Intelligent calculation network structure and construction method of intelligent calculation network structure

Abstract

The invention discloses an intelligent calculation network structure and a construction method of the intelligent calculation network structure. The intelligent computing network structure provided by the specification comprises a core layer, a convergence layer, an access layer and a client, in the intelligent computing network structure, each core layer is connected with each convergence layer, each convergence layer is connected with at least part of access layers according to a preset rule, each access layer is connected with at least part of convergence layers according to the preset rule, and each client is connected with one access layer; wherein the core layer is connected with the convergence layer in an electric exchange mode, the convergence layer is connected with the access layer in an electric exchange mode, and the access layer is connected with the client side in an optical transmission mode. By adopting the structure, the fault rate of the optical link can be reduced, and the construction cost of the intelligent computing network structure is greatly reduced.

Description

Technical Field

[0001] The present invention relates to the field of communication technology, and in particular to an intelligent computing network structure and a method for constructing the intelligent computing network structure. Background Art

[0002] The development of artificial intelligence (AI) technology has driven the upgrade of large industrial models, but also brought about the demand for massive computing power and energy. Intelligent computing network is an AI-oriented application that takes intelligent computing as its core and combines advanced network technology to achieve the infrastructure for efficient scheduling, allocation and utilization of computing resources.

[0003] In the current construction of intelligent computing centers, clusters of thousands or tens of thousands of cards are already common, and even large-scale clusters of more than ten thousand cards are being actively deployed. However, the current intelligent computing network not only has huge deployment and use costs, but also has a high failure rate, which has caused considerable difficulties for the further development and breakthrough of AI-related technical applications.

[0004] Based on this, how to build a more stable and reliable intelligent computing network at low cost is an urgent problem to be solved. Summary of the invention

[0005] This specification provides an intelligent computing network structure and a method for constructing an intelligent computing network structure to at least partially solve the above-mentioned problems existing in the prior art.

[0006] This manual adopts the following technical solutions:

[0007] This specification provides an intelligent computing network structure, which includes a core layer, a convergence layer, an access layer, and a client;

[0008] In the intelligent computing network structure, each core layer is interconnected with each aggregation layer, each aggregation layer is interconnected with at least part of the access layer according to a preset rule, each access layer is interconnected with at least part of the aggregation layer according to the preset rule, and each client is interconnected with an access layer;

[0009] The core layer and the convergence layer are connected as well as the convergence layer and the access layer are connected in the form of electrical exchange, and the access layer and the client are connected in the form of optical transmission.

[0010] Optionally, the core layer, the convergence layer and the access layer constitute an electrical exchange domain, and the convergence layer and the access layer constitute an optical transmission domain;

[0011] For any two network layers that need to be connected to each other in the electrical switching domain, if the distance between the two network layers is not greater than a specified distance, the two network layers are connected in the form of electrical switching;

[0012] If the distance between the two network layers is greater than the specified distance, the two network layers are connected in the form of optical transmission.

[0013] Optionally, the access layer includes a retimer component;

[0014] At least one end of the retimer assembly is used for an electrical switching connection between the access layer and the convergence layer, and at least one end is used for an optical transmission connection between the access layer and the client.

[0015] Optionally, in the optical transmission connection between the access layer and the client, one end of the optical cable used in the optical transmission connection is fixed to the access layer;

[0016] In the electrical switching connection between the access layer and the convergence layer, the electrical switching connection is connected to the access layer in a pluggable interface form.

[0017] This specification provides a method for constructing an intelligent computing network structure, wherein the intelligent computing network structure includes a core layer, a convergence layer, an access layer, and a client, and the method includes:

[0018] For each core layer, the core layer is connected to each aggregation layer in the form of electrical switching;

[0019] For each convergence layer, the convergence layer is connected to at least part of the access layer in the form of electrical switching according to a preset rule;

[0020] For each access layer, optical transmission is used to connect the access layer and the client covered by the access layer.

[0021] Optionally, the core layer and each convergence layer are connected in the form of electrical switching, specifically including:

[0022] If the distance between the core layer and a convergence layer is not greater than a specified distance, the core layer and the convergence layer are connected in the form of electrical switching;

[0023] If the distance between the core layer and a convergence layer is greater than the specified distance, the core layer and the convergence layer are connected in the form of optical transmission;

[0024] Connecting the convergence layer and at least part of the access layer in the form of electrical switching according to a preset rule specifically includes:

[0025] Determine the access layer that needs to be connected to the core layer as the target access layer according to a preset rule;

[0026] If the distance between the convergence layer and a target access layer is not greater than the specified distance, connecting the convergence layer and the target access layer in the form of electrical switching;

[0027] If the distance between the convergence layer and a target access layer is greater than the specified distance, the convergence layer and the target access layer are connected in the form of electrical switching.

[0028] Optionally, connecting the convergence layer and at least part of the access layer in the form of electrical switching according to a preset rule specifically includes:

[0029] Connecting the convergence layer and at least part of the access layer in a pluggable electrical exchange format according to preset rules;

[0030] The access layer is connected to the client covered by the access layer in the form of optical transmission, specifically including:

[0031] The access layer is connected with the client covered by the access layer in the form of optical transmission in which one end of the optical cable is fixed on the access layer.

[0032] This specification provides a device for constructing an intelligent computing network structure, the intelligent computing network structure includes a core layer, a convergence layer, an access layer and a client, and the device includes:

[0033] The core building module is used to connect each core layer with each aggregation layer by means of electrical switching;

[0034] A convergence construction module, used for connecting each convergence layer with at least part of the access layer in the form of electrical switching according to a preset rule;

[0035] The access construction module is used to connect each access layer with the client covered by the access layer in the form of optical transmission.

[0036] This specification provides a computer-readable storage medium, which stores a computer program. When the computer program is executed by a processor, it implements the method for constructing the above-mentioned intelligent computing network structure.

[0037] This specification provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the method for constructing the above-mentioned intelligent computing network structure when executing the program.

[0038] At least one of the above technical solutions adopted in this specification can achieve the following beneficial effects:

[0039] The intelligent computing network structure provided in the present specification includes a core layer, an aggregation layer, an access layer, and a client; in the intelligent computing network structure, each core layer is interconnected with each aggregation layer, each aggregation layer is interconnected with at least part of the access layer according to a preset rule, each access layer is connected to at least part of the aggregation layer according to the preset rule, and each client is connected to an access layer; wherein the core layer and the aggregation layer and the aggregation layer and the access layer are connected in the form of electrical exchange, and the access layer and the client are connected in the form of optical transmission.

[0040] When using the intelligent computing network structure provided in this specification, based on the concept of "electrical in, optical out", the connection between the core layer and the aggregation layer, as well as between the aggregation layer and the access layer, is optimized to an electrical exchange form of connection. Compared with the all-optical interconnection solution, this intelligent computing network structure reduces the number of optical modules used exponentially, which not only reduces the failure rate of the optical link, but also greatly reduces the construction cost of the system structure. At the same time, since the signal processing is done in an electrical way, there is no need for energy conversion in the electrical interconnection, which improves power utilization and reduces the power consumption of the entire system. In addition, the exchange area of ​​this intelligent computing network structure is concentrated, which is convenient for building a global management network. At the same time, due to the increased reliability of the system, the frequency of operation and maintenance is reduced, and it can adapt to the networking solutions of PODs of various sizes. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] The drawings described herein are used to provide a further understanding of this specification and constitute a part of this specification. The illustrative embodiments and descriptions of this specification are used to explain this specification and do not constitute an improper limitation on this specification. In the drawings:

[0042] Figure 1 A schematic diagram of a smart computing network structure provided in this specification;

[0043] Figure 2 A flowchart of a method for constructing an intelligent computing network structure provided in this specification;

[0044] Figure 3 A schematic diagram of a device for constructing an intelligent computing network structure provided in this specification;

[0045] Figure 4 The corresponding Figure 2 Schematic diagram of electronic equipment. DETAILED DESCRIPTION

[0046] The current intelligent computing center's GPU servers generally use the Spine-Leaf network architecture based on RoCE V2. In the Spine-leaf architecture, any server traffic across leafs will only go through two hops, three devices, to ensure that the delay is predictable. Among them, each leaf switch is connected to each spine switch to achieve full interconnection. The Spine-leaf architecture uses a unified three-layer network and equal-cost multi-path routing, which can fully utilize all link bandwidths and achieve load balancing. This design is very suitable for combination with cloud controllers and other management tools to perform unified configuration distribution and other automated network operation and maintenance management, reducing network complexity and management difficulty.

[0047] In actual deployment, the RoCE V2 protocol is usually used, and the switch adopts a traditional Ethernet switch. According to different networking solutions, frame-type or box-type switches can be used for combination. The switches at different layers are interconnected by optical fiber.

[0048] Due to the "synchronous" communication characteristics in large model applications, the short board effect will be greatly magnified. Therefore, the intelligent computing cluster is required to have high reliability. However, the high-performance RoCE network based on optical transmission has become the main source of failure in large-scale clusters, especially the failure of optical modules, which accounts for about 30% of the entire cluster failure. This is mainly because optical transmission is extremely sensitive to temperature, but the intelligent computing cluster is a scene with huge power consumption and concentrated heat dissipation, which will inevitably lead to aging and failure problems of optical modules. On the other hand, due to the flexible plug-in and unplug-out characteristics of optical modules, and the high-precision requirements of optical transmission on the optical fiber system, factors such as manual plug-in and unplugging, vibration in the computer room, and wiring interference may cause failure problems in the optical transmission system.

[0049] On the other hand, the photoelectric conversion efficiency is currently only about 40%. Huge energy losses will be caused in the photoelectric conversion process. And most of this lost energy is converted into heat energy, which poses a great challenge to the heat dissipation of the optical module itself. Since optical modules are widely used in intelligent computing clusters, their low energy efficiency affects the PUE index of the entire data center. In addition, optical modules have a high cost. As the scale of intelligent computing clusters increases, the number of optical modules increases by several times, becoming the main source of cost in intelligent computing networks.

[0050] To solve the above problems, this specification proposes a low-cost, high-reliability intelligent computing network structure.

[0051] In order to make the purpose, technical solutions and advantages of this specification more clear, the technical solutions of this specification will be clearly and completely described below in combination with the specific embodiments of this specification and the corresponding drawings. Obviously, the described embodiments are only part of the embodiments of this specification, not all of the embodiments. Based on the embodiments in this specification, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0052] The technical solutions provided by the embodiments of this specification are described in detail below in conjunction with the accompanying drawings.

[0053] Figure 1 A schematic diagram of a smart computing network structure provided in this specification, such as Figure 1 As shown, the intelligent computing network structure provided in this specification includes a core layer, a convergence layer, an access layer and a client.

[0054] The intelligent computing network structure provided in this specification is also an intelligent computing network structure based on Spine-Leaf networking. It includes the core layer L3 (Core), the aggregation layer L2 (Spine), the access layer L1 (Leaf) and the client (Server). In the Spine-Leaf networking, the core layer carries the transmission function between the network server and each application port, and is the supporting backbone and data transmission channel of the entire network. The core switch provides high-speed forwarding for packets entering and leaving the data center, and provides connectivity for multiple aggregation layers. The core switch usually provides a flexible routing network for the entire network. The aggregation layer connects the core layer of the network and each access application layer, and plays the role of "media transmission" between the two layers. Before the application accesses the core layer, it is first processed by the aggregation layer to reduce the load on the core layer equipment. The aggregation layer provides services such as content switching, firewall, Secure Sockets Layer (SSL) unloading, intrusion detection, and network analysis. The access layer is mainly oriented to end customers and provides access functions for end customers. The main function of the access layer is to plan the number of workstations in the same network segment and increase the bandwidth of each access terminal.

[0055] like Figure 1 As shown, in the intelligent computing network structure, each core layer is interconnected with each aggregation layer, each aggregation layer is interconnected with at least part of the access layer according to a preset rule, each access layer is interconnected with at least part of the aggregation layer according to the preset rule, and each client is connected to an access layer; wherein the core layer and the aggregation layer and the aggregation layer and the access layer are connected in the form of electrical exchange, and the access layer and the client are connected in the form of optical transmission.

[0056] In the above connection mode, there may be multiple preset rules according to different specific requirements. Figure 1 The star connection method based on centralized switching is adopted, that is, every two aggregation layers and two access layers form a group, and the two aggregation layers in a group are connected to the two access layers in pairs. Through the above connection method, it is guaranteed that each client can communicate with each other, and the size of the topology structure is reduced to a certain extent.

[0057] At the same time, the intelligent computing network structure provided in this specification adopts the form of electrical exchange to connect between the core layer and the aggregation layer, and between the aggregation layer and the access layer, and adopts the form of optical transmission to connect between the access layer and the client. Compared with the traditional method of using optical transmission for connection, the method adopted in this specification saves the deployment cost of the intelligent computing network structure on the one hand. Under normal circumstances, the cost consumption of electrical modules is often only 1 / 10 of that of optical modules. Using electrical modules instead of optical modules can greatly save expenses; on the other hand, the method adopted in this specification also greatly improves the reliability of the intelligent computing network structure. As introduced in the background technology, under the influence of various reasons, the failure rate of optical modules is extremely high, generally reaching about 2% to 3%. Under the same environment, the failure rate of electrical modules is usually only one in ten thousand, which is much lower than that of optical modules.

[0058] In addition, under current technical conditions, the efficiency of photoelectric conversion is generally only about 40%. In other words, a large amount of energy is lost in the process of photoelectric conversion. The heat converted from the lost energy also poses a challenge to the heat dissipation of the entire intelligent computing network. This problem does not exist at all under electrical exchange.

[0059] It is undeniable that compared with optical transmission, the transmission efficiency of electrical switching is indeed lower. However, under normal circumstances, the core layer, aggregation layer, and access layer in the intelligent computing network structure are usually deployed in the same computer room, and the distance between them is very close. Under the condition of short-distance transmission, the speed advantage of optical transmission is no longer obvious. Compared with the various advantages brought by electrical switching described above, the impact of the tiny efficiency gap is completely acceptable.

[0060] In addition, for more special situations, the intelligent computing network structure provided in this specification also makes corresponding considerations. Specifically, the core layer, the convergence layer and the access layer constitute an electrical switching domain, and the convergence layer and the access layer constitute an optical transmission domain; for any two network layers that need to be interconnected in the electrical switching domain, if the distance between the two network layers is not greater than the specified distance, the two network layers are connected in the form of electrical switching; if the distance between the two network layers is greater than the specified distance, the two network layers are connected in the form of optical transmission.

[0061] In practical applications, the core layer, aggregation layer and access layer are all network layers in the intelligent computing network structure. In the electrical switching domain composed of the core layer, aggregation layer and access layer, for any two network layers that need to be connected to each other, the connection method can be changed according to their specific distance. When the two network layers are close to each other, the electrical switching method advocated by this specification is adopted for connection; when the two network layers are far apart, optical transmission can still be used to effectively ensure communication efficiency. Among them, the specific standard for judging the distance is whether the distance between the two is not greater than the specified distance. The specified distance can also be set according to specific needs. For example, it can generally be set to 5 meters.

[0062] At the same time, in the intelligent computing network structure provided in this specification, the access layer and the client are still connected by optical transmission. Clients are usually deployed in various places covered by the intelligent computing network structure. Therefore, the distance between each client and the access layer will be far. Based on this, the long-distance advantage of optical transmission is still used between the access layer and the client to solve the physical distribution requirements of the intelligent computing cluster.

[0063] In the intelligent computing network structure provided in the present specification, the access layer needs to be connected to the aggregation layer in the form of electrical switching, and also needs to be connected to the client in the form of optical transmission, which is equivalent to the boundary between the electrical switching domain and the optical transmission domain. Therefore, in the intelligent computing network structure provided in the present specification, it can also be specifically that the access layer includes a retimer component; at least one end of the retimer component is used for the electrical switching connection between the access layer and the aggregation layer, and at least one end is used for the optical transmission connection between the access layer and the client. Adding a retimer component to the access layer can play a role in converting between electrical switching and optical transmission, and can also extend the distance of electrical switching transmission.

[0064] Furthermore, due to the high-precision requirements of the optical fiber system, a large number of plugging and unplugging of interface components will inevitably cause problems such as failures in the optical module. To solve this problem, in the intelligent computing network structure provided in this specification, specifically, in the optical transmission connection between the access layer and the client, one end of the optical cable used for the optical transmission connection is fixed to the access layer; in the electrical switching connection between the access layer and the convergence layer, the electrical switching connection is connected to the access layer in the form of a pluggable interface. At the hardware level of the access layer, the interface of the electrical switching is designed as a pluggable interface, and the interface of the optical transmission is designed as a fixed, non-pluggable interface or a semi-fixed interface. In this way, the loss caused by frequent plugging and unplugging can be greatly reduced, thereby reducing the failure rate.

[0065] When using the intelligent computing network structure provided in this specification, based on the concept of "electrical in, optical out", the connection between the core layer and the aggregation layer, as well as between the aggregation layer and the access layer, is optimized to an electrical exchange form of connection. Compared with the all-optical interconnection solution, this intelligent computing network structure reduces the number of optical modules used exponentially, which not only reduces the failure rate of the optical link, but also greatly reduces the construction cost of the system structure. At the same time, since the signal processing is done in an electrical way, there is no need for energy conversion in the electrical interconnection, which improves power utilization and reduces the power consumption of the entire system. In addition, the exchange area of ​​this intelligent computing network structure is concentrated, which is convenient for building a global management network. At the same time, due to the increased reliability of the system, the frequency of operation and maintenance is reduced, and it can adapt to the networking solutions of PODs of various sizes.

[0066] The above is the intelligent computing network structure provided in this specification. Based on the same idea, this specification also provides a method for constructing a corresponding intelligent computing network structure, such as Figure 2 shown.

[0067] Figure 2 The following is a flow chart of a method for constructing an intelligent computing network structure in this specification. The intelligent computing network structure includes a core layer, a convergence layer, an access layer and a client. The method specifically includes the following steps:

[0068] S100: For each core layer, the core layer is connected to each aggregation layer in the form of electrical switching.

[0069] All steps in the method for constructing an intelligent computing network structure provided in this specification can be implemented by any electronic device with computing functions, such as terminals, servers and other devices.

[0070] Among them, when connecting the core layer and each convergence layer in the form of electrical exchange, specifically, if the distance between the core layer and a convergence layer is not greater than a specified distance, the core layer and the convergence layer are connected in the form of electrical exchange; if the distance between the core layer and a convergence layer is greater than the specified distance, the core layer and the convergence layer are connected in the form of optical transmission.

[0071] S102: For each convergence layer, connect the convergence layer and at least part of the access layer in the form of electrical switching according to a preset rule.

[0072] Among them, when connecting the aggregation layer and at least part of the access layer in the form of electrical exchange according to preset rules, specifically, the access layer that needs to be connected to the core layer can be determined according to the preset rules as the target access layer; if the distance between the aggregation layer and a target access layer is not greater than the specified distance, the aggregation layer and the target access layer are connected in the form of electrical exchange; if the distance between the aggregation layer and a target access layer is greater than the specified distance, the aggregation layer and the target access layer are connected in the form of electrical exchange.

[0073] The convergence layer and at least part of the access layer are connected in the form of electrical switching according to preset rules. Specifically, the convergence layer and at least part of the access layer are connected in the form of electrical switching with pluggable interfaces according to preset rules.

[0074] S104: For each access layer, connect the access layer and the client covered by the access layer by using optical transmission.

[0075] Wherein, when connecting the access layer and the client covered by the access layer in the form of optical transmission, the access layer and the client covered by the access layer can be connected in the form of optical transmission in which one end of the optical cable is fixed on the access layer.

[0076] Since the method for constructing the intelligent computing network structure provided in this specification is a supplementary explanation of the intelligent computing network structure provided in this specification, its specific content has been elaborated in detail in the intelligent computing network structure, and this specification will not repeat it here.

[0077] The above is the intelligent computing network structure and the method for constructing the intelligent computing network structure provided in this specification. Based on the same idea, this specification also provides a corresponding intelligent computing network structure construction device, such as Figure 3 shown.

[0078] Figure 3 A schematic diagram of a device for constructing an intelligent computing network structure provided in this specification, wherein the intelligent computing network structure includes a core layer, a convergence layer, an access layer, and a client, and the device includes:

[0079] The core building module 200 is used to connect each core layer with each convergence layer in the form of electrical switching;

[0080] The convergence construction module 202 is used to connect each convergence layer with at least part of the access layer in the form of electrical switching according to a preset rule;

[0081] The access construction module 204 is used to connect each access layer with the client covered by the access layer in the form of optical transmission.

[0082] Optionally, the core building module 200 is specifically configured to connect the core layer and the convergence layer in the form of electrical exchange if the distance between the core layer and the convergence layer is not greater than a specified distance; and connect the core layer and the convergence layer in the form of optical transmission if the distance between the core layer and the convergence layer is greater than the specified distance;

[0083] The aggregation construction module 202 is specifically used to determine the access layer that needs to be connected to the core layer as the target access layer according to a preset rule; if the distance between the aggregation layer and a target access layer is not greater than the specified distance, the aggregation layer and the target access layer are connected in the form of electrical switching; if the distance between the aggregation layer and a target access layer is greater than the specified distance, the aggregation layer and the target access layer are connected in the form of electrical switching.

[0084] Optionally, the convergence construction module 202 is specifically configured to connect the convergence layer and at least part of the access layer in a pluggable electrical exchange form according to a preset rule;

[0085] The access construction module 204 is specifically used to connect the access layer and the client covered by the access layer in the form of optical transmission in which one end of the optical cable is fixed on the access layer.

[0086] This specification also provides a computer-readable storage medium, which stores a computer program, which can be used to execute the above Figure 2 The provided method for constructing the intelligent computing network structure.

[0087] This manual also provides Figure 4 The schematic structure diagram of the electronic device shown in FIG. Figure 4 As mentioned above, at the hardware level, the electronic device includes a processor, an internal bus, a network interface, a memory, and a non-volatile memory, and may also include other hardware required for the business. The processor reads the corresponding computer program from the non-volatile memory into the memory and then runs it to achieve the above Figure 2 The method for constructing the intelligent computing network structure. Of course, in addition to the software implementation, this specification does not exclude other implementation methods, such as logic devices or a combination of software and hardware, etc., that is, the execution subject of the following processing flow is not limited to each logic unit, but can also be hardware or logic devices.

[0088] For the improvement of a technology, it can be clearly distinguished whether it is a hardware improvement (for example, improvement of the circuit structure of diodes, transistors, switches, etc.) or a software improvement (improvement of the method flow). However, with the development of technology, many improvements of the method flow today can be regarded as direct improvements of the hardware circuit structure. Designers almost always obtain the corresponding hardware circuit structure by programming the improved method flow into the hardware circuit. Therefore, it cannot be said that the improvement of a method flow cannot be implemented with a hardware entity module. For example, a programmable logic device (PLD) (such as a field programmable gate array (FPGA)) is such an integrated circuit whose logical function is determined by the user's programming of the device. Designers can "integrate" a digital system on a PLD by programming themselves, without having to ask chip manufacturers to design and make dedicated integrated circuit chips. Moreover, nowadays, instead of manually making integrated circuit chips, this kind of programming is mostly implemented by "logic compiler" software, which is similar to the software compiler used when developing and writing programs, and the original code before compilation must also be written in a specific programming language, which is called hardware description language (HDL). There is not only one HDL, but many kinds, such as ABEL (Advanced Boolean Expression Language), AHDL (Altera Hardware Description Language), Confluence, CUPL (Cornell University Programming Language), HDCal, JHDL (Java Hardware Description Language), Lava, Lola, MyHDL, PALASM, RHDL (Ruby Hardware Description Language), etc. The most commonly used ones are VHDL (Very-High-Speed ​​Integrated Circuit Hardware Description Language) and Verilog. Those skilled in the art should also know that it is only necessary to program the method flow slightly in the above-mentioned hardware description languages ​​and program it into the integrated circuit, and then it is easy to obtain the hardware circuit that implements the logic method flow.

[0089] The controller can be implemented in any appropriate manner, for example, the controller can take the form of a microprocessor or processor and a computer-readable medium storing a computer-readable program code (such as software or firmware) that can be executed by the (micro)processor, a logic gate, a switch, an application-specific integrated circuit (ASIC), a programmable logic controller, and an embedded microcontroller. Examples of controllers include, but are not limited to, the following microcontrollers: ARC 625D, Atmel AT91SAM, Microchip PIC18F26K20, and Silicone Labs C8051F320. The memory controller can also be implemented as part of the control logic of the memory. Those skilled in the art also know that in addition to implementing the controller in a purely computer-readable program code manner, the controller can be implemented in the form of a logic gate, a switch, an application-specific integrated circuit, a programmable logic controller, and an embedded microcontroller by logically programming the method steps. Therefore, this controller can be considered as a hardware component, and the devices included therein for implementing various functions can also be regarded as structures within the hardware component. Or even, the devices for implementing various functions can be regarded as both software modules for implementing the method and structures within the hardware component.

[0090] The systems, devices, modules or units described in the above embodiments may be implemented by computer chips or entities, or by products with certain functions. A typical implementation device is a computer. Specifically, the computer may be, for example, a personal computer, a laptop computer, a cellular phone, a camera phone, a smart phone, a personal digital assistant, a media player, a navigation device, an email device, a game console, a tablet computer, a wearable device, or a combination of any of these devices.

[0091] For the convenience of description, the above device is described in various units according to their functions. Of course, when implementing this specification, the functions of each unit can be implemented in the same or multiple software and / or hardware.

[0092] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment in combination with software and hardware. Moreover, the present application may adopt the form of a computer program product implemented in one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that include computer-usable program code.

[0093] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0094] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0095] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0096] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0097] The memory may include non-permanent storage in a computer-readable medium, random access memory (RAM) and / or non-volatile memory in the form of read-only memory (ROM) or flash RAM. The memory is an example of a computer-readable medium.

[0098] Computer readable media include permanent and non-permanent, removable and non-removable media that can be implemented by any method or technology to store information. Information can be computer readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, compact disk read-only memory (CD-ROM), digital versatile disk (DVD) or other optical storage, magnetic cassettes, magnetic tape magnetic disk storage or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device. As defined herein, computer readable media does not include temporary computer readable media (transitory media), such as modulated data signals and carrier waves.

[0099] It should also be noted that the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, commodity or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, commodity or device. In the absence of more restrictions, the elements defined by the sentence "comprises a ..." do not exclude the existence of other identical elements in the process, method, commodity or device including the elements.

[0100] Those skilled in the art will appreciate that the embodiments of this specification may be provided as methods, systems or computer program products. Therefore, this specification may take the form of a complete hardware embodiment, a complete software embodiment or an embodiment combining software and hardware. Moreover, this specification may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0101] This specification may be described in the general context of computer-executable instructions executed by a computer, such as program modules. Generally, program modules include routines, programs, objects, components, data structures, etc. that perform specific tasks or implement specific abstract data types. This specification may also be practiced in distributed computing environments where tasks are performed by remote processing devices connected through a communication network. In a distributed computing environment, program modules may be located in local and remote computer storage media, including storage devices.

[0102] Each embodiment in this specification is described in a progressive manner, and the same or similar parts between the embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the system embodiment, since it is basically similar to the method embodiment, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiment.

[0103] The above description is only an embodiment of this specification and is not intended to limit this specification. For those skilled in the art, this specification may have various changes and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of this specification should be included in the scope of the claims of this application.

Claims

1. An intelligent computing network structure, characterized in that: The intelligent computing network structure includes a core layer, a convergence layer, an access layer, and a client; In the intelligent computing network structure, each core layer is interconnected with each aggregation layer, each aggregation layer is interconnected with at least part of the access layer according to a preset rule, each access layer is interconnected with at least part of the aggregation layer according to the preset rule, and each client is interconnected with an access layer; The core layer and the convergence layer are connected as well as the convergence layer and the access layer are connected in the form of electrical exchange, and the access layer and the client are connected in the form of optical transmission.

2. The intelligent computing network structure according to claim 1, characterized in that: The core layer, the convergence layer and the access layer constitute an electrical exchange domain, and the convergence layer and the access layer constitute an optical transmission domain; For any two network layers that need to be connected to each other in the electrical switching domain, if the distance between the two network layers is not greater than a specified distance, the two network layers are connected in the form of electrical switching; If the distance between the two network layers is greater than the specified distance, the two network layers are connected in the form of optical transmission.

3. The intelligent computing network structure according to claim 1, characterized in that: The access layer includes a retimer component; At least one end of the retimer assembly is used for an electrical switching connection between the access layer and the convergence layer, and at least one end is used for an optical transmission connection between the access layer and the client.

4. The intelligent computing network structure according to claim 1, characterized in that: In the optical transmission connection between the access layer and the client, one end of the optical cable used in the optical transmission connection is fixed to the access layer; In the electrical switching connection between the access layer and the convergence layer, the electrical switching connection is connected to the access layer in a pluggable interface form.

5. A method for constructing an intelligent computing network structure, characterized in that: The intelligent computing network structure includes a core layer, a convergence layer, an access layer and a client, and the method includes: For each core layer, the core layer is connected to each aggregation layer in the form of electrical switching; For each convergence layer, the convergence layer is connected to at least part of the access layer in the form of electrical switching according to a preset rule; For each access layer, optical transmission is used to connect the access layer and the client covered by the access layer.

6. The intelligent computing network structure according to claim 5, characterized in that: The core layer and each aggregation layer are connected in the form of electrical switching, including: If the distance between the core layer and a convergence layer is not greater than a specified distance, the core layer and the convergence layer are connected in the form of electrical switching; If the distance between the core layer and a convergence layer is greater than the specified distance, the core layer and the convergence layer are connected in the form of optical transmission; Connecting the convergence layer and at least part of the access layer in the form of electrical switching according to a preset rule specifically includes: Determine the access layer that needs to be connected to the core layer as the target access layer according to a preset rule; If the distance between the convergence layer and a target access layer is not greater than the specified distance, connecting the convergence layer and the target access layer in the form of electrical switching; If the distance between the convergence layer and a target access layer is greater than the specified distance, the convergence layer and the target access layer are connected in the form of electrical switching.

7. The intelligent computing network structure according to claim 5, characterized in that: Connecting the convergence layer and at least part of the access layer in the form of electrical switching according to a preset rule specifically includes: Connecting the convergence layer and at least part of the access layer in a pluggable electrical exchange format according to preset rules; The access layer is connected to the client covered by the access layer in the form of optical transmission, specifically including: The access layer is connected with the client covered by the access layer in the form of optical transmission in which one end of the optical cable is fixed on the access layer.

8. A device for constructing an intelligent computing network structure, characterized in that: The intelligent computing network structure includes a core layer, a convergence layer, an access layer and a client, and the device includes: The core building module is used to connect each core layer with each aggregation layer by means of electrical switching; A convergence construction module, used for connecting each convergence layer with at least part of the access layer in the form of electrical switching according to a preset rule; The access construction module is used to connect each access layer with the client covered by the access layer in the form of optical transmission.

9. A computer-readable storage medium, characterized in that: The storage medium stores a computer program, and when the computer program is executed by a processor, the method according to any one of claims 5 to 7 is implemented.

10. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the program, the method according to any one of claims 5 to 7 is implemented.

Images