contact (or otherwise thermally couple) with the heat spreader 112, and also allows the flat bottom surface to be thermally coupled (e.g., through thermal interface material 136) to the low-power components (e.g., memory 808A, 808B) cooled by the side wing heat sinks 108A, 108B.

[0090] The thermal interface material 136 enhances thermal conductivity and ensures thermal coupling between the storage modules 808A, 808B or other low-power components that are operatively connected to the side wing heat sinks 108A, 108B. In some embodiments, the thermal interface material 136 may be, for example, a gap pad (rather than a phase change material). Using a gap pad instead of a phase change material simplifies the maintenance process because the gap pad is easier to remove and replace, thereby reducing the complexity of maintenance and consequently reducing the time required for maintenance.

[0091] Multiple vertical fins 120 can be directly mounted to the upper surface of the side fin base 132, or mounted to the upper surface of the side fin base 132 via a thermal interface material, such as thermal interface material 136. In some embodiments, additional thermal interface material 136 can be used between the reinforcing base 116 of the central radiator 104 and the vertical fins 120. The thermal interface material 136 can include any suitable type of thermal interface material (TIM), which may include, for example, thermal grease, spacer pads, phase change materials, thermal putty, combinations thereof, etc. In some embodiments, the thermal interface material 136 can be flexible (e.g., compressible and / or soft) to help compensate for tolerance differences between at least one of the reinforcing base 116, the first radiator segment 104, one or more second radiator segments 108, the processing unit 804, and / or the storage module 808. The thermal interface material 136 can also be configured to provide substantially uniform cooling between the components as if the two components were physically in contact with each other.

[0092] One or more heat pipes 128B may be mounted to and / or embedded in each side wing base 132. The heat pipes 128B are configured to extend upward through a plurality of horizontal fins 124, which may be stacked above the side wing base 132 at a uniform or irregular vertical spacing along the height of the heat pipes 128B.

[0093] The vertical fins 120 used in the central radiator 104 and the first and second side fin radiators 108A and 108B can be made of copper, aluminum, other metals, and / or any metal alloy. The thickness of each vertical fin 120, the spacing between any two vertical fins 120, and the length and height of each vertical fin 120 (or, alternatively, the surface area of ​​each vertical fin 120) can be selected to provide a desired level of heat dissipation. Furthermore, the vertical fins 120 can be oriented on the reinforcing base 116 and the side fin base 132 such that air flowing through the split radiator 100 can pass between the vertical fins 120 and flow along the length of each vertical fin 120.

[0094] Similarly, the horizontal fins 124 used in the central radiator 104 and the first and second side fin radiators 108A and 108B can be made of copper, aluminum, other metals, and / or any metal alloy. The thickness of each horizontal fin 124, the spacing between any two horizontal fins 124 (e.g., the pitch of the horizontal fins 124), and the length and width of each horizontal fin 124 (or, alternatively, the surface area of ​​each horizontal fin 124) can be selected to provide the desired level of heat dissipation.

[0095] As described in this disclosure, the design of thermally coupling the side fin 108A, 108B to the central radiator 104 advantageously allows for a wider pitch between adjacent horizontal fins 124 on each segment of the split radiator 100, compared to using only the central radiator 104 to extract and dissipate heat from the thermally coupled high-power components. This advantageously reduces the pressure drop across the radiator, enabling the maintenance of a constant airflow through the radiator without the need for a high-power fan, and reducing the cost, complexity, and energy consumption of the entire cooling system.

[0096] In some embodiments, the relative edges of one or more horizontal fins 124 of the first radiator segment 104, the second radiator segment 108A, and / or the third radiator segment 108B may be bent downwards (e.g., toward the next lower horizontal fin 124) or upwards (e.g., toward the next higher horizontal fin 124), respectively, to ensure that air flowing between adjacent horizontal fins 124 flows along the length of the split radiator 100. For example, on the first side fin radiator 108A and the second side fin radiator 108B, the edges of the horizontal fins 124 near the central radiator 104 may be bent upwards or downwards as described above, and the relative edges of the same horizontal fins 124 may also be bent upwards or downwards as described above. Similarly, on the central radiator 104, the edges of the horizontal fins 124 facing the first side fin radiator 108A may be bent upwards or downwards as described above, and the relative edges of the same horizontal fins 124 (e.g., the edges facing the second side fin radiator 108B) may also be bent upwards or downwards as described above.

[0097] Heat pipes 128A and 128B can be any type of heat pipe suitable for computing and / or data center hardware, and can have the same or different cross-sectional shapes. In some embodiments, each heat pipe 128A and / or 128B can be made of copper, aluminum, or other metals or metal alloys, and can include a sealed chamber containing a working fluid (e.g., water, ammonia, methanol, ethanol). Heat pipes 128A and 128B are configured to transfer heat generated by computing components (e.g., CPU 804, one or more storage modules 808A, 808B) cooled by the split heat sink 100 upwards and into horizontal fins 124, which are configured to absorb the heat from the heat pipes and dissipate the heat (e.g., by convection to air or other fluid flowing through the horizontal fins 124). When the side radiators 108A and 108B are assembled to the central radiator 104, the heat pipes 128B in the side radiators 108A and 108B are thermally coupled to the vapor chamber 112 and configured to conduct heat from the first radiator section (e.g., the central radiator 104) to the second and third radiator sections (e.g., the side radiators 108A and 108B) to improve the dissipation of heat generated by high-power components operatively connected to the first radiator section. In some embodiments, the central radiator 104 and the side radiators 108A and 108B may have more or fewer heat pipes 128A and 128B respectively, and / or the heat pipes 128A and 128B may be arranged in different configurations within the central radiator 104 and the side radiators 108A and 108B respectively.

[0098] Now refer to Figure 2A and Figure 2B Each side wing heatsink 108A, 108B can be independently assembled and disassembled with the central heatsink 104, with the PCB (e.g., PCB 812) mounted on the split heatsink 100, and / or with the rack or other data center structural components mounted on the split heatsink 100. Disassembly can be achieved by removing mounting fasteners (e.g., screws, bolts, pins), such as mounting fasteners 148, from the mounting holes 144 located near the corners of the side wing heatsinks 108A or 108B. As described above, in some embodiments, two of the mounting holes 144 are located in the side wing base 132, and the other two of the mounting holes 144 are located in mounting ears 140 extending from the side wing base 132. However, in different embodiments, the arrangement of the mounting holes 144 may differ, and different embodiments may have more or fewer mounting holes 144.

[0099] Still refer to Figure 1 for the overall picture. Figure 8 and specifically as follows Figure 2AAs shown, the opposing edges 122 of each horizontal fin 124 of the central radiator 104 and the side fin radiators 108A, 108B are curved downwards to contact or nearly contact the next lower horizontal fin 124 (or, for the lowest horizontal fin 124, such that the opposing edges 122 of that horizontal fin 124 extend downwards toward the reinforcing base 116 or the side fin base 132). Thus, the stacked horizontal fins 124 form a substantially enclosed channel through which air flows, helping to guide airflow in the same direction and ensuring airflow over the entire surface of each horizontal fin 124. Similarly, as... Figure 2A As shown, all vertical fins 120 and horizontal fins 124 are arranged to facilitate airflow through the split radiator 100 in the same direction.

[0100] In embodiments of this disclosure that include more than three segments, the arrangement of the vertical fins 120 and the arrangement of the horizontal fins 124 (specifically the curved edges 122 of the horizontal fins 124) on a given segment can be configured differently from other segments, so that when the second segment and the additional segment are both assembled to the first segment, the entire split radiator is configured to guide airflow in the same direction.

[0101] Refer to Figure 1 for the overall layout. Figure 8 and specifically as follows Figure 2B As shown, the heat pipes 128B of each side fin radiator 108A, 108B pass through the heat pipe cutouts 134 in the side fin base 132 and extend toward the heat spreader 112 of the central radiator 104. When the side fin radiators 108A, 108B are assembled to the central radiator 104, a portion of each heat pipe 128B abuts against the heat spreader 112, thereby facilitating the transfer of heat from the heat spreader 112 to the horizontal fins 124 of the side fin radiators 108A, 108B. The thermal interface material 136 located below each side fin heatsink 108A, 108B also facilitates the transfer of heat from storage modules (e.g., storage modules 808A, 808B) or other low-power computing components operatively connected to the side fin heatsinks 108A, 108B to the heat pipes 128B of the side fin heatsinks 108A or 108B, thereby enabling the heat to be transferred to the horizontal fins 124 of the side fin heatsinks 108A or 108B for dissipation into the surrounding air or other fluids.

[0102] Figure 3A A bottom view of a split-type heat sink 100 according to at least some embodiments of the present disclosure is depicted, wherein the thermal interface material 136 of each side fin heat sink 108A, 108B is removed. With the thermal interface material 136 removed, the arrangement of the heat pipes 128B is more readily observable. It can be seen that the heat pipes 128B are configured to maintain a large surface area in contact with the thermal interface material 136 located thereon, thereby facilitating the transfer of heat from the thermal interface material 136 to the heat pipes 128B. Figure 3AThe cutout 134 into which the heating pipe 128B is embedded can also be seen on the side wing base 132.

[0103] The reinforcing base 116 includes four holes 152, each configured to receive a mechanical fastener for securing the central heatsink 104 of the split heatsink 100 to an underlying PCB (e.g., PCB 812) and / or a data center rack or other structural element. In other embodiments, the reinforcing base 116 may include more or fewer than four holes for this purpose.

[0104] Figure 3B and Figure 4 Another bottom view of a split-type heat sink 100 according to at least some embodiments of the present disclosure is shown, in which the heat spreader 112 has been removed from the central heat sink 104, and the thermal interface material 136 of each side fin heat sink 108A, 108B has also been removed. In this view, the portion of the heat pipe 128B in direct contact with the heat spreader 112 is visible, as is the bottom surface of the heat pipe 128A, which is also in direct contact with the heat spreader 112 during operation of the split-type heat sink 104. In some embodiments, a thermal interface material (such as thermal interface material 136) may be used to fill any gaps between the heat pipe 128A and the heat spreader 112, and / or to ensure thermal coupling between the heat pipe 128A and the heat spreader 112. The heat pipe 128A can be seen through a cutout 160 in the reinforcing base 116, in which the heat pipe 128A is embedded. The notch 160 can be configured to accommodate heat pipes 128A, with minimal or no gap between each heat pipe 128A and the adjacent wall of the corresponding notch 160, to enable and improve heat transfer (e.g., by conduction) from the reinforcing base 116 to the heat pipe 128A. Each heat pipe 128A has a flat bottom surface to increase the surface area in contact with the heat spreader 112 and / or the thermal interface material 136, thereby facilitating heat transfer into and through the heat pipe 128A to the horizontal fins 124.

[0105] After removing the vapor chamber 112, holes 144 for securing the side fin heatsinks 108A, 108B to the central heatsink 104 and holes 156 for mounting the vapor chamber 112 to the reinforcing base 116 (or vice versa) are also visible in the reinforcing base 116. In various embodiments of this disclosure, the vapor chamber 112 can be mounted to the reinforcing base 116 using mechanical fasteners (e.g., screws, bolts, pins, rivets), adhesives, thermal interface materials, and / or any other suitable mounting system, material, and / or method. When using mechanical fasteners, these fasteners can be the same as or similar to the mechanical fastener 148 described herein. Furthermore, in some embodiments, the number of holes 144 and 156 in the reinforcing base 116 may be more or less than four.

[0106] like Figure 5A and Figure 5B As shown in the perspective view, the reinforcing base 116 includes a side cutout 164 that exposes a portion of the upper surface of the heat spreader 112 and forms a connection interface for the side wing heat sinks 108A and 108B. More specifically, the side cutout 164 accommodates a portion of the first side wing heat sink 108A and the second side wing heat sink 108B (e.g., a portion of the side wing base 132 of each side wing heat sink 108A and 108B) such that when the side wing heat sinks 108A and 108B are fixed to the central heat sink 104, a portion of the bottom surface of each heat pipe 128B abuts against the exposed upper surface of the heat spreader 112. In some embodiments, the bottom surface of each heat pipe 128B may be in direct contact with the heat spreader 112 (e.g., a portion of the upper surface of the heat spreader exposed by the side cutout 164). In other embodiments, thermal interface material 136 may be located between heat pipe 128B and vapor chamber 112 to ensure thermal coupling between these components and facilitate heat transfer from vapor chamber 112 to heat pipe 128B. As described herein, thermal interface material (e.g., thermal interface material 136) may be used between any two components of the split radiator 100 to improve thermal coupling between these components, whether they are permanently or removably fixed to each other or arranged adjacent to each other.

[0107] Figure 6 Only the heat pipes 128A and 128B of the split-type heat sink 100 are depicted. These heat pipes are carefully constructed and arranged to extract heat from the vapor chamber 112 and transfer it away through the horizontal fins of the central heat sink 104 and the side heat sinks 108A and 108B. Thus, the split-type heat sink 100 can provide sufficient cooling for the processing unit (e.g., CPU 804) cooled by the split-type heat sink 100, while still allowing the side heat sinks 108A and 108B to be removed for maintenance of one or more low-power computing components (e.g., storage modules 808A and 808B) below them. Furthermore, the heat pipes 128B are configured to receive heat from the aforementioned low-power computing components (e.g., storage modules 808A and 808B) and transfer this heat, along with the heat received from the vapor chamber 112, to the horizontal fins 124 of the side heat sinks 108A and 108B.

[0108] although Figure 6 (and more broadly, Figure 1-) Figure 8Figure 1 illustrates the specific structure of heat pipes 128A and 128B, but embodiments of this disclosure also include other heat pipe structures. In some embodiments, one or more segments of the first segment (e.g., central radiator 104), the second segment (e.g., first side fin radiator 108A), and the third segment (e.g., second side fin radiator 108B) include only a single heat pipe 128A or 128B. Similarly, in some embodiments, the number of heat pipes 128A and 128B included in one or more segments of the first, second, and third segments may be more or less than that in Figure 1. Figure 8 The quantities shown. Furthermore, regardless of whether the number of heat pipes 128A and 128B included in the first, second, and / or third sections is the same or different, the arrangement or structure of these heat pipes can be consistent with... Figure 6 The cross-sections differ as shown. In some embodiments, one or more heat pipes 128A in the central radiator 104 may have different cross-sections from one or more other heat pipes 128A in the central radiator 104. Similarly, in some embodiments, one or more heat pipes 128B in either or both of the first side fin radiator 108A and the second side fin radiator 108B may have different cross-sections from one or more other heat pipes 128B in the same radiator segment.

[0109] Still refer to Figure 1 for the overall picture. Figure 8 And refer to the specific details. Figure 7 The vapor chamber 112 can be, for example, a heat transfer device containing a fluid that rapidly diffuses heat to the surface of the vapor chamber 112 through repeated evaporation and condensation. The vapor chamber 112 can be made of copper, copper alloys, other metals or metal alloys, or other thermally conductive materials. In some embodiments, the vapor chamber 112 may include a sealed chamber 114 containing a working fluid. The working fluid may be, for example, water, ammonia, methanol, or ethanol. The sealed chamber 114 may extend below the bottom surface 115 of the vapor chamber 112. The shape of the sealed chamber 114 can be designed to precisely cover the entire processing unit or processing unit package, such as a CPU 804.

[0110] The reinforcing base 116 can be made of copper, aluminum, copper alloys, aluminum alloys, other metals, and / or other metal alloys. The material and / or thickness of the reinforcing base 116 can be selected to ensure sufficient structural rigidity for the central radiator 104 and / or the entire split radiator 100. The reinforcing base 116 includes multiple heat pipe cutouts 160 configured to secure heat pipes 128A, allowing the flat bottom surface of the heat pipes 128A to directly contact the heat spreader 112 when it is mounted to or within the heat spreader 116 (or vice versa). The reinforcing base 116 also includes one or more side cutouts 164 exposing a portion of the upper surface of the heat spreader 112 (opposite to the bottom surface 115), thereby enabling efficient heat transfer from high-power components thermally coupled to the heat spreader 112 to the heat pipes 128B of the side radiators 108A, 108B.

[0111] Figure 8 A side view of a heat sink 100 according to at least some embodiments of the present disclosure is depicted. The heat sink is arranged in a functional assembly manner and mates (e.g., thermally coupled) with a CPU 804, a memory module 808A, and a memory module 808B, each of which is mounted on a PCB 812. The PCB 812 carries electrical signals between the CPU 804, the memory module 808A, and / or the memory module 808B. Figure 8 As shown, the CPU 804 is in direct contact with the heat spreader 112 (although in some embodiments, a thermal interface material (such as thermal interface material 136) may be used to ensure thermal coupling between the CPU 804 and the heat spreader 112). Similarly, a portion of the side wing base 132 of each side wing heatsink 108A, 108B is accommodated in the side cutout 164 of the reinforcing base 116 (in Figure 8 (Not visible in the center) This allows the heat pipes 128B of the side wing radiators 108A and 108B to contact the heat spreader 112 and transfer heat from the heat spreader 112.

[0112] If an operator of a data center including CPU 804, storage modules 808A and 808B, and heatsink 100 needs to replace or otherwise maintain, for example, storage module 808A, the split heatsink 100 according to embodiments of this disclosure allows the operator to remove only the side wing heatsink 108A from the central heatsink 104 and / or PCB 812, remove the side wing heatsink 108A, remove the storage module 808A, install the maintained storage module 808A (or install a new storage module 808A), and then reassemble the side wing heatsink 108A to the central heatsink 104 and / or PCB 812, without disassembling (and subsequently reinstalling) the rest of the split heatsink 100. The storage module 808B can be maintained using essentially the same process by removing and subsequently reassembling only the second side wing heatsink 108B. Therefore, the embodiments of this disclosure improve maintainability, reduce wear and tear on computing components (e.g., wear and tear from repeated installation / removal cycles), and shorten data center downtime, thereby reducing costs, reducing maintenance workload, and improving performance.

[0113] CPU 804 represents any suitable processing unit, processing unit package, or other high-power component that can be used in, for example, a data center environment, including any processing unit, processing unit package, and / or high-power component described herein. For the avoidance of doubt, CPU 804 may, for example, represent a GPU, a quantum processing unit (QPU), a parallel processing unit (PPU), an application-specific integrated circuit (ASIC), and / or any other processing unit or similar silicon package that generates heat during use.

[0114] Each of the storage modules 808A and 808B can be used, for example, as a storage module or other low-power component in a data center environment, and can be operatively connected to the CPU 804 (or other processing unit, processing unit package, or high-power component) via PCB 812. Storage modules 808A and 808B can be, for example, small form factor compact-fit memory modules (SOCAMM), high bandwidth memory (HBM), hybrid memory cubes (HMC), dual in-line memory modules (DIMM), network modules, optical modules, retimers, onboard optical modules (MBOM), any other memory, storage module, or low-power component described herein, and / or any other storage module or low-power component.

[0115] Although this disclosure discusses the use of a split heatsink (such as split heatsink 100) in conjunction with a processing unit and one or more memories, the split heatsink according to embodiments of this disclosure can be used in any application scenario where cooling of a high-power component and one or more adjacent low-power components is required.

[0116] Figure 9 A thermal diagram is shown illustrating two different split-type heatsink structures, similar to the aforementioned heatsink 100, in operation. The first structure uses a split-type heatsink 900 (as shown in Figure 1-). Figure 8 The second structure is essentially similar to the heatsink 100 in Figure 1, which has a central heatsink 904, two side heatsinks 908A and 908B, and multiple heat pipes 912 (e.g., for transferring heat from the computing components below the split heatsink 900 to horizontal fins, such as horizontal fins 124). The second structure employs a split heatsink 950 (also similar to Figure 1-). Figure 8 The heatsink 900 is substantially similar to the heatsink 950, having a central heatsink 954, two side heatsinks 958A and 958B, and multiple heat pipes 912 (e.g., for transferring heat from the computing components below the split heatsink 950 to horizontal fins, such as horizontal fins 124). The difference between the split heatsink 900 and the split heatsink 950 lies in the arrangement of the heat pipes 912, wherein the split heatsink 950 includes two additional heat pipes that extend upward through the horizontal fins of the side heatsinks 958A and 958B.

[0117] from Figure 9 It is evident that the split-type heatsink design of the 900 and 950 series enables a uniform heat distribution between the CPU heatsink (e.g., the central heatsink 904, 954) and the side heatsinks (e.g., side heatsinks 908A, 908B, 958A, 958B). Even under worst-case tolerance accumulation conditions for the side heatsinks—i.e., one side heatsink being at its maximum possible height and the other at its minimum possible height (e.g., due to manufacturing errors during side heatsink production / assembly)—the temperature distribution downstream of the heatsink remains relatively uniform. Figure 9 In the heat map test shown, the height difference between the two side wings is 0.6 mm, resulting in a thermal interface material layer thickness of 1.3 mm between one side wing (e.g., side wing radiators 908A, 958A) and the central radiator (e.g., central radiators 904, 954), and a thermal interface material layer thickness of 0.7 mm between the other side wing (e.g., side wing radiators 908B, 958B) and the central radiator (e.g., central radiators 904, 954). Even with inhomogeneities and tolerances in the thermal interface material itself, it is still possible to obtain... Figure 9 The results are shown.

[0118] In order to obtain Figure 9The results show a 300W CPU used below the central heatsinks 904 and 954, and a 40W SOCAMM module used below each side heatsink 908A, 908B, 958A, and 958B. The inlet air temperature is 35 degrees Celsius, and the airflow rate through the split heatsinks 900 and 950 is 75 cubic feet per minute. For the split heatsink 900, the maximum air temperature fluctuation along the length of the heatsink (e.g., from one end of side heatsink 908A to the opposite end of side heatsink 908B) is 4.1 degrees Celsius (range 40.5 to 41.6 degrees Celsius). For the split heatsink 950, the maximum air temperature fluctuation along the length of the heatsink (e.g., from one end of side heatsink 958A to the opposite end of side heatsink 958B) is 2.1 degrees Celsius (range 41.2 to 43.3 degrees Celsius). These results demonstrate that the split-type heat sink according to embodiments of the present disclosure can still achieve uniform heat dissipation between its fins even in the event of adverse tolerance accumulation.

[0119] Now refer to Figures 10-13 This describes various aspects and embodiments of data centers that can use the split-type heat sinks described above, as well as components of such data centers. Figure 10 A computer system 1000 according to at least one embodiment is shown. In at least one embodiment, the systems and devices described throughout this disclosure can be implemented in conjunction with a computer system such as computer system 1000.

[0120] In at least one embodiment, the computer system 1000 includes, but is not limited to, at least one central processing unit (“CPU”) 1002 connected to a communication bus 1010, which is implemented using any suitable protocol, such as PCI (“Peripheral Component Interconnect”), High-Speed ​​Peripheral Component Interconnect (“PCI-Express”), AGP (“Accelerated Graphics Port”), HyperTransport, or any other bus or one or more point-to-point communication protocols. In at least one embodiment, the computer system 1000 includes, but is not limited to, main memory 1004. Control logic (e.g., implemented in hardware, software, or a combination thereof) and data are stored in main memory 1004, which may be in the form of random access memory (“RAM”). In at least one embodiment, a network interface subsystem (“network interface”) 1022 provides an interface to other computing devices and networks for data transmission and reception between the computer system 1000 and other systems.

[0121] In at least one embodiment, the computer system 1000 includes, but is not limited to, an input device 1008, a parallel processing system 1012, and a display device 1006, which may be implemented using conventional cathode ray tube (“CRT”), liquid crystal display (“LCD”), light-emitting diode (“LED”), plasma display, or other suitable display technologies. In at least one embodiment, user input is received from the input device 1008 (e.g., a keyboard, mouse, touchpad, microphone, etc.). In at least one embodiment, each of the foregoing modules may reside on a single semiconductor platform to form the processing system.

[0122] In at least one embodiment, a computer program in the form of machine-readable executable code or computer control logic algorithms is stored in main memory 1004 and / or secondary storage. When executed by one or more processors, the computer program enables system 1000 to perform various functions according to at least one embodiment. Memory 1004, secondary storage, and / or any other storage device are possible examples of computer-readable media. In at least one embodiment, secondary storage can refer to any suitable storage device or system, such as hard disk drives and / or removable storage drives, representing floppy disk drives, magnetic tape drives, optical disk drives, digital versatile optical disc (“DVD”) drives, recording devices, Universal Serial Bus (“USB”) flash memory, etc. In at least one embodiment, the architecture and / or functionality of the various figures described herein are implemented in the context of CPU 1002, parallel processing system 1012, integrated circuits capable of implementing at least a portion of the capabilities of both CPU 1002 and parallel processing system 1012, chipsets (e.g., a set of integrated circuits designed to work together and sold as a unit to perform related functions), and any suitable combination of one or more integrated circuits.

[0123] In at least one embodiment, the architecture and / or functionality of the various figures described herein can be implemented in the following environments: general-purpose computer systems, circuit board systems, game console systems designed for entertainment purposes, and dedicated systems. In at least one embodiment, the computer system 1000 can take the form of: desktop computers, laptop computers, tablet computers, servers, supercomputers, smartphones (e.g., wireless handheld devices), personal digital assistants (“PDAs”), digital cameras, vehicles, head-mounted displays, handheld electronic devices, mobile phone devices, televisions, workstations, game consoles, embedded systems, and / or any other type of logic.

[0124] In at least one embodiment, the parallel processing system 1012 includes, but is not limited to, multiple parallel processing units (“PPUs”) 1016 and associated memory 1016. In at least one embodiment, the PPUs 1016 are connected to a host processor or other peripheral devices via interconnects 1018 and switches 1020 or multiplexers. In at least one embodiment, the parallel processing system 1012 distributes computational tasks to parallelizable PPUs 1016—for example, as part of distributing computational tasks across multiple graphics processing units (“GPUs”) thread blocks. In at least one embodiment, memory is shared and accessible (e.g., read and / or write access) among some or all of the PPUs 1016, although such shared memory may result in a performance penalty compared to using local memory and registers residing in the PPUs 1016. In at least one embodiment, the operation of the PPUs 1016 is synchronized using commands such as syncthreads(), which require all threads in a block (e.g., executing across multiple PPUs 1016) to reach a certain point in code execution before continuing.

[0125] Figure 10 Each of the CPU 1002 and PPU 1014 shown can be cooled, for example, using embodiments of this disclosure.

[0126] According to at least some embodiments of this disclosure, the following can be employed: Figure 11 The exemplary data center 1300 shown has a cooling system that follows the improvements described herein. Any one or more of the features described above can be utilized in data centers such as data center 1300. Many specific details are set forth herein to provide a thorough understanding, but the concepts herein can be implemented without one or more of these specific details.

[0127] In at least one embodiment, the data center cooling system can handle sudden surges in heat demand caused by varying computing loads in today's computing components. In at least one embodiment, because these demands are variable or typically fall within a range of minimum to maximum cooling requirements, a suitable cooling system must be employed to meet these demands economically. In at least one embodiment, a liquid cooling system can be used for medium to high cooling demands. In at least one embodiment, partial immersion cooling can economically meet high cooling demands. In at least one embodiment, these different cooling demands also reflect the different thermal characteristics of the data center. In at least one embodiment, the heat generated by these components, servers, and racks is collectively referred to as thermal characteristics or cooling demands, because the cooling demands must be perfectly adapted to the thermal characteristics.

[0128] In at least one embodiment, a split-type heatsink system is disclosed. In at least one embodiment, the split-type heatsink system adapts to the thermal characteristics of associated computing or data center equipment, including graphics processing units (GPUs), central processing units (CPUs), switches, and dual in-line memory modules (DIMMs). In at least one embodiment, these components may be referred to herein as high-heat-density computing components or simply high-power components. Furthermore, in at least one embodiment, the associated computing or data center equipment may be a processing card having one or more GPUs, switches, or CPUs. In at least one embodiment, each of the GPU, switch, and CPU may be a thermal characteristic of the computing device. In at least one embodiment, the GPU, CPU, or switch may have one or more cores, and each core may be a thermal characteristic.

[0129] In at least one embodiment, the split-type heat sink described above, more specifically the central heat sink or first heat sink segment described above, can be thermally coupled to one or more of the high-power components described above. In some embodiments, a liquid cooling system may be used instead of the split-type heat sink, or used in conjunction with the split-type heat sink (or only in conjunction with other components of the data center).

[0130] In at least one embodiment, the following can be employed: Figure 11The exemplary data center 1300 shown herein has a cooling system following the improvements described herein. In at least one embodiment, the data center 1300 may be or include one or more server rooms 1302 having racks 1310 and auxiliary equipment for accommodating one or more servers on one or more server trays. In at least one embodiment, the data center 1300 is supported by a cooling tower 1304 located outside the data center 1300. In at least one embodiment, the cooling tower 1304 dissipates heat inside the data center 1300 by acting on a main cooling loop 1306. In at least one embodiment, a cooling distribution unit (CDU) 1312 is disposed between the main cooling loop 1306 and a second or auxiliary cooling loop 1308 for absorbing heat from the second or auxiliary cooling loop 1308 to the main cooling loop 1306. In at least one embodiment, in one aspect, the auxiliary cooling loop 1308 may connect to various pipes in the server trays as needed. In at least one embodiment, loops 1306, 1308 are shown as lines in the figures, but those skilled in the art will recognize that one or more pipe features may be used. In at least one embodiment, flexible polyvinyl chloride (PVC) piping can be used in conjunction with matching conduits to deliver fluid in each provided loop 1306, 1308. In at least one embodiment, one or more coolant pumps can be used to maintain a pressure differential within the coolant loops 1306, 1308 to drive coolant flow based on temperature sensors at various locations, including within a server room, within one or more racks 1310, and / or within server chassis or server trays within one or more racks 1310.

[0131] In at least one embodiment, the coolant in the main cooling circuit 1306 and the auxiliary cooling circuit 1308 may contain at least water and an additive. In at least one embodiment, the additive may be ethylene glycol or propylene glycol. In at least one embodiment, each of the main cooling circuit and the auxiliary cooling circuit may have its own coolant during operation. In at least one embodiment, the coolant in the auxiliary cooling circuit may be a dedicated coolant according to the requirements of the components within the server tray or mounting rack 1310. In at least one embodiment, the CDU 1312 is capable of independent or synchronous fine control of the coolant in the coolant circuits 1306 and 1308. In at least one embodiment, the CDU may be configured to control the flow rate of the coolant so that the coolant can be appropriately distributed to absorb the heat generated within the mounting rack 1310. In at least one embodiment, a plurality of flexible conduits 1314 extend from the auxiliary cooling circuit 1308 into each server tray to provide coolant to the electrical and / or computing components therein.

[0132] Heat transfer fluids typically include water, aqueous solutions (e.g., propylene glycol-water solutions), brine, antifreeze, mixtures of antifreeze and water, oil, alcohol, mercury, or any other suitable heat-conducting fluid. The heat transfer fluid can be a conductive coolant and can include water, deionized water, or a coolant (e.g., R-134a), a mixture of water and additives (e.g., a mixture of water and ethylene glycol or a mixture of water and propylene glycol, such as a 25% propylene glycol solution in deionized water). The heat transfer fluid can also be a pure dielectric fluid (e.g., water-free for the purposes of this disclosure) or a combination of water and additives including at least one dielectric fluid, such as one or more of deionized water, ethylene glycol, and propylene glycol. In at least one embodiment, the heat transfer fluid can be applied to an absorption chiller, the working fluid of which is a mixed solution containing lithium bromide as the absorbent material and water as the carrier. The heat transfer fluid can also be a two-phase coolant with a boiling point below the expected operating temperature of the electronic device. Exemplary two-phase coolants include 2,3,3,3-tetrafluoropropylene, 1,1,1,2-tetrafluoroethane, and water.

[0133] In at least one embodiment, tubing 1318 forming part of auxiliary cooling loop 1308 may be referred to as a room manifold. Alternatively, in at least one embodiment, additional tubing 1316 may extend from row manifold tubing 1318 and may also be part of auxiliary cooling loop 1308, but may be referred to as a row manifold. In at least one embodiment, coolant tubing 1314, as part of auxiliary cooling loop 1308, extends into each rack, but may be referred to as a rack cooling manifold within one or more racks. In at least one embodiment, row manifold 1316 extends to all racks arranged in a row within data center 1300. In at least one embodiment, the tubing system of auxiliary cooling loop 1308 (including coolant manifolds 1318, 1316, and 1314) may be improved by at least one embodiment herein. In at least one embodiment, a chiller 1320 may be provided in the main cooling loop within data center 1300 to support cooling before the cooling tower. In at least one embodiment, for the purposes of this disclosure, an additional cooling circuit that may exist in the main control circuit and provide cooling outside the rack and outside the auxiliary cooling circuit may be classified as a class of the main cooling circuit and distinguished from the auxiliary cooling circuit.

[0134] In at least one embodiment, during operation, heat generated within the server trays of the rack 1310 can be transferred via flexible piping of the manifold 1314 of the auxiliary cooling circuit 1308 to coolant flowing out of one or more racks 1310. In at least one embodiment, auxiliary coolant from CDU 1312, used for cooling the racks 1310 (located in the auxiliary cooling circuit 1308), flows to one or more racks 1310 via provided piping. In at least one embodiment, auxiliary coolant from CDU 1312 flows from one side of the data center manifold with piping 1318, through manifold 1316 to one side of the rack 1310, and through different piping 1314 across one side of the server tray. In at least one embodiment, the heat-absorbing or recirculated auxiliary coolant (or outflow-type auxiliary coolant carrying heat from the computing components) flows out from the other side of the server tray (e.g., for a server tray, the coolant flows in from the left side of the rack, circulates through the server tray or components on the server tray, and then flows out from the right side of the rack). In at least one embodiment, the heat-absorbing auxiliary coolant flowing out of the server tray or rack 1310 flows out from the other side (e.g., the outflow side) of the conduit 1314 and flows into a parallel conduit in the manifold 1316, also on the outflow side. In at least one embodiment, the heat-absorbing auxiliary coolant flows from the manifold 1316 into a parallel conduit portion of the data center manifold 1318 and flows in the opposite direction to the inflow-type auxiliary coolant (which may also be a newer auxiliary coolant) towards the CDU 1312.

[0135] In at least one embodiment, the heat-absorbing auxiliary coolant exchanges heat with the main coolant in the main cooling circuit 1306 via CDU 1312. In at least one embodiment, the heat-absorbing auxiliary coolant can be refreshed (e.g., relatively cooled compared to the temperature of the heat-absorbing auxiliary coolant stage) and prepared to be circulated back to one or more computing components via auxiliary cooling circuit 1308. In at least one embodiment, various flow and temperature control mechanisms in CDU 1312 are capable of controlling the heat exchanged from the heat-absorbing auxiliary coolant or controlling the flow rate of the auxiliary coolant in and out of CDU 1312. In at least one embodiment, CDU 1312 is also capable of controlling the flow rate of the main coolant in the main cooling circuit 1306.

[0136] Even in such Figure 11 In the liquid-cooled data center 1300, a split-type heat sink (such as split-type heat sink 100) can also be used to facilitate the cooling of high-power components and adjacent or nearby low-power components. This is particularly advantageous in improving the maintainability of adjacent or nearby low-power components. Removing a single section of the split-type heat sink 100 to maintain the underlying components is far more efficient than removing, for example... Figure 11The aforementioned liquid-cooled radiator or cold plate is simpler.

[0137] In integrated circuit (IC) chip packaging, various packaging schemes are employed, including traditional two-dimensional (2D) IC packaging and the recently introduced 2.5D and 3D IC packaging. In 2D IC packaging, multiple chips are mounted on a printed circuit board, with high-performance logic, low-performance logic, memory, analog / radio frequency (RF) functions, and other functional components distributed as discrete devices within separate chip packages. In contrast, in 2.5D and 3D IC packaging, multiple IC chips are mounted on a silicon interposer, rather than a traditional packaging substrate. The silicon interposer, typically a silicon wafer, allows for the formation of very small and high-density conductive traces between multiple IC chips because the manufacturing process for forming these conductive traces is the same as the process for forming the metal interconnects in the metallization layer of the silicon chip.

[0138] Compared to 2.5D IC packages and 3D IC packages, circuit boards with individually packaged chips, such as 2D IC packages, have several disadvantages. For example, 2D IC packages are typically larger, heavier, and consume more power. Furthermore, 2D IC packages are slower than their equivalent 2.5D or 3D IC packages because signals travel relatively slowly from one chip to another on the circuit board. Additionally, 2D IC packages have more potential points of failure because solder joints on the circuit board are more prone to failure than electrical connections formed within the interposer. Nevertheless, troubleshooting 2D IC packages is relatively straightforward once the individual chips are mounted on the circuit board. Specifically, the conductive traces transmitting I / O signals between the chips on the circuit board are easily accessible and can therefore be used to measure specific I / O signals during troubleshooting.

[0139] In contrast, troubleshooting 2.5D or 3D IC packages is far more difficult because I / O signals transmitted between different chips are typically embedded in the silicon interposer, making physical contact impossible. Furthermore, due to the high bandwidth and high density of 2.5D and 3D IC packages, typical implementations can include thousands of conductive traces routed between different chips. An example of such an implementation is a memory bus between a processor and a high-bandwidth memory chip. In such implementations, even if physical contact with the traces from the silicon interposer is possible via probes, accurately and reliably selecting a particular conductive trace or combination of conductive traces for troubleshooting the IC package is extremely difficult, if not entirely impossible.

[0140] In at least one embodiment, one or more parallel processors (e.g., parallel processing unit 1014 of parallel processing system 1012) include circuitry optimized for graphics and video processing, including, for example, video output circuitry, and constitute a graphics processing unit (“GPU”). In at least one embodiment, one or more parallel processors include circuitry optimized for general-purpose processing. In at least one embodiment, components of a computing system (e.g., computer system 1000) may be integrated with one or more other system elements on a single integrated circuit. For example, in at least one embodiment, one or more parallel processors, a storage hub (e.g., main memory 1004 of computer system 1000), one or more processors (e.g., CPU 1002 of computer system 1000), and an I / O hub (used to enable communication with, for example, one or more display devices 1006, one or more input devices 1008, and / or network interface 1022 of a system such as computer system 1000) may be integrated into a system-on-a-chip (SoC) integrated circuit. In at least one embodiment, components of such a computing system may be integrated into a single package to form a system-in-package (“SIP”) structure. In at least one embodiment, at least some components of such a computing system may be integrated into a multi-chip module (“MCM”), which may interconnect with other MCMs to form a modular computing system. In at least one embodiment, such a computing system may omit an I / O subsystem and any display device.

[0141] Figure 12 This is a schematic block diagram illustrating a computing system 1600 (e.g., a data center or high-performance computing (HPC) cluster) according to embodiments described herein. According to at least one embodiment, system 1600 includes multiple subsystems, such as multiple processing devices, multiple network devices, and multiple networks coupled to each other. The computing system 1600 is designed with multiple integrated circuits (referred to as processing devices), each of which may include one or more CPUs and GPUs, forming a powerful and flexible architecture.

[0142] The various processing devices are interconnected via NVLink or other high-speed interconnects to enable high-speed communication between subsystems, and are also connected via NICs or DPUs to ensure efficient data transmission within the computing system 1600 and with one or more external networks 1630, 1636. In this example, system 1600 includes a packet switch 1648 that connects NIC / DPU 1628 to network 1630, and a packet switch 1650 that connects NIC / DPU 1632 to network 1636.

[0143] Seamless data exchange and parallel processing are enabled through NVLink coupling of the processing device, improving overall computing performance. The processing device connects to multiple networks via one or more Network Interface Controllers (NICs) or Data Processing Units (DPUs), enabling the system to handle complex multi-network tasks with high bandwidth and low latency. This configuration is ideal for applications requiring significant processing power, such as artificial intelligence (AI), machine learning (ML), and data-intensive computing, while ensuring robust connectivity and scalability across various network environments. The integrated circuits of the Computing System 1600 may include one or more CPUs and one or more GPUs.

[0144] Figure 12 An example architecture for a multi-GPU architecture is also shown. As illustrated, computing system 1600 includes a processing device 1602 with a multi-GPU architecture. Specifically, processing device 1602 may be a system-on-a-chip and includes multiple subsystems, such as CPU 1606, GPU 1608, and GPU 1610. CPU 1606 may be coupled to GPU 1608 via die-to-die (D2D) or chip-to-chip (C2C) interconnects 1612 (e.g., ground reference signal interconnects (GRS interconnects)). CPU 1606 may be coupled to GPU 1610 via D2D or C2C interconnects 1614. CPU 1606 may also be coupled to GPU 1608 and GPU 1610 via PCIe interconnects.

[0145] The CPU 1606 can be coupled to one or more NICs or DPUs, which in turn are coupled to one or more networks. For example, Figure 12 As shown, CPU 1606 is coupled to a first NIC / DPU 1626, which is coupled to network 1630. CPU 1606 is also coupled to a second NIC / DPU 1628, which is coupled to network 1630 via switch 1648. For example, NIC / DPU 1626 and NIC / DPU 1628 can be coupled to network 1630 via Ethernet (ETH), NVLINK, or InfiniBand (IB) connections.

[0146] The computing system 1600 also includes a processing device 1604 with a multi-GPU architecture. Specifically, the processing device 1604 includes multiple subsystems, including a CPU 1616, a GPU 1618, and a GPU 1620. The CPU 1616 can be coupled to the GPU 1618 via a D2D or C2C interconnect 1622. The CPU 1616 can be coupled to the GPU 1620 via a D2D or C2C interconnect 1624. The CPU 1616 can also be coupled to the GPU 1618 and GPU 1620 via a PCIe interconnect. The CPU 1616 can be coupled to one or more NICs or DPUs, which are then coupled to one or more networks. For example, as... Figure 12 As shown, CPU 1616 is coupled to a first NIC / DPU 1632, which is coupled to network 1636. CPU 1616 is also coupled to a second NIC / DPU 1634, which is coupled to network 1636 via switch 1650. NIC / DPU 1632 and NIC / DPU 1634 can be coupled to network 1636 via Ethernet (ETH), NVLINK, or InfiniBand (IB) connections.

[0147] In at least one embodiment, processing device 1602 and processing device 1604 can communicate with each other via NIC / DPU 1638, for example, via PCIe interconnect. Processing device 1602 and processing device 1604 can also communicate with each other via high-bandwidth communication interconnect 1640 (e.g., NVLink interconnect or other high-speed interconnect). Figure 12 The packet switches in the diagram may include, for example, Nvidia Quantum-2 switches. The NIC / DPU in the diagram may include, for example, Nvidia Bluefield DPUs.

[0148] One or more components of system 1600 may be cooled using a cooling system (such as system 100) described herein. For example, such a cooling system may be used to cool one or more of CPUs 1606 and 1616, one or more of GPUs 1608, 1610, 1618 and 1620, the entire superchip including CPUs 1606 and GPUs 1608 and 1610, and / or the entire superchip including CPUs 1616 and GPUs 1618 and 1620, and / or any other physical, heat-generating component of system 1600.

[0149] Figure 13An example computing environment 1700 according to at least one embodiment of the present disclosure is shown, in which forward pass offloading to available memory can be performed. It should be understood that embodiments of the present disclosure can also be used in conjunction with alternative environments, specific discussions of components are provided in a non-limiting manner, and equivalents may be included. Furthermore, some features have been omitted for clarity and brevity. Additionally, the systems and methods can be applied to a variety of different architectures. The example computing environment 1700 may include a server 1702, which can be used to perform HPC workloads, such as AI training or machine learning model training. In one embodiment, server 1702 may be an application instance or a compute node. Server 1702 may include a CPU 1710 associated with a switch 1720 (e.g., a high-speed peripheral component interconnect (PCIe) switch) that can control at least some data transmissions on communication paths interconnecting the components. In one embodiment, CPU 1710 may include a root complex processor.

[0150] PCIe switch 1720 may also be associated with GPU 1730 and DPU 1740, and may transfer data between at least some of CPU 1710, GPU 1730, DPU 1740, and other components. In one embodiment, PCIe switch 1720 may be associated with more than one GPU or more than one DPU. In another embodiment, PCIe switch 1720 may be located within DPU 1740. PCIe switch 1720 may manage the transfer of at least some data between CPU 1710, GPU 1730, and DPU 1740. In another embodiment, the number of GPUs associated with PCIe switch 1720 may be equal to the number of DPUs associated with PCIe switch 1720. In at least one embodiment, server 1702 may include, but is not limited to, any number of CPU 1710, PCIe switch 1720, GPU 1730, and / or DPU 1740, in any combination. For example, in at least one embodiment, server 1702 may include eight, sixteen, thirty-two, and / or more GPUs 1730. In at least one embodiment, Figure 13 The communication paths of the interconnected components (including but not limited to CPU 1710, PCIe switch 1720, GPU 1730 and DPU 1740) can be implemented using any suitable protocol, such as peripheral component interconnect (PCI) based protocols (e.g. PCIe), other buses, or point-to-point communication interfaces and / or protocols, such as NV-Link high-speed interconnect, or interconnect protocols.

[0151] The DPU 1740 may include a network interface controller (NIC) 1742, DDR memory 1744, and a high-speed non-volatile memory (NVMe) device 1746. The NIC 1742 is capable of interfaced with a network 1704, which may also interface with additional NVMe devices available to the DPU 1740, for example, through a structure. In one embodiment, the DPU 1740 may not include the NVMe device 1746. In another embodiment, the NVMe device 1746 may reside on a server 1702 instead of the DPU 1740. In yet another embodiment, the computing environment 1700 may include more than one NVMe device 1746, such as a first NVMe device located in the DPU 1740 and a second NVMe device located on the server 1702 and directly associated with the PCIe switch 1720. In one embodiment, the DPU 1740 may not include the DDR memory 1744, but may instead include a compute storage service (CSS) to replace or share with the DDR memory 1744. For example, computing environment 1700 may include DPU compute storage (CS) memory 1706, which is available to DPU 1740 as part of the CSS. Network 1704 may interface with DPU CS memory 1706 via NIC 1742, using any suitable interface protocol, such as Remote Direct Memory Access over Ethernet (RDMA), InfiniBand, Fibre Channel, etc.

[0152] By using DPU 1740 on system nodes, the total available memory for data storage in compute environment 1700 can be expanded. DPU 1740 can access pool 1750 of memory already available on server 1702, such as Double Data Rate (DDR) memory, onboard NVMe devices, on-structure NVMe devices, and CS. Memory pool 1750 may include at least one of DDR memory 1744, NVMe 1746, and DPU CS memory 1706. DPU 1740 can also access available memory from other DPUs and include it as part of pool 1750, and other DPUs can also access available memory from DPU 1740, such as pool 1750. This available memory can be accessed and used for data storage without increasing compute resources (e.g., compute nodes) as in other scenarios. The available pool 1750 accessible to DPU 1740 can be configured for server 1702 to expand the total available memory for data storage, for example, reducing the data storage load on CPU 1710 or GPU 1730, which can instead increase their memory utilization for processing. For example, during AI training, model states, residual states, activation functions, and checkpoints can be stored or unloaded onto pool 1750 accessible by DPU 1740.

[0153] As described above, system 100 can be configured to support air cooling and / or liquid cooling solutions. General Reference Figures 14-24 And refer to the specific details Figure 14 , Figure 15 and Figure 16 System 100 (which may be referred to as a liquid cooling system or liquid-cooled radiator) is shown as including a first split radiator assembly 1800A and a second split radiator assembly 1800B. The split radiator assemblies 1800A and 1800B can be uniformly cooled by a cooling fluid supplied through fluid inlet 1820. As will be discussed in further detail herein, each split radiator assembly 1800A and 1800B may have a section directly cooled by the cooling fluid, and other sections indirectly cooled by the cooling fluid (e.g., through thermal coupling with the section directly cooled by the cooling fluid).

[0154] System 100 is also shown to include a fluid outlet 1824 for receiving cooling fluid flowing through a first split radiator assembly 1800A and a second split radiator assembly 1800B. In some embodiments, a first inlet pipe 1828 may deliver cooling fluid from a fluid inlet 1820 to a first central radiator 1804A, which may also be referred to as a first radiator section of the first split radiator assembly 1800A. Cooling fluid may enter the first central radiator 1804A at a first chamber 1904A and then flow through the cold plate assembly of the first central radiator 1804A.

[0155] like Figure 15 and Figure 16 As shown, cooling fluid entering the first chamber 1904A can flow along the fluid flow path / channel 2004 into the first cold plate assembly of the first central radiator 1804A. The cooling fluid can then exit the first cold plate assembly of the first central radiator 1804A at the first coupling manifold 1908A. The first coupling manifold 1908A can be in fluid communication with a second coupling manifold 1908B, which is part of the second cold plate assembly of the second central radiator 1804B. The cooling fluid can then flow through the cold plate assembly of the second central radiator 1804B. After flowing through the cold plate assembly of the second central radiator 1804B, the cooling fluid can exit at the second chamber 1904B and then be delivered to the fluid outlet 1824 via the outlet pipe 1832.

[0156] In some embodiments, the first central radiator 1804A and the second central radiator 1804B are thermally coupled to heat-generating components (e.g., high-power components). Since the central radiators 1804A and 1804B are directly cooled by a cooling fluid, the efficiency of the central radiators 1804A and 1804B in extracting heat from the high-power components is significantly improved compared to radiator sections that are not directly cooled by a cooling fluid.

[0157] It should be understood that tubes 1820, 1824, and 1836 can be connected to the corresponding components of the radiator using any suitable connection mechanism. In some embodiments, O-rings can be used to seal the connection between tubes 1820, 1824, and 1836 and the components to which tubes 1820, 1824, and 1832 are inserted / connected. This connection allows for a relatively flexible design, allowing tubes 1820, 1824, and 1836 to rotate and / or move up and down for relatively easy maintenance. When using one or more O-rings, a leak detection sensor can advantageously be placed at or near the location of the O-ring. For example, the leak detection sensor can be integrated into a tube near the O-ring, or it can be inserted in series between the tube and the O-ring to help quickly detect leaks at the O-ring. As an alternative design to flexible tubes and O-rings, more rigid tubes can also be used, and these tubes can be connected to the corresponding components using a more robust mechanism (e.g., brazing, soldering, etc.).

[0158] In some embodiments, cooling fluid may be directed through the split heatsink assemblies 1800A and 1800B in a manner that helps to cool all components as efficiently as possible. As a non-limiting example, the cooling fluid may first be directed to the portions of the split heatsink assemblies 1800A and 1800B that are thermally coupled to low-power components. The cooling fluid may then be directed to cool high-power components. This sequential flow path logic helps protect low-power components (e.g., memory) from the heat generated by high-power components (e.g., the CPU).

[0159] like Figure 22 As shown, the coupling tube 1836 provides fluid communication between the first central radiator 1804A and the second central radiator 1804B. The coupling tube 1836 can be formed in a serpentine shape to accommodate a flexible connection between the central radiators 1804A and 1804B. In other words, the shape design of the coupling tube 1836 allows for communication between the central radiators 1804A and 1804B without relying on precision machining tolerances between them. Therefore, in the context of system 100, the installation / positioning of one central radiator (e.g., the first central radiator 1804A) does not directly determine the precise installation / positioning of the other central radiator (e.g., the second central radiator 1804B). The serpentine shape of the coupling tube 1836 allows for deviations in the installation height and distance between the central radiators 1804A and 1804B. In some embodiments, the coupling pipe 1836 may be made of a metal (e.g., copper) and may be brazed to the coupling manifolds 1908A, 1908B to provide a fluid seal between the coupling pipe 1836 and each coupling manifold 1908A, 1908B. Alternatively, the coupling pipe 1836 may be made of a rigid or semi-rigid plastic (e.g., polyvinyl chloride (PVC), high-density polyethylene (HDPE), nylon, and engineering plastics such as PEEK or PVDF), in which case the coupling pipe 1836 may be secured to the coupling manifolds 1908A, 1908B by adhesives, friction fits, or similar methods. The inlet pipe 1828 and outlet pipe 1832 may also be made of the same metal, plastic, or similar material as the coupling pipe 1836. In some embodiments, the materials selected for pipes 1828, 1832, and / or 1836 may be used to ensure that cooling fluid can flow through the components of system 100 at pressures of 15 PSI or higher.

[0160] Figure 14The diagram further illustrates that the handle 1840 can be used to connect the first central radiator 1804A to the second central radiator 1804B. The handle 1840 can correspond to a rigid or semi-rigid structure that can be connected to the two central radiators 1804A, 1804B via one or more support posts or spacers 1844. In some embodiments, the handle 1840 can be used to move or manipulate the components of the system 100 (e.g., for the purpose of moving the system 100, for the purpose of installing the system 100, and / or for the purpose of physically adjusting the position of the system 100). Compared to the coupling tube 1836, the handle 1840 can provide a more rigid connection between the central radiators 1804A, 1804B. The handle 1840 and spacers 1844 can be made of the same metal, plastic, or similar material used for one or more tubes 1828, 1832, 1836.

[0161] Overall reference Figures 14-24 And refer to more specifically Figure 14 , Figure 15 , Figure 17 and Figure 21 Further details regarding the modularity of the split radiator assemblies 1800A and 1800B will be described according to at least some embodiments of the present disclosure. Each split radiator assembly 1800 may include a central radiator section and a plurality of side radiators 1808. As a non-limiting example, a first central radiator 1804A is shown connected to a first side radiator 1808A and a second side radiator 1808B. In some embodiments, each of the first side radiator 1808A and the second side radiator 1808B may be connected to the first central radiator 1804A using at least one screw 1848. In some embodiments, each of the first side radiator 1808A and the second side radiator 1808B is capable of being fitted to and removed from the first central radiator 1804A. The assemblability and removability of the side radiators 1808A and 1808B facilitate modular maintenance of the components below one of the sections (e.g., below the side radiators 1808A and 1808B) without removing other sections (e.g., the first central radiator 1804A) from the heat-generating components below the first central radiator 1804A.

[0162] Similar to the first split-type radiator assembly 1800A, the second split-type radiator assembly 1800B may include a set of modular elements. The second central radiator 1804B is shown connected to a third side radiator 1808C and a fourth side radiator 1808D. In some embodiments, each of the third side radiator 1808C and the fourth side radiator 1808D may be connected to the second central radiator 1804B using at least one screw 1848. In some embodiments, each of the third side radiator 1808C and the fourth side radiator 1808D can be fitted onto and removed from the second central radiator 1804B. The assemblability and removability of the side radiators 1808C and 1808D facilitate modular maintenance of the components below one of the sections (e.g., below the side radiators 1808C and 1808D) without removing other sections (e.g., the second central radiator 1804B) from the heat-generating components below the second central radiator 1804B.

[0163] In addition to a physical connection between the side radiators 1808 and the central radiator 1804, a thermal connection can also be established between them. In some embodiments, each side radiator 1808 may include one or more heat pipes 2108 for transferring heat between the side radiator 1808 and the central radiator 1804. One or more heat pipes 2108 may extend between the thermal coupling member 2104 and the lower section 2504 of the side radiator 1808. The lower surface of the thermal coupling member 2104 may abut or physically contact the top surface of the central radiator 1804 to establish a thermal connection between the components. At least one screw 1848 may be used to secure the physical connection between the side radiators 1808 and the central radiator 1804.

[0164] One or more additional screws 1848 may be provided on the mounting boss 2508 of the side wing heat sink 1808. The additional screws 1848 provided on the mounting boss 2508 may be configured to connect the side wing heat sink 1808 to the underlying substrate or plate on which heat-generating components (e.g., high-power components and / or low-power components) are mounted. By securing the side wing heat sink 1808 to the underlying substrate or plate, a direct thermal connection can be established between the bottom surface of the lower segment 2504 and the top surface of the heat-generating component (e.g., low-power component).

[0165] In some embodiments, the side radiator 1808 can be easily and safely removed from the central radiator 1804 without removing the central radiator 1804 from its mounting base or plate. As an example, after removing the screws connecting the two segments of the radiator, the side radiator 1808 can be easily removed from the central radiator 1804 by a relatively simple sliding motion. The segments of the radiator can be configured with sufficient mechanical clearance to allow for removal of the heat diffuser using a first rearward sliding motion and a second lateral sliding motion. As a more specific but non-limiting configuration, a "rearward and leftward sliding" motion can be used to remove the side radiator 1808 from the central radiator 1804. In some embodiments, the side radiator 1808 may also include one or more handles or brackets mounted on its upper surface to facilitate user gripping and make the sliding removal motion easier to apply to the side radiator 1808.

[0166] Now refer to Figures 18A-18B and Figures 23A-24 Further details of the liquid-cooled cold plate or cold plate assembly will be described based on at least some embodiments of this disclosure. Figure 24 As shown, a liquid-cooled cold plate may include a cold plate thermal contact 2008, a cold plate body 2304, and a cold plate cover 2700. The cold plate thermal contact 2008 may be fitted into a base 2404 formed on a first side of the cold plate body 2304. A second side of the cold plate body 2304 (e.g., opposite its first side) may mate with the cold plate cover 2700. Thus, the cold plate body 2304 is configured to be located between the cold plate thermal contact 2008 and the cold plate cover 2700. One, some, or all of the components of the liquid-cooled cold plate may be made of metal or a similar type of thermally conductive material. The liquid-cooled cold plate may be configured to extract heat from a heat-generating component (e.g., a high-power component) that is in physical contact with the cold plate thermal contact 2008.

[0167] like Figure 18A and Figure 18BAs shown, the cold plate thermal contact 2008 may include a first side 2216 and a second side 2220. The first side 2216 may be configured to directly or indirectly (e.g., with a TIM between them) contact a heat-generating component. A plurality of heat dissipation fins 2208 may be formed on the second side 2220. The heat dissipation fins 2208 may be formed on the second side 2220 of the cold plate thermal contact 2008 by shaving, machining, extrusion, or other means. In some embodiments, forming the heat dissipation fins 2208 by a shaving process can create substantially uniform fins on the second side 2220. The shaving process can use precise, specialized blades to scrape and lift a thin layer of metal (e.g., aluminum or copper) from a solid substrate, thereby forming a high-performance, monolithic cold plate thermal contact 2008. The shaving process can help form dense, thin fins 2208 that are part of the cold plate thermal contact 2008, thereby eliminating thermal interface resistance and significantly improving thermal conductivity compared to extruded radiators. However, it should be understood that the fin 2208 can be formed using an extrusion process without departing from the scope of this disclosure.

[0168] The cold plate heat contact component 2008 also includes a flange 2204 having an edge 2212 that seals with a corresponding flange on the base 2404. For example... Figure 20A and Figure 20B As shown, the cold plate body 2304 includes a base 2404, which simultaneously forms a structure for engaging the cold plate thermal contact member 2008 and a fluid cavity 2408 for the flow of cooling fluid. More specifically, the cold plate body 2304 is shown to include a fluid hole 2312 within the area defined by the base 2404. The fluid flow 2004 can pass through the fluid cavity 2408 and flow through the fins 2208 of the cold plate thermal contact member 2008. Cooling fluid can enter and / or exit the fluid cavity 2408 via the fluid hole 2312. Although the cold plate body 2304 is shown to have a single fluid hole 2312, it should be understood that more than one fluid hole 2312 may be provided in the cold plate body 2304 without departing from the scope of this disclosure.

[0169] The cold plate body 2304 may also include a plurality of mounting holes 2308 for receiving screws, bolts, or other suitable fasteners to secure the components of the cold plate together and to the underlying substrate or plate. In some embodiments, the cold plate cover 2700 may have a plurality of mating mounting holes 2308 substantially aligned with the mounting holes 2308 of the cold plate body 2304. Thus, screws 1848 or similar fasteners can pass through the aligned mounting holes 2308 to secure the cold plate to the substrate or plate.

[0170] like Figure 19A and Figure 19BAs shown, the cold plate body 2304 may also have a flow channel 2324 formed on the side facing the cold plate cover 2700. The flow channel 2324 can be used to define and maintain the fluid flow 2004 through the cold plate. In some embodiments, the flow channel 2324 may extend through a large portion (e.g., more than 50%) of the cold plate body 2304 to define the path of the fluid flow 2004 through the thermally conductive material of the cold plate body 2304. Providing an extended flow channel 2324 helps to improve the thermal performance of the liquid-cooled cold plate. In some embodiments, the flow channel 2324 may be formed as a groove, recess, or a series of grooves on the side of the cold plate body 2304 facing the cold plate cover 2700.

[0171] The cold plate cover 2700 can be used to seal the flow channel 2324, and also provides a location for cooling fluid to enter and exit the flow channel 2324. Specifically, the cold plate cover 2700 can be brazed or otherwise attached to the cold plate body 2304 to press the first surface 2716 of the cold plate cover 2700 against the side of the cold plate body 2304 facing the cold plate cover 2700. The cold plate cover 2700 may also include a fluid orifice 2712 that provides a passage between chambers 1904, to which fluid conduits 1828, 1832 are connected. Around the fluid orifice 2712, the cold plate cover 2700 may also have a ridge 2704 that provides a mating structure for the chambers 1904. In addition, a seal 2316 can be provided between the cold plate cover 2700 and the cold plate body 2304, substantially aligned with the fluid orifice 2712. Seal 2316 may mate with ridge 2704 to maintain a fluid seal between cold plate cover 2700 and chamber 1904. In the illustrated embodiment, chamber 1904 is provided as a separate component from cold plate cover 2700.

[0172] The cold plate cover 2700 may further include a coupling manifold 1908 formed therein. In some embodiments, the cold plate cover 2700 may be machined, molded, or otherwise formed to include the coupling manifold 1908 as an integral component therein. Both the coupling manifold 1908 and the ridge 2704 are shown located on the second surface 2720 of the cold plate cover 2700, while the first surface 2716 of the cold plate cover 2700 may be coupled to the cold plate body 2304. In some embodiments, the first surface 2716 of the cold plate cover 2700 may further include a plurality of fins 2708 formed therein. The fins 2708 may be formed in a manner similar to that of fins 2208. In some embodiments, the fins 2708 may extend downward from the first surface 2716 into the flow channel 2324 to enhance heat exchange between the cold plate and the cooling fluid flowing therethrough. One, some, or all of the components of the cold plate assembly (e.g., cold plate thermal contact 2008, cold plate body 2304, and / or cold plate cover 2700) may be made of metal or similar thermally conductive material. The cold plate assembly constitutes a major part of the central radiator 1804A or 1804B for cooling heat-generating components (e.g., high-power components) in thermal communication with it. The cold plate assembly may also include multiple mounting holes facilitating connection to a substrate or plate, and further facilitating modular connection to one or more side fin radiators 1808A, 1808B, 1808C, 1808D. As further described above, a thermal coupling element 2104 for the side fin radiator may be provided on the second surface 2720 of the cold plate cover 2700, in which heat pipes 2108 may be arranged. Heat can be transferred between the cold plate assembly and the side fin radiators thermally coupled thereto via one or more heat pipes 2108 to efficiently transfer heat from the heat-generating components mounted below the central radiator and the heat-generating components mounted below the side fin radiators.

[0173] In some embodiments, the flow channels 2324, manifold 1908, and other components of the cold plate assembly can be configured to ensure that heat generated by one component being cooled does not inadvertently damage other components of the system. For example, the cold plate body 2304 and manifold 1908 can be configured in parallel or in series. Continuing with this example, employing a parallel configuration, using parallel flow paths before or within the cold plate body 2304, and / or using flow channels 2324 with branch manifolds all contribute to improved heat dissipation under low-flow conditions where series flow paths would lead to overheating. In some embodiments, the manifold 1908 can be configured to guide two independent fluid flow paths from each of the side radiators 1808A, 1808B, 1808C, and 1808D, and merge these two independent fluid flow paths (e.g., from two different pipes) into a single fluid flow, which is then supplied to the central radiator 1804 and the cold plate body 2304 contained therein. In embodiments where the side fin heat sinks 1808A, 1808B, 1808C, and 1808D receive cooling first, system 100 can be designed to have sufficient thermal tolerance to allow the use of thicker and more compliant materials (e.g., system 100 can tolerate greater height tolerances between components).

[0174] Now refer to Figure 25 A method 2500 for maintaining system 100 will be described according to at least some embodiments. Method 2500 may include physically separating a side fin heatsink 1808 from a central heatsink 1804, thereby enabling access to components beneath one segment without removing both segments from their mounting base or plate. The side fin heatsink 1808 can be separated from the central heatsink 1804 by removing at least one fastener coupling the side fin heatsink 1808 (step 2504). In some embodiments, the fastener may correspond to screw 1848, multiple screws, etc.

[0175] Once the fasteners used to physically couple the radiator segments have been removed, method 2500 can proceed by sliding the side radiator 1808 along a first direction without moving the central radiator 1804 (step 2508). As an example, the side radiator 1808 can slide rearward relative to the central radiator 1804 while maintaining fluid communication between the central radiator 1804 and one or more pipes supplying cooling fluid to it.

[0176] Method 2500 may further include the step of sliding the side wing radiator 1808 along a second direction (step 2512). As an example, after the side wing radiator 1808 has slid along a first direction (e.g., rearward), the side wing radiator 1808 may be slid along a second direction (e.g., rightward or leftward) that is different from or orthogonal to the first direction.

[0177] After the side radiator 1808 is properly slid to disconnect it from the central radiator 1804, the side radiator 1808 can be removed from the system 100 to expose the underlying components, allowing maintenance of these components without removing the central radiator 1804 (step 2516). Thus, components beneath parts of one radiator segment (e.g., the central radiator 1804) can remain in place, while other components beneath parts of another radiator segment (e.g., the side radiator 1808) can be accessed for maintenance.

[0178] Embodiments of this disclosure can be made from any suitable materials known in the art (e.g., carbon steel, aluminum, polymers, ceramics, etc.), particularly materials with high thermal conductivity. As an example, the components described herein can be manufactured using extrusion and / or machining processes. In such examples, a monolithic body with a fixed cross-sectional area can be manufactured using an extrusion process.

[0179] The embodiments of this disclosure may be described in view of the following terms: 1. A radiator, comprising: The first section is used to extract heat generated by high-power components; and At least a second segment is capable of being assembled to and detached from the first segment, the at least second segment being used to extract heat generated by the low-power component, wherein, when the at least second segment is assembled to the first segment, the at least second segment is thermally coupled to the first segment, such that the heat generated by the high-power component is further dissipated through the at least second segment.

[0180] 2. The heat sink according to Clause 1, wherein the high-power component includes at least one of a graphics processing unit (GPU) and a central processing unit (CPU).

[0181] 3. The heat sink according to Clause 1 or 2, wherein the low-power component includes a pluggable module.

[0182] 4. The heat sink according to any one of Clauses 1-3, wherein the storage module comprises at least one of the following: small form factor compact laminated memory module (SOCAMM), dual in-line memory module (DIMM), network module, optical module, retimer, and onboard optical module (MBOM).

[0183] 5. The radiator according to any one of Clauses 1-4, wherein the low-power component can be maintained without removing the high-power component from the first segment.

[0184] 6. The radiator according to any one of clauses 1-5, wherein the first section is assembled to the at least second section by one or more screws.

[0185] 7. The radiator according to any one of clauses 1-6, wherein the at least second section accommodates one or more heat pipes.

[0186] 8. The radiator according to any one of Clauses 1-7, wherein the one or more heat pipes are thermally coupled to the first segment and conduct heat from the first segment to the at least the second segment.

[0187] 9. The radiator according to any one of clauses 1-8, wherein the first segment and the at least the second segment each comprise one or more fin structures.

[0188] 10. The heat sink according to any one of Clauses 1-9, wherein the at least second segment comprises a second segment thermally coupled to a first low-power component and a third segment thermally coupled to a second low-power component.

[0189] 11. The radiator according to any one of clauses 1-10, wherein the first segment is located between the second segment and the third segment.

[0190] 12. A system comprising: Processing unit; First storage module; A first heat sink section is used to extract heat generated by the processing unit; and A second heat sink section is used to extract heat generated by the first storage module, wherein the second heat sink section can be assembled to and removed from the first heat sink section, and wherein the first heat sink section and the second heat sink section, in the assembled state, jointly transfer heat from the processing unit.

[0191] 13. The system according to Clause 12, further comprising: A printed circuit board (PCB) is connected to both the processing unit and the first storage module.

[0192] 14. The system according to Clause 12 or 13, wherein the PCB transmits electrical signals between the processing unit and the first storage module.

[0193] 15. The system according to any one of clauses 12-14 further includes: Second storage module; and A third heat sink section is used to extract heat generated by the second storage module, wherein the third heat sink section can be assembled to and removed from the first heat sink section, and wherein the first heat sink section and the third heat sink section jointly transfer heat from the processing unit in the assembled state.

[0194] 16. The system according to any one of Clauses 12-15, wherein the first storage module can be maintained by removing the first heat sink section from the second heat sink section.

[0195] 17. The system according to any one of clauses 12-16 further includes: A reinforcing base is provided to provide structural support for the first heat sink segment and to provide a connection interface between the first heat sink segment and the second heat sink segment.

[0196] 18. The system according to any one of clauses 12-17 further includes: Thermal interface material (TIM).

[0197] 19. The system according to any one of clauses 12-18, wherein the TIM is disposed between the reinforcing base and the second heat sink section, or between the second heat sink section and the first storage module, wherein the TIM comprises at least one of thermal grease, spacer pad, phase change material, and thermal putty, wherein the TIM provides substantially uniform cooling between the reinforcing base, the first heat sink section, and / or the second heat sink section, and wherein the TIM adapts to tolerance differences between the reinforcing base, the first heat sink section, the second heat sink section, the processing unit, and / or the storage module.

[0198] 20. A split-type radiator, comprising: A central radiator is configured to extract heat generated by the processing unit; and A first side wing heat sink is configured to extract heat generated by the storage module, the first side wing heat sink including at least one heat pipe and at least one fin. The first side wing heat sink is detachably mounted to the central heat sink and is configured to extract a portion of the heat generated by the processing unit when mounted to the central heat sink.

[0199] With the help of the teachings in the foregoing description and the accompanying drawings, those skilled in the art will conceive of many modifications and other embodiments of the disclosure set forth herein. Although the drawings show only some components of the apparatus and systems described herein, it should be understood that various other components (e.g., components of printed circuit boards, transceivers, cables, etc.) may be used in conjunction with the components described herein. Therefore, it should be understood that the disclosure is not limited to the specific embodiments disclosed, and that modifications and other embodiments are intended to be included within the scope of the appended claims.

[0200] It should be understood that any feature described herein may be combined with any other one or more features described herein to form a claim, regardless of whether such features are derived from embodiments of the same description.

[0201] Specific details are set forth in this specification to provide a thorough understanding of the embodiments. However, those skilled in the art will understand that these embodiments can be practiced without these specific details. In other instances, well-known circuits, processes, algorithms, structures, and techniques may be shown without including unnecessary details in order to avoid obscuring the embodiments.

[0202] While illustrative embodiments of this disclosure have been described in detail herein, it should be understood that the inventive concept may be embodied and applied in other ways, and the appended claims are intended to be construed as including such variations, unless limited by the prior art.

Claims

1. A radiator, comprising: The first section is used to extract heat generated by high-power components; as well as At least a second segment is capable of being assembled to and detached from the first segment, the at least second segment being used to extract heat generated by the low-power component, wherein, when the at least second segment is assembled to the first segment, the at least second segment is thermally coupled to the first segment, such that the heat generated by the high-power component is further dissipated through the at least second segment.

2. The radiator according to claim 1, wherein a flow channel for conveying cooling fluid is provided between the first section and the high-power component.

3. The radiator according to claim 2, wherein the first section includes a cold plate assembly, the cold plate assembly having a cold plate body, a cold plate cover, and a cold plate thermal contact element, wherein, The cold plate body is located between the cold plate cover and the cold plate thermal contact, and the flow channel passes between the cold plate body and the cold plate thermal contact to allow the heat generated by the high-power component to be transferred to the cooling fluid via the cold plate thermal contact.

4. The radiator according to claim 3, wherein the cold plate thermal contact member comprises a plurality of fins, and wherein the cooling fluid flows through the plurality of fins.

5. A system comprising: Processing unit; First storage module; The first heat sink section is used to extract the heat generated by the processing unit; as well as A second heat sink section is used to extract heat generated by the first storage module, wherein the second heat sink section can be assembled to and removed from the first heat sink section, and wherein the first heat sink section and the second heat sink section, in the assembled state, jointly transfer heat from the processing unit.

6. The system according to claim 5, further comprising: A printed circuit board (PCB) is connected to both the processing unit and the first storage module.

7. The system of claim 6, wherein the PCB transmits electrical signals between the processing unit and the first storage module.

8. The system according to claim 5, further comprising: Second storage module; as well as A third heat sink section is used to extract heat generated by the second storage module, wherein the third heat sink section can be assembled to and removed from the first heat sink section, and wherein the first heat sink section and the third heat sink section jointly transfer heat from the processing unit in the assembled state.

9. The system of claim 5, wherein the first storage module can be maintained by removing the first heat sink segment from the second heat sink segment.

10. The system according to claim 5, further comprising: A reinforcing base is provided to provide structural support for the first heat sink segment and to provide a connection interface between the first heat sink segment and the second heat sink segment.

11. The system of claim 10, further comprising: Thermal interface material (TIM).

12. The system of claim 11, wherein the TIM is disposed between the reinforcing base and the second heat sink segment, or between the second heat sink segment and the first storage module, wherein the TIM comprises at least one of thermal grease, spacer pad, phase change material, and thermal putty, wherein the TIM provides substantially uniform cooling between the reinforcing base, the first heat sink segment, and / or the second heat sink segment, and wherein the TIM adapts to tolerance differences between the reinforcing base, the first heat sink segment, the second heat sink segment, the processing unit, and / or the storage module.

13. The system according to claim 5, further comprising: fluid inlet; Fluid outlet; An inlet pipe that delivers cooling fluid between the fluid inlet and the first radiator section; as well as The flow channel contained within the first radiator section, wherein the cooling fluid flows through the flow channel to extract the heat generated by the processing unit.

14. The system according to claim 5, wherein the first heat sink section is liquid-cooled.

15. The system according to claim 5, wherein the first heat sink section is air-cooled.

16. A system comprising: The first split-type radiator assembly includes: The first central radiator is liquid-cooled; and At least one side fin radiator, thermally coupled to the first central radiator; and The second split-type radiator assembly includes: The second central radiator is liquid-cooled; and At least one additional side fin radiator is thermally coupled to the second central radiator.

17. The system according to claim 16, wherein, The first central radiator is connected to the at least one side radiator by at least a first screw, wherein the second central radiator is connected to the at least one additional side radiator by at least a second screw, wherein the at least one side radiator can be mounted to and removed from the first central radiator, and wherein the at least one additional side radiator can be mounted to and removed from the second central radiator.

18. The system of claim 17, wherein the first central radiator is coupled to the second central radiator via a coupling pipe, the coupling pipe conveying cooling fluid between the first central radiator and the second central radiator.

19. The system of claim 18, wherein the at least one side fin heatsink is thermally coupled to the first central heatsink via at least a first heat pipe, and wherein the at least one additional side fin heatsink is thermally coupled to the second central heatsink via at least a second heat pipe.

20. The system of claim 16, wherein the first central radiator extracts heat generated by the first high-power component, wherein the second central radiator extracts heat generated by the second high-power component, wherein the at least one side radiator extracts heat generated by the first low-power component, wherein the at least one additional side radiator extracts heat generated by the second low-power component, and wherein, The cooling fluid flows along a flow path that allows it to pass through the first low-power component and the second low-power component before passing through the first high-power component and the second high-power component, in order to prevent the heat generated by the first high-power component and the second high-power component from damaging the first low-power component and the second low-power component.

Images