Processing and switching die, communication chip, and data processing apparatus and method

Abstract

A processing and switching die, a communication chip, and a data processing apparatus and method, which relate to the technical field of communications, and are used for greatly reducing the number of processing units in a device, reducing the device cost, improving the energy efficiency of the device, and also saving on the heat dissipation and power supply costs and deployment space. Provided is a communication ship. The communication chip comprises a processing and switching die, and a plurality of network processing dies, which are coupled to the processing and switching die, wherein the processing and switching die comprises at least one processing unit and a plurality of switching units, the at least one processing unit is configured to control and manage the plurality of switching units and the plurality of network processing dies, and the plurality of network processing dies are configured to provide capabilities of a network interface, network processing, etc.

Description

Processing exchange chip, communication chip, data processing device and method

[0001] This application claims priority to the Chinese patent application filed with the State Intellectual Property Office on September 27, 2023, with application number 202311282067.4 and application name “Processing Switching Chip, Communication Chip, Data Processing Device and Method”, all contents of which are incorporated by reference into this application. Technical Field

[0002] The present application relates to the field of communication technology, and in particular to a processing switching die, a communication chip, a data processing device and a method. Background Art

[0003] Currently, data processing devices such as routers and switches, or the system-on-chip (SoC) used in routers and switches, typically include multiple different functional modules. Each functional module includes a processing unit for configuring and managing the hardware module. The processing units in different functional modules jointly implement the control and management functions of the data processing device.

[0004] For example, taking a router as an example, the router may include a main processing unit (MPU), a switch fabric unit (SFU) and an interface board (LPU). The MPU, the SFU and the LPU all include a central processing unit (CPU). Each CPU is responsible for the configuration and management of the module in which it is located. The CPU in the MPU manages the CPU in the SFU and the CPU in the LPU through a multi-level management strategy.

[0005] However, the CPU resources in the above-mentioned different functional modules cannot be shared, so a large number of CPUs need to be consumed, and the data processing device will also have problems such as high power consumption, high cost and large size.

[0006] Summary of the Invention

[0007] The present application provides a processing switching die, a communication chip, a data processing device and a method, which relate to the field of communication technology and are used to significantly reduce the number of processing units in a device, reduce device costs, improve device energy efficiency, and save heat dissipation and power supply costs and deployment space.

[0008] To achieve the above objectives, the embodiments of the present application adopt the following technical solutions:

[0009] In a first aspect, a processing switch die is provided, which can be used in a communications chip. The processing switch die is configured to couple with multiple network processing die in the communications chip. The die may also be referred to as a bare die or die. The processing switch die includes at least one processing unit and multiple switching units. The at least one processing unit is configured to control and manage the multiple switching units and the multiple network processing die. The multiple switching units can be configured to provide data exchange functions, and the multiple network processing die can be configured to provide network interface and network processing capabilities. The control and management can include control and management of the control plane and / or the management plane.

[0010] In the above technical solution, the processing switching crystal can be used to control and manage the multiple switching units and the multiple network processing crystals, and provide computing capabilities. The multiple network processing crystals can be used to provide network interfaces and network processing capabilities, so that a network-based computing (i.e., network-based and computing-assisted) communication chip can be provided based on the processing switching crystal. The communication chip can be used in network equipment with network processing capabilities as the core, and can significantly reduce the number of processing units in the network equipment, reduce equipment costs, improve equipment energy efficiency, and save heat dissipation and power supply costs and deployment space, thereby achieving green, low-carbon and sustainable development.

[0011] In one possible implementation of the first aspect, the at least one processing unit includes a processing unit coupled to the multiple switching units; the processing unit is configured to send first control and management information to the multiple switching units; and the processing unit is further configured to send second control and management information to the multiple network processing crystals via the multiple switching units. In this possible implementation, a centralized processing unit is provided in the processing switching crystal, and control and management of the multiple switching units and the multiple network processing crystals is implemented via the processing unit, thereby significantly reducing the number of processing units.

[0012] In one possible implementation of the first aspect, a first switching unit among the multiple switching units is directly connected to a first network processing crystal among the multiple network processing crystals; and the processing unit is further configured to send second control and management information to the first network processing crystal through the first switching unit. In this possible implementation, the first switching unit is directly connected to the first network processing crystal, and the processing unit sends the second control and management information to the first network processing crystal through the first switching unit, thereby enabling control and management of the first network processing crystal.

[0013] In one possible implementation of the first aspect, a second network processing grain among the multiple network processing grains is directly connected to the first network processing grain, and the second network processing grain is not directly connected to the multiple switching units; the processing unit is further configured to send second control and management information to the second network processing grain via the first switching unit and the first network processing grain. In the above possible implementation, the first switching unit is directly connected to the first network processing grain, and the second network processing grain is directly connected to the first network processing grain. The processing unit sends the second control and management information to the second network processing grain via the first switching unit and the first network processing grain, thereby enabling control and management of the second network processing grain.

[0014] In one possible implementation of the first aspect, a second switching unit among the multiple switching units is directly connected to a third network processing grain among the multiple network processing grains, and the third network processing grain is directly connected to the second network processing grain; the processing unit is further configured to send second control and management information to the second network processing grain via the second switching unit and the third network processing grain. In the above possible implementation, the second switching unit is directly connected to the third network processing grain, and the third network processing grain is directly connected to the second network processing grain. The processing unit sends the second control and management information to the second network processing grain via the second switching unit and the third network processing grain, thereby also implementing control and management of the second network processing grain.

[0015] In one possible implementation of the first aspect, the processing unit is further configured to distribute the second control management information to the second network processing crystal through the first switching unit and the second switching unit. In this possible implementation, when the second control management information of a network processing crystal exists over at least two paths, the processing unit distributes the second control management information to the network processing crystal, thereby increasing control management flexibility and ensuring load balancing across the at least two paths.

[0016] In a possible implementation of the first aspect, the at least one processing unit includes multiple processing units, the multiple processing units and the multiple switching units have a one-to-one correspondence and are integrated into multiple processing switching units, and any two of the multiple processing switching units are coupled; the multiple processing switching units are configured to send second control and management information to the multiple network processing crystals. In the above possible implementation, multiple distributed processing units are provided in the processing switching crystals, the multiple processing units and the multiple switching units have a one-to-one correspondence and are integrated into multiple processing switching units. Control and management of the multiple network processing crystals are implemented through the multiple processing switching units, which can significantly reduce the number of processing units while improving the flexibility of control and management and ensuring load balancing of the multiple processing switching units.

[0017] In one possible implementation of the first aspect, a first processing exchange unit among the multiple processing exchange units is directly connected to a first network processing grain among the multiple network processing grains; the first processing exchange unit is configured to send second control and management information to the first network processing grain. In this possible implementation, the first processing exchange unit is directly connected to the first network processing grain, and the first processing exchange unit sends the second control and management information to the first network processing grain, thereby enabling control and management of the first network processing grain.

[0018] In one possible implementation of the first aspect, a second network processing grain among the multiple network processing grains is directly connected to the first network processing grain, and the second network processing grain is not directly connected to the multiple processing and switching units; the first processing and switching unit is further configured to send second control and management information to the second network processing grain via the first network processing grain. In this possible implementation, the first processing and switching unit is directly connected to the first network processing grain, the second network processing grain is directly connected to the first network processing grain, and the first processing and switching unit sends the second control and management information to the second network processing grain via the first network processing grain, thereby enabling control and management of the second network processing grain.

[0019] In one possible implementation of the first aspect, a second processing exchange unit among the multiple processing exchange units is directly connected to a third network processing grain among the multiple network processing grains, and the third network processing grain is directly connected to the second network processing grain; the second processing exchange unit is further configured to send second control and management information to the second network processing grain via the third network processing grain. In the above possible implementation, the second processing exchange unit is directly connected to the third network processing grain, and the second network processing grain is directly connected to the third network processing grain. The second processing exchange unit sends the second control and management information to the second network processing grain via the third network processing grain, thereby also achieving control and management of the second network processing grain.

[0020] Optionally, when there are at least two paths for the second control management information of a network processing grain, different processing switching units among the multiple processing switching units can also send the second control management information to the network processing grain in a shared manner, thereby improving the flexibility of control management and ensuring load balancing of at least two paths.

[0021] In one possible implementation of the first aspect, the multiple processing and switching units include a first processing and switching unit and a second processing and switching unit; the processing unit in the first processing and switching unit is configured to send the second control and management information to the processing unit in the second processing and switching unit. In this possible implementation, any one of the multiple processing and switching units can reuse the processing unit in another processing and switching unit, thereby improving the performance of the processing and switching crystal and ensuring normal execution of control and management functions even when a processing and switching crystal fails.

[0022] In one possible implementation of the first aspect, the multiple processing and switching units include a first processing and switching unit and a second processing and switching unit; the processing unit in the first processing and switching unit is further configured to send first control and management information to the switching unit in the second processing and switching unit. In this possible implementation, any one of the multiple processing and switching units can reuse a switching unit in another processing and switching unit, thereby improving the performance of the processing and switching crystal and ensuring normal execution of control and management functions even when a processing and switching crystal fails.

[0023] In a second aspect, a communication chip is provided, comprising: at least one processing switch crystal, and a plurality of network processing crystals coupled to the at least one processing switch crystal, wherein the at least one processing switch crystal is the processing switch crystal as provided in the first aspect or the first aspect.

[0024] In a possible implementation manner of the second aspect, the at least one process exchange die includes a first process exchange die and a second process exchange die, and the first process exchange die is directly connected to the second process exchange die.

[0025] In a third aspect, a data processing device is provided, comprising: a backplane, and at least one communication chip arranged on the backplane, wherein the communication chip is the communication chip provided by the second aspect or any possible implementation of the second aspect.

[0026] In a fourth aspect, a control management method is provided, which is applied to a communication chip, wherein the communication chip includes a processing switching grain and multiple network processing grains coupled to the processing switching grain, and the processing switching grain includes at least one processing unit and multiple switching units. The method includes: the at least one processing unit sends first control management information to the multiple switching units; the at least one processing unit sends second control management information to the multiple network processing grains through the multiple switching units.

[0027] In a possible implementation of the fourth aspect, the first control management information of each switching unit includes first configuration information for configuring the switching unit, and the second control management information of each network processing crystal includes second configuration information for configuring the network processing crystal.

[0028] In a possible implementation of the fourth aspect, the second control management information also includes transmission control information of the data packet, and the method also includes: any network processing grain among the multiple network processing grains sends a signaling message to the at least one processing unit through the multiple switching units, and the signaling message is used to trigger the at least one processing unit to send the transmission control information.

[0029] In a possible implementation of the fourth aspect, the second control management information also includes device management information, and the method also includes: any network processing grain among the multiple network processing grains sends device status information to the at least one processing unit through the multiple switching units, and the device status information is used for the at least one processing unit to send the device management information.

[0030] In a possible implementation of the fourth aspect, the at least one processing unit includes a processing unit, the multiple switching units are all coupled to the processing unit, and the first switching unit among the multiple switching units is directly connected to the first network processing grain among the multiple network processing grains; the at least one processing unit sends the second control management information to the multiple network processing grains through the multiple switching units, including: the processing unit sends the second control management information to the first network processing grain through the first switching unit.

[0031] In a possible implementation of the fourth aspect, a second network processing grain among the multiple network processing grains is directly connected to the first network processing grain, and the second network processing grain is not directly connected to the multiple switching units; the at least one processing unit sends the second control management information to the multiple network processing grains through the multiple switching units, and also includes: the processing unit sends the second control management information to the second network processing grain through the first switching unit and the first network processing grain.

[0032] In a possible implementation of the fourth aspect, a second switching unit among the multiple switching units is directly connected to a third network processing grain among the multiple network processing grains, and the third network processing grain is directly connected to the second network processing grain; the at least one processing unit sends the second control management information to the multiple network processing grains through the multiple switching units, and also includes: the processing unit sends the second control management information to the second network processing grain through the second switching unit and the third network processing grain.

[0033] In a possible implementation of the fourth aspect, the at least one processing unit includes multiple processing units, the multiple processing units and the multiple switching units correspond one-to-one and are integrated into multiple processing switching units, and any two of the multiple processing switching units are coupled; the at least one processing unit sends the second control management information to the multiple network processing grains through the multiple switching units, including: the multiple processing switching units send the second control management information to the multiple network processing grains.

[0034] In a possible implementation of the fourth aspect, a first processing exchange unit among the multiple processing exchange units is directly connected to a first network processing grain among the multiple network processing grains; the multiple processing exchange units send second control management information to the multiple network processing grains, including: the first processing exchange unit sends the second control management information to the first network processing grain.

[0035] In a possible implementation of the fourth aspect, a second network processing grain among the multiple network processing grains is directly connected to the first network processing grain, and the second network processing grain is not directly connected to the multiple processing switching units; the multiple processing switching units send the second control management information to the multiple network processing grains, and also includes: the first processing switching unit sends the second control management information to the second network processing grain through the first network processing grain.

[0036] In a possible implementation of the fourth aspect, a second processing exchange unit among the multiple processing exchange units is directly connected to a third network processing grain among the multiple network processing grains, and the third network processing grain is directly connected to the second network processing grain; the multiple processing exchange units send second control management information to the multiple network processing grains, and also include: the second processing exchange unit sends the second control management information to the second network processing grain through the third network processing grain.

[0037] In a possible implementation of the fourth aspect, the multiple processing switching units include a first processing switching unit and a second processing switching unit; the at least one processing unit sends first control management information to the multiple switching units, including: the processing unit in the first processing switching unit sends the first control management information to the switching unit in the second processing switching unit.

[0038] In a possible implementation manner of the fourth aspect, the control management information includes control management information of a control plane and / or a management plane.

[0039] In another aspect of the present application, a computer-readable storage medium is provided, in which a computer program or instruction is stored. When the computer program or instruction is executed, the control management method provided by the fourth aspect or any possible implementation of the fourth aspect is implemented.

[0040] In another aspect of the present application, a computer program product is provided, which includes: a computer program (also referred to as code, or instructions), which, when executed, enables the computer to execute the control management method provided in the fourth aspect or any possible implementation of the fourth aspect.

[0041] It can be understood that the beneficial effects that can be achieved by any of the communication chips, data processing devices, control management methods, computer-readable storage media and computer program products provided above can correspond to the beneficial effects in the control management method provided above, and will not be repeated here. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] FIG1 is a schematic diagram of an application scenario of a router provided in an embodiment of the present application;

[0043] FIG2 is a schematic diagram of the structure of a network device provided in an embodiment of the present application;

[0044] FIG3 is a schematic diagram of the structure of another network device provided in an embodiment of the present application;

[0045] FIG4 is a schematic diagram of the front and back of a network device provided in an embodiment of the present application;

[0046] FIG5 is a schematic diagram of control management in a network device provided in an embodiment of the present application;

[0047] FIG6 is a schematic diagram of the structure of a DPU provided in an embodiment of the present application;

[0048] FIG7 is a schematic diagram of the structure of another DPU provided in an embodiment of the present application;

[0049] FIG8 is a schematic diagram of a structure for processing exchange die according to an embodiment of the present application;

[0050] FIG9 is a schematic diagram of another structure for processing exchange die according to an embodiment of the present application;

[0051] FIG10 is a schematic diagram of a structure of another method for processing exchange die provided in an embodiment of the present application;

[0052] FIG11 is a schematic diagram of the structure of a communication chip provided in an embodiment of the present application;

[0053] FIG12 is a schematic diagram of the structure of another communication chip provided in an embodiment of the present application;

[0054] FIG13 is a schematic structural diagram of another communication chip provided in an embodiment of the present application;

[0055] FIG14 is a schematic diagram of the structure of another communication chip provided in an embodiment of the present application;

[0056] FIG15 is a schematic structural diagram of another communication chip provided in an embodiment of the present application;

[0057] FIG16 is a schematic diagram of the structure of another communication chip provided in an embodiment of the present application;

[0058] FIG17 is a schematic structural diagram of another communication chip provided in an embodiment of the present application;

[0059] FIG18 is a schematic diagram of the structure of another communication chip provided in an embodiment of the present application;

[0060] FIG19 is a schematic diagram of the structure of another communication chip provided in an embodiment of the present application. DETAILED DESCRIPTION

[0061] The following sections discuss the making and use of various embodiments in detail. However, it should be understood that many applicable inventive concepts provided herein can be implemented in a variety of specific contexts. The specific embodiments discussed are merely illustrative of specific ways to implement and use this description and technology and do not limit the scope of this application.

[0062] Unless defined otherwise, all technical and scientific terms used herein have the same meanings as commonly understood by one of ordinary skill in the art.

[0063] Various circuits or other components may be described or referred to as being "configured to" perform one or more tasks. In this case, "configured to" is used to imply structure by indicating that the circuit / component includes structure (e.g., circuitry) that performs the one or more tasks during operation. Thus, even when a specified circuit / component is not currently operational (e.g., not turned on), the circuit / component may be referred to as being configured to perform the task. Circuits / components used with the phrase "configured to" include hardware, such as circuitry that performs an operation, etc.

[0064] The technical solutions in the embodiments of the present application will be described below with reference to the drawings in the embodiments of the present application. In the present application, "at least one" refers to one or more, and "more" refers to two or more. "And / or" describes the association relationship of associated objects, indicating that there may be three relationships. For example, A and / or B can represent: the existence of A alone, the existence of A and B at the same time, and the existence of B alone, where A and B can be singular or plural. The character " / " generally indicates that the previous and next associated objects are in an "or" relationship. "At least one of the following" or similar expressions refers to any combination of these items, including any combination of single or plural items. For example, at least one of a, b or c can represent: a, b, c, a and b, a and c, b and c, a, b and c; where a, b and c can be single or multiple.

[0065] The embodiments of this application use terms such as "first" and "second" to distinguish objects with similar names, functions, or effects. Those skilled in the art will understand that terms such as "first" and "second" do not limit the quantity or order of execution. The term "coupled" is used to indicate an electrical connection, including direct connection via wires or connectors or indirect connection via other devices. Therefore, "coupling" should be considered a broadly defined electronic communication connection.

[0066] It should be noted that, in this application, words such as "exemplary" or "for example" are used to indicate examples, illustrations, or descriptions. Any embodiment or design described in this application as "exemplary" or "for example" should not be construed as being preferred or advantageous over other embodiments or designs. Rather, the use of words such as "exemplary" or "for example" is intended to present the relevant concepts in a concrete manner.

[0067] Before introducing the embodiments of the present application, the relevant application scenarios involved in the present application are first introduced and explained.

[0068] Currently, data processing devices such as routers and switches, or the system-on-chip (SoC) used in routers and switches, typically include multiple functional modules. Each functional module includes a processing unit for configuring and managing the hardware module. The processing units in the different functional modules collectively implement the control and management functions of the data processing device. The following uses a router as an example to illustrate this data processing device.

[0069] A router is a network device used to connect two or more networks. For example, as shown in Figure 1, it can be used to connect a local area network (LAN) and the internet. The LAN can include devices such as mobile phones, tablets, and personal computers, and the internet can include one or more servers. A router acts as a gateway between networks, reading the destination address in data packets and deciding how to route them based on that address. Routers can recognize different network protocols, such as the Ethernet protocol used by a local area network and the TCP / IP protocol used by the internet. This allows them to analyze the destination addresses of data packets from various networks, converting addresses from non-TCP / IP networks into TCP / IP addresses and vice versa. Then, based on a selected routing algorithm, the router routes data packets with different destination addresses along the optimal route to their destination.

[0070] As shown in Figures 2 and 3, the network device may include a frame, and a main processing unit (MPU), a switch fabric unit (SFU), an interface board (LPU), a power module (PM), and a fan module (FM) installed on the frame. The main control board MPU and the interface board LPU can exchange data (i.e., a data plane channel) through the switch fabric unit SFU. The power module PM can be used to supply power to the main control board MPU, the switch fabric unit SFU, the interface board LPU, and the fan module FM, and the fan module FM can be used to provide heat dissipation. The interface board LPU can also be called a line processing unit. The frame, power module PM, and fan module FM are not shown in Figure 3.

[0071] In this router, the main control unit (MPU), the switching fabric unit (SFU), and the interface unit (LPU) all include a central processing unit (CPU). Each CPU is responsible for configuring and managing its own module. The CPUs in these modules collectively control and manage the router. The following describes the functions and structures of the MPU, SFU, and LPU, respectively.

[0072] The main control board MPU: can be used to be responsible for the control plane and management plane of the entire system. The control plane can be used to complete the system's protocol processing, business processing, routing operations, forwarding control, business scheduling, traffic statistics, system security and other functions. The management plane can be used to complete the system's operating status monitoring, environmental monitoring, log and alarm information processing, system loading, system upgrade and other functions. Among them, the main control board MPU mainly includes a CPU chip and a network processing (NP) chip. In a possible example, as shown in Figure 3, the main control board MPU may include a network interface (fabric interface, FIC), an NP chip, a switching (SW) unit, a CPU chip and a system monitoring component, etc. The switching unit may be a local area network switching (LAN switch, LSW) unit.

[0073] The Switching Fabric Unit (SFU) is responsible for the data plane of the entire system. This data plane provides high-speed data channels and enables service exchange between various service modules. For example, the MPU and LPU can communicate through the SFU. The SFU primarily consists of a CPU chip and a switch (SW) chip. Figure 3 shows only the SW chip.

[0074] The interface board LPU is used to provide data forwarding functions and can also provide interfaces of different types and different rates. For example, the interface may include an optical interface and an electrical interface. The interface board LPU can forward data through a distributed data plane. The interface board LPU mainly includes a CPU chip and an NP chip. In a possible example, as shown in Figure 3, the interface board LPU may include a CPU chip, a network interface FIC, a traffic manager (TM), an NP chip, a peripheral interface (PIC), and a monitoring component. In the figure, the traffic manager TM and the network interface FIC are represented as TM+FIC.

[0075] In actual applications, the same network device may include multiple main control boards (MPUs), multiple switching fabric boards (SFUs), and multiple interface boards (LPUs). For example, as shown in Figure 4 , the network device may include two main control boards (MPUs), eight switching fabric boards (SFUs), and eight interface boards (LPUs). If each main control board (MPU) includes one CPU chip and one NP chip, each interface board (LPU) includes one CPU chip and one NP chip, and each switching fabric board includes one CPU chip and one SW chip, the network device requires a total of 18 CPU chips, 10 NP chips, and eight SW chips. Figure 4 illustrates the example of two main control boards (MPUs) and eight interface boards (LPUs) installed on the front of the network device, and eight switching fabric boards (SFUs) installed on the back of the network device.

[0076] In the above-mentioned network device, the CPU chip on the main control board MPU centrally manages each interface board LPU and switching network board SFU through the NP chip and other means. Each interface board LPU and switching network board SFU has its own CPU responsible for the configuration and management of this module. For example, as shown in Figure 5, the SW unit in the main control board MPU is connected to the CPU chip via the Ethernet interface GE (for example, 10GE) and to the NP chip via 40GE; wherein, the CPU chip in the main control board MPU manages the CPU chip in the interface board LPU and the CPU chip in the switching network board SFU through the high-speed inter-board communication channel in LSW and 10GE, the CPU chip in the interface board LPU manages the NP chip via the PCIe interface, and the CPU chip in the switching network board SFU manages the SW chip via the PCIe interface.

[0077] However, the CPU resources in the above-mentioned main control board MPU, switching network board SFU and interface board LPU cannot be shared, so a large number of CPU chips need to be consumed. At the same time, the network equipment will have problems of high power consumption, high cost and large size, which will also waste the deployment space of the network equipment to a certain extent.

[0078] In addition, the data processing unit (DPU) is a data processing device designed and implemented based on a computing-network fusion architecture. The DPU primarily integrates a computing cluster and a network unit. The computing cluster can be used for data processing, and the network unit can be used for data exchange. The computing cluster and the network unit are also used together to implement the control and management of the DPU. The structure of the DPU is illustrated below using Figures 6 and 7.

[0079] In one possible example, as shown in FIG6 , the data processing unit DPU may include: multiple data clusters, an on-chip network, a control cluster, a work scheduler, a double-bit random access memory DDR4, a high-bandwidth memory (HBM), a host unit, and a network unit. The multiple data clusters may communicate with other components via the on-chip network; the control cluster and the work scheduler may be used to control and schedule the data processing unit DPU; the DDR4 and the HBM may be used to store data of the data processing unit DPU; the host unit may be used to support the connection between the data processing unit DPU and a host; and the network unit may be used to support the data processing unit DPU in data exchange and network processing.

[0080] In one possible example, as shown in FIG7 , the data processing unit DPU may include: a CPU subsystem, a system level cache coupled to the CPU subsystem, a data accelerator, PCIe switch 1, PCIe switch 2, and a network interface card (NIC) subsystem coupled to PCIe switch 1 and PCIe switch 2. In addition, the system level cache may be coupled to DRAM; the PCIe switch 1 may be coupled to solid state drives (SSDs) or graphics processing units (GPUs), with SSDs and GPUs represented as SSDs / GPU; the NIC subsystem may include a network interface (e.g., an Ethernet interface or an InfiniBand interface), a data path accelerator (DPA), and accelerated switch and packet processing (ASAP) 2 ) module, etc., the network interface can be used to connect to the network; the PCIe switch 2 can also be connected to the host.

[0081] The core of the aforementioned DPUs is a computing-network fusion architecture, primarily based on computing power, supplemented by network capabilities. Therefore, these DPUs have limitations in network interfaces and processing capabilities, making them unsuitable for network devices centered around network processing.

[0082] Based on this, the embodiments of the present application provide a communication chip that integrates network and computing (i.e., network-based and computing-assisted). The communication chip includes a processing and switching crystal and multiple NP crystals. The processing and switching crystal can be used to control and manage the communication chip and provide computing power, and the multiple NP crystals can be used to provide network interfaces and network processing capabilities. Therefore, the communication chip provided in the embodiments of the present application can be applied to network devices with network processing capabilities as the core. Compared with the existing technology, it can significantly reduce the number of processing units in the network device, reduce equipment costs, improve equipment energy efficiency, and save heat dissipation and power supply costs and deployment space, thereby achieving green, low-carbon and sustainable development.

[0083] The following describes the specific contents of the processing switch die and communication chip provided in the embodiments of the present application.

[0084] FIG8 is a schematic diagram of the structure of a process exchange die provided in an embodiment of the present application. The process exchange die 10 can be used to couple with multiple NP die 20. For example, any one of the multiple NP die 20 can be directly coupled with the process exchange die 10, or indirectly coupled with the process exchange die 10 through another NP die.

[0085] The aforementioned die may also be referred to as a bare die or a die, and thus the process exchange die 10 may also be referred to as a process exchange bare die or a process exchange die, and the plurality of NP die 20 may also be referred to as a plurality of NP bare die or a plurality of NP die. Optionally, the process exchange die 10 and the plurality of NP die 20 may be a plurality of independent die cut from a wafer, or may be a plurality of uncut die located on the same wafer.

[0086] Optionally, in actual applications, the multiple NP chips 20 may be replaced with other chips for use in devices other than routers and switches. In one possible example, these other chips may include, but are not limited to, image processing chips, signal processing chips, neural network processing chips, and the like. This embodiment of the present application does not impose any specific limitations on this, and the description herein uses these multiple NP chips as an example.

[0087] In an embodiment of the present application, the processing switch die 10 may include at least one processing unit (PU) 11 and multiple switch units (SWU) 12. The multiple switch units 12 are coupled to the at least one processing unit 11. The at least one processing unit 11 may be used to control and manage the multiple switch units 12 and the multiple NP die 20. The control and management may include control and management of the control plane and / or the management plane. This control and management approach may be referred to as single-level management of processing units and computing resource sharing.

[0088] Exemplarily, the at least one processing unit 11 can be used to control and manage the protocol processing, business processing, routing calculations, message forwarding, traffic statistics and other functions of the multiple switching units 12 and the multiple NP grains 20, and can also be used to control and manage the operation status monitoring, environmental monitoring, log and alarm information processing, system loading, system upgrade and other functions of the communication chip.

[0089] Optionally, the processing units in the at least one processing unit 11 may include but are not limited to: a central processing unit (CPU), a digital signal processing unit (DSP), a neural processing unit (NPU), a graphics processing unit (GPU), an image signal processor (ISP), a microcontroller or microprocessor, an application-specific integrated circuit, a field programmable gate array or other programmable logic devices, etc.

[0090] In one possible embodiment, as shown in FIG9 , the at least one processing unit 11 in the processing switch die 10 includes a processing unit PU, and the multiple switch units 12 are all coupled to the processing unit. Exemplarily, the multiple switch units 12 include four switch units, represented as switch units SWU1 through SWU4, each of which is coupled to the processing unit PU. FIG9 illustrates an example in which the multiple NP die 20 includes eight NP die, represented as NP die 21 through NP die 28.

[0091] The processing unit PU is configured to send first control management information to the multiple switching units 12, that is, to send the first control management information corresponding to each switching unit in the multiple switching units 12, so as to control and manage the switching unit using the first control management information. For example, the processing unit PU may be configured to send the first control management information to each switching unit in the switching units SWU1 to SWU4, respectively.

[0092] Exemplarily, the first control management information may include at least one of the following: first configuration information for configuring the switching unit, data packet transmission control information, and device management information sent based on multiple received device status information. The data packet transmission control information may be obtained based on configuration or generated based on a triggering signaling message. The signaling message may be received by the NP crystal from the outside and forwarded to the processing unit PU via the switching unit. The multiple device status information may be status information of devices monitored by the NP crystal and, for example, fan modules and power supply modules coupled to the NP crystal. This device status information may also be referred to as maintenance information.

[0093] The processing unit PU is also used to: send second control management information to the multiple NP grains 20 through the multiple switching units 12, that is, send the second control management information corresponding to the NP grain to each of the multiple NP grains 20, so as to control and manage the NP grain through the second control management information. Exemplarily, the second control management information may include at least one of the following information: second configuration information for configuring the NP grain, transmission control information of the data message, and device management information sent based on the received multiple device status information. The transmission control information of the data message can be obtained based on the configuration, or can be generated based on the triggering of the signaling message. The signaling message can be received by the NP grain from the outside and forwarded to the processing unit PU through the switching unit. The multiple device status information can be the status information of the NP grain, and the devices monitored by the fan module and power supply module coupled to the NP grain. The device status information can also be called maintenance information.

[0094] Exemplarily, the above-mentioned configuration may include configuration, debugging, and querying of operational data, specifically for configuring registers and table entry spaces of switching units and NP crystals. The above-mentioned signaling messages may include routing protocol messages. Because signaling messages cannot directly act on the data plane to guide forwarding, they must first be parsed by the at least one processing unit 11 before they can be recognized by the data plane. For example, when the signaling message is used for routing calculation, the at least one processing unit 11 may send the destination IP address and next-hop address to the NP crystal's routing table, so that the data message is forwarded according to the destination IP address and next-hop address in the routing table.

[0095] In one example, if the switching unit SWU1 among the multiple switching units 12 is the first switching unit, and the first switching unit SWU1 is directly connected to the first NP die 21 among the multiple NP die 20, then the processing unit PU sends the second control management information to the first NP die 21. Specifically, the processing unit PU sends the second control management information to the first NP die 21 via the first switching unit SWU1. That is, the processing unit PU sends the second control management information to the first switching unit SWU1; upon receiving the second control management information, the first switching unit SWU1 sends the second control management information to the first NP die 21.

[0096] In another example, if the second NP grain 22 in the multiple NP grains 20 is directly connected to the first NP grain 21, but is not directly connected to the multiple switching units 12, the processing unit PU sends the second control management information corresponding to the second NP grain 22 to the second NP grain 22. Specifically, the processing unit PU sends the second control management information to the second NP grain 22 via the first switching unit SWU1 and the first NP grain 21. That is, the processing unit PU sends the second control management information to the first switching unit SWU1; upon receiving the second control management information, the first switching unit SWU1 sends the second control management information to the first NP grain 21; and upon receiving the second control management information, the first NP grain 21 sends the second control management information to the second NP grain 22.

[0097] In another example, if the switching unit SWU2 among the multiple switching units 12 is the second switching unit, the second switching unit SWU2 is directly connected to the third NP grain 23 among the multiple NP grains 20, and the third NP grain 23 is directly connected to the second NP grain 22, then the processing unit PU sends the second control management information corresponding to the second NP grain 22 to the second NP grain 22. Specifically, the processing unit PU sends the second control management information to the second NP grain 22 via the second switching unit SWU2 and the third NP grain 23. That is, the processing unit PU sends the second control management information to the second switching unit SWU2; upon receiving the second control management information, the second switching unit SWU2 sends the second control management information to the third NP grain 23; upon receiving the second control management information, the third NP grain 23 sends the second control management information to the second NP grain 22.

[0098] Optionally, when there are multiple paths for sending the second control management information corresponding to a certain NP grain, the processing unit PU can distribute the second control management information to at least two of the multiple paths for sending, or specify one of the multiple paths for sending.

[0099] For example, taking the example of the processing unit PU sending the second control management information corresponding to the second NP grain 22 to the second NP grain 22, the processing unit PU can be used to distribute the second control management information to the second NP grain 22 through the first path where the first switching unit SWU1 and the first NP grain 21 are located, and the second path where the second switching unit SWU2 and the third NP grain 23 are located.

[0100] Among them, when the processing unit PU distributes and sends the second control management information corresponding to the above-mentioned second NP grain 22 through at least two paths, the processing unit PU can distribute and send it through load balancing, or it can adaptively distribute and send it according to the transmission conditions of different paths. The embodiment of the present application does not impose specific restrictions on this.

[0101] In another possible embodiment, as shown in FIG10 , the at least one processing unit 11 in the processing exchange die 10 includes a plurality of processing units. The plurality of processing units and the plurality of exchange units 12 correspond one-to-one and are integrated into a plurality of processing exchange units, and any two of the plurality of processing exchange units are coupled. Exemplarily, the plurality of processing exchange units include four processing exchange units, represented as processing exchange unit PSWU1 to processing exchange unit PSWU4, and any two of the processing exchange units from processing exchange unit PSWU1 to processing exchange unit PSWU4 are directly connected. FIG10 illustrates an example in which the plurality of NP die 20 includes eight NP die, represented as NP die 21 to NP die 28.

[0102] The plurality of processing and switching units (e.g., PSWU1 to PSWU4) are configured to send the second control and management information to the plurality of NP die 20. Specifically, the plurality of processing units in the plurality of processing and switching units may be configured to send the second control and management information to the plurality of NP die 20. Furthermore, the plurality of processing units in the plurality of processing and switching units may also be configured to respectively send the first control and management information to the corresponding switching units to control and manage the corresponding switching units.

[0103] Exemplarily, the second control management information may specifically include, but is not limited to, second configuration information for configuring the NP chip, data packet transmission control information, and device management information sent based on multiple received device status information. The data packet transmission control information may be obtained based on configuration or generated based on a triggering signaling message. The signaling message may be received by the NP chip from the outside and sent to the multiple processing and switching units. The multiple device status information may include status information of devices monitored by the NP chip and devices coupled to the NP chip, such as fan modules and power supply modules. This device status information may also be referred to as maintenance information.

[0104] In one example, if the processing exchange unit PSWU1 among the multiple processing exchange units is the first processing exchange unit, and the first processing exchange unit PSWU1 is directly connected to the first NP grain 21 among the multiple NP grains 20, then the first processing exchange unit PSWU1 can be used to: send the second control management information to the first NP grain 21, specifically, the second control management information can be sent to the first NP grain 21 by the processing unit in the first processing exchange unit PSWU1.

[0105] In another example, the second NP die 22 in the plurality of NP die 20 is directly connected to the first NP die 21, but the second NP die 22 is not directly connected to the plurality of processing and switching units; then the first processing and switching unit PSWU1 sending the second control and management information corresponding to the second NP die 22 to the second NP die 22 may include: the first processing and switching unit PSWU1 sending the second control and management information to the second NP die 22 via the first NP die 21. For example, the processing units in the first processing and switching unit PSWU1 may send the second control and management information to the second NP die 22 via the first NP die 21.

[0106] In another example, the processing exchange unit PSWU2 among the multiple processing exchange units is the second processing exchange unit, the second processing exchange unit PSWU2 is directly connected to the third NP grain 23 among the multiple NP grains 20, and the third NP grain 23 is directly connected to the second NP grain 22, then the second processing exchange unit PSWU2 can be used to: send second control management information to the second NP grain 22 through the third NP grain 23.

[0107] Optionally, any one of the multiple processing and switching units may also reuse a processing unit or a switching unit in another processing and switching unit.

[0108] For example, for the first processing and switching unit PSWU1 and the second processing and switching unit PSWU2, the processing units in the first processing and switching unit PSWU1 can be used to generate second control and management information that the processing units in the second processing and switching unit PSWU2 need to send, and send the information to the processing units in the second processing and switching unit PSWU2. In other words, the processing units in the second processing and switching unit PSWU2 can reuse the processing units in the first processing and switching unit PSWU1 to generate the second control and management information.

[0109] Alternatively, for the first processing and switching unit PSWU1 and the second processing and switching unit PSWU2, the processing unit in the first processing and switching unit PSWU1 can also be used to send the second control and management information to the switching unit in the second processing and switching unit PSWU2. In other words, the processing unit in the first processing and switching unit PSWU1 can reuse the switching unit in the second processing and switching unit PSWU2 to send the second control and management information.

[0110] Optionally, when there are multiple paths for sending the second control management information corresponding to a certain NP grain, the multiple processing and switching units can distribute the second control management information among at least two of the multiple paths for sending, or specify one of the multiple paths for sending.

[0111] For example, taking the example of sending the second control management information corresponding to the second NP grain 22 to the second NP grain 22, the first processing exchange unit PSWU1 sends the second control management information to the second NP grain 22 through the first path where the first NP grain 21 is located, and the second processing exchange unit PSWU2 sends the second control management information to the second NP grain 22 through the second path where the third NP grain 23 is located.

[0112] Among them, when the multiple processing and switching units share and send the second control management information corresponding to the above-mentioned second NP grain 22 through at least two paths, the multiple processing and switching units can share and send it in a load balancing manner, or can adaptively share and send it according to the transmission conditions of different paths. The embodiment of the present application does not impose specific restrictions on this.

[0113] Furthermore, as shown in Figures 11 and 12, any of the aforementioned multiple NP die 20 may include a selection switch. Taking the first NP die 21 and the second NP die 22 as examples, the selection switch may include a connection terminal Q0 and at least one selection terminal. For example, the at least one selection terminal includes a first selection terminal Q1 and a second selection terminal Q2. In the first NP die 21, the connection terminal Q0 of the selection switch is connected to the processing switch die 10, and the first selection terminal Q1 and the second selection terminal Q2 of the selection switch are respectively connected to the external port of the first NP die 21 and the second NP die 22. In the second NP die 22, the connection terminal Q0 of the selection switch is connected to the first NP die 21, and the first selection terminal Q1 and the second selection terminal Q2 of the selection switch are respectively connected to the external port of the second NP die 22 and the third NP die 23. Figure 11 takes the example of the processing switch crystal 10 including a processing unit PU and multiple switching units 12 (for example, switching unit SWU1 to switching unit SWU4) for explanation, and Figure 12 takes the example of the processing switch crystal 10 including multiple processing switch units (for example, processing switch unit PSWU1 to processing switch unit PSWU4) for explanation.

[0114] The following takes the example of the processing exchange chip 10 sending the second control management information to the first NP chip 21 and the second NP chip 22 respectively as an example to illustrate the control process of the selection switch in the NP chip.

[0115] For example, as shown in Figures 11 and 12, when the processing exchange grain 10 is used to send the second control management information to the first NP grain 21, the first NP grain 21 can control the selection switch in the first NP grain 21 to turn on the first selection end Q1, so that the processing exchange grain 10 sends the second control management information to the first NP grain 21; when the processing exchange grain 10 is used to send the second control management information to the second NP grain 22, the first NP grain 21 can control the selection switch in the first NP grain 21 to turn on the second selection end Q2 corresponding to the second NP grain 22, and at the same time, the second NP grain 22 can control the selection switch in the second NP grain 22 to turn on the connection end Q0, so that the processing exchange grain 10 sends the second control management information to the second NP grain 22 through the first NP grain 21.

[0116] Optionally, the multiple NP crystals 20 may include other functional units in addition to the selection switch. Exemplarily, the first NP crystal 21 and the second NP crystal 22 may also include a media access control (MAC) aggregation unit, a network processor (NP), and a traffic manager (TM). The MAC aggregation unit may be referred to as a MAG. The MAG may be used to sense optical or electrical signals, forward the optical or electrical signals into data frames (i.e., messages), and perform a validity check on the data frames. The NP's uplink processing may include message parsing, data flow analysis, message forwarding, and mapping MAC / IP addresses to TM queue numbers; the NP's downlink processing may include message encapsulation and egress processing. The TM may be used to provide functions such as large-capacity memory, congestion avoidance, traffic shaping, multi-stage traffic scheduling, and traffic congestion control. It may also be used to map the TM queue number to the identifier of the destination NP crystal.

[0117] Furthermore, in practical applications, the plurality of NP grains 20 may include a greater number of other NP grains in addition to the first NP grain 21 and the second NP grain 22. FIG. 11 and FIG. 12 illustrate an example in which the plurality of NP grains 20 includes eight NP grains, each of which includes a selector switch, a MAG, an NP, and a TM.

[0118] Optionally, when the NP grain includes multiple different functional units, such as a selector switch, a media access control convergence unit (MAG), a network processor (NP), and a traffic manager (TM), the NP grain can also be split into multiple different grains, each of which is used to correspond to one or more of the above functional units. For example, as shown in FIG13 in conjunction with FIG11 , each NP grain can include two grains, including one corresponding to the MAG and one corresponding to the NP, TM, and selector switch.

[0119] The structures of the multiple NP grains in the embodiments of the present application may be identical or different, and the corresponding processes may be identical or different. When the structures of the multiple NP grains are identical, the multiple NP grains may be identical NP grains obtained through a single tape-out or the same tape-out process. By arranging and rotating the multiple NP grains, NP grains at different locations in the communication chip can be obtained, which can significantly reduce the development cost of the NP grains.

[0120] Similarly, the processing exchange die 10 can be split into multiple different dies, and the processes used in the different dies can be the same or different. For example, the input and output interfaces for communication can be integrated into a separate die, and other functions can be integrated into another die. This embodiment of the present application does not impose specific limitations on this.

[0121] Furthermore, optionally, the control and management functions implemented by at least one processing unit 11 in the processing switching crystal 10 provided above are merely exemplary. In actual applications, the control and management functions in the LPU and SFU may be offloaded to at least one processing unit in the processing switching crystal 10, while retaining the control and management functions in the MPU.

[0122] In addition, at least one processing unit 11 in the above-mentioned processing and switching grain 10 may include a CPU, and may also include DPU, DPU, DSP and processing units related to AI services, etc.; in addition to modules such as MAG, NP and TM, the above-mentioned NP grain may also deploy modules corresponding to capabilities such as flexible Ethernet (FlexE) and media access control security (MACSEC) as needed, and may also deploy DPU, DPU, DSP and processing units related to AI services, etc. as needed according to scenario requirements.

[0123] An embodiment of the present application further provides a communication chip, as shown in FIG14 , which includes at least one processing switch die 10 as provided above, and a plurality of NP die 20 coupled to the at least one processing switch die 10. FIG14 illustrates an example in which the at least one processing switch die 10 includes one processing switch die 10, and the plurality of NP die 20 includes eight NP die, which can be represented as NP die 21 to NP die 28, respectively.

[0124] Each of the at least one processing switch die 10 may be configured to control and manage its own switch unit, as well as control and manage some or all of the NP die 20. For example, each processing switch die may be configured to control and manage one or more NP die directly or indirectly coupled to the processing switch die 10.

[0125] Optionally, when the at least one process exchange die 10 includes a plurality of process exchange die, any one of the plurality of process exchange die may reuse another process exchange die.

[0126] In one possible embodiment, as shown in FIG15 , the at least one process exchange die 10 includes a first process exchange die 101 and a second process exchange die 102, and the first process exchange die 101 is directly connected to the second process exchange die 102. Exemplarily, the plurality of NP die 20 may include four NP die, which may be represented as NP die 21 to NP die 24, respectively.

[0127] In another possible embodiment, as shown in FIG16 , the at least one process exchange die 10 includes four process exchange die, which are represented as process exchange die 101 to process exchange die 104, respectively. Any two adjacent process exchange die among the four process exchange die are coupled. For example, the plurality of NP die 20 may include 12 NP die, which may be represented as NP die 21 to NP die 32, respectively.

[0128] Furthermore, the communication chip may also include: at least one storage grain, any one of the multiple NP grains is connected to one or more storage grains in the at least one storage grain. Exemplarily, in combination with Figure 14, as shown in Figure 17, the at least one storage grain includes 8 storage grains, and each of the 8 NP grains is connected to a storage grain. Optionally, the storage medium type of the storage grain includes but is not limited to: dynamic random access memory (DRAM), static random access memory (SRAM), ferroelectric random access memory (FeRAM) or magnetic random access memory (MRAM), etc. The figure takes the storage medium type of the storage grain as DRAM as an example for illustration.

[0129] Among them, when the at least one storage grain includes multiple storage grains, the storage medium types of the multiple storage grains can be the same or different; when an NP grain is coupled with multiple storage grains, the storage medium types of the multiple storage grains coupled to the NP grain can be the same or different, and the embodiments of the present application do not impose specific restrictions on this.

[0130] Optionally, the storage grains can be connected through 3D stacking or external connection, for example, through 2.5D sealing, board-level interconnection, or a hybrid of 2.5D and 3D. The embodiments of the present application do not impose specific restrictions on this.

[0131] In another embodiment of the present application, a communication chip is provided. As shown in FIG18 , the communication chip may include a processing switch die 10 and multiple NP die 20 (e.g., NP die 21 to NP die 28). The processing switch die 10 may include a processing unit PU and multiple switching units. Each of the multiple NP die may also include a processing unit PU. The processing unit PU in the processing switch die 10 and the processing units PU in the multiple NP die may be used together to control and manage the communication chip. This control and management method may be referred to as multi-level processing unit management and exclusive computing resource sharing.

[0132] In one possible example, the processing unit PU in the processing switch die 10 can be used to control and manage the multiple switching units, and the processing units PU in the multiple NP die can be used to control and manage the multiple NP die. Optionally, the processing switch die 10 can also manage the processing units PU in the multiple NP die.

[0133] In another embodiment of the present application, a communication chip is provided. As shown in FIG19 , the communication chip may include a switch die 30 and multiple NP die 20 (e.g., NP die 21 to NP die 28). The switch die 30 may include multiple switch units. Each of the multiple NP die may also include a processing unit (PU). The processing units (PU) in the multiple NP die may be collectively used to control and manage the communication chip. This control and management method can be referred to as single-level management of processing units and a distributed exclusive use of computing resources.

[0134] In one possible example, the processing units PU in the multiple NP grains can be used to control and manage the multiple switching units and the multiple NP grains. For example, the processing unit PU of any one of the multiple NP grains can be used to control and manage its own NP grain; further, it can also be used to control and manage other NP grains and the switching units in the switching grain 30. Optionally, the processing units PU in any two adjacent NP grains in the multiple NP grains can communicate with each other, for example, to transmit control and management information.

[0135] It is understandable that the detailed description above regarding the control and management of the processing switch die 10 can be referred to in the corresponding embodiments of the communication chip, and the embodiments of the present application will not be repeated here.

[0136] Based on this, embodiments of the present application further provide a data processing device, comprising: a backplane; and at least one communication chip disposed on the backplane. The communication chip can be any of the communication chips provided above. Optionally, the data exchange device can be a router, a switching box, a switching frame, or a switch, etc., which embodiments of the present application do not specifically limit.

[0137] It can be understood that the detailed descriptions of the processing exchange die and the NP die above can be referred to in the corresponding embodiments of the communication chip and the data processing device, and the embodiments of the present application will not be repeated here.

[0138] Based on this, embodiments of the present application also provide a control management method that can be applied to a communication chip, the communication chip including a processing switch die and multiple network processing die coupled to the processing switch die, the processing switch die including at least one processing unit and multiple switching units. The method includes: the at least one processing unit sending first control management information to the multiple switching units; and the at least one processing unit sending second control management information to the multiple network processing die via the multiple switching units.

[0139] The control and management information includes control and management information of the control plane and / or the management plane. Exemplarily, the at least one processing unit can be used to control and manage functions such as protocol processing, service processing, routing operations, message forwarding, and traffic statistics of the multiple switching units and the multiple NP grains, and can also be used to control and manage functions such as operation status monitoring, environmental monitoring, log and alarm information processing, system loading, and system upgrades of the communication chip.

[0140] Exemplarily, the first control management information may include at least one of the following: first configuration information for configuring the switching unit, data packet transmission control information, and device management information sent based on multiple received device status information. The data packet transmission control information may be obtained based on configuration or generated based on a triggering signaling message. The signaling message may be received by the NP crystal from the outside and forwarded to the processing unit PU via the switching unit. The multiple device status information may be status information of devices monitored by the NP crystal and, for example, fan modules and power supply modules coupled to the NP crystal. This device status information may also be referred to as maintenance information.

[0141] Exemplarily, the second control management information may include at least one of the following: second configuration information for configuring the NP grain, data packet transmission control information, and device management information sent based on multiple received device status information. The data packet transmission control information may be obtained based on configuration or generated based on a triggering signaling message. The signaling message may be received by the NP grain from the outside and forwarded to the processing unit PU via a switching unit. The multiple device status information may include status information of the NP grain and devices monitored by the NP grain, such as fan modules and power supply modules coupled to the NP grain. This device status information may also be referred to as maintenance information.

[0142] Optionally, in an example, the first control management information of each switching unit includes first configuration information for configuring the switching unit, and the second control management information of each network processing crystal includes second configuration information for configuring the network processing crystal.

[0143] Optionally, in another example, the second control management information also includes transmission control information of the data packet, and the method also includes: any one of the multiple network processing grains sends a signaling message to the at least one processing unit through the multiple switching units, and the signaling message is used to trigger the at least one processing unit to send the transmission control information.

[0144] Optionally, in another example, the second control management information also includes device management information, and the method also includes: any one of the multiple network processing grains sends device status information to the at least one processing unit through the multiple switching units, and the device status information is used for the at least one processing unit to send the device management information.

[0145] Further, in a possible embodiment, the at least one processing unit includes a processing unit, the multiple switching units are all coupled to the processing unit, and the first switching unit among the multiple switching units is directly connected to the first network processing grain among the multiple network processing grains; the at least one processing unit sends the second control management information to the multiple network processing grains through the multiple switching units, including: the processing unit sends the second control management information to the first network processing grain through the first switching unit.

[0146] In another possible embodiment, the second network processing grain among the multiple network processing grains is directly connected to the first network processing grain, and the second network processing grain is not directly connected to the multiple switching units; the at least one processing unit sends the second control management information to the multiple network processing grains through the multiple switching units, and also includes: the processing unit sends the second control management information to the second network processing grain through the first switching unit and the first network processing grain.

[0147] In another possible embodiment, the second switching unit among the multiple switching units is directly connected to the third network processing grain among the multiple network processing grains, and the third network processing grain is directly connected to the second network processing grain; the at least one processing unit sends the second control management information to the multiple network processing grains through the multiple switching units, and also includes: the processing unit sends the second control management information to the second network processing grain through the second switching unit and the third network processing grain.

[0148] Further, in a possible embodiment, the at least one processing unit includes multiple processing units, the multiple processing units and the multiple switching units correspond one-to-one and are integrated into multiple processing switching units, and any two of the multiple processing switching units are coupled; the at least one processing unit sends the second control management information to the multiple network processing grains through the multiple switching units, including: the multiple processing switching units send the second control management information to the multiple network processing grains.

[0149] In another possible embodiment, the first processing exchange unit among the multiple processing exchange units is directly connected to the first network processing grain among the multiple network processing grains; the multiple processing exchange units send second control management information to the multiple network processing grains, including: the first processing exchange unit sends the second control management information to the first network processing grain.

[0150] In another possible embodiment, the second network processing grain among the multiple network processing grains is directly connected to the first network processing grain, and the second network processing grain is not directly connected to the multiple processing switching units; the multiple processing switching units send the second control management information to the multiple network processing grains, and also includes: the first processing switching unit sends the second control management information to the second network processing grain through the first network processing grain.

[0151] Optionally, the second processing exchange unit among the multiple processing exchange units is directly connected to the third network processing grain among the multiple network processing grains, and the third network processing grain is directly connected to the second network processing grain; the multiple processing exchange units send second control management information to the multiple network processing grains, and also include: the second processing exchange unit sends the second control management information to the second network processing grain through the third network processing grain.

[0152] Optionally, the multiple processing and switching units include a first processing and switching unit and a second processing and switching unit; the at least one processing unit sends first control management information to the multiple switching units, including: the processing unit in the first processing and switching unit sends the first control management information to the switching unit in the second processing and switching unit.

[0153] In an embodiment of the present application, a control and management method is provided for controlling and managing a communication chip based on network-based computing (i.e., network-based and computing-assisted). The communication chip includes a processing and switching crystal and multiple NP crystals. In this method, the processing and switching crystal controls and manages the communication chip and provides computing power, and the multiple NP crystals can be used to provide network interfaces and network processing capabilities. Therefore, the communication chip and control and management method provided in the embodiment of the present application can be applied to network devices with network processing capabilities as the core. Compared with the existing technology, they can significantly reduce the number of processing units in the network device, reduce equipment costs, improve equipment energy efficiency, and save heat dissipation and power supply costs and deployment space, thereby achieving green, low-carbon and sustainable development.

[0154] The method or device provided in the embodiment of the present application can be implemented in whole or in part by software, hardware or a combination thereof. When implemented using software, it can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, the process or function described in accordance with the embodiment of the present application is generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, a network device or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions can be transmitted from a website, computer, server or data center to another website, computer, server or data center by wired (e.g., coaxial cable, optical fiber, twisted pair) or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any medium that a computer can access or a data storage device such as a server or data center that includes one or more media integrations. The medium can be a magnetic medium (e.g., a floppy disk, a hard disk, a tape), an optical medium (e.g., an optical disk), or a semiconductor medium (e.g., a solid-state drive (SSD)).

[0155] In another aspect of the present application, a computer-readable storage medium is provided, which includes computer instructions. When the computer instructions are executed by a device, the device executes the step of processing the exchange grain in the control management method provided by the above method embodiment.

[0156] In another aspect of the present application, a computer-readable storage medium is provided, which includes computer instructions. When the computer instructions are executed by a device, the device executes the NP grain steps in the control management method provided in the above method embodiment.

[0157] In another aspect of the present application, a computer program product comprising instructions is provided, which, when executed on a computer, enables the computer to execute the step of processing the exchange of dies in the control management method provided in the above method embodiment.

[0158] In another aspect of the present application, a computer program product comprising instructions is provided, which, when executed on a computer, enables the computer to execute the steps of controlling NP grains in the control management method provided in the above method embodiment.

[0159] Finally, it should be noted that the above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any changes or substitutions within the technical scope disclosed in this application should be included in the scope of protection of this application. Therefore, the scope of protection of this application should be based on the scope of protection of the claims.

Claims

1. A method for processing and exchanging grains, characterized in that: The processing exchange crystal grain is used to couple with a plurality of network processing crystal grains; The processing and switching crystal includes at least one processing unit and a plurality of switching units, and the at least one processing unit is used to control and manage the plurality of switching units and the plurality of network processing crystals.

2. The process exchange die according to claim 1, characterized in that: The at least one processing unit includes a processing unit, and the plurality of switching units are coupled to the processing unit; The processing unit is used to send first control management information to the multiple switching units; The processing unit is further configured to send second control management information to the plurality of network processing grains through the plurality of switching units.

3. The process exchange die according to claim 2, characterized in that: A first switching unit among the plurality of switching units is directly connected to a first network processing crystal grain among the plurality of network processing crystal grains; The processing unit is further configured to send the second control management information to the first network processing chip through the first switching unit.

4. The process exchange die according to claim 3, characterized in that: A second network processing crystal grain among the plurality of network processing crystal grains is directly connected to the first network processing crystal grain, and the second network processing crystal grain is not directly connected to the plurality of switching units; The processing unit is further configured to send the second control management information to the second network processing crystal grain through the first switching unit and the first network processing crystal grain.

5. The process exchange die according to claim 4, characterized in that: A second switching unit among the plurality of switching units is directly connected to a third network processing crystal grain among the plurality of network processing crystal grains, and the third network processing crystal grain is directly connected to the second network processing crystal grain; The processing unit is further configured to send the second control management information to the second network processing crystal grain through the second switching unit and the third network processing crystal grain.

6. The process exchange die according to claim 5, characterized in that: The processing unit is further used to distribute and send the second control management information to the second network processing grain through the first switching unit and the second switching unit.

7. The process exchange die according to claim 1, characterized in that: The at least one processing unit includes a plurality of processing units, the plurality of processing units correspond to the plurality of switching units one by one and are integrated into a plurality of processing switching units, and any two processing switching units among the plurality of processing switching units are coupled; The plurality of processing exchange units are used to send second control management information to the plurality of network processing grains.

8. The process exchange die according to claim 7, characterized in that: A first processing exchange unit among the plurality of processing exchange units is directly connected to a first network processing crystal grain among the plurality of network processing crystal grains; The first processing exchange unit is used to send the second control management information to the first network processing grain.

9. The process exchange die according to claim 8, characterized in that: A second network processing crystal grain among the plurality of network processing crystal grains is directly connected to the first network processing crystal grain, and the second network processing crystal grain is not directly connected to the plurality of processing switching units; The first processing exchange unit is further configured to send the second control management information to the second network processing crystal grain through the first network processing crystal grain.

10. The process exchange die according to claim 9, characterized in that: A second processing exchange unit among the plurality of processing exchange units is directly connected to a third network processing crystal grain among the plurality of network processing crystal grains, and the third network processing crystal grain is directly connected to the second network processing crystal grain; The second processing exchange unit is further configured to send the second control management information to the second network processing crystal grain through the third network processing crystal grain.

11. The process for exchanging dies according to any one of claims 7 to 10, characterized in that: The plurality of processing switching units include a first processing switching unit and a second processing switching unit; The processing unit in the first processing and switching unit is used to send the second control management information to the processing unit in the second processing and switching unit.

12. The process exchange die according to any one of claims 7 to 11, characterized in that: The plurality of processing switching units include a first processing switching unit and a second processing switching unit; The processing unit in the first processing and switching unit is further configured to send first control management information to the switching unit in the second processing and switching unit.

13. The process exchange die according to any one of claims 1 to 12, characterized in that: The control management includes control management of a control plane and / or a management plane.

14. A communication chip, characterized in that: The communication chip includes: at least one processing switch crystal, and a plurality of network processing crystals coupled to the at least one processing switch crystal, wherein the at least one processing switch crystal is the processing switch crystal according to any one of claims 1-13.

15. The communication chip according to claim 14, characterized in that: The at least one process exchange die includes a first process exchange die and a second process exchange die, and the first process exchange die is directly connected to the second process exchange die.

16. A data processing device, characterized in that: The data processing device comprises: a backplane, and at least one communication chip arranged on the backplane, and the communication chip is the communication chip according to any one of claims 14-15.

17. A control management method, characterized in that: Applied in a communication chip, the communication chip includes a processing switch crystal grain and a plurality of network processing crystal grains coupled to the processing switch crystal grain, the processing switch crystal grain includes at least one processing unit and a plurality of switching units, the method includes: The at least one processing unit sends first control management information to the plurality of switching units; The at least one processing unit sends second control management information to the plurality of network processing die through the plurality of switching units.

18. The method according to claim 17, characterized in that The first control management information of each switching unit includes first configuration information for configuring the switching unit, and the second control management information of each network processing crystal includes second configuration information for configuring the network processing crystal.

19. The method according to claim 17 or 18, characterized in that The second control management information also includes transmission control information of the data message, and the method further includes: Any network processing crystal of the plurality of network processing crystals sends a signaling message to the at least one processing unit through the plurality of switching units, where the signaling message is used to trigger the at least one processing unit to send the transmission control information.

20. The method according to any one of claims 17 to 19, characterized in that: The second control management information also includes device management information, and the method further includes: Any network processing crystal of the plurality of network processing crystals sends device status information to the at least one processing unit through the plurality of switching units, and the device status information is used for the at least one processing unit to send the device management information.

21. The method according to any one of claims 17 to 20, characterized in that: The at least one processing unit includes a processing unit, the multiple switching units are coupled to the processing unit, and a first switching unit among the multiple switching units is directly connected to a first network processing grain among the multiple network processing grains; The at least one processing unit sends the second control management information to the plurality of network processing grains through the plurality of switching units, including: The processing unit sends the second control management information to the first network processing chip through the first switching unit.

22. The method according to claim 21, characterized in that A second network processing crystal grain among the plurality of network processing crystal grains is directly connected to the first network processing crystal grain, and the second network processing crystal grain is not directly connected to the plurality of switching units; The at least one processing unit sends the second control management information to the plurality of network processing grains through the plurality of switching units, further comprising: The processing unit sends the second control management information to the second network processing crystal grain through the first switching unit and the first network processing crystal grain.

23. The method according to claim 21 or 22, characterized in that A second switching unit among the plurality of switching units is directly connected to a third network processing crystal grain among the plurality of network processing crystal grains, and the third network processing crystal grain is directly connected to the second network processing crystal grain; The at least one processing unit sends the second control management information to the plurality of network processing grains through the plurality of switching units, further comprising: The processing unit sends the second control management information to the second network processing crystal grain through the second switching unit and the third network processing crystal grain.

24. The method according to any one of claims 17 to 20, characterized in that: The at least one processing unit includes a plurality of processing units, the plurality of processing units correspond to the plurality of switching units one by one and are integrated into a plurality of processing switching units, and any two processing switching units among the plurality of processing switching units are coupled; The at least one processing unit sends the second control management information to the plurality of network processing grains through the plurality of switching units, including: The plurality of processing switching units send second control management information to the plurality of network processing die.

25. The method according to claim 24, characterized in that The first processing exchange unit among the plurality of processing exchange units is directly connected to the first network processing crystal grain among the plurality of network processing crystal grains; the plurality of processing exchange units send second control management information to the plurality of network processing crystal grains, including: The first processing switch unit sends the second control management information to the first network processing die.

26. The method according to claim 25, characterized in that A second network processing crystal grain among the plurality of network processing crystal grains is directly connected to the first network processing crystal grain, and the second network processing crystal grain is not directly connected to the plurality of processing switching units; The plurality of processing switching units sending second control management information to the plurality of network processing grains further includes: The first processing exchange unit sends the second control management information to the second network processing crystal grain through the first network processing crystal grain.

27. The method according to claim 25 or 26, characterized in that The second processing exchange unit among the plurality of processing exchange units is directly connected to the third network processing crystal grain among the plurality of network processing crystal grains, and the third network processing crystal grain is directly connected to the second network processing crystal grain; the plurality of processing exchange units send second control management information to the plurality of network processing crystal grains, further comprising: The second processing exchange unit sends the second control management information to the second network processing crystal grain through the third network processing crystal grain.

28. The method according to any one of claims 24 to 27, characterized in that: The multiple processing and switching units include a first processing and switching unit and a second processing and switching unit; the at least one processing unit sends the first control management information to the multiple switching units, including: The processing unit in the first processing and switching unit sends the first control management information to the switching unit in the second processing and switching unit.

29. The method according to any one of claims 17 to 28, characterized in that: The control management information includes control management information of a control plane and / or a management plane.

Images