Cable module, server, server configuration method, and product

Abstract

The present application relates to the technical field of servers, and discloses a cable module, a server, a server configuration method, and a product. The cable module comprises: a board end and a first cable; two surfaces of the board end are electrically connected, wherein a first connector is fixed to one surface of the board end, and a soldering surface is provided on the other surface of the board end; one end of the first cable is soldered on the soldering surface, and a second connector is fixed to the other end of the first cable; the first connector and the second connector are respectively connected to a first data processing unit and a second data processing unit; and the first data processing unit and the second data processing unit are configured to execute an artificial intelligence task. The cable module provided by the present application is used as a relay device in a server, thereby reducing costs, improving the stability of a communication link, enabling the server to adapt to more usage scenarios, and enhancing user experience.

Description

A cable module, a server, a server configuration method, and a product.

[0001] Cross-references to related applications

[0002] This application claims priority to Chinese Patent Application No. 202411533595.7, filed on October 30, 2024, entitled "A cable module, server, server configuration method and product", the entire contents of which are incorporated herein by reference. Technical Field

[0003] This application relates to the field of server technology, and in particular to a cable module, a server, a server configuration method, and a product. Background Technology

[0004] With the development of artificial intelligence, AI (Artificial Intelligence) servers integrating powerful data processing units are in high demand. In AI scenarios, servers are typically deployed in separate chassis, consisting of a first chassis and a second chassis, each housing a separate data processing unit. For example, the first chassis might be a general-purpose server with a CPU (Central Processing Unit), while the second chassis only houses a GPU (Graphics Processing Unit), without a CPU. To achieve high-speed interconnection between different chassis and extend signal transmission distance, repeater devices are typically used in the link. Since repeater devices have insertion loss, adding repeater devices to the link increases the total insertion loss of the link. To ensure signal integrity, the total insertion loss of the link (including cable insertion loss, pad insertion loss, and device insertion loss) must typically be kept within a limited range.

[0005] In related technologies, retimer cards or passive cards are used as repeater devices. Retimers are active devices, requiring lengthy parameter adjustments when used in a link, a complex and costly process. Furthermore, retimers suffer from excessive latency, impacting link transmission performance. While passive cards do not involve parameter adjustments, their internal circuitry is complex, resulting in significant insertion loss. When using passive cards as repeater devices, to ensure the total insertion loss remains within a limited range, only short communication cables can be used, affecting link transmission performance, restricting server usage scenarios, and impacting user experience.

[0006] Given that the relay devices used in related technologies suffer from problems such as difficulty in parameter tuning, wasted costs, and high insertion loss. Summary of the Invention

[0007] In view of this, this application aims to propose a cable module, a server, a server configuration method and product to improve the transmission performance of the server communication link and enable the server to adapt to more usage scenarios.

[0008] To achieve the above objectives, the technical solution of this application is as follows:

[0009] The first aspect of this application provides a cable module, including: a board end and a first cable;

[0010] The two sides of the board end are electrically connected. One side of the board end is fixed with a first connector, and the other side of the board end is provided with a soldering surface. One end of the first cable is soldered to the soldering surface.

[0011] The other end of the first cable is fixed with a second connector; the first connector and the second connector are respectively connected to the first data processing unit and the second data processing unit; the first data processing unit and the second data processing unit are configured to perform artificial intelligence tasks.

[0012] Optionally, the welding surface is located within the projection area of ​​the first connector on the board end;

[0013] At least one via is provided on the end of the plate, and the via is located in the projection area of ​​the welding surface on the end of the plate.

[0014] Electrical connections are formed on both sides of the board end through vias.

[0015] Optionally, the first connector is configured to plug into a communication device with the first connector.

[0016] Optionally, the second connector is configured to plug into a communication device that has a second connector.

[0017] According to a second aspect of the embodiments of this application, a server is provided, including: a first chassis and a second chassis, wherein the first chassis communicates with the second chassis via a second cable;

[0018] Inside the first chassis, a central processing unit and a cable module provided in the first aspect of the embodiments of this application are deployed; wherein, the central processing unit is deployed on the motherboard and is connected to the second connector of the cable module through the second interface on the motherboard;

[0019] Inside the second chassis, a graphics processor and an adapter board are deployed; the graphics processor is connected to the adapter board via a high-speed connector; the adapter board is connected to the cable module via a second cable.

[0020] Optionally, the adapter card is provided with a first interface;

[0021] The two ends of the second cable are fixed with first connectors, which are respectively connected to the first interface on the adapter board and the first connector of the cable module.

[0022] Optionally, the length of the second cable is not greater than the first threshold.

[0023] The first threshold is determined based on the insertion loss of the link between the central processing unit and the graphics processing unit.

[0024] Optionally, the length of the second cable is not less than the second threshold; the second threshold is determined based on the first position where the first chassis is located and the second position where the second chassis is located.

[0025] Optionally, the first connector is connected to the second cable, which in turn connects to the graphics processor; the second connector is connected to the central processing unit.

[0026] Optionally, the second cable is a fifth-generation high-speed serial computer expansion bus; the adapter card is a high-speed serial computer expansion bus standard card.

[0027] Optionally, the server includes: a first chassis and a second chassis;

[0028] The adapter board includes at least one uplink interface; the adapter board is connected to a corresponding number of second cables via at least one uplink interface;

[0029] The first chassis includes at least one cable module, each cable module being connected to a second cable;

[0030] The central processing unit is connected to a cable module via multiple secondary interfaces on the motherboard.

[0031] Optionally, the server includes: a first chassis and at least two second chassis;

[0032] Each adapter board includes at least one uplink interface; each adapter board is connected to a corresponding number of second cables via at least one uplink interface;

[0033] The first chassis includes at least two cable modules, each of which is connected to a second cable.

[0034] The central processing unit is connected to a cable module via multiple secondary interfaces on the motherboard.

[0035] Optionally, in the server, the number of second cables is not greater than the maximum value of the cable modules that the central processing unit can connect to.

[0036] Optionally, the server includes: a first chassis and two second chassis;

[0037] The first chassis is equipped with 8 cable modules, each of which is connected to a second cable.

[0038] All second cables are connected to adapter boards in the two second chassis in an evenly distributed manner, with each adapter board connected to four second cables via four uplink interfaces.

[0039] Optionally, the server includes: one first chassis and three second chassis;

[0040] The first chassis is equipped with 8 cable modules, each of which is connected to a second cable.

[0041] All the second cables are connected to the adapter boards in the three second chassis respectively. One adapter board in the second chassis is connected to the four second cables through four uplink interfaces, and the adapter boards in the two second chassis are connected to the two second cables through two uplink interfaces respectively.

[0042] According to a third aspect of the embodiments of this application, a server configuration method is provided, the method comprising:

[0043] A target server is configured based on the first chassis and the second chassis in the server provided in the second aspect of the embodiments of this application; the target server includes: a first chassis and at least one second chassis;

[0044] Detect and acquire the insertion loss of the cable module in the first chassis;

[0045] Based on the insertion loss of the cable module, a first threshold is calculated as the upper limit of the length of the second cable; the second cable is configured to connect the first chassis and the second chassis.

[0046] Based on the distance between the first chassis and the second chassis, a second threshold is calculated as the lower limit of the length of the second cable;

[0047] The length of the second cable is determined based on the first threshold and the second threshold.

[0048] Optionally, the first chassis and at least one second chassis are arranged in a vertically stacked configuration;

[0049] When the number of second chassis is not less than 2, the first chassis is located between the two second chassis.

[0050] According to a fourth aspect of the present application, a computer program product is provided, including a computer program that, when executed by a processor, implements the steps in the server configuration method provided in the third aspect of the present application.

[0051] According to a fifth aspect of the present application, a computer-readable storage medium is provided, on which a computer program is stored, which, when executed by a processor, implements the steps in the server configuration method provided in the third aspect of the present application.

[0052] According to a sixth aspect of the present application, an electronic device is provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the steps in the server configuration method provided in the third aspect of the present application.

[0053] Compared to relay devices such as re-timers and passive cards, the cable module provided in this application uses welding to fix the first connector, the first cable, and the second connector together. In the server, the cable module acts as a relay device, connecting different data processing units through the first connector and the second connector respectively. This reduces the number of detachable interfaces in the link, lowers the probability of link failure, and improves the stability of the link.

[0054] Meanwhile, compared to devices such as retimers and passive cards, cable modules save on complex internal wiring. Therefore, using cable modules as relay devices in server communication links greatly reduces insertion loss and improves link transmission performance. While ensuring signal integrity, servers can use longer communication cables, improving the server's adaptability to different usage scenarios and enhancing user experience.

[0055] In addition, replacing components such as timers and passive cards with inexpensive cable modules reduces the cost of the server. Attached Figure Description

[0056] To more clearly illustrate the technical solutions of the embodiments of this application, the drawings used in the description of the embodiments of this application will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0057] Figure 1 is a schematic diagram of the structure of a cable module according to an embodiment of this application;

[0058] Figure 2 is one of the schematic diagrams of the internal architecture of a server in related technologies;

[0059] Figure 3 is the second schematic diagram of the internal architecture of the server in the related technology;

[0060] Figure 4 is a schematic diagram of the server architecture proposed in an embodiment of this application;

[0061] Figure 5 is a flowchart of a server configuration method proposed in an embodiment of this application;

[0062] Figure 6(a) is one of the layout diagrams of the first and second server chassis;

[0063] Figure 6(b) is the second layout diagram of the first and second server chassis;

[0064] Figure 6(c) is the third of the layout diagrams of the first and second server chassis;

[0065] Figure 6(d) is the fourth of the layout diagrams of the first and second server chassis;

[0066] Figure 7 is a schematic diagram of an electronic device according to an embodiment of this application. Detailed Implementation

[0067] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0068] It should be understood that the phrase "one embodiment" or "an embodiment" throughout the specification means that a specific feature, structure, or characteristic related to the embodiment is included in at least one embodiment of this application. Therefore, "in one embodiment" or "in an embodiment" appearing throughout the specification does not necessarily refer to the same embodiment. Furthermore, these specific features, structures, or characteristics can be combined in any suitable manner in one or more embodiments.

[0069] In the various embodiments of this application, it should be understood that the sequence number of each process described below does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application.

[0070] Exemplary embodiments will now be described in detail, examples of which are illustrated in the accompanying drawings. When the following description relates to the drawings, unless otherwise indicated, the same numbers in different drawings denote the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with this application. Rather, they are merely examples of apparatuses and methods consistent with some aspects as detailed in this application.

[0071] It should be noted that, unless otherwise specified, the embodiments and features described in this application can be combined with each other.

[0072] Figure 2 is one of the schematic diagrams of the internal architecture of a server in related technologies. As shown in Figure 2, the server includes a first chassis and a second chassis. The first chassis houses a motherboard containing a central processing unit (CPU) and a retimer card. The second chassis houses a graphics processing unit (GPU) and a PCIe (Peripheral Component Interconnect Express, a high-speed serial computer expansion bus standard) card. The interface on the CPU motherboard connects to the retimer card's connector via an internal high-speed cable. The other end of the retimer connects to the PCIe card (which carries a switching chip) via an external high-speed cable. A high-density connector board establishes a high-speed link between the CPU and the GPU.

[0073] In server architecture, achieving gradual server stability requires adjusting the equalization parameters of the retimer, increasing server configuration complexity and operating costs. Furthermore, relay devices have a certain failure rate, increasing security risks to the communication link. Additionally, the retimer introduces latency, making it unsuitable for time-sensitive AI computations.

[0074] Figure 3 is a second schematic diagram of the internal architecture of a server in related technologies. As shown in Figure 3, the server includes a first chassis and a second chassis. The first chassis houses a motherboard with a central processing unit (CPU) and passive cards. The second chassis houses a graphics processor (GPU) and a PCIe board with a switching chip. The PCIe board and the GPU are connected via a high-speed connector. The first chassis also contains internal high-speed cables, and an external high-speed cable connects the first and second chassis. One end of the internal high-speed cable connects to the first internal port of the passive card, and the other end connects to the second internal port on the motherboard. One end of the external high-speed cable connects to the connector of the passive card via a connector, and the other end connects to the interface on the PCIe board via a connector. This solution uses passive cards as relay devices in the link. Compared to retimers, this solution eliminates the need for equalization parameter adjustments for the passive cards, reducing operating costs. However, due to the higher insertion loss of passive cards, only short external high-speed cables can be used in the link, limiting the server's application scenarios and impacting user experience.

[0075] The present application will now be described in detail with reference to the accompanying drawings and embodiments.

[0076] Figure 1 is a structural schematic diagram of a cable module according to an embodiment of this application. As shown in Figure 1, the cable module includes: a board end and a first cable;

[0077] The two sides of the board end are electrically connected. One side of the board end is fixed with a first connector, and the other side of the board end is provided with a soldering surface. One end of the first cable is soldered to the soldering surface.

[0078] The other end of the first cable is fixed with a second connector; the first connector and the second connector are respectively connected to the first data processing unit and the second data processing unit; the first data processing unit and the second data processing unit are configured to perform artificial intelligence tasks.

[0079] As shown in Figure 1, in this embodiment, the cable module includes a board end and a first cable. A first connector is located on the front of the board end, which can be configured to connect to a communication device equipped with the first connector. A soldering surface with solder pads is located on the back of the board end, and one end of the first cable is soldered to the soldering surface via these pads. The front and back of the board end form an electrical connection, thereby establishing an electrical connection between the first cable and the first connector. The other end of the first cable leads to a second connector, which can be configured to connect to a communication device equipped with the second connector. The first connector and the second connector can be of the same type or different types.

[0080] In practical applications, the first and second connectors can be configured according to the specific needs of the application. For example, in a PCIe (Peripheral Component Interconnect Express, a high-speed serial computer expansion bus standard) high-speed communication link, the cable module can be deployed as a relay device in the link. An eASM (Extended Advanced Switching Matrix) interface connector can be used as the first connector of the cable module, and an MCIO (Modular Communications Input / Output) interface connector can be used as the second connector.

[0081] Because the first and second connectors in this cable module are fixedly connected to both ends of the cable, compared to the method in related technologies where retimers and passive cards are plugged into the internal cables of the link, this saves detachable connectors and a large amount of wiring in the link, reducing the total insertion loss of the link and thus improving the transmission performance of the link. When using cable modules to replace retimers or passive cards, servers can use longer communication cables, thus adapting to more usage scenarios and improving the user experience. Furthermore, reducing detachable connectors in the link also reduces the probability of link failure, improving link stability. Moreover, using inexpensive cable modules to replace retimers and passive cards also reduces the cost of the server.

[0082] As one embodiment of this application, the welding surface is disposed within the projection area of ​​the first connector on the board end;

[0083] At least one via is provided on the end of the plate, and the via is located in the projection area of ​​the welding surface on the end of the plate.

[0084] Electrical connections are formed on both sides of the board end through vias.

[0085] In one embodiment, a first connector located on the front side of the cable module board is positioned opposite a soldering surface located on the back side of the board, and electrical connection is achieved through a via on the board. Optionally, the first connector is fixed to the front side of the board by crimping and soldering, with the soldering surface located on the back side of the board and within the projection area of ​​the first connector on the board. At least one via is provided within the common projection area of ​​the first connector and the soldering surface on the board, and electrical connection is achieved between the first connector on the front side of the board and the first cable soldered to the soldering surface on the back side of the board through the via.

[0086] In this embodiment, by aligning the soldering surface with the shadow area of ​​the first connector on the board end, and setting a via in the shadow area of ​​the soldering surface on the board end, the shortest path from the first connector to the first cable is achieved on the board end, thereby reducing the insertion loss of the cable module.

[0087] Based on the same inventive concept, one embodiment of this application provides a server, including: a first chassis and a second chassis, wherein the first chassis communicates with the second chassis via a second cable;

[0088] Inside the first chassis, a central processing unit and a cable module as described in the above embodiment are deployed; wherein, the central processing unit is deployed on the motherboard and is connected to the second connector of the cable module through a second interface on the motherboard;

[0089] Inside the second chassis, a graphics processor and an adapter board are deployed; the graphics processor is connected to the adapter board via a high-speed connector; the adapter board is connected to the cable module via a second cable.

[0090] Figure 4 is a schematic diagram of the server architecture according to an embodiment of this application. As shown in Figure 4, in this embodiment, the server includes a first chassis and a second chassis, which communicate with each other via a second cable located outside the first and second chassis. The first chassis houses a central processing unit (CPU), and at least one cable module is provided inside the first chassis. The motherboard on which the CPU resides has a second interface adapted to the second connector, and the CPU connects to the second connector of the cable module through the second interface on the motherboard. Optionally, when there are multiple cable modules, the motherboard on which the CPU resides has multiple second interfaces, the same number as the number of cable modules, and each interface connects to the second connector of one cable module.

[0091] The first connector of the cable module connects to a second cable outside the chassis, establishing a communication link between the first and second chassis. The second chassis houses a graphics processing unit (GPU), an adapter board, and a high-speed connector. The GPU and the adapter board communicate via the high-speed connector. The adapter board carries a switching chip and has a first interface adapted to the first connector. The first connector at one end of the second cable connects to the second chassis through this first interface, enabling communication between the central processing unit in the first chassis and the GPU in the second chassis.

[0092] Optionally, the first connector and the second connector can be connectors with different interface types, such as eASM connectors, CDFP (C-form-factor Dual-Port) connectors, MCIO connectors, etc. In this embodiment, an eASM connector is used as the first connector, and an MCIO connector is used as the second connector.

[0093] Furthermore, different data processing units can be deployed in the first and second chassis according to actual application needs. For example, central processing units, graphics processing units, field-programmable gate arrays, and application-specific integrated circuits can be configured as data processing units for AI computing. A variety of high-speed connectors are available for connecting the adapter card and the GPU, including PCIe connectors, NVLink (NVIDIA Link) connectors, and CXL (Compute Express Link) connectors.

[0094] As one embodiment of this application, the adapter board is provided with a first interface;

[0095] The two ends of the second cable are fixed with first connectors, which are respectively connected to the first interface on the adapter board and the first connector of the cable module.

[0096] In one embodiment, both ends of the second cable are provided with the same type of first connector, and the adapter board inside the second chassis is provided with a first interface that is compatible with the first connector. The second cable is plugged into the first interface of the adapter board through the first connector, thereby establishing a communication link between the central processing unit in the first chassis and the graphics processing unit in the second chassis.

[0097] As one embodiment of this application, the length of the second cable is not greater than the first threshold;

[0098] The first threshold is determined based on the insertion loss of the link between the central processing unit and the graphics processing unit.

[0099] In one embodiment, the length of the second cable between the first and second chassis of the server is determined based on the insertion loss of the link between the central processing unit (CPU) in the first chassis and the graphics processing unit (GPU) in the second chassis. To ensure signal integrity of the link transmission between the first and second chassis, the insertion loss of the entire link cannot exceed the specified value. Therefore, in this embodiment, a first threshold is calculated based on the overall insertion loss of the transmission link between the two processors, thus determining the upper limit of the achievable length of the second cable. The following explanation uses a PCIe high-speed link as an example.

[0100] The insertion loss of a PCIe Gen5 high-speed link is required to be no more than the specification value -36dB. When using a passive card, the calculated first threshold is 300mm, meaning the second cable length can be up to 300mm. In this case, the layout of the first and second chassis is very limited and cannot adapt to many application scenarios. Figure 6(a) is one of the schematic diagrams of the layout of the server's first and second chassis. As shown in Figure 6(a), in the scenario using a passive card, the first chassis must be placed adjacent to the second chassis, and the placement of the first and second chassis must ensure that the actual wiring length between the first connectors of the two chassis does not exceed 300mm.

[0101] In this embodiment, the actual insertion loss of the passive card was calculated through actual measurements. This included the internal trace insertion loss (2 inches) of -1.87 dB / 16 GHz, and the insertion loss of vias and pads of -2.1 dB / 16 GHz. After replacing the passive card with a cable module, the device insertion loss decreased to -0.3 dB / 16 GHz. That is, by using a cable module to replace the passive card in this embodiment, an insertion loss of 3.67 dB / 16 GHz within the specified range can be used to increase the length of the second cable. The calculated insertion loss of a 300 mm long cable was -3.65 dB / 16 GHz. A rough calculation showed that the length of the second cable could be increased to approximately 601 mm. Actual testing confirmed that the length of the second cable reached 605 mm (insertion loss value -7.32 dB / 16 GHz).

[0102] In this embodiment, the maximum length of the second cable is determined based on the insertion loss of the link between the central processing unit of the first chassis and the graphics processor of the second chassis. This ensures the integrity of the signal during transmission while increasing the length of the second cable, making the layout of the first and second chassis more flexible and adaptable to different application scenarios.

[0103] In one embodiment, the server includes a first chassis and a second chassis. The first chassis houses a motherboard containing a central processing unit (CPU) and a cable module. The second chassis houses a graphics processing unit (GPU) and a PCIe card carrying a switching chip. The PCIe card and the GPU are connected via a high-speed connector. The PCIe card is configured to enable communication between the CPU and the GPU. One end of the cable module in the first chassis connects to the MCIO interface on the motherboard via an MCIO connector, and the other end connects to the high-speed cable located between the first and second chassis via an eASM connector. The other end of the high-speed cable, also using an eASM connector, connects to the eASM interface on the PCIe card in the second chassis, thereby interconnecting the CPU and GPU and establishing a communication link between the first and second chassis.

[0104] In one embodiment, the server includes a first chassis and multiple second chassis. The first chassis houses a motherboard containing a central processing unit (CPU) and cable modules. Each second chassis contains a graphics processing unit (GPU) and a PCIe card carrying a switching chip. The PCIe card and the GPU are connected via a high-speed connector. The PCIe card is configured to enable communication between the CPU and the GPU. Multiple cable modules are deployed within the first chassis. One end of each cable module connects to an MCIO interface on the motherboard via an MCIO connector, and the other end connects to a high-speed cable located between the first and second chassis via an eASM connector. The other end of each high-speed cable also uses an eASM connector to connect to an eASM interface on a PCIe card within a second chassis, thereby interconnecting the CPU and GPU and establishing multiple communication links between the first and second chassis.

[0105] As one embodiment of this application, the length of the second cable is not less than a second threshold; the second threshold is determined based on the first position where the first chassis is located and the second position where the second chassis is located.

[0106] In one embodiment, the length of the second cable between the first and second chassis of the server is determined based on the link insertion loss between the central processing unit in the first chassis and the graphics processing unit in the second chassis, as well as the first location of the first chassis and the second location of the second chassis. In this embodiment, determining the length of the second cable requires ensuring both the integrity of the transmitted signal in the link and the layout of the server in the application scenario.

[0107] Optionally, to ensure signal integrity in the link transmission between the first and second chassis, the insertion loss of the entire link must not exceed the specified value. Therefore, a first threshold is calculated based on the overall insertion loss of the transmission link between the two processors, thus determining the maximum possible length of the second cable.

[0108] To adapt to the device layout requirements of the current application scenario, it is also necessary to determine the minimum length of the second cable between the first and second chassis, i.e., the second threshold, based on the location of the first chassis (i.e., the first position) and the location of the second chassis (i.e., the second position). In practical applications, determining the second threshold based on the distance between the first and second chassis using the second cable length also requires considering the interface locations on the server, the routing of other external cables, and reserving redundant lengths.

[0109] Finally, the length of the second cable is determined based on the first threshold and the second threshold. Optionally, the first threshold is used as the upper limit of the length of the second cable, and the second threshold is used as the lower limit of the length of the second cable, thereby determining the actual length of the second cable.

[0110] In this embodiment, the upper limit of the length of the second cable is determined based on the overall insertion loss on the communication link between the central processing unit in the first chassis and the graphics processor in the second chassis, and the lower limit of the length of the second cable is determined based on the actual layout requirements of the first chassis and the second chassis in the application scenario. This allows for flexible adjustment of the length of the second cable while ensuring the integrity of the transmitted signal in the link and the layout requirements of different application scenarios.

[0111] In one embodiment of this application, the second cable is a fifth-generation high-speed serial computer expansion bus; the adapter board is a high-speed serial computer expansion bus standard board.

[0112] In one embodiment, a high-speed PCIe Gen5 standard cable is selected as the second cable connecting the first chassis and the second chassis. Correspondingly, a PCIe adapter card of the corresponding type is selected as the adapter card in the second chassis. Accordingly, the high-speed connector between the adapter card and the CPU is a PCIe connector.

[0113] As one embodiment of this application, the server includes: a first chassis and a second chassis;

[0114] The adapter board includes at least one uplink interface; the adapter board is connected to a corresponding number of second cables via at least one uplink interface;

[0115] The first chassis includes at least one cable module, each cable module being connected to a second cable;

[0116] The central processing unit is connected to a cable module via multiple secondary interfaces on the motherboard.

[0117] In one embodiment, the server includes a first chassis and a second chassis. Multiple communication links exist between the first chassis and the second chassis. Optionally, a central processing unit (CPU) and multiple cable modules are deployed in the first chassis. The motherboard housing the CPU has multiple second interfaces, the same number as the cable modules, and each interface connects to a second connector of one cable module.

[0118] Each cable module's first connector connects to a second cable outside the chassis, establishing a communication link between the first and second chassis. The second chassis houses a graphics processor, an adapter board, and a high-speed connector. The graphics processor and the adapter board communicate via the high-speed connector. Within the second chassis, the adapter board includes multiple uplink interfaces (first interfaces adapted to the first connectors), each of which connects to a second cable. This first interface connects to the second chassis, enabling communication between the central processing unit (CPU) in the first chassis and the graphics processor (GPU) in the second chassis.

[0119] In this embodiment, multiple communication links are established between the central processing unit and the graphics processing unit in the server to improve the server's data transmission performance. By deploying cable modules on each communication link to replace relay devices such as timers or passive cards, the link insertion loss is reduced, thereby improving transmission performance.

[0120] As one embodiment of this application, the server includes: a first chassis and at least two second chassis;

[0121] Each adapter board includes at least one uplink interface; each adapter board is connected to a corresponding number of second cables via at least one uplink interface;

[0122] The first chassis includes at least two cable modules, each of which is connected to a second cable.

[0123] The central processing unit is connected to a cable module via multiple secondary interfaces on the motherboard.

[0124] In the above embodiments, the server includes a first chassis and multiple second chassis. One or more communication links exist between the first chassis and each of the second chassis. Optionally, a central processing unit (CPU) and multiple cable modules are deployed in the first chassis. The motherboard housing the CPU has multiple second interfaces, the same number as the cable modules, and each interface connects to a second connector of one cable module.

[0125] Each cable module's first connector connects to a second cable outside the chassis, establishing a communication link between the first chassis and each of the second chassis. The second chassis houses a graphics processor, an adapter board, and high-speed connectors. The graphics processor and the adapter board communicate via the high-speed connectors. In each second chassis, the adapter board includes multiple uplink interfaces (first interfaces adapted to the first connector). In each second chassis, the adapter board connects to one or more uplink interfaces via a second cable, enabling communication between the central processing unit in the first chassis and the graphics processors in the multiple second chassis.

[0126] In this embodiment, multiple communication links are established between the central processing unit (CPU) and multiple graphics processors (GPUs) in the server to improve the server's parallel data processing performance. By deploying cable modules on each communication link to replace relay devices such as timers or passive cards, link insertion loss is reduced, transmission performance is improved, and data processing efficiency is increased.

[0127] In one embodiment of this application, the number of second cables in the server is no greater than the maximum value of the cable modules that the central processing unit can connect to.

[0128] In the above embodiments, the number of cable modules that the central processing unit of the first chassis in the server can connect to has an upper limit. Since the number of second cables in the server is no greater than the number of cable modules deployed in the first chassis, the upper limit of the number of cable modules can be used as the upper limit of the number of second cables. The total number of uplink interfaces occupied by all adapter cards in the second chassis of the server is the same as the number of second cables, that is, it is also no greater than this upper limit. For example, if there are 8 cable modules deployed in the first chassis of the server, the number of second cables between the first chassis and the second chassis is no greater than 8.

[0129] Based on this, the communication links between the first chassis and each second chassis can be flexibly constructed in an even or uneven manner, depending on the number of second chassis in the server, thereby forming different server architectures to adapt to diverse application scenarios.

[0130] For example, in a server configuration comprising one first chassis and two second chassis, the first chassis houses eight cable modules, each connected to one of the eight second cables. All second cables are evenly distributed to adapter boards in the two second chassis, with each adapter board connected to four second cables via four uplink interfaces.

[0131] In a server configuration comprising one primary chassis and three secondary chassis, the primary chassis houses eight cable modules, each connected to one of the eight secondary cables. All secondary cables are then connected to adapter boards in the three secondary chassis. Specifically, one secondary chassis' adapter board connects to four secondary cables via four uplink interfaces, while the other two secondary chassis' adapter boards each connect to two secondary cables via two uplink interfaces.

[0132] In one embodiment, the server includes a first chassis and at least a plurality of second chassis. The plurality of second chassis constitute at least one computing unit, and each computing unit includes at least two first chassis. The first chassis are connected to any of the second chassis in each computing unit via at least one second cable, and the second chassis in each computing unit are connected to each other via at least one second cable in each pair.

[0133] In this embodiment, the server is configured to perform large-scale AI calculations. The central processing unit (CPU) inside the first chassis, upon receiving the target computation task, processes the task based on the number of computing units within the server, generating subtasks equal to the number of computing units. These subtasks are then distributed to the respective computing units. The graphics processing units (GPUs) within each computing unit perform parallel computations on the received subtasks and return the results to the CPU. By using cable modules instead of retimers or passive cards in the large-scale server architecture, the overall insertion loss of the link is significantly reduced, server performance is improved, and the resources consumed by the device's computation are greatly saved, reducing device development and usage costs.

[0134] Based on the same inventive concept, one embodiment of this application provides a server configuration method. Referring to FIG5, FIG5 is a flowchart of a server configuration method proposed in an embodiment of this application. As shown in FIG5, the method includes:

[0135] S1: Based on the first chassis and the second chassis provided in the above embodiments, configure the target server; the target server includes: a first chassis and at least one second chassis;

[0136] S2: Detect and acquire the insertion loss of the cable module in the first chassis;

[0137] S3: Based on the insertion loss of the cable module, calculate a first threshold as the upper limit of the length of the second cable; the second cable is configured to connect the first chassis and the second chassis;

[0138] S4: Calculate the second threshold based on the distance between the first chassis and the second chassis, and use it as the lower limit of the length of the second cable;

[0139] S5: Determine the length of the second cable based on the first threshold and the second threshold.

[0140] In this embodiment, when configuring the server, the number of second chassis to be deployed in the server is first determined. The insertion loss of the cable modules used is determined, and the overall insertion loss of the links between the first chassis and each of the second chassis is calculated when the cable modules are deployed. Based on the overall insertion loss of each link, a first threshold is calculated as the upper limit of the second cable length. Then, based on the application scenario of the server, the server layout is determined. The positions of the first chassis and each of the second chassis are determined according to the server layout, and the distances between the first chassis and each of the second chassis are determined. Based on the distances between the first chassis and each of the second chassis, a second threshold is calculated as the lower limit of the second cable length. Finally, based on the upper and lower limits of the second cable length, the length of the second cable is determined.

[0141] In one embodiment of this application, the first chassis and at least one second chassis are arranged vertically.

[0142] When the number of second chassis is not less than 2, the first chassis is located between the two second chassis.

[0143] In one embodiment, the server is placed in a rack, with the first and second chassis arranged vertically. Given that in practical applications, the first and second chassis may belong to different vendors, the locations of the external cable (second cable) connectors may differ. An example is provided below for illustration.

[0144] This embodiment illustrates four different layouts of the first and second chassis that can be implemented according to the present application. Optionally, Figure 6(a) is one of the layout diagrams of the first and second server chassis; Figure 6(b) is another layout diagram of the first and second server chassis; Figure 6(c) is a third layout diagram of the first and second server chassis; and Figure 6(d) is a fourth layout diagram of the first and second server chassis. Among them, Figure 6(a) shows the layout method used in related technologies, in which the shorter first chassis is placed on top of the taller second chassis, the socket positions of the first and second chassis are deployed on the top of the chassis, the socket positions of the two chassis for plugging in the external second cable are close to each other, and the required length of the second cable is short.

[0145] Compared to the layout shown in Figure 6(a), the layouts shown in Figures 6(b), 6(c), and 6(d) are as follows:

[0146] In Figure 6(b), the taller second chassis is placed on top of the shorter first chassis, and the ports of both the first and second chassis are located on the top of the chassis.

[0147] In Figure 6(c), the first chassis with a smaller height is placed on top of the second chassis with a larger height. The sockets of the first chassis are located at the top of the chassis, while the sockets of the second chassis are located at the lower part of the chassis.

[0148] In Figure 6(d), the smaller first chassis is placed on top of the two larger second chassis. The ports of both the first and second chassis are located at the top of the chassis. If there are more second chassis, they are placed on top of the currently topmost second chassis or below the bottommost second chassis.

[0149] Since the socket positions on the chassis are far apart in Figures 6(b), 6(c), and 6(d), their layout requires a longer second cable. Related technologies use repeater devices such as timers and passive cards, but these three layouts cannot be implemented due to the overall insertion loss of the link. However, the cable module provided in this application can reduce the insertion loss of the link, and thus a longer second cable can be used, enabling the server to adapt to the application scenarios shown in Figures 6(b), 6(c), and 6(d).

[0150] Based on the same inventive concept, one embodiment of this application provides a computer program product, including a computer program that, when executed by a processor, implements the steps in the server configuration method of any of the above embodiments of this application.

[0151] Based on the same inventive concept, one embodiment of this application provides a readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps in the server configuration method of any of the above embodiments of this application.

[0152] Based on the same inventive concept, one embodiment of this application provides an electronic device. Figure 7 is a schematic diagram of an electronic device according to an embodiment of this application. As shown in Figure 7, the electronic device includes: a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the steps in the server configuration method as described in any of the above embodiments of this application.

[0153] Regarding the products in the above embodiments, the way in which each module performs operations has been described in detail in the embodiments related to the method, and will not be elaborated here.

[0154] The above are merely preferred embodiments of this application and are not intended to limit this application. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the protection scope of this application.

[0155] For the sake of simplicity, the method embodiments are described as a series of actions. However, those skilled in the art should understand that this application is not limited to the described order of actions, as some steps may be performed in other orders or simultaneously. Furthermore, those skilled in the art should also understand that the embodiments described in the specification are optional, and the actions and components involved are not necessarily essential to this application.

[0156] Those skilled in the art will understand that embodiments of this application can be provided as methods, apparatus, or computer program products. Therefore, embodiments of this application can take the form of entirely hardware embodiments, entirely software embodiments, or embodiments combining software and hardware aspects. Furthermore, embodiments of this application can take the form of computer program products implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0157] This application describes embodiments with reference to flowchart illustrations and / or block diagrams of methods, terminal devices (systems), and computer program products according to embodiments of this application. It should be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing terminal device to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing terminal device, produce means configured to implement the functions specified in one or more blocks of the flowchart illustrations and / or one or more blocks of the block diagrams.

[0158] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing terminal device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means that implement the functions specified in one or more flowcharts and / or one or more block diagrams.

[0159] These computer program instructions may also be loaded onto a computer or other programmable data processing terminal equipment to cause a series of operational steps to be performed on the computer or other programmable terminal equipment to produce a computer-implemented process, such that the instructions, which execute on the computer or other programmable terminal equipment, provide steps for implementing the functions specified in one or more flowcharts and / or one or more block diagrams.

[0160] Although optional embodiments of the present application have been described, those skilled in the art, once they understand the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, this application is to be interpreted as including the optional embodiments as well as all changes and modifications falling within the scope of the embodiments of this application.

[0161] Finally, it should be noted that in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or terminal device that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or terminal device. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or terminal device that includes the element.

[0162] The cable module, server, server configuration method, and product provided in this application have been described in detail above. Optional examples have been used to illustrate the principles and implementation methods of this application. The descriptions of the above embodiments are only for the purpose of helping to understand the methods and core ideas of this application. At the same time, for those skilled in the art, there will be changes in the optional implementation methods and application scope based on the ideas of this application. Therefore, the content of this specification should not be construed as a limitation of this application.

Claims

1. A cable module, characterized in that, include: Board end and first cable; The two sides of the board end are electrically connected, wherein a first connector is fixed on one side of the board end, and a soldering surface is provided on the other side of the board end; one end of the first cable is soldered to the soldering surface. The other end of the first cable is fixed with a second connector; the first connector and the second connector are respectively connected to the first data processing unit and the second data processing unit; the first data processing unit and the second data processing unit are configured to perform artificial intelligence tasks.

2. The cable module according to claim 1, characterized in that, The welding surface is located within the projection area of ​​the first connector on the end of the board. At least one through hole is provided on the end of the plate, and the through hole is located in the projection area of ​​the welding surface on the end of the plate. The two sides of the plate end are electrically connected through the through holes.

3. The cable module according to claim 1, characterized in that, The first connector is configured to plug into a communication device that has the first connector.

4. The cable module according to claim 1, characterized in that, The second connector is configured to plug into a communication device that has a second connector.

5. A server, characterized in that, include: A first chassis and a second chassis, wherein the first chassis communicates with the second chassis via a second cable; The first chassis houses a central processing unit and a cable module as described in claim 1 or 2; wherein the central processing unit is mounted on a motherboard and is connected to the second connector of the cable module via a second interface on the motherboard. The second chassis houses a graphics processor and an adapter board; the graphics processor is connected to the adapter board via a high-speed connector; the adapter board is connected to the cable module via the second cable.

6. The server according to claim 5, characterized in that, The adapter board is provided with a first interface; The two ends of the second cable are fixed with first connectors, which are respectively connected to the first interface on the adapter board and the first connector of the cable module.

7. The server according to claim 5, characterized in that, The length of the second cable is no greater than the first threshold. The first threshold is determined based on the insertion loss of the link between the central processing unit and the graphics processing unit.

8. The server according to claim 7, characterized in that, The length of the second cable is not less than the second threshold; the second threshold is determined based on the first location of the first chassis and the second location of the second chassis.

9. The server according to claim 5, characterized in that, The first connector is connected to the second cable, which in turn connects to the graphics processor. The second connector is connected to the central processing unit.

10. The server according to claim 5, characterized in that, The second cable is a fifth-generation high-speed serial computer expansion bus; the adapter board is a high-speed serial computer expansion bus standard board.

11. The server according to claim 5, characterized in that, The server includes: a first chassis and a second chassis; The adapter board includes at least one uplink interface; the adapter board is connected to a corresponding number of second cables via at least one uplink interface; The first chassis includes at least one cable module, and each cable module is connected to a second cable. The central processing unit is connected to a cable module through multiple second interfaces on the motherboard.

12. The server according to claim 5, characterized in that, The server includes: a first chassis and at least two second chassis; Each adapter board includes at least one uplink interface; each adapter board is connected to a corresponding number of second cables via at least one uplink interface; The first chassis includes at least two cable modules, each cable module being connected to a second cable; The central processing unit is connected to a cable module through multiple second interfaces on the motherboard.

13. The server according to claim 11 or 12, characterized in that, In the server, the number of second cables is not greater than the maximum value of the cable modules that the central processing unit can connect to.

14. The server according to claim 5, characterized in that, The server includes: one first chassis and two second chassis; The first chassis is equipped with 8 cable modules, each of which is connected to a second cable. All second cables are connected to adapter boards in the two second chassis in an evenly distributed manner, with each adapter board connected to four second cables via four uplink interfaces.

15. The server according to claim 5, characterized in that, The server includes: one first chassis and three second chassis; The first chassis is equipped with 8 cable modules, each of which is connected to a second cable. All the second cables are connected to the adapter boards in the three second chassis respectively. One adapter board in the second chassis is connected to the four second cables through four uplink interfaces, and the adapter boards in the two second chassis are connected to the two second cables through two uplink interfaces respectively.

16. A server configuration method, characterized in that, include: Configure the target server based on the first chassis and the second chassis as described in any one of claims 5-15; The target server includes: a first chassis and at least one second chassis; Detect and acquire the insertion loss of the cable module in the first chassis; Based on the insertion loss of the cable module, a first threshold is calculated as the upper limit of the length of the second cable; the second cable is configured to connect the first chassis and the second chassis. Based on the distance between the first chassis and the second chassis, a second threshold is calculated as the lower limit of the length of the second cable; The length of the second cable is determined based on the first threshold and the second threshold.

17. The server configuration method according to claim 16, characterized in that, The first chassis and the at least one second chassis are arranged vertically. When the number of second chassis is not less than 2, the first chassis is located between the two second chassis.

18. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by a processor, it implements the steps of the method of claim 11 or 12.

19. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the method described in claim 16 or 17.

20. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the steps in the method as described in claim 16 or 17.

Images