A calculation method and device for assisting the networking of an intelligent computing center

Abstract

The present application provides a calculation method and device for assisting in the networking of an intelligent computing center. The method provided by the present application includes: determining input parameters based on the networking structure type of the intelligent computing center; determining the number of network structure layers corresponding to the networking structure type based on the relationships between the input parameters; determining the dimension of the networking parameters according to the networking structure type, and calculating the networking parameter values corresponding to each networking parameter dimension based on the number of network structure layers and the input parameters; wherein, determining the layer dimension parameter and the layer connection dimension parameter corresponding to the networking structure type according to the number of network structure layers, the layer dimension parameter being the configuration parameter of each layer in the networking structure, and the layer connection dimension parameter being the connection configuration parameter between the layers in the networking structure. The calculation method and device for assisting in the networking of an intelligent computing center provided by the present application realize the function of automatic networking calculation to assist in the networking of the intelligent computing center and have strong versatility.

Description

Technical Field

[0001] This application relates to the field of artificial intelligence technology, and in particular, to a calculation method and device for assisting in the networking of intelligent computing centers. Background Art

[0002] Due to the booming development of artificial intelligence, the demand for computing power has increased exponentially. The construction demand for intelligent computing centers is very strong, and the construction scale is also getting larger and larger. This has resulted in the need for a large number of servers and matching network devices to form an intelligent computing center. When building an intelligent computing center, a customized networking solution needs to be provided to customers. Then, the quantity of hardware devices required for networking, including servers, switches, optical modules, optical fibers or optical cables, etc., all need to be listed in detail in the solution. Using manual calculation is time-consuming and laborious, and due to differences in understanding and capabilities among different solution writers, errors are extremely likely to occur, which will have a very adverse impact during the reporting to customers or the bidding process. Therefore, there is an urgent need for a software to implement the networking calculation function.

[0003] Currently, no publicly available networking calculation software has been found in the market. Therefore, there is an urgent need for a method to achieve the function of automatic networking calculation to assist in the networking of intelligent computing centers and have strong versatility. Summary of the Invention

[0004] In view of this, this application provides a calculation method and device for assisting in the networking of intelligent computing centers to achieve the function of automatic networking calculation to assist in the networking of intelligent computing centers and have strong versatility.

[0005] Specifically, this application is implemented through the following technical solutions:

[0006] The first aspect of this application provides a calculation method for assisting in the networking of intelligent computing centers, and the method includes:

[0007] Based on the networking structure type of the intelligent computing center, determine input parameters, where the dimensions of the input parameters are different for different networking structure types;

[0008] Based on the relationship between the input parameters, determine the number of network structure layers corresponding to the networking structure type;

[0009] According to the networking structure type, determine the dimension of the networking parameters, and calculate the networking parameter values corresponding to each dimension of the networking parameters based on the number of network structure layers and the input parameters;

[0010] Wherein, according to the number of network structure layers, determine the layer dimension parameters and layer connection dimension parameters corresponding to the networking structure type. The layer dimension parameters are the configuration parameters of each layer in the networking structure, and the layer connection dimension parameters are the connection configuration parameters between layers in the networking structure.

[0011] The second aspect of the present application provides a computing device for assisting in the networking of an intelligent computing center. The device includes: a determination module and a calculation module; wherein,

[0012] The determination module is configured to determine input parameters based on the networking structure type of the intelligent computing center. Among them, when the networking structure types are different, the dimensions of the input parameters are different;

[0013] The determination module is further configured to determine the number of network structure layers corresponding to the networking structure type based on the relationship between the input parameters;

[0014] The calculation module is configured to determine the dimension of the networking parameters according to the networking structure type, and calculate the networking parameter values corresponding to each dimension of the networking parameters based on the number of network structure layers and the input parameters;

[0015] Among them, the layer dimension parameter and the layer connection dimension parameter corresponding to the networking structure type are determined according to the number of network structure layers. The layer dimension parameter is the configuration parameter of each layer in the networking structure, and the layer connection dimension parameter is the connection configuration parameter between the layers in the networking structure.

[0016] For the calculation method and device for assisting in the networking of an intelligent computing center provided by the present application, in the first aspect, based on the networking structure type of the intelligent computing center, input parameters are determined. Only a few necessary networking parameters need to be input to generate the quantity of hardware devices required for networking, realizing networking. The input parameters all contain simple and easy-to-understand descriptions, without the need for in-depth understanding of the architecture of the intelligent computing center network, reducing the usage threshold of users. In the second aspect, by digitizing the hardware devices required for networking, the differences between different manufacturers and different product models are ignored, and the universality is strong, making it easier to be accepted by users. In the third aspect, machines are used to replace the previous manual calculation and verification. There is no need for cumbersome and error-prone paper and pen calculation and verification. When facing the requirements of multiple projects, it saves time and effort and is not prone to errors, improving work efficiency and reducing the probability of human errors. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 It is a flowchart of the first embodiment of the calculation method for assisting in the networking of an intelligent computing center provided by the present application;

[0018] Figure 2 It is a schematic structural diagram of the first embodiment of the calculation device for assisting in the networking of an intelligent computing center provided by the present application. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0019] Exemplary embodiments will be described in detail herein, and examples thereof are shown in the accompanying drawings. When the following description refers to the accompanying drawings, unless otherwise indicated, the same numbers in different drawings represent the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the present application.

[0020] The terms used in this application are for the purpose of describing specific embodiments only and are not intended to limit this application. The singular forms "a", "the", and "said" used in this application are also intended to include the plural forms unless the context clearly indicates otherwise. It should also be understood that the term "and / or" as used herein refers to and includes any or all possible combinations of one or more of the associated listed items.

[0021] It should be understood that although the terms first, second, third, etc. may be used in this application to describe various information, such information should not be limited to these terms. These terms are only used to distinguish the same type of information from each other. For example, without departing from the scope of this application, the first information may also be referred to as the second information, and similarly, the second information may also be referred to as the first information. Depending on the context, the word "if" as used herein may be interpreted as "when" or "while" or "in response to a determination".

[0022] Specific embodiments are given below to introduce the technical solutions of this application in detail.

[0023] Figure 1 This is a flowchart of the first embodiment of the calculation method for assisting the networking of the intelligent computing center provided by this application. Please refer to Figure 1 , the method provided in this embodiment may include:

[0024] S101. Determine input parameters based on the networking structure type of the intelligent computing center, where different networking structure types have different dimensions of input parameters.

[0025] Specifically, the networking structure type of the intelligent computing center refers to the network architecture and layout method used to connect and manage computing resources, storage devices, network devices, and other related components in the intelligent computing center. Common networking structure types include centralized structure, distributed structure, hybrid structure, multi-layer structure, leaf-spine structure, hyper-converged structure, and edge computing structure.

[0026] Furthermore, different networking structure types have different dimensions of input parameters. Common input parameters at least include the number of servers, the number of network cards per server, the number of ports per network card, the number of switch ports, the networking convergence ratio, and the type of connection medium between devices.

[0027] Specifically, when implemented, determining the input parameters based on the networking structure type of the intelligent computing center includes:

[0028] (1) Determine the structure parameters corresponding to the networking structure type of the intelligent computing center, and determine the dimension of the input parameters according to the structure parameters and the configuration attributes of the intelligent computing center.

[0029] Specifically, the structure parameters refer to the key technical indicators and configuration parameters related to the networking structure type. The structure parameters correspond to the networking structure type. When the networking structure type is a multi-layer structure (including the core layer, aggregation layer, and access layer), correspondingly, the structure parameters include core layer bandwidth, aggregation layer bandwidth, access layer bandwidth, redundancy, and latency. When the networking structure type is a leaf-spine structure, correspondingly, the structure parameters include the number of ports of the leaf layer switches, the number of ports of the spine layer switches, the topology structure, and latency.

[0030] Furthermore, the configuration attributes of the intelligent computing center involve various technical and architectural parameters, which directly affect the computing power, network performance, storage capacity, reliability, and scalability of the intelligent computing center. The configuration attributes of the intelligent computing center usually include resource configuration attributes, network configuration attributes, storage configuration attributes, virtualization and containerization attributes, security configuration attributes, management and monitoring attributes, scalability and flexibility attributes, energy efficiency attributes, and reliability and high availability attributes.

[0031] In specific implementation, based on the networking structure type of the intelligent computing center, perform requirement analysis to determine the structure parameters of the intelligent computing center under this networking structure type. Furthermore, based on the determined structure parameters, refine the configuration attributes of the intelligent computing center. Based on the refined configuration attributes, determine the specific input parameters and their dimensions.

[0032] (2) Based on the networking requirements of the intelligent computing center, input the input parameters for each dimension.

[0033] Specifically, the networking requirements of the intelligent computing center include computing requirements, network requirements, storage requirements, virtualization requirements, security requirements, management requirements, scalability requirements, energy efficiency requirements, and reliability requirements.

[0034] In specific implementation, based on a specific requirement among the networking requirements of the intelligent computing center, input the input parameters for the dimension corresponding to this requirement. Based on the same method, input the input parameters for each dimension.

[0035] For example, in one embodiment, the input parameters for each dimension include: the number of servers is 96, the number of network cards per server is 4, the number of ports per network card is 1, the number of switch ports is 32, the networking convergence ratio is 1:1, one port of the switch is connected to one optical fiber and then to one network card (connection type between the server and the switch), and one port of the switch is connected to one optical fiber and then to one network card (connection type between switches).

[0036] Optionally, after the input parameters for each dimension are entered, the method further includes: determining the value range of the input parameters for each dimension based on the performance requirements of the intelligent computing center and the type of network architecture; traversing the input parameters for each dimension in sequence to check whether the input parameters for each dimension meet the value range; when the input parameters for any dimension do not meet the value range, re-determining the input parameters until the input parameters for each dimension meet the value range.

[0037] Specifically, based on the type of network architecture of the intelligent computing center, determine the input parameters for each dimension. Further, by considering the performance requirements of the intelligent computing center, limit the input parameters for the corresponding dimension (for example, in combination with the above example, the value range of the number of network card ports per server is 1-2), and determine the value range of the input parameters for each dimension. By traversing the input parameters for each dimension, check whether their values meet the value range, and adjust the input parameters whose values do not meet the value range until the values of the input parameters for each dimension meet the value range.

[0038] S102. Based on the relationship between the input parameters, determine the number of network structure layers corresponding to the type of network architecture.

[0039] Specifically, as described above, the input parameters at least include the number of servers, the number of network cards per server, the number of network card ports per server, the number of switch ports, and the network convergence ratio.

[0040] Further, the number of network structure layers corresponds to the type of network architecture. For example, as described above, when the type of network architecture is a multi-layer structure (including the core layer, aggregation layer, and access layer), correspondingly, the number of network structure layers is three. When the type of network architecture is a leaf-spine structure (including the leaf layer and the spine layer), correspondingly, the number of network structure layers is two.

[0041] When specifically implemented, the determining the number of network structure layers corresponding to the type of network architecture based on the relationship between the input parameters includes:

[0042] (1) Calculate the number of leaf layer switches based on the input parameters; where the input parameters at least include the number of servers, the number of network cards per server, the number of network card ports per server, the number of switch ports, and the network convergence ratio.

[0043] Specifically, the leaf layer switch is a key component connecting the computing nodes and the spine layer switches, and the number of leaf layer switches affects the throughput, latency, and reliability of the network.

[0044] When specifically implemented, the calculating the number of leaf layer switches based on the input parameters includes:

[0045] (i)Determine the total number of ports based on the number of servers, the number of network cards per server, and the number of ports per network card.

[0046] Specifically, the total number of ports refers to the sum of the port numbers of all devices on the server. The total number of ports is equal to the product of the number of servers, the number of network cards per server, and the number of ports per network card.

[0047] When specifically implemented, the total number of ports can be determined according to the following formula:

[0048] Total number of ports = Number of servers × Number of network cards per server × Number of ports per network card

[0049] (ii)Determine the uplink port number of the leaf switches based on the number of switch ports and the network convergence ratio.

[0050] Specifically, the uplink port number of the leaf switches refers to the number of ports on each leaf switch connected to the upper-layer network device (usually the spine switch or the core switch).

[0051] When specifically implemented, the uplink port number of the leaf switches can be determined according to the following formula:

[0052] Uplink port number of leaf switches = Number of switch ports / (Network convergence ratio + 1)

[0053] (iii)Determine the downlink port number of the leaf switches based on the uplink port number of the leaf switches and the number of switch ports.

[0054] Specifically, the downlink port number of the leaf switches refers to the number of ports on each leaf switch connected to the computing nodes or other terminal devices.

[0055] When specifically implemented, the downlink port number of the leaf switches can be determined according to the following formula:

[0056] Downlink port number of leaf switches = Number of switch ports — Uplink port number of leaf switches

[0057] (iiii)Determine the number of leaf switches based on the total number of ports and the downlink port number of the leaf switches.

[0058] When specifically implemented, the number of leaf switches can be determined according to the following formula:

[0059] Number of leaf switches = Total number of ports / Downlink port number of leaf switches

[0060] (2)Determine the number of network structure layers corresponding to the network structure type based on the size relationship between the number of leaf switches and the number of switch ports.

[0061] Specifically, the relationship between the number of leaf switches and the number of switch ports determines the number of network structure layers.

[0062] In specific implementation, determining the number of network structure layers corresponding to the network structure type based on the relationship between the number of leaf switches and the number of switch ports includes: when the number of leaf switches is not greater than the number of switch ports, determining that the number of network structure layers is two; when the number of leaf switches is greater than the number of switch ports, determining that the number of network structure layers is three.

[0063] Specifically, when the number of leaf switches exceeds the port number limit of each switch (i.e., the number of switch ports), it is usually necessary to design the network structure as three layers (for example, adding an aggregation layer to connect multiple leaf switches) to effectively manage and connect all devices. By adding an aggregation layer, it can help manage a large number of leaf switches, so that each leaf switch does not have to be directly connected to the core layer, but is connected through the aggregation layer. This design helps to improve the flexibility and scalability of the network, while reducing the burden on the core layer switches.

[0064] The method provided in this embodiment gives a scientific method for determining the number of network structure layers: first, determine the number of leaf switches according to the relevant parameters in the input parameters and the calculation formula of the number of leaf switches, and then determine the number of network structure layers by comparing the calculated number of leaf switches with the number of switch ports in the input parameters. After determining the number of network structure layers corresponding to the network structure type, various networking parameters and their values can be calculated based on the determined number of network structure layers and the input parameters in the subsequent process, that is, determine the number of hardware devices required for networking, so as to achieve networking.

[0065] S103. According to the network structure type, determine the networking parameter dimensions, and calculate the networking parameter values corresponding to each of the networking parameter dimensions based on the number of network structure layers and the input parameters; wherein, determine the layer dimension parameter and the layer connection dimension parameter corresponding to the network structure type according to the number of network structure layers, the layer dimension parameter is the configuration parameter of each layer in the network structure, and the layer connection dimension parameter is the connection configuration parameter between each layer in the network structure.

[0066] Specifically, the networking parameter dimensions are related to the number of network structure layers and the input parameters, and the networking parameter dimensions include the layer dimension parameter and the layer connection dimension parameter. The layer dimension parameter is the configuration parameter of each layer in the network structure, and the layer dimension parameter at least includes the number of switches and the number of ports. The layer connection dimension parameter is the connection configuration parameter between each layer in the network structure, and the layer connection dimension parameter at least includes the number of optical modules and the number of optical fibers and optical cables.

[0067] In specific implementation, determining the layer dimension parameters and layer connection dimension parameters corresponding to the networking structure type according to the number of network structure layers includes:

[0068] (1) Based on the number of network structure layers, determine the functions of each layer of the network in the network structure corresponding to the number of network structure layers.

[0069] Specifically, the functions of each layer of the network in different network structures are different. For example, in combination with the above description, when the number of network structure layers is two (including the core layer and the leaf layer), the function of the core layer is to be responsible for network aggregation and routing, connecting to the upper-level network or edge network, and processing the aggregation and distribution of internal traffic in the data center. The function of the leaf layer is to connect to computing nodes and other terminal devices, provide high-density port connections to support a large number of computing nodes, carry most of the internal data traffic in the data center, and provide an uplink connection to the core layer. Another example is that when the number of network structure layers is three (including the core layer, the aggregation layer, and the leaf layer), the function of the core layer is to be responsible for cross-network routing and aggregation, processing traffic from the aggregation layer and other core devices, and performing security and performance optimizations. The function of the aggregation layer is to connect multiple leaf layer switches, aggregate the energy of the leaf layer switches to the core layer, provide redundant paths and load balancing, and connect and distribute data. The function of the leaf layer is to connect to computing nodes and other terminal devices, provide a large number of downlink ports to support connections to all computing nodes and other devices, and process communications between internal devices in the data center.

[0070] In specific implementation, based on the number of network structure layers, determine the network structure corresponding to the number of network structure layers. Conduct a functional analysis of each layer of the determined network structure to determine the functions of each layer of the network.

[0071] (2) Based on the functions of each layer of the network in the network structure, determine the layer dimension parameters and layer connection dimension parameters corresponding to each layer of the network.

[0072] Specifically, the layer dimension parameters and layer connection dimension parameters corresponding to each layer of the network are different. For example, in combination with the above description, when the number of network structure layers is two (including the core layer and the leaf layer), the layer dimension parameters corresponding to the core layer may include the number of ports, the number of switches, the size of the routing table, processing capabilities, etc., and the layer connection dimension parameters corresponding to the core layer may include the number of optical modules, the number of optical fibers, the number of optical cables, bandwidth, latency, and redundancy level. The layer dimension parameters corresponding to the leaf layer may include the number of ports, the number of switches, port rate, port utilization, etc., and the layer connection dimension parameters corresponding to the leaf layer may include the number of optical modules, the number of optical fibers, the number of optical cables, bandwidth, latency, and redundancy level.

[0073] In specific implementation, based on the functions of each layer of the network in the determined network structure, determine the layer dimension parameters and layer connection dimension parameters corresponding to each layer of the network.

[0074] (3) Combine the layer dimension parameters and layer connection dimension parameters corresponding to each layer of the network to obtain the layer dimension parameters and layer connection dimension parameters corresponding to the network architecture type.

[0075] In specific implementation, add the layer dimension parameters and layer connection dimension parameters corresponding to each layer of the network, and determine the combined layer dimension parameters and layer connection dimension parameters as the layer dimension parameters and layer connection dimension parameters corresponding to the network architecture type.

[0076] For example, in combination with the above example, when the number of network structure layers is two (including the core layer and the leaf layer), in this step, the layer dimension parameters corresponding to the core layer are combined and added with the layer dimension parameters corresponding to the leaf layer as the layer dimension parameters corresponding to the network architecture type. The layer connection dimension parameters corresponding to the core layer are combined and added with the layer connection dimension parameters corresponding to the leaf layer as the layer connection dimension parameters corresponding to the network architecture type.

[0077] As a preferred embodiment, calculating the network parameter values corresponding to each network parameter dimension based on the number of network structure layers and the input parameters includes:

[0078] (1) Based on the number of network structure layers, determine the layer dimension parameters and layer connection dimension parameters corresponding to the network architecture type.

[0079] Specifically, the specific implementation process and implementation principle of this step can refer to the description in the above embodiment and will not be elaborated here.

[0080] (2) Based on the number of servers, the number of network cards per server, the number of ports per network card, the number of switch ports, and the network convergence ratio in the input parameters, determine the layer dimension parameters.

[0081] Specifically, the layer dimension parameters correspond to the network architecture type. In this embodiment, the leaf-spine structure type of the network architecture is described. In combination with the above description, the layer dimension parameters at least include the number of switches and the number of ports. The number of switches includes the number of leaf layer switches and the number of spine layer switches. The number of ports includes the number of uplink ports of the leaf layer switches, the number of downlink ports of the leaf layer switches, the number of redundant ports of the leaf layer switches, and the number of redundant ports of the spine layer switches.

[0082] Method for determining the number of leaf layer switches:

[0083] Number of leaf layer switches = Total number of ports / Number of downlink ports of leaf layer switches

[0084] Total number of ports = Number of servers × Number of network cards per server × Number of ports per network card

[0085] Determination method for the number of spine switches:

[0086] Number of spine switches = Number of leaf switches × Number of uplink ports of leaf switches / Number of switch ports

[0087] Determination method for the number of uplink ports of leaf switches:

[0088] Number of uplink ports of leaf switches = Number of switch ports / (Networking convergence ratio + 1)

[0089] Determination method for the number of downlink ports of leaf switches:

[0090] Number of downlink ports of leaf switches = Number of switch ports — Number of uplink ports of leaf switches

[0091] Determination method for the number of redundant ports of leaf switches:

[0092] Number of redundant ports of leaf switches = Number of leaf switches × (Number of switch ports — (Number of uplink ports of leaf switches + Number of downlink ports of leaf switches))

[0093] Determination method for the number of redundant ports of spine switches:

[0094] Number of redundant ports of spine switches = Number of spine switches × Number of switch ports — Number of leaf switches × Number of uplink ports of leaf switches

[0095] (3) Based on the number of servers, the number of network cards per server, the number of ports per network card, and the connection medium type between devices in the input parameters, determine the layer connection dimension parameter.

[0096] Specifically, the layer connection dimension parameter corresponds to the network structure type. In this embodiment, the description is for the network structure type of leaf-spine structure. The determining the layer connection dimension parameter based on the number of servers, the number of network cards per server, the number of ports per network card, and the connection medium type between devices in the input parameters includes:

[0097] (1) Based on the connection medium type between devices in the input parameters, determine the first coefficient.

[0098] Specifically, the connection medium type between devices refers to the physical transmission medium used to connect different devices (such as servers, switches) in the intelligent computing center. Common connection medium types include optical fiber, copper cable, hybrid optical fiber and copper cable connection, and wireless connection.

[0099] It should be noted that the first coefficients corresponding to different connection medium types are different, and the first coefficient represents how many ports each optical module corresponds to under this connection medium type. For example, when the connection medium type is optical fiber, the first coefficient is 1.

[0100] Based on the number of servers, the number of network cards per server, and the number of ports per network card in the input parameters, determine the total number of ports.

[0101] Specifically, in combination with the above description, the total number of ports is equal to the product of the number of servers, the number of network cards per server, and the number of ports per network card.

[0102] In specific implementation, based on the input number of servers, the number of network cards per server, and the number of ports per network card, multiply the three to obtain the total number of ports.

[0103] Based on the first coefficient and the total number of ports, determine the number of optical modules and the number of optical fibers and optical cables.

[0104] Specifically, the number of optical modules and the number of optical fibers and optical cables are equal to the product of the first coefficient and the total number of ports.

[0105] In specific implementation, multiply the calculated total number of ports by the determined first coefficient to obtain the number of optical modules and the number of optical fibers and optical cables, that is, obtain the layer connection dimension parameter.

[0106] The method provided in this embodiment gives a scientific method for determining the networking parameter dimension and the networking parameter value: when determining the networking parameter dimension, first, based on the determined number of network structure layers, determine the functions of each layer of the network in the network structure corresponding to this number of network structure layers. Based on the functions of each layer of the network, determine the layer dimension parameter and the layer connection dimension parameter corresponding to each layer of the network, and then determine the networking parameter dimension corresponding to this networking structure type. When determining the values of each networking parameter, determine the layer dimension parameter based on the relevant parameters in the input parameters and the calculation formula of the layer dimension parameter, and determine the layer connection dimension parameter based on the relevant parameters in the input parameters and the calculation formula of the layer connection dimension parameter, and then determine the values of each networking parameter. After determining the networking parameter dimension and the networking parameter value, the number of hardware devices required for networking is determined, and networking can be conveniently implemented subsequently.

[0107] The calculation method for assisting the networking of the intelligent computing center provided in this embodiment, on the first hand, determines the input parameters based on the networking structure type of the intelligent computing center. Only a few necessary networking parameters need to be input to generate the quantity of hardware devices required for networking, thus realizing networking. The input parameters all contain simple and easy-to-understand descriptions, without the need for in-depth understanding of the architecture of the intelligent computing center network, reducing the usage threshold for users. On the second hand, by digitizing the hardware devices required for networking, the differences between different manufacturers and different product models are ignored, with strong versatility and being more easily accepted by users. On the third hand, machines are used to replace the previous manual calculation and verification. There is no need for cumbersome and error-prone paper-and-pencil calculation and verification. When facing the requirements of multiple projects, it saves time and effort, is not prone to errors, improves work efficiency, and reduces the probability of human errors.

[0108] Corresponding to the foregoing embodiment of a calculation method for assisting the networking of the intelligent computing center, the present application also provides an embodiment of a calculation device for assisting the networking of the intelligent computing center.

[0109] Figure 2 It is a schematic structural diagram of the first embodiment of the calculation device for assisting the networking of the intelligent computing center provided by the present application. Please refer to Figure 2 , the device provided in this embodiment, the device includes: a determination module 210 and a calculation module 220; wherein,

[0110] The determination module 210 is configured to determine input parameters based on the networking structure type of the intelligent computing center, wherein, different networking structure types result in different dimensions of the input parameters;

[0111] The determination module 210 is further configured to determine the number of network structure layers corresponding to the networking structure type based on the relationship between the input parameters;

[0112] The calculation module 220 is configured to determine the dimension of the networking parameters according to the networking structure type, and calculate the networking parameter values corresponding to each dimension of the networking parameters based on the number of network structure layers and the input parameters;

[0113] Wherein, the layer dimension parameter and the layer connection dimension parameter corresponding to the networking structure type are determined according to the number of network structure layers. The layer dimension parameter is the configuration parameter of each layer in the networking structure, and the layer connection dimension parameter is the connection configuration parameter between each layer in the networking structure.

[0114] The computing device for assisting the networking of the intelligent computing center. In the first aspect, based on the networking structure type of the intelligent computing center, input parameters are determined. Only a few necessary networking parameters need to be input, and then the quantity of the hardware devices required for networking can be generated to achieve networking. The input parameters all contain simple and easy-to-understand descriptions, without the need for in-depth understanding of the architecture of the intelligent computing center network, which reduces the usage threshold for users. In the second aspect, by digitizing the hardware devices required for networking, the differences between different manufacturers and different product models are ignored, and the versatility is relatively strong, making it more acceptable to users. In the third aspect, machines are used to replace the previous manual calculation and verification. There is no need to perform cumbersome and error-prone paper-and-pencil calculations and verifications. When facing the requirements of multiple projects, it saves time and effort and is not prone to errors, improving work efficiency and reducing the probability of human errors.

[0115] The device of this embodiment can be used to execute Figure 1 the steps of the method embodiment shown. The specific implementation principle and process are similar and will not be elaborated here.

[0116] For the implementation process of the functions and roles of each unit in the above device, please refer to the implementation process of the corresponding steps in the above method for details and will not be elaborated here.

[0117] For the device embodiment, since it basically corresponds to the method embodiment, the relevant parts can refer to the partial description of the method embodiment. The device embodiments described above are only illustrative. The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed to multiple network units. Some or all of the modules can be selected according to actual needs to achieve the purpose of the solution of this application. Those of ordinary skill in the art can understand and implement it without creative efforts.

[0118] The above are only the preferred embodiments of this application and are not intended to limit this application. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principle of this application shall be included in the scope of protection of this application.

Claims

1. A calculation method for assisting the networking of an intelligent computing center, characterized in that The method includes: Determining input parameters based on the networking structure type of the intelligent computing center, where different networking structure types have different dimensions of the input parameters; Determining the number of network structure layers corresponding to the networking structure type based on the relationship between the input parameters; Among them, performing requirement analysis based on the networking structure type of the intelligent computing center to determine the structure parameters of the intelligent computing center under the networking structure type; based on the structure parameters, refining the configuration attributes of the intelligent computing center, and determining the input parameters and dimensions based on the refined configuration attributes; according to the networking requirements of the intelligent computing center, inputting the input parameters of each dimension, and the input parameters at least include the number of servers, the number of network cards per server, the number of ports per network card, the number of switch ports, and the networking convergence ratio, and the total number of ports represents the sum of the port numbers of all devices on the server; Determining the total number of ports based on the product of the number of servers and the number of network cards per server and the number of ports per network card, determining the uplink port number of the leaf layer switch based on the quotient of the number of switch ports divided by the sum of the networking convergence ratio and 1, and determining the downlink port number of the leaf layer switch based on the difference between the number of switch ports and the uplink port number of the leaf layer switch; determining the number of leaf layer switches based on the quotient of the total number of ports divided by the downlink port number of the leaf layer switch; determining the number of network structure layers corresponding to the networking structure type based on the size relationship between the number of leaf layer switches and the number of switch ports; Determining the dimension of the networking parameters according to the networking structure type, and calculating the values of the networking parameters corresponding to each dimension of the networking parameters based on the number of network structure layers and the input parameters; Among them, determining the layer dimension parameter and the layer connection dimension parameter corresponding to the networking structure type according to the number of network structure layers, where the layer dimension parameter is the configuration parameter of each layer in the networking structure, and the layer connection dimension parameter is the connection configuration parameter between each layer in the networking structure.

2. The method according to claim 1, wherein After inputting the input parameters of each dimension, the method further includes: Determining the value range of the input parameters of each dimension based on the performance requirements of the intelligent computing center and the networking structure type; Traversing the input parameters of each dimension in sequence to check whether the input parameters of each dimension meet the value range; When the input parameters of any dimension do not meet the value range, re-determining the input parameters until the input parameters of each dimension meet the value range.

3. The method according to claim 1, wherein The determining the number of network structure layers corresponding to the networking structure type based on the size relationship between the number of leaf layer switches and the number of switch ports includes: When the number of leaf layer switches is not greater than the number of switch ports, determining that the number of network structure layers is two; When the number of leaf layer switches is greater than the number of switch ports, determining that the number of network structure layers is three.

4. The method according to claim 1, wherein The layer dimension parameters at least include the number of switches and the number of ports. The layer connection dimension parameters at least include the number of optical modules, the number of optical fibers, and the number of optical cables. The number of switches includes the number of leaf switches and the number of spine switches. The number of ports includes the number of uplink ports of the leaf switches, the number of downlink ports of the leaf switches, the number of redundant ports of the leaf switches, and the number of redundant ports of the spine switches; The determining the layer dimension parameters and the layer connection dimension parameters corresponding to the network architecture type according to the number of network architecture layers includes: Based on the number of network architecture layers, determining the functions of each layer of network in the network architecture corresponding to the number of network architecture layers; Based on the functions of each layer of network in the network architecture, determining the layer dimension parameters and the layer connection dimension parameters corresponding to each layer of network; Combining the layer dimension parameters and the layer connection dimension parameters corresponding to each layer of network to obtain the layer dimension parameters and the layer connection dimension parameters corresponding to the network architecture type.

5. The method according to claim 1, characterized in that, The calculating the network parameter values corresponding to each network parameter dimension based on the number of network architecture layers and the input parameters includes: Based on the number of network architecture layers, determining the layer dimension parameters and the layer connection dimension parameters corresponding to the network architecture type; Based on the number of servers, the number of network cards per server, the number of ports per network card, the number of switch ports, and the network convergence ratio in the input parameters, determining the layer dimension parameters; Based on the number of servers, the number of network cards per server, the number of ports per network card, and the type of connection medium between devices in the input parameters, determining the layer connection dimension parameters.

6. The method according to claim 5, wherein The determining the layer connection dimension parameters based on the number of servers, the number of network cards per server, the number of ports per network card, and the type of connection medium between devices in the input parameters includes: Based on the type of connection medium between devices in the input parameters, determining a first coefficient; the first coefficient represents the number of ports corresponding to each optical module under the connection medium type; Based on the number of servers, the number of network cards per server, and the number of ports per network card in the input parameters, determining the total number of ports; Based on the first coefficient and the total number of ports, determining the number of optical modules and the number of optical fibers and optical cables.

7. A computing device for assisting in the networking of an intelligent computing center group, characterized in that, The device includes: a determination module and a calculation module; wherein, The determination module is used to determine the input parameters based on the network architecture type of the intelligent computing center. Among them, the dimensions of the input parameters are different for different network architecture types; The determination module is further used to determine the number of network architecture layers corresponding to the network architecture type based on the relationship between the input parameters; Among them, based on the networking structure type of the intelligent computing center, the requirement analysis is carried out to determine the structural parameters of the intelligent computing center under the networking structure type; based on the structural parameters, the configuration attributes of the intelligent computing center are refined, and the input parameters and dimensions are determined based on the refined configuration attributes; based on the networking requirements of the intelligent computing center, the input parameters of each dimension are input, and the input parameters at least include the number of servers, the number of network cards per server, the number of ports per network card, the number of switch ports, and the networking convergence ratio, and the total number of ports represents the sum of the port numbers of all devices on the server. Based on the product of the number of servers and the number of network cards per server and the number of ports per network card, the total number of ports is determined. Based on the quotient of the number of switch ports divided by the sum of the networking convergence ratio and 1, the uplink port number of the leaf switch is determined. Based on the difference between the number of switch ports and the uplink port number of the leaf switch, the downlink port number of the leaf switch is determined. Based on the quotient of the total number of ports divided by the downlink port number of the leaf switch, the number of leaf switches is determined. Based on the size relationship between the number of leaf switches and the number of switch ports, the number of network structure layers corresponding to the networking structure type is determined. The calculation module is used to determine the networking parameter dimensions according to the networking structure type, and calculate the networking parameter values corresponding to each of the networking parameter dimensions based on the number of network structure layers and the input parameters. Among them, according to the number of network structure layers, the layer dimension parameters and layer connection dimension parameters corresponding to the networking structure type are determined. The layer dimension parameters are the configuration parameters of each layer in the networking structure, and the layer connection dimension parameters are the connection configuration parameters between the layers in the networking structure.

Images