Electronic assembly and cabinet server

Abstract

An electronic assembly and a whole-cabinet server are provided. The electronic assembly includes a pair of processing elements, a chip cooling element, and a power supply cooling element. The pair of processing elements are disposed in a stacked manner in a thickness direction. Each processing element includes a carrier board, a chip, and a power supply unit. The carrier board has a first side and a second side opposite to each other in the thickness direction. The chip is disposed on the first side of the carrier board. The power supply unit is electrically connected with the carrier board. The chip cooling element is in contact with the chip. The power supply cooling element is in contact with the power supply unit, wherein the chip cooling element or the power supply cooling element is disposed between the pair of processing elements.

Description

Technical Field

[0001] The examples in this article generally relate to the field of electronic equipment technology, and in particular to an electronic component and rack server. Background Technology

[0002] With the continuous increase in chip power consumption and current density, data center and server products are placing higher demands on the integration density of electronic components. Traditional power supply architectures typically employ a planar layout, placing chips, power modules, and their heat dissipation structures on the same side of the carrier board. This approach makes it difficult to achieve high-density chip integration within the limited height of the server rack. Furthermore, high-power chips and power modules generate significant heat during operation, and traditional heat dissipation solutions often struggle to simultaneously meet the cooling requirements of both chips and power units, thus limiting the overall performance and integration of electronic components. Summary of the Invention

[0003] In some cases, an electronic assembly is provided. The electronic assembly includes a pair of processing elements, a chip cooling element, and a power cooling element. The pair of processing elements are stacked in the thickness direction, and each processing element includes a carrier board, a chip, and a power supply unit. The carrier board has a first side and a second side opposite to each other in the thickness direction. The chip is disposed on the first side of the carrier board, and the power supply unit is electrically connected to the carrier board. The chip cooling element contacts the chip. The power cooling element contacts the power supply unit, wherein the chip cooling element or the power cooling element is disposed between the pair of processing elements.

[0004] In some cases, a rack-mount server is provided. This rack-mount server includes a rack and the aforementioned electronic components, which are housed within the rack.

[0005] According to the examples in this paper, by stacking pairs of processing elements in the thickness direction and placing chip cooling elements or power cooling elements between the pairs of processing elements, the integration density of electronic components can be increased, while achieving effective thermal management of chips and power units.

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

[0007] The above and other features, advantages, and aspects of the examples herein will become more apparent when taken in conjunction with the accompanying drawings and the following detailed description. In the accompanying drawings, the same or similar reference numerals denote the same or similar elements, wherein: Figure 1 An example usage environment for the electronic component is shown; Figure 2Schematic diagrams of electronic components in some scenarios are shown, in which the power supply unit is located on the second side of the carrier board; Figure 3 Schematic diagrams of electronic components in some scenarios are shown, in which the power supply unit is located within the first opening of the carrier board; Figure 4 Schematic diagrams of electronic components in some scenarios are shown, in which a power supply cooling element (30) is in contact with two power supply units; and Figure 5 Schematic diagrams of electronic components are shown in some scenarios, where paired bus power boards are positioned between paired processing elements.

[0008] Explanation of reference numerals in the attached figures: X, thickness direction; 1000. Example usage environment; 100. Electronic components; 10. Processing element; 11. Carrier board; 110. First opening; 111. First side; 112. Second side; 12. Chip; 13. Power supply unit; 131. Power supply carrier board; 132. Power supply module; 20. Chip cooling element; 201. Heat sink; 21. First chip cooling element; 30. Power supply cooling element; 31. First power supply cooling element; 40. Busbar power supply board; 51. First connector; 52. Second connector; 60. Voltage conversion unit; 700, server rack. Detailed Implementation

[0009] The examples described herein will now be described in more detail with reference to the accompanying drawings. While examples of this document are shown in the drawings, it should be understood that this document can be implemented in various forms and should not be limited to the examples set forth herein. Rather, these examples are provided to make this document more thorough and complete, and to fully convey the scope of this document to those skilled in the art.

[0010] The term “comprising” and its variations, as used herein, signify open inclusion, i.e., “including but not limited to.” Unless otherwise stated, the term “or” means “and / or.” The term “based on” means “at least partially based on.” The term “one example” means “at least one example.” The term “another example” means “at least one additional example.” The terms “first,” “second,” etc., may refer to different or the same objects.

[0011] As data centers and server products continue to demand higher computing performance, the current density of high-power chips is constantly increasing. At the same time, high-power products are placing higher demands on chip density on single boards. Within the 1U height (approximately 40mm or more) and 1-meter depth of a standard rack chassis, several or even dozens of high-power chips often need to be integrated. This dual requirement of high current density and high integration poses a severe challenge to traditional power supply architectures.

[0012] Traditional power supply architectures typically employ a planar layout, placing chips, power modules, and their heat dissipation structures on the same side of the carrier board. In this layout, chips and power modules occupy separate horizontal spaces, limiting the number of chips that can be integrated per unit area. When products aim for higher density, high-power chip integration, traditional planar layouts struggle to achieve the desired integration goals within the limited height of the server rack.

[0013] Besides integration limitations, heat dissipation is also a bottleneck in traditional power supply architectures. High-power chips and power modules generate a lot of heat during operation, and traditional heat dissipation solutions often struggle to meet the heat dissipation requirements of both the chips and the power units simultaneously. In traditional architectures, the chip-side power supply may use a vertical power supply method, such as core voltage power supply (CoreVoltage, VCORE), but the bus power supply of the power module itself still uses the traditional solution. This hybrid architecture presents inconsistencies in thermal management.

[0014] The combination of the above factors may lead to the following situations: the power supply and heat dissipation architecture cannot meet the power supply and heat dissipation requirements of the chip, resulting in the product's failure; or the limitations of the power supply and heat dissipation system may force the sacrifice of product performance in exchange for stable system operation.

[0015] To address the aforementioned challenges, this paper proposes a vertical power supply scheme employing a sandwich structure. This scheme, by stacking pairs of processing elements along the thickness direction and placing chip cooling or power supply cooling elements between the pairs of processing elements, can increase the integration density of electronic components within standard rack height constraints, while simultaneously achieving effective thermal management of the chip and power supply units.

[0016] To facilitate understanding of the technical solutions presented in this paper, the following explanations are provided for the terms used in this paper.

[0017] The processing element described herein refers to a functional unit that carries a chip and its power supply components. A processing element may include a carrier board, a chip mounted on the carrier board, and a power supply unit electrically connected to the carrier board. In some cases, the processing element may also include voltage conversion devices such as voltage conversion units. In some cases, the processing element may also be referred to as a server processing node or a server computing unit.

[0018] The carrier board described herein refers to a circuit board structure used to carry and electrically connect chips and power supply units. The carrier board has a first side and a second side that are opposite to each other in the thickness direction. The chip can be mounted on the first side of the carrier board. The carrier board can be made of a printed circuit board (PCB) or a substrate material similar to that of the chip.

[0019] The term "chip" as used herein refers to an integrated circuit device that performs computational or data processing functions. A chip can be the computing core of a product, consuming electrical energy and generating heat during operation. In some cases, each processing element may include multiple chips.

[0020] The power supply unit mentioned in this document refers to the power supply module that provides the operating voltage to the chip. The power supply unit may include a power carrier board and a power module. The power module is electrically connected to the power carrier board, and the power carrier board is electrically connected to a corresponding carrier board. The power supply unit can convert the intermediate voltage from the voltage conversion unit into the final voltage required for the chip to operate.

[0021] The chip cooling element described in this article refers to a thermal management structure that contacts the chip to dissipate heat. The chip cooling element can take the form of a cold plate, which conducts and dissipates the heat generated by the chip through thermal contact. In some cases, the chip cooling element can be placed between a pair of processing elements and shared by the chips of both processing elements.

[0022] The power supply cooling element described in this article refers to a thermal management structure that contacts the power supply unit to dissipate heat. The power supply cooling element can take the form of an ultra-thin heat dissipation plate to dissipate the heat generated by the power supply unit during operation.

[0023] The thickness direction mentioned in this article refers to the direction in which paired processing components are stacked in an electronic assembly. When the electronic assembly is installed in a cabinet, the thickness direction may correspond to the height direction of the cabinet.

[0024] The voltage conversion unit described in this article may include an Intermediate Bus Converter (IBC). The IBC is a power device that performs the conversion of the bus voltage from high to low. In the power supply architecture described in this article, the power conversion can employ a two-stage conversion process: the IBC performs the first stage conversion, converting the high bus voltage (e.g., 54V) to an intermediate low voltage (e.g., 12V or 6.75V); the power module performs the second stage conversion, converting the intermediate low voltage to the final voltage required for chip operation. In this way, the two-stage conversion architecture can meet the chip's power supply requirements while reducing transmission losses and improving power supply efficiency.

[0025] Example usage environment Figure 1An example usage environment 1000 for electronic component 100 is shown. For example... Figure 1 As shown, example environment 1000 illustrates the configuration relationship between electronic component 100 and cabinet 700.

[0026] Electronic components 100 can be in the form of flat rectangular units. Multiple electronic components 100 can be stacked for storage and transportation. The cabinet 700 can be a rectangular frame structure extending along the height direction, with internal mounting rails and shelves for receiving and securing the electronic components 100.

[0027] In some cases, the electronic components 100 are installed horizontally within the rack 700. The rack 700 has multiple mounting positions along its height, each capable of accommodating one or more electronic components 100. In this way, multiple electronic components 100 can be sequentially arranged along the height of the rack 700 to form a server system within the rack 700.

[0028] As previously mentioned, the paired processing elements in electronic component 100 are stacked in the thickness direction. When electronic component 100 is installed in rack 700, the thickness direction of electronic component 100 can correspond to the height direction of rack 700. Since the height space of rack 700 is limited (for example, the height of a standard rack 1U is approximately 40 mm or more), this paper adopts a sandwich structure vertical power supply scheme, which can improve the integration density of electronic component 100 under the limited rack height constraint.

[0029] In some cases, rack 700 can be installed in a data center or server room environment. After the electronic components 100 are installed inside rack 700, they can be connected to external systems through the power and network interfaces provided by rack 700, thereby realizing computing and data processing functions.

[0030] Example electronic components Figure 2 The overall architecture of electronic component 100 is shown. For example... Figure 2 As shown, the electronic component 100 adopts a sandwich structure configuration, including a pair of processing elements 10, a chip cooling element 20, and a power supply cooling element 30.

[0031] Pairs of processing elements 10 are stacked in the thickness direction X, forming a symmetrical configuration with respect to the chip cooling element 20. The thickness direction X refers to the direction in which the pairs of processing elements 10 are stacked in the electronic assembly 100. In this sandwich structure, the two processing elements 10 are arranged back-to-back, sharing the centrally located chip cooling element 20. In this way, the dual computing cores can share a common heat dissipation structure, thereby achieving higher integration density within a limited thickness space.

[0032] like Figure 2 As shown, each processing element 10 may include a carrier board 11, a chip 12, and a power supply unit 13. The carrier board 11 is a circuit board structure for carrying and electrically connecting the chip 12 and the power supply unit 13. The carrier board 11 has a first side 111 and a second side 112 that are opposite to each other in the thickness direction X. The chip 12 is disposed on the first side 111 of the carrier board 11 and faces the central chip cooling element 20. The power supply unit 13 is electrically connected to the carrier board 11 and provides operating voltage to the chip 12.

[0033] In some cases, the paired processing elements 10 are identical, with each processing element 10 having the same model, size, specifications, and electrical parameters for its carrier board 11, chip 12, and power supply unit 13. The two processing elements 10 are symmetrically arranged in the thickness direction X relative to the centrally located cooling element, with each chip 12 facing and in thermal contact with the cooling element. The power supply units 13 face opposite outward directions. This symmetrical arrangement allows for more uniform heat dissipation of the electronic components 100, avoiding localized overheating. Furthermore, the interchangeability of the identical processing elements 10 simplifies the manufacturing process and reduces material procurement and inventory management costs.

[0034] In some cases, the paired processing elements 10 maintain only the aforementioned sandwich stacking architecture. The model, size, specifications, and electrical parameters of their respective carrier boards 11, chips 12, and power supply units 13 can be uniquely configured according to actual application requirements. The two processing elements 10 are stacked in the thickness direction X, sharing a centrally located cooling element. Each chip 12 faces and thermally contacts the cooling element. This arrangement enhances the design freedom of the electronic components 100, allowing for the selection of chips 12 with different performance levels to meet varying computing power requirements. Power supply units 13 with appropriate power consumption are matched to the chip power consumption, and suitable carrier board 11 layers and wiring schemes are selected, thereby retaining the advantages of high integration density while adapting to diverse application scenarios and differentiated functional requirements.

[0035] exist Figure 2 In the configuration shown, the chips 12 of the paired processing elements 10 are located on opposite sides of the chip cooling element 20 in the thickness direction X. The two chips 12 are arranged back-to-back on opposite sides of the chip cooling element 20, allowing the chip cooling element 20 to simultaneously contact the chips 12 of both processing elements 10. Therefore, the two processing elements 10 can share the same chip cooling element 20 for heat dissipation, reducing the number of heat dissipation structures and lowering the overall size of the electronic component 100 in the thickness direction X.

[0036] like Figure 2As shown, the power supply unit 13 can be disposed on the second side 112 of the carrier board 11, that is, the side opposite to the chip 12. Two power cooling elements 30 are respectively disposed on the side of the paired processing elements 10 away from the chip cooling element 20, and respectively contact the power supply unit 13 of the corresponding processing element 10 to dissipate the heat generated by the power supply unit 13 during operation. In this way, the chip 12 and the power supply unit 13 are thermally managed by independent cooling elements, which can meet the heat dissipation requirements of high-power chips and power modules.

[0037] Continue to refer to Figure 2 In this example, a chip cooling element 20 is disposed between a pair of processing elements 10. A first side 111 of each carrier plate 11 faces the chip cooling element 20, such that the chips 12 of both processing elements 10 are oriented towards the central chip cooling element 20. The chip cooling element 20 contacts the chips 12 of the pair of processing elements 10 on opposite sides in the thickness direction X. In this manner, the chip cooling element 20 can simultaneously receive heat generated by the chips 12 from two directions and dissipate the heat through its internal cooling medium.

[0038] In some cases, the power supply unit 13 of each processing element 10 is located on the second side 112 of the carrier 11. Since the chip 12 is disposed on the first side 111 of the carrier 11, the power supply unit 13 and the chip 12 are respectively located on opposite sides of the carrier 11 in the thickness direction X. In this way, the chip 12 and the power supply unit 13 are arranged in a layered manner in the thickness direction X, so that the chip cooling element 20 and the power cooling element 30 can perform thermal management on the chip 12 and the power supply unit 13 from different directions.

[0039] In this configuration, the power cooling element 30 is in contact with the power supply unit 13 to dissipate the heat generated by the power supply unit 13 during operation. Since the power supply unit 13 is located on the second side 112 of the carrier plate 11, the power cooling element 30 can be disposed on the outside of the paired processing elements 10, that is, on the side away from the chip cooling element 20 in the thickness direction X. Thus, the chip cooling element 20 is located in the center of the electronic assembly 100, while the power cooling element 30 is located on the outside of the electronic assembly 100, forming a sandwich structure with the chip 12 in the center and the power supply unit 13 on the outside.

[0040] The two processing elements 10 are arranged back to back, with the chip 12 facing the central chip cooling element 20 and the power supply unit 13 facing the outer power cooling element 30. In this way, the heat generated by the chip 12 is conducted to the central chip cooling element 20, and the heat generated by the power supply unit 13 is conducted to the outer power cooling element 30, thereby achieving bidirectional heat dissipation.

[0041] In some cases, the chip cooling element 20 can be in the form of a cold plate, with channels for the flow of cooling medium inside. The cooling medium can be a liquid or a gas, which carries away the heat generated by the chip 12 by flowing inside the cold plate. Alternatively or additionally, the chip cooling element 20 can also be in the form of a heat pipe, a vapor chamber, or other heat-conducting structure.

[0042] In some cases, the two side surfaces of the chip cooling element 20 may be provided with contact areas that match the chip 12. The shape and size of the contact areas can be designed according to the package form of the chip 12 to achieve good thermal contact between the chip 12 and the chip cooling element 20.

[0043] In some cases, the power supply unit 13 includes a power carrier board 131 and a power module 132. The power carrier board 131 is a carrier board structure connecting the chip 12 and the power module 132, and can be made of a substrate material or a printed circuit board (PCB) similar in material to the carrier board 11. The power module 132 is electrically connected to the power carrier board 131, and the power carrier board 131 is electrically connected to the corresponding carrier board 11. In this way, the power module 132 can establish an electrical connection with the carrier board 11 through the power carrier board 131, thereby transmitting the converted voltage to the chip 12.

[0044] In some cases, the power module 132 may have a housing made of a non-conductive material. The electrical connections of the power module 132 are located at its bottom and are made electrically to the power carrier plate 131 via solder balls. Because the housing of the power module 132 is made of a non-conductive material, the top surface of the power module 132 can directly contact the power cooling element 30 without causing an electrical short circuit. In this way, the power cooling element 30 can form good thermal contact with the housing surface of the power module 132, thereby effectively dissipating the heat generated by the power module 132 during operation.

[0045] In this configuration, the power module 132 performs a voltage conversion function, converting the intermediate voltage from the voltage conversion unit 60 into the final voltage required for the chip 12 to operate. The power carrier board 131 serves as an electrical bridge structure between the power module 132 and the carrier board 11, carrying the power module 132 and providing a current transmission path. Thus, the power unit 13 can achieve power supply functionality within a compact space, while achieving effective thermal management through the power cooling element 30.

[0046] In some cases, the power cooling element 30 includes a pair of first power cooling elements 31. The pair of first power cooling elements 31 are respectively disposed on the side of the pair of processing elements 10 away from the chip cooling element 20 in the thickness direction X, and respectively contact the power supply unit 13 of the corresponding processing element 10. Since the power supply unit 13 is located on the second side 112 of the carrier 11, the first power cooling element 31 forms thermal contact with the power supply unit 13 from the outside direction of the electronic assembly 100. In this way, the pair of first power cooling elements 31 can independently manage the thermal of the power supply units 13 of the two processing elements 10.

[0047] In this configuration, the chip cooling element 20 is located at the center of the electronic assembly 100, while the pair of first power cooling elements 31 are located at both ends of the electronic assembly 100 in the thickness direction X. Thus, the electronic assembly 100 forms a stacked structure in the thickness direction X consisting of the first power cooling element 31, the processing element 10, the chip cooling element 20, the processing element 10, and the first power cooling element 31. This stacked structure allows heat generated by the chip 12 to be conducted centrally to the chip cooling element 20, while heat generated by the power supply unit 13 is conducted outwards to the first power cooling element 31, thereby achieving bidirectional heat dissipation.

[0048] In some cases, the first power supply cooling element 31 can take the form of a heat dissipation plate. The thickness of the heat dissipation plate can be designed according to the overall height constraints of the electronic component 100, so as to reduce the overall size of the electronic component 100 in the thickness direction X while meeting the heat dissipation requirements. Alternatively or additionally, the first power supply cooling element 31 can also take the form of a heat pipe, a heat spreader, or other heat-conducting structures.

[0049] In some cases, the surface of the first power cooling element 31 facing the power unit 13 may be provided with a contact area that matches the power unit 13. The shape and size of the contact area can be designed according to the shape of the power unit 13 to achieve good thermal contact between the power unit 13 and the first power cooling element 31. Thus, the heat generated by the power unit 13 during operation can be dissipated through the first power cooling element 31, thereby maintaining the power unit 13 operating within a suitable operating temperature range.

[0050] In some cases, each processing element 10 also includes a voltage conversion unit 60. The voltage conversion unit 60 is disposed on a first side 111 of the carrier 11 and in contact with the chip cooling element 20. The voltage conversion unit 60 is configured to convert a first voltage on the carrier 11 into a second voltage and supply the second voltage to the corresponding power supply unit 13 via the carrier 11. In this way, the voltage conversion unit 60 can perform voltage conversion close to the chip 12, shortening the power supply path, thereby reducing transmission losses and improving power supply efficiency.

[0051] like Figure 2 As shown, the voltage conversion unit 60 can be soldered to the carrier board 11 to establish an electrical connection. One end of the voltage conversion unit 60 is soldered to the carrier board 11, thereby receiving a first voltage (e.g., high bus voltage) from the carrier board 11 and transmitting the converted second voltage (e.g., intermediate low voltage) to the power supply unit 13 via the carrier board 11. After receiving the second voltage, the power supply unit 13 can further convert the second voltage into the final voltage required for the operation of the chip 12. Thus, the voltage conversion unit 60 and the power supply unit 13 together constitute a two-stage voltage conversion architecture, which can improve the overall power supply efficiency while meeting the power supply requirements of the chip 12.

[0052] In some cases, the voltage conversion unit 60 includes an intermediate bus converter. As previously described, the intermediate bus converter is a power device that performs the conversion of the bus voltage from high voltage to low voltage. The intermediate bus converter generates heat during operation, so thermal contact with the chip cooling element 20 can effectively dissipate heat from the voltage conversion unit 60.

[0053] In some cases, since the height of the voltage conversion unit 60 in the thickness direction X may be greater than the height of the chip 12, the chip cooling element 20 may include a heat sink 201 to accommodate the voltage conversion unit 60. In some cases, the chip cooling element 20 includes a heat sink 201 disposed on the side facing the voltage conversion unit 60. The heat sink 201 is an opening structure formed on the surface of the chip cooling element 20 for receiving a portion of the voltage conversion unit 60. The end of the voltage conversion unit 60 away from the carrier plate 11 is inserted into the heat sink 201. In this way, the heat sink 201 can accommodate the height of the voltage conversion unit 60, so that the overall height of the electronic component 100 in the thickness direction X can be controlled within the system height constraints.

[0054] In some cases, the heat sink 201 can also increase the contact area between the voltage conversion unit 60 and the chip cooling element 20. When the voltage conversion unit 60 is inserted into the heat sink 201, the side of the voltage conversion unit 60 can form thermal contact with the inner wall of the heat sink 201, thereby increasing the heat dissipation area. Thus, the heat generated by the voltage conversion unit 60 during operation can be conducted to the chip cooling element 20 through the inner wall of the heat sink 201 and dissipated through the cooling medium inside the chip cooling element 20.

[0055] In some cases, a thermally conductive interface material may be provided between the voltage conversion unit 60 and the heat sink 201. The thermally conductive interface material may include silicone grease or other thermally conductive materials, used to fill the gap between the voltage conversion unit 60 and the inner wall of the heat sink 201. In this way, the thermally conductive interface material can reduce the thermal resistance between the voltage conversion unit 60 and the chip cooling element 20, thereby improving the heat dissipation effect of the voltage conversion unit 60.

[0056] In this configuration, the voltage conversion unit 60 establishes an electrical connection with the carrier board 11 via soldering, and simultaneously establishes thermal contact with the chip cooling element 20 via the insertion of the heat sink 201. Thus, the voltage conversion unit 60 can achieve effective thermal management while performing its voltage conversion function, thereby maintaining the voltage conversion unit 60 within a suitable operating temperature range.

[0057] Figure 3 Another configuration of the electronic assembly 100 is shown, wherein the power supply unit 13 is disposed within the first opening 110 of the carrier plate 11. (See diagram below.) Figure 3 As shown, the electronic component 100 adopts a sandwich structure configuration, including a pair of processing elements 10, a chip cooling element 20, and a power supply cooling element 30.

[0058] In some cases, the carrier 11 of each processing element 10 includes a first opening 110. The first opening 110 extends through the carrier 11 in the thickness direction X, forming a cutout area on the carrier 11. A power supply unit 13 is disposed within the first opening 110 and is electrically connected to the chip 12. In this way, the power supply unit 13 can correspond to the chip 12 through the first opening 110, thereby shortening the electrical path length between the power supply unit 13 and the chip 12.

[0059] In this configuration, the carrier board 11 can be made of a printed circuit board (PCB), and the first opening 110 is formed by hollowing out the PCB. The shape and size of the first opening 110 can be designed according to the shape of the power supply unit 13 to accommodate at least a portion of the power supply unit 13. Since the first opening 110 penetrates the carrier board 11 in the thickness direction X, the power supply unit 13 can pass through the thickness of the carrier board 11 and be directly electrically connected to the chip 12.

[0060] Compared to the configuration where the power supply unit 13 is located on the second side 112 of the carrier board 11, the first opening 110 further shortens the power supply path between the power supply unit 13 and the chip 12. In the conventional configuration, when current is transmitted from the power supply unit 13 to the chip 12, it may be transmitted horizontally via the wiring layer of the carrier board 11. However, in the configuration of the first opening 110, the power supply unit 13 is directly positioned below the chip 12, and the current can be transmitted vertically along the thickness direction X, thereby reducing the length of the transmission path.

[0061] Shortening the power supply path length can reduce resistance loss during transmission. Since resistive heat loss occurs when current travels through a conductor, the longer the transmission path, the greater the resistance loss. By placing the power supply unit 13 within the first opening 110 and positioning it below the chip 12, the electrical path length is shortened, thereby reducing resistance loss and improving power supply efficiency. In some cases, the vertical power supply structure using the first opening 110 configuration can improve power supply efficiency by approximately 5% compared to a structure without this configuration.

[0062] Continue to refer to Figure 3 A chip cooling element 20 is disposed between a pair of processing elements 10. The first side 111 of each carrier plate 11 faces the chip cooling element 20, such that the chips 12 of both processing elements 10 face the central chip cooling element 20. The chip cooling element 20 includes a heat sink 201 for accommodating a voltage conversion unit 60. The voltage conversion unit 60 is disposed on the first side 111 of the carrier plate 11 and inserted into the heat sink 201, thereby forming thermal contact with the chip cooling element 20.

[0063] Continue to refer to Figure 3 The power cooling element 30 may include a pair of first power cooling elements 31. The pair of first power cooling elements 31 are in contact with the power supply unit 13 of the corresponding processing element 10, and are used to dissipate the heat generated by the power supply unit 13 during operation. Since the power supply unit 13 is disposed in the first opening 110, the first power cooling element 31 can form thermal contact with the power supply unit 13 from the second side 112 direction of the carrier plate 11.

[0064] In this configuration, the electronic component 100 may further include a bus power board 40, a first connector 51, and a second connector 52. The bus power board 40 provides a distribution path for bus power. The first connector 51 is electrically connected between the carrier board 11 and the corresponding bus power board 40. The second connector 52 is electrically connected between the power supply unit 13 and the corresponding bus power board 40. In this way, the carrier board 11 and the power supply unit 13 can be connected via the first connector 51 and the second connector 52, thereby enabling power supply to the chip 12.

[0065] In some cases, the electronic component 100 also includes a pair of bus power boards 40. The pair of bus power boards 40 are respectively disposed on the side of the pair of first power cooling elements 31 away from the chip cooling element 20 in the thickness direction X, and are electrically connected to the carrier board 11 of the corresponding processing element 10 and the power supply unit 13, respectively. The bus power boards 40 provide a distribution path for the bus power supply, transmitting the intermediate voltage output from the voltage conversion unit 60 to the power supply unit 13.

[0066] In this configuration, since the power supply unit 13 is located within the first opening 110, there may be no direct printed circuit board (PCB) connection path between the voltage conversion unit 60 and the power supply unit 13. The bus power board 40 can establish an electrical connection between the output terminal of the voltage conversion unit 60 and the input terminal of the power supply unit 13, thereby enabling voltage transmission. In this way, after the voltage conversion unit 60 completes voltage conversion, the converted intermediate voltage can be transmitted to the power supply unit 13 via the bus power board 40 for subsequent voltage conversion by the power supply unit 13.

[0067] In some cases, the electronic component 100 also includes a first connector 51 and a second connector 52. The first connector 51 is electrically connected between the carrier board 11 and the corresponding bus power board 40. The second connector 52 is electrically connected between the power supply unit 13 and the corresponding bus power board 40. In this way, the bus power board 40 can establish electrical connections with the carrier board 11 and the power supply unit 13 via the first connector 51 and the second connector 52, respectively.

[0068] In some cases, the first connector 51 and the second connector 52 can be in the form of spring-loaded pins. Spring-loaded pin connectors provide contact pressure through a spring structure, thereby establishing a reliable electrical connection between the two electrical contact surfaces. When there is no direct PCB connection path between the output of the voltage conversion unit 60 and the input of the power supply unit 13, the spring-loaded pin connector can provide an alternative electrical connection. In this way, the intermediate voltage output by the voltage conversion unit 60 can be transmitted to the bus power board 40 via the spring-loaded pin connector, and then to the power supply unit 13 via the bus power board 40.

[0069] The first connector 51 electrically connects the bus power board 40 to the carrier board 11, transmitting the output voltage of the voltage conversion unit 60 to the bus power board 40. The second connector 52 electrically connects the bus power board 40 to the power supply unit 13, transmitting the intermediate voltage converted by the voltage conversion unit 60 to the power supply unit 13. Thus, the bus power board 40, acting as an electrical bridge between the voltage conversion unit 60 and the power supply unit 13, enables voltage transmission even without a direct PCB path.

[0070] In this configuration, the bus power board 40, the first connector 51, and the second connector 52 together form a power distribution network. After the bus power supply is input from the outside, it undergoes a first-stage voltage conversion via the voltage conversion unit 60 on the carrier board 11. The intermediate voltage output from the voltage conversion unit 60 is transmitted to the power supply unit 13 via the bus power board 40 and the second connector 52, where it undergoes a second-stage voltage conversion to provide the operating voltage for the chip 12. In this way, the two-stage voltage conversion architecture can improve overall power supply efficiency while meeting the power supply requirements of the chip 12.

[0071] Figure 4 Another configuration of the electronic assembly 100 is shown, wherein the power supply cooling element 30 is disposed between the pair of processing elements 10. (See diagram below.) Figure 4 As shown, the electronic component 100 adopts a sandwich structure configuration, including a pair of processing elements 10, a chip cooling element 20, and a power supply cooling element 30.

[0072] In this configuration, pairs of processing elements 10 are stacked in the thickness direction X, forming a symmetrical arrangement with respect to the power cooling element 30. Unlike the aforementioned configuration where the chip cooling element 20 is centered, this configuration places the power cooling element 30 between the pairs of processing elements 10, while the pair of chip cooling elements 20 is placed on the side of the pairs of processing elements 10 away from the power cooling element 30. In this way, the electronic assembly 100 forms a sandwich structure with the power unit 13 centered and the chip 12 on the outside.

[0073] like Figure 4 As shown, each processing element 10 includes a carrier board 11, a chip 12, and a power supply unit 13. The carrier board 11 has a first side 111 and a second side 112 that are opposite to each other in the thickness direction X. The chip 12 is disposed on the first side 111 of the carrier board 11. The power supply unit 13 is electrically connected to the carrier board 11 and is used to provide operating voltage to the chip 12.

[0074] In some cases, a power cooling element 30 is disposed between a pair of processing elements 10, with the second side 112 of each carrier plate 11 facing the power cooling element 30. Since the power supply unit 13 is located on the second side 112 of the carrier plate 11, the power supply units 13 of both processing elements 10 face the central power cooling element 30. The power cooling element 30 contacts the power supply units 13 of the pair of processing elements 10 on opposite sides in the thickness direction X. In this way, the power cooling element 30 can simultaneously receive heat generated by the power supply units 13 from two directions and dissipate the heat through internal cooling medium.

[0075] In this configuration, the two processing elements 10 are arranged face-to-face, with the power supply unit 13 facing the central power cooling element 30 and the chip 12 facing the outer chip cooling element 20. In this way, the heat generated by the power supply unit 13 is conducted towards the center to the power cooling element 30, and the heat generated by the chip 12 is conducted outward to the chip cooling element 20, thereby achieving bidirectional heat dissipation.

[0076] This configuration features the following characteristics: the power cooling element 30 is located in the center of the electronic assembly 100, allowing it to simultaneously form thermal contact with the power supply units 13 of both processing elements 10. Since the heat generated by the power supply units 13 during operation is concentrated in the central region of the electronic assembly 100, the power cooling element 30 can achieve thermal management of both power supply units 13 through a single cooling structure. In this way, the number of power cooling elements 30 in the electronic assembly 100 can be reduced, thereby simplifying the heat dissipation structure and reducing the overall size of the electronic assembly 100 in the thickness direction X.

[0077] In some cases, the power supply cooling element 30 may be in the form of a cold plate, with channels for the flow of cooling medium inside. The cooling medium may be a liquid or a gas, which carries away the heat generated by the power supply unit 13 by flowing inside the cold plate. Alternatively or additionally, the power supply cooling element 30 may also be in the form of a heat pipe, a heat spreader, or other heat-conducting structures.

[0078] In some cases, the two side surfaces of the power supply cooling element 30 may be provided with contact areas that match the power supply unit 13. The shape and size of the contact areas can be designed according to the shape of the power supply unit 13 to achieve good thermal contact between the power supply unit 13 and the power supply cooling element 30. As a result, the heat generated by the power supply unit 13 during operation can be dissipated through the power supply cooling element 30, thereby maintaining the power supply unit 13 in operation within a suitable operating temperature range.

[0079] Since the chip 12 is disposed on the first side 111 of the carrier 11, and the first side 111 faces outward of the electronic assembly 100, the chip cooling element 20 can be disposed on the side of the paired processing element 10 away from the power cooling element 30 to form thermal contact with the chip 12. In this configuration, the chip cooling element 20 may include a pair of first chip cooling elements 21, which are respectively disposed on the side of the paired processing element 10 away from the power cooling element 30 in the thickness direction X, and respectively contact the chip 12 of the corresponding processing element 10. In this way, the pair of first chip cooling elements 21 can perform independent thermal management on the chips 12 of the two processing elements 10.

[0080] In this configuration, the electronic component 100 forms a stacked structure in the thickness direction X, consisting of a first chip cooling element 21, a processing element 10, a power cooling element 30, and the first chip cooling element 21. This stacked structure allows heat generated by the power supply unit 13 to be conducted centrally to the power cooling element 30, while heat generated by the chip 12 is conducted outwards to the first chip cooling element 21, thereby achieving bidirectional heat dissipation.

[0081] In some cases, each processing element 10 may further include a voltage conversion unit 60. The voltage conversion unit 60 is disposed on a first side 111 of the carrier plate 11 and is in contact with the chip cooling element 20. The voltage conversion unit 60 is configured to convert a first voltage of the carrier plate 11 into a second voltage and supply the second voltage to a corresponding power supply unit 13 via the carrier plate 11. Because the voltage conversion unit 60 is disposed on the first side 111 of the carrier plate 11, the voltage conversion unit 60 can form thermal contact with the first chip cooling element 21, thereby achieving heat dissipation for the voltage conversion unit 60.

[0082] In some cases, the first chip cooling element 21 may include a heat sink 201 for accommodating the voltage conversion unit 60. The end of the voltage conversion unit 60 furthest from the carrier plate 11 can be inserted into the heat sink 201, thereby increasing the contact area between the voltage conversion unit 60 and the first chip cooling element 21. In this way, the heat generated by the voltage conversion unit 60 during operation can be conducted to the first chip cooling element 21 through the inner wall of the heat sink 201 and dissipated via the cooling medium inside the first chip cooling element 21.

[0083] In some cases, the chip cooling element 20 includes a pair of first chip cooling elements 21. The pair of first chip cooling elements 21 are respectively disposed on the side of the pair of processing elements 10 away from the power supply cooling element 30 in the thickness direction X, and respectively contact the chip 12 of the corresponding processing element 10. Since the chip 12 is disposed on the first side 111 of the carrier 11, and the first side 111 faces outwards from the electronic assembly 100, the first chip cooling element 21 forms thermal contact with the chip 12 from the outward direction of the electronic assembly 100. In this way, the pair of first chip cooling elements 21 can provide independent thermal management for the chips 12 of the two processing elements 10.

[0084] like Figure 4 As shown, chip 12 faces outwards, while power supply units 13 are located in the central region and share a power cooling element 30. In this configuration, the power supply units 13 of both processing elements 10 face the central power cooling element 30, thus sharing the same cooling structure. In this way, the power cooling element 30 can simultaneously manage the thermal of both power supply units 13 through a single cooling structure, reducing the number of cooling elements.

[0085] In some cases, the first chip cooling element 21 can be in the form of a cold plate, with channels for the flow of cooling medium inside. The cooling medium can be a liquid or a gas, which carries away the heat generated by the chip 12 by flowing inside the cold plate. Alternatively or additionally, the first chip cooling element 21 can also be in the form of a heat pipe, a vapor chamber, or other heat-conducting structures.

[0086] In some cases, the surface of the first chip cooling element 21 facing the chip 12 may have a contact area that matches the chip 12. The shape and size of the contact area can be designed according to the package form of the chip 12 to achieve good thermal contact between the chip 12 and the first chip cooling element 21. Thus, the heat generated by the chip 12 during operation can be dissipated through the first chip cooling element 21, thereby maintaining the chip 12 in a suitable operating temperature range.

[0087] This configuration exhibits different heat flow path characteristics compared to the configuration where the chip cooling element 20 is centered. In the configuration where the chip cooling element 20 is centered, the heat from the two chips 12 converges towards the shared chip cooling element 20. However, in the configuration where the first chip cooling element 21 is located outside, the heat from the two chips 12 is conducted outwards to their respective first chip cooling elements 21. In this way, the heat generated by the chips 12 can be dissipated through independent heat dissipation paths, preventing the heat from the two chips 12 from accumulating in the central area.

[0088] Figure 5 Another configuration of the electronic assembly 100 is shown, wherein a pair of first power cooling elements 31 are disposed between a pair of processing elements 10. (See diagram below.) Figure 5 As shown, the electronic component 100 adopts a sandwich structure configuration, including a pair of processing elements 10, a chip cooling element 20, and a power supply cooling element 30.

[0089] In this configuration, pairs of processing elements 10 are stacked in the thickness direction X, forming a symmetrical arrangement about a central region. Each processing element 10 includes a carrier board 11, a chip 12, and a power supply unit 13. The carrier board 11 has a first side 111, and the chip 12 is disposed on the first side 111 of the carrier board 11. The power supply unit 13 is electrically connected to the carrier board 11 and provides an operating voltage to the chip 12.

[0090] In some cases, the power cooling element 30 includes a pair of first power cooling elements 31. The pair of first power cooling elements 31 are disposed between the pair of processing elements 10 and are in contact with the power supply units 13 of the pair of processing elements 10, respectively. Unlike the aforementioned configuration where the first power cooling elements 31 are disposed outside the pair of processing elements 10, this configuration places the pair of first power cooling elements 31 in the central region of the electronic assembly 100. In this way, the power supply units 13 of both processing elements 10 face the central first power cooling element 31, thereby sharing the cooling structure located in the central region.

[0091] In this configuration, the chip 12 faces outward from the electronic assembly 100. The chip cooling element 20 may include a pair of first chip cooling elements 21, which are respectively disposed on the side of the pair of processing elements 10 away from the power cooling element 30 in the thickness direction X, and respectively contact the chip 12 of the corresponding processing element 10. Thus, the electronic assembly 100 forms a stacked structure of first chip cooling element 21, processing element 10, first power cooling element 31, bus power board 40, first power cooling element 31, processing element 10, and first chip cooling element 21 in the thickness direction X.

[0092] In some cases, the electronic assembly 100 also includes a pair of bus power boards 40. The pair of bus power boards 40 are disposed between a pair of first power cooling elements 31 and are electrically connected to the carrier board 11 and power supply unit 13 of the respective processing element 10. The pair of bus power boards 40 are located at the center of the electronic assembly 100, forming a symmetrical central bus layout between the pair of first power cooling elements 31. In this way, the bus power boards 40 can serve as a central hub for power distribution, providing power to the processing elements 10 on both sides.

[0093] like Figure 5 As shown, the paired bus power boards 40 are located in the central area of ​​the electronic assembly 100, and the power modules and chips 12 of the two processing elements 10 are symmetrically arranged on both sides of the bus power boards 40. In this way, the bus power boards 40 can transmit power to the processing elements 10 on both sides with equal path lengths, thereby achieving balanced power distribution.

[0094] In this configuration, the electronic component 100 may further include a first connector 51 and a second connector 52. The first connector 51 is electrically connected between the carrier board 11 and the corresponding bus power board 40. The second connector 52 is electrically connected between the power supply unit 13 and the corresponding bus power board 40. In this way, an electrical connection between the carrier board 11 and the power supply unit 13 can be achieved through the first connector 51 and the second connector 52, thereby enabling power supply to the chip 12.

[0095] In some cases, such as Figure 5 As shown, the carrier board 11 may include a first opening 110. The first opening 110 extends through the carrier board 11 in the thickness direction X. A power supply unit 13 is disposed within the first opening 110 and is electrically connected to the chip 12. Because the power supply unit 13 is disposed within the first opening 110, the power supply unit 13 can be positioned below the chip 12, thereby shortening the electrical path length between the power supply unit 13 and the chip 12.

[0096] In this configuration, the first power cooling element 31 can form thermal contact with the power unit 13 from the direction of the first opening 110 of the carrier plate 11. Since the pair of first power cooling elements 31 are disposed between the pair of processing elements 10, the first power cooling element 31 can contact the surface of the power unit 13 facing the central region. In this way, the heat generated by the power unit 13 during operation can be conducted towards the center to the first power cooling element 31 and dissipated through the cooling medium inside the first power cooling element 31.

[0097] This configuration exhibits different heat flow path characteristics compared to the configuration where the first power cooling element 31 is located outside the paired processing elements 10. In the configuration where the first power cooling element 31 is located outside, the heat generated by the power supply unit 13 is conducted outwards to its respective first power cooling element 31. In the configuration where the first power cooling element 31 is located in the center, the heat from the two power supply units 13 converges towards the paired first power cooling element 31. In this way, the heat generated by the power supply unit 13 can be concentrated and dissipated in the central region of the electronic component 100.

[0098] In some cases, the first power supply cooling element 31 can take the form of an ultra-thin heat dissipation plate. The thickness of the ultra-thin heat dissipation plate can be designed according to the overall height constraints of the electronic component 100, so as to reduce the overall size of the electronic component 100 in the thickness direction X while meeting the heat dissipation requirements. Alternatively or additionally, the first power supply cooling element 31 can also take the form of a heat pipe, a heat spreader, or other heat-conducting structures.

[0099] In some cases, the chip cooling element 20 may include a heat sink 201. The heat sink 201 is used to accommodate components disposed on a first side 111 of the carrier plate 11. In this way, the heat sink 201 can provide a receiving space between the chip cooling element 20 and the carrier plate 11, thereby accommodating component arrangements of different heights.

[0100] In some cases, the electronic component 100 also includes a first thermally conductive element and / or a second thermally conductive element. The first thermally conductive element is disposed between the chip cooling element 20 and the corresponding chip 12. The second thermally conductive element is disposed between the power supply cooling element 30 and the corresponding power supply unit 13. In this way, the first and second thermally conductive elements can establish a low thermal resistance heat conduction path between the heat-generating element and the cooling element, thereby improving heat transfer efficiency.

[0101] In some cases, the first thermally conductive element can be in the form of a thermal interface material (TIM). The TIM is used to fill the microscopic gaps between the chip 12 and the chip cooling element 20. Because the surfaces of the chip 12 and the chip cooling element 20 have uneven areas at the microscale, air gaps will form between the two surfaces when they are in direct contact. Air has a low thermal conductivity, and the presence of these air gaps will increase the thermal resistance between the chip 12 and the chip cooling element 20. The first thermally conductive element can fill these microscopic gaps, eliminate air, and establish a continuous heat conduction path, thereby reducing the interfacial thermal resistance.

[0102] In some cases, the first thermal conductive element may include silicone grease. Silicone grease is a paste-like material made with silicone oil as the base oil and added thermally conductive fillers. Silicone grease has good fluidity and can fill the microscopic unevenness between the chip 12 and the chip cooling element 20 during application. In some cases, the first thermal conductive element may also be in the form of a thermal pad. A thermal pad is a pre-formed sheet of thermally conductive material that can be cut to size according to the dimensions of the chip 12 and placed between the chip 12 and the chip cooling element 20.

[0103] The second heat-conducting element can be made of the same or different thermal interface material as the first heat-conducting element. The second heat-conducting element is disposed between the power supply cooling element 30 and the corresponding power supply unit 13, and is used to fill the microscopic gap between the power supply unit 13 and the power supply cooling element 30. In this way, the second heat-conducting element can reduce the interfacial thermal resistance between the power supply unit 13 and the power supply cooling element 30, thereby improving the heat dissipation effect of the power supply unit 13.

[0104] In some cases, the second thermal conductive element may include thermal grease or a thermal pad. When there is a significant height difference between the surface of the power supply unit 13 and the surface of the power supply cooling element 30, a thicker thermal pad can be used to accommodate this height difference. When the gap between the surface of the power supply unit 13 and the surface of the power supply cooling element 30 is small, thermal grease can be used to achieve a lower interfacial thermal resistance. Additionally or alternatively, the second thermal conductive element may also be in the form of a phase change material, liquid metal, or other thermally conductive material.

[0105] As mentioned above, a thermally conductive interface material may also be provided between the voltage conversion unit 60 and the chip cooling element 20. The thermally conductive interface material may include silicone grease or other thermally conductive materials, used to fill the gap between the voltage conversion unit 60 and the inner wall of the heat sink 201. In this way, the thermally conductive interface material can reduce the thermal resistance between the voltage conversion unit 60 and the chip cooling element 20, thereby improving the heat dissipation effect of the voltage conversion unit 60.

[0106] In this configuration, the first heat-conducting element, the second heat-conducting element, and the thermal interface material between the voltage conversion unit 60 and the heat sink 201 together constitute the thermal interface management system of the electronic component 100. By setting thermal interface materials between each heat-generating element and its corresponding cooling element, the electronic component 100 can establish a low thermal resistance thermal conduction path between the chip 12, the power supply unit 13, and the voltage conversion unit 60 and their respective cooling elements, thereby achieving effective thermal management of each heat-generating element.

[0107] In some cases, the power cooling element 30 includes a second opening on the side facing the power unit 13. The second opening is an opening structure formed on the surface of the power cooling element 30 for receiving a portion of the power unit 13. At least part of the end of the power unit 13 away from the corresponding chip 12 is disposed within the second opening. In this way, the second opening can provide accommodating space for the power unit 13, allowing the height of the power unit 13 in the thickness direction X to extend into the interior of the power cooling element 30.

[0108] In the sandwich structure configuration of electronic component 100, the overall height of electronic component 100 in the thickness direction X is limited by the system height constraint. When the height of power supply unit 13 in the thickness direction X exceeds the available space between carrier plate 11 and power cooling element 30, the second opening can provide additional accommodating space to accommodate the height of power supply unit 13. Thus, electronic component 100 can accommodate power supply unit 13 with a relatively large height while meeting the system height constraint.

[0109] The depth of the second opening can be designed according to the height of the power supply unit 13. When the height of the power supply unit 13 exceeds the available space by a large amount, the depth of the second opening can be increased accordingly to accommodate the protruding part of the power supply unit 13. When the height of the power supply unit 13 exceeds the available space by a small amount, the depth of the second opening can be decreased accordingly. In this way, the depth of the second opening can be adjusted according to the actual height of the power supply unit 13, thereby reducing the amount of material removed from the power cooling element 30 while meeting the accommodation requirements.

[0110] The second opening also increases the contact area between the power supply unit 13 and the power cooling element 30. When one end of the power supply unit 13 is inserted into the second opening, the side of the power supply unit 13 can form thermal contact with the inner wall of the second opening, thereby increasing the heat dissipation area. Thus, the heat generated by the power supply unit 13 during operation can be conducted to the power cooling element 30 through the inner wall of the second opening and dissipated through the cooling medium inside the power cooling element 30.

[0111] In some cases, the shape of the second opening can be designed according to the external shape of the power supply unit 13. The second opening can be rectangular, circular, or other shapes that match the external shape of the power supply unit 13. In this way, the second opening can form a good fit with the outer surface of the power supply unit 13, thereby achieving good thermal contact between the power supply unit 13 and the power cooling element 30.

[0112] In some cases, a thermally conductive interface material may be provided between the second opening and the power supply unit 13. The thermally conductive interface material may include silicone grease or other thermally conductive materials, used to fill the gap between the power supply unit 13 and the inner wall of the second opening. In this way, the thermally conductive interface material can reduce the thermal resistance between the power supply unit 13 and the power cooling element 30, thereby improving the heat dissipation effect of the power supply unit 13.

[0113] This configuration shares a similar design concept with the aforementioned chip cooling element 20, which includes a heat sink 201 to accommodate the voltage conversion unit 60. In the configuration of the chip cooling element 20, the heat sink 201 is used to accommodate the protruding portion of the voltage conversion unit 60, thereby controlling the overall height of the electronic component 100 in the thickness direction X. In the configuration of the power supply cooling element 30, the second opening is used to accommodate the protruding portion of the power supply unit 13, which similarly enables effective thermal management of the power supply unit 13 while meeting system height constraints.

[0114] In some cases, each processing element 10 includes multiple chips 12. The multiple chips 12 may be disposed on a first side 111 of the carrier 11 and respectively form thermal contact with the chip cooling element 20. In this way, a single processing element 10 can carry multiple computing cores, thereby achieving higher computing density within a single sandwich structure unit.

[0115] In some cases, each processing element 10 includes multiple power supply units 13. These power supply units 13 are electrically connected to the carrier board 11 and provide operating voltages to their respective chips 12. When the processing element 10 includes multiple chips 12, the multiple power supply units 13 can provide independent power supply paths to each chip 12, thereby meeting the power supply requirements of the multiple chips 12.

[0116] As previously mentioned, the power supply unit 13 may include a power supply carrier board 131 and a power supply module 132. The power supply carrier board 131 may be made of a substrate material similar to that of the carrier board 11 to achieve a match in thermal expansion coefficient and mechanical properties. In this way, the power supply carrier board 131 and the chip 12 can maintain a stable connection during temperature changes, thereby improving the reliability of the electronic component 100.

[0117] The aforementioned sandwich structure can be extended as a repeatable basic unit. This basic unit includes a pair of processing elements 10, a chip cooling element 20, and a power cooling element 30. By stacking the pair of processing elements 10 in the thickness direction X and sharing the cooling element, dual computing cores can be integrated within a single basic unit. In this way, compared to traditional planar layouts, the sandwich structure can achieve approximately 100% higher integration density under the same rack height constraints.

[0118] This sandwich structure employs a configuration where an intermediate bus converter is embedded in a cold plate. As previously described, the voltage conversion unit 60 may include an intermediate bus converter, which is disposed on the first side 111 of the carrier board 11 and inserted into the heat sink 201 of the chip cooling element 20. In this manner, the intermediate bus converter can achieve effective heat dissipation through the chip cooling element 20 while performing voltage conversion functions. This vertical power supply structure can achieve an efficiency improvement of approximately 7% compared to conventional architectures.

[0119] In the example employing the first opening 110 configuration, the power supply unit 13 is disposed within the first opening 110 and positioned below the chip 12, allowing current to be transmitted vertically along the thickness direction X. As previously mentioned, this configuration further improves power supply efficiency by approximately 5% compared to the structure without the first opening 110 configuration. Thus, combining the configuration of the intermediate bus converter embedded in the cold plate with the first opening 110 configuration, this sandwich structure can achieve an overall efficiency improvement of approximately 7% to 12%.

[0120] The examples described above are exemplary and not exhaustive, nor are they limited to the disclosed examples. Many modifications and variations will be apparent to those skilled in the art without departing from the scope and spirit of the illustrated examples. The terminology used herein is chosen to best explain the principles, practical applications, or technological improvements to the examples in the market, or to enable others skilled in the art to understand the examples disclosed herein.

Claims

1. An electronic component (100), comprising: A pair of processing elements (10) are stacked in the thickness direction (X). Each processing element (10) includes a carrier board (11), a chip (12), and a power supply unit (13). The carrier board (11) has a first side (111) and a second side (112) that are opposite to each other in the thickness direction (X). The chip (12) is disposed on the first side (111) of the carrier board (11). The power supply unit (13) is electrically connected to the carrier board (11). A chip cooling element (20) is in contact with the chip (12); and A power cooling element (30) is in contact with the power unit (13), wherein the chip cooling element (20) or the power cooling element (30) is disposed between the pair of processing elements (10).

2. The electronic assembly (100) according to claim 1, wherein the power supply unit (13) of each processing element (10) is located on the second side (112) of the carrier plate (11).

3. The electronic assembly (100) according to claim 1, wherein the carrier plate (11) of each processing element (10) includes a first opening (110) extending through the carrier plate (11) in the thickness direction (X), and the power supply unit (13) is disposed within the first opening (110) and electrically connected to the chip (12).

4. The electronic assembly (100) according to claim 2 or 3, wherein the chip cooling element (20) is disposed between the pair of processing elements (10), and the first side (111) of each carrier plate (11) is disposed facing the chip cooling element (20), wherein the chip cooling element (20) contacts the chip (12) of the pair of processing elements (10) on opposite sides in the thickness direction (X).

5. The electronic assembly (100) according to claim 4, wherein the power supply cooling element (30) comprises: A pair of first power supply cooling elements (31) are respectively disposed on the side of the pair of processing elements (10) away from the chip cooling element (20) in the thickness direction (X), and respectively contact the power supply unit (13) of the corresponding processing element (10).

6. The electronic component (100) according to claim 5 further comprises: The paired bus power boards (40) are respectively disposed on the side of the paired first power cooling element (31) away from the chip cooling element (20) in the thickness direction (X), and are electrically connected to the carrier board (11) of the corresponding processing element (10) and the power unit (13).

7. The electronic assembly (100) according to claim 2 or 3, wherein the power cooling element (30) is disposed between the paired processing elements (10), and the second side (112) of each carrier plate (11) is disposed facing the power cooling element (30), wherein the power cooling element (30) contacts the power unit (13) of the paired processing elements (10) on opposite sides in the thickness direction (X).

8. The electronic component (100) according to claim 7, wherein the chip cooling element (20) comprises: A pair of first chip cooling elements (21) are respectively disposed on the side of the pair of processing elements (10) away from the power supply cooling element (30) in the thickness direction (X), and respectively contact the chip (12) of the corresponding processing element (10).

9. The electronic assembly (100) according to claim 7, wherein the power supply cooling element (30) comprises: A pair of first power cooling elements (31) are disposed between the pair of processing elements (10) and are in contact with the power supply unit (13) of the pair of processing elements (10).

10. The electronic component (100) according to claim 9, further comprising: A pair of bus power supply boards (40) are disposed between the pair of first power cooling elements (31) and are electrically connected to the carrier plate (11) of the corresponding processing element (10) and the power supply unit (13), respectively.

11. The electronic component (100) according to claim 6 or 10, further comprising: First (51), an electrical connection is made between the carrier plate (11) and the corresponding bus power supply plate (40); as well as The second connector (52) is electrically connected between the power supply unit (13) and the corresponding bus power board (40).

12. The electronic assembly (100) according to claim 1, wherein each processing element (10) further comprises: A voltage conversion unit (60) is disposed on the first side (111) of the carrier plate (11) and in contact with the chip cooling element (20). The voltage conversion unit (60) is configured to convert a first voltage of the carrier plate (11) into a second voltage and provide the second voltage to the corresponding power supply unit (13) via the carrier plate (11).

13. The electronic assembly (100) according to claim 12, wherein the voltage conversion unit (60) includes an intermediate bus converter.

14. The electronic component (100) according to claim 1, further comprising: A first heat-conducting element is disposed between the chip cooling element (20) and the corresponding chip (12); and / or The second heat-conducting element is disposed between the power cooling element (30) and the corresponding power unit (13).

15. The electronic assembly (100) according to claim 1, wherein each processing element (10) comprises a plurality of the chips (12).

16. The electronic assembly (100) according to claim 1, wherein each processing element (10) includes a plurality of the power supply units (13).

17. The electronic component (100) according to claim 1, wherein the power supply unit (13) includes a power supply carrier board (131) and a power supply module (132), the power supply module (132) being electrically connected to the power supply carrier board (131), and the power supply carrier board (131) being electrically connected to a corresponding carrier board (11).

18. The electronic assembly (100) according to claim 1, wherein the power cooling element (30) includes a second opening disposed on one side facing the power unit (13), wherein at least part of the end of the power unit (13) away from the corresponding chip (12) is disposed within the second opening.

19. A rack-mount server, comprising: Server rack (700); as well as The electronic component (100) according to any one of claims 1 to 18 is disposed within the cabinet (700).

Images