Key agreement method, apparatus, and related device

Abstract

The application discloses a key negotiation method and device and related equipment. The method is applied to an artificial intelligence (AI) acceleration device, the AI acceleration device comprises a plurality of AI acceleration cards, and the method comprises the following steps: for any first target AI acceleration card, determining a plurality of second target AI acceleration cards based on a key negotiation instruction sent by an electronic device and received by the first target AI acceleration card; sending the key negotiation instruction to the plurality of second target AI acceleration cards respectively; after receiving at least one negotiation algorithm sent by any second target AI acceleration card through the first target AI acceleration card, determining a target negotiation algorithm supported by the first target AI acceleration card from the at least one negotiation algorithm; and determining a key between the first target AI acceleration card and any second target AI acceleration card by using the target negotiation algorithm. Therefore, the key negotiation between the plurality of AI acceleration cards is realized, information leakage is reduced, and data security is improved.

Description

Technical Field

[0001] This invention relates to the field of data processing technology, and in particular to key negotiation methods, apparatus and related equipment. Background Technology

[0002] With the widespread adoption of AI (Artificial Intelligence) technology in cloud computing, a secure architecture for confidential computing has emerged to address a range of security issues in AI applications. This security model divides the computing system into trusted and untrusted domains. Third-party AI acceleration devices, typically not belonging to the CPU (Central Processing Unit) or the vendor's default trusted domain, require additional encryption and decryption to ensure the security of data channels when accessing confidential virtual machines.

[0003] In existing technologies, only single-card computing scenarios exist. In these scenarios, confidential computing technologies involve key negotiation between the CPU and the accelerator card to generate a shared key for encryption and decryption during data transmission. However, as model size increases, so does the size of AI accelerator card clusters. In large-scale model training scenarios, there are significant data exchange needs between multiple AI accelerator cards. This data exchange uses dedicated high-speed channels. If plaintext data is transmitted directly through these channels, there is a risk of confidential information being leaked, making it impossible to guarantee data security for large-scale multi-card cluster training. Therefore, a key negotiation method between multiple AI accelerator cards is urgently needed. Summary of the Invention

[0004] This invention provides a key negotiation method, apparatus, and related equipment for enabling key negotiation between multiple AI accelerator cards, reducing information leakage, and improving data security.

[0005] According to a first aspect of the present invention, a key negotiation method is provided, applied in an artificial intelligence (AI) acceleration device, the AI ​​acceleration device including multiple AI acceleration cards, the method comprising: For any one of the plurality of AI accelerator cards as a first target AI accelerator card, after receiving a key negotiation instruction sent by an electronic device through the first target AI accelerator card, a plurality of second target AI accelerator cards are determined based on the key negotiation instruction, wherein the second target AI accelerator card is an AI accelerator card that needs to perform key negotiation with the first target AI accelerator card; And send key negotiation instructions to the multiple second target AI acceleration cards respectively; After receiving at least one negotiation algorithm from any second target AI acceleration card through the first target AI acceleration card, the target negotiation algorithm supported by itself is determined from the at least one negotiation algorithm; The target negotiation algorithm is then used to determine the key between the first target AI accelerator card and any one of the second target AI accelerator cards.

[0006] According to another aspect of the present invention, a key negotiation method is provided, applied in an electronic device, the method comprising: Obtain a pre-constructed network topology map, wherein the network topology map includes the connection relationship between multiple AI accelerator cards in the artificial intelligence (AI) acceleration device; For any one of the multiple AI accelerator cards as a first target AI accelerator card, the network topology map is used to determine the second target AI accelerator card corresponding to the first target AI accelerator card, wherein the second target AI accelerator card is the AI ​​accelerator card that needs to perform key negotiation with the first target AI accelerator card. Based on the first target AI accelerator card and the second target AI accelerator card, a key negotiation instruction is constructed and sent to the first target AI accelerator card so that the first target AI accelerator card can perform key negotiation with the second target AI accelerator card based on the key negotiation instruction.

[0007] According to another aspect of the present invention, a key negotiation device is provided, applied in an artificial intelligence (AI) acceleration device, the AI ​​acceleration device including multiple AI acceleration cards, the device comprising: The first accelerator card determination module is used to determine multiple second target AI accelerator cards based on the key negotiation instruction received by the first target AI accelerator card from the electronic device, for any one of the multiple AI accelerator cards as a first target AI accelerator card. The second target AI accelerator cards are AI accelerator cards that need to perform key negotiation with the first target AI accelerator card. The instruction sending module is used to send key negotiation instructions to the plurality of second target AI accelerator cards respectively; The negotiation algorithm determination module is used to determine the target negotiation algorithm it supports from the at least one negotiation algorithm after receiving at least one negotiation algorithm sent by any second target AI acceleration card through the first target AI acceleration card; A key determination module is used to determine, using the target negotiation algorithm, the key between the first target AI accelerator card and any one of the second target AI accelerator cards.

[0008] According to another aspect of the present invention, a key negotiation apparatus is provided, the apparatus comprising: The acquisition module is used to acquire a pre-built network topology map, wherein the network topology map includes the connection relationship between multiple AI acceleration cards in the artificial intelligence AI acceleration device; The second accelerator card determination module is used to determine, for any one of the multiple AI accelerator cards, the second target AI accelerator card corresponding to the first target AI accelerator card using the network topology map, wherein the second target AI accelerator card is an AI accelerator card that needs to perform key negotiation with the first target AI accelerator card. The instruction construction module is used to construct a key negotiation instruction based on the first target AI accelerator card and the second target AI accelerator card, and send the key negotiation instruction to the first target AI accelerator card so that the first target AI accelerator card can perform key negotiation with the second target AI accelerator card based on the key negotiation instruction.

[0009] According to another aspect of the present invention, a computer program product is also provided, including a computer program that, when executed by a processor, implements the steps of the key negotiation method as described in any embodiment of the present invention.

[0010] The technical solution of this invention addresses a plurality of AI accelerator cards. For any one first target AI accelerator card, after receiving a key negotiation instruction from an electronic device, the first target AI accelerator card determines a plurality of second target AI accelerator cards that need to negotiate keys with it. Key negotiation instructions are then sent to each of the second target AI accelerator cards. After receiving at least one negotiation algorithm from any one of the second target AI accelerator cards, the first target AI accelerator card determines its supported target negotiation algorithm from the at least one negotiation algorithm. Finally, the target negotiation algorithm is used to determine the key between the first target AI accelerator card and any one of the second target AI accelerator cards. Therefore, this embodiment of the application achieves key negotiation between multiple AI accelerator cards, and data transmission between any two AI accelerator cards is encrypted, reducing information leakage and improving data security.

[0011] It should be understood that the description in this section is not intended to identify key or essential features of the embodiments of the present invention, nor is it intended to limit the scope of the invention. Other features of the invention will become readily apparent from the following description. Attached Figure Description

[0012] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0013] Figure 1 This is a schematic diagram illustrating an application scenario of the key negotiation method provided in an embodiment of the present invention; Figure 2 This is one of the flowcharts illustrating a key negotiation method provided according to an embodiment of the present invention; Figure 3 This is a second schematic flowchart of the key negotiation method provided in the embodiment of the present invention; Figure 4 This is a schematic diagram of a key negotiation method between multiple AI accelerator cards according to an embodiment of the present invention; Figure 5 This is one of the schematic diagrams of a key negotiation device provided according to an embodiment of the present invention; Figure 6 This is a second schematic diagram of a key negotiation device provided according to an embodiment of the present invention; Figure 7 This is a schematic diagram of an electronic device provided according to an embodiment of the present invention. Detailed Implementation

[0014] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0015] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0016] In existing technologies, only single-card computing scenarios exist. In these scenarios, confidential computing technologies involve key negotiation between the CPU and the accelerator card to generate a shared key for encryption and decryption during data transmission. However, as model size increases, so does the size of AI accelerator card clusters. In large-scale model training scenarios, there are significant data exchange needs between multiple AI accelerator cards. This data exchange uses dedicated high-speed channels. If plaintext data is transmitted directly through these channels, there is a risk of confidential information being leaked, making it impossible to guarantee data security for large-scale multi-card cluster training. Therefore, a key negotiation method between multiple AI accelerator cards is urgently needed.

[0017] In view of this, this application provides a key negotiation method. For any one of multiple AI accelerator cards as a first target AI accelerator card, after receiving a key negotiation instruction from an electronic device, the first target AI accelerator card determines multiple second target AI accelerator cards that need to negotiate keys with it based on the key negotiation instruction; and sends key negotiation instructions to each of the multiple second target AI accelerator cards respectively; after receiving at least one negotiation algorithm from any one of the second target AI accelerator cards, the first target AI accelerator card determines its supported target negotiation algorithm from the at least one negotiation algorithm; and uses the target negotiation algorithm to determine the key between the first target AI accelerator card and any one of the second target AI accelerator cards. Therefore, this embodiment of the application realizes key negotiation between multiple AI accelerator cards, and data transmission between any two AI accelerator cards is encrypted, reducing information leakage and improving data security.

[0018] Before introducing the key negotiation method provided in the embodiments of this application, for ease of understanding, the technical background of the embodiments of this application will be described in detail below.

[0019] Before introducing the key negotiation method in this application, we will first introduce the application scenarios corresponding to the key negotiation method in this application. Figure 1 This is a diagram illustrating the application scenario, from Figure 1 As can be seen, this application scenario includes server 110 and AI acceleration device 120. The server 110 in this embodiment can be implemented as a physical server or a virtual server. The electronic device in this embodiment can be a server, but this embodiment does not limit the electronic device. The AI ​​acceleration device in this embodiment includes multiple AI acceleration cards.

[0020] In one possible embodiment, server 110 obtains a pre-constructed network topology map, wherein the network topology map includes the connection relationships between multiple AI accelerator cards in AI acceleration device 120; for any one of the multiple AI accelerator cards as a first target AI accelerator card, server 110 uses the network topology map to determine a second target AI accelerator card corresponding to the first target AI accelerator card, wherein the second target AI accelerator card is an AI accelerator card that needs to perform key negotiation with the first target AI accelerator card; server 110 then constructs a key negotiation instruction based on the first target AI accelerator card and the second target AI accelerator card, and sends the key negotiation instruction to the first target AI accelerator card, so that the first target AI accelerator card can perform key negotiation with the second target AI accelerator card based on the key negotiation instruction. The AI ​​acceleration device 120, for any one of the plurality of AI acceleration cards, a first target AI acceleration card, after receiving a key negotiation instruction from an electronic device through the first target AI acceleration card, determines a plurality of second target AI acceleration cards based on the key negotiation instruction, wherein the second target AI acceleration card is the AI ​​acceleration card that needs to negotiate the key with the first target AI acceleration card; and sends key negotiation instructions to each of the plurality of second target AI acceleration cards; after receiving at least one negotiation algorithm from any one of the second target AI acceleration cards through the first target AI acceleration card, it determines the target negotiation algorithm it supports from the at least one negotiation algorithm; and uses the target negotiation algorithm to determine the key between the first target AI acceleration card and any one of the second target AI acceleration cards.

[0021] in, Figure 1 The server 110 and the AI ​​acceleration device 120 can interact via a communication network. The communication network can use either wireless or wired communication.

[0022] For example, server 110 can access the network via cellular mobile communication technology to communicate with AI acceleration device 120, wherein the cellular mobile communication technology includes, for example, 5th generation mobile networks (5G) technology.

[0023] Optionally, the server 110 can access the network and communicate with the AI ​​acceleration device 120 via short-range wireless communication, wherein the short-range wireless communication method includes, for example, Wireless Fidelity (Wi-Fi) technology.

[0024] Furthermore, the description in this application focuses only on a single server 110 and a single AI acceleration device 120. However, those skilled in the art should understand that the illustrated server 110 and AI acceleration device 120 are intended to indicate that the technical solution of this application involves the operation of the server 110 and AI acceleration device 120, and do not imply any limitation on the number, type, or location of the server 110 and AI acceleration device 120. It should be noted that adding additional modules to or removing individual modules from the illustrated environment will not change the underlying concept of the exemplary embodiments of this application.

[0025] The following describes an exemplary implementation of the key negotiation method of this application in conjunction with the application scenarios described above and with reference to the accompanying drawings. It should be noted that the above application scenarios are only shown to facilitate understanding of the methods and principles of this application, and the implementation of this application is not limited in any way in this respect.

[0026] Combination Figure 1 In the application scenarios, the key negotiation method in the embodiments of this application will be described in detail below. Figure 2 A flowchart of a key negotiation method provided in an embodiment of the present invention, applied in an electronic device, is shown below. Figure 2 As can be seen from this, it may specifically include the following steps: S210: Obtain a pre-constructed network topology diagram, wherein the network topology diagram includes the connection relationship between multiple AI acceleration cards in the artificial intelligence (AI) acceleration device; In this embodiment of the application, before executing S210, high-speed channels on all AI accelerator cards are created and initialized, and key negotiation between any two AI accelerator cards is carried out through the already created high-speed channels.

[0027] S220: For any one of the multiple AI accelerator cards, determine the second target AI accelerator card corresponding to the first target AI accelerator card using the network topology diagram, wherein the second target AI accelerator card is the AI ​​accelerator card that needs to perform key negotiation with the first target AI accelerator card. In one possible implementation, S220 can be specifically implemented as follows: for any first target AI accelerator card, search for other AI accelerator cards connected to the first target AI accelerator card from the network topology graph; form key negotiation sets with the first target AI accelerator card respectively; for any key negotiation set, compare the stored key negotiation sets corresponding to the plurality of AI accelerator cards with the arbitrary key negotiation set; if there is a set in the stored key negotiation set that is the same as the arbitrary key negotiation set, delete the key negotiation set; if there is no set in the stored key negotiation set that is the same as the arbitrary key negotiation set, determine the other AI accelerator cards in the key negotiation set as the second target AI accelerator card corresponding to the first target AI accelerator card.

[0028] For example, if the key negotiation set stored in the AI ​​accelerator card includes (a, b), (a, c), and (a, d), and the determined key negotiation set is (b, a), then the determined key negotiation set (b, a) will be deleted. If the determined key negotiation set is (b, d), then the key accelerator card d will be determined as the second target AI accelerator card corresponding to the first target AI accelerator card a.

[0029] S230: Based on the first target AI accelerator card and the second target AI accelerator card, construct a key negotiation instruction and send the key negotiation instruction to the first target AI accelerator card so that the first target AI accelerator card can perform key negotiation with the second target AI accelerator card based on the key negotiation instruction.

[0030] The key negotiation instructions in this application embodiment include a second target AI computing card.

[0031] This application also provides a key negotiation method, applied in AI acceleration devices, such as... Figure 3 The diagram shown illustrates the key negotiation method, which may include the following steps: S310: For any one of the plurality of AI accelerator cards, after receiving a key negotiation instruction sent by an electronic device through the first target AI accelerator card, a plurality of second target AI accelerator cards are determined based on the key negotiation instruction, wherein the second target AI accelerator card is an AI accelerator card that needs to perform key negotiation with the first target AI accelerator card; The key negotiation instructions in this application embodiment include multiple second target AI accelerator cards.

[0032] S320: Send key negotiation instructions to the plurality of second target AI accelerator cards respectively through the first target AI accelerator card; In one possible implementation, before executing S320, an authentication request is sent from the first target AI accelerator card to the second target AI accelerator card; after receiving the digital certificate sent by the second target AI accelerator card, the first target AI accelerator card performs authentication based on the digital certificate; upon successful authentication, the first target AI accelerator card sends an authentication success instruction to the second target AI accelerator card; upon receiving the authentication request from the second target AI accelerator card, the first target AI accelerator card sends its own digital certificate to the second target AI accelerator card, so that the second target AI accelerator card can authenticate the first target AI accelerator card based on the digital certificate; upon receiving the authentication success instruction from the second target AI accelerator card, the step of sending key negotiation instructions to the plurality of second target AI accelerator cards is executed.

[0033] The authentication method in this application embodiment is the authentication method in the prior art, and will not be described in detail here.

[0034] S330: After receiving at least one negotiation algorithm sent by any second target AI acceleration card through the first target AI acceleration card, determine the target negotiation algorithm supported by itself from the at least one negotiation algorithm; In one possible implementation, after executing S330, the target negotiation algorithm is sent to the second target AI accelerator card so that the second target AI accelerator card can use the target negotiation algorithm to determine the key between the first target AI accelerator card and the second target AI accelerator card, and broadcast the key to the port of the second target AI accelerator card, wherein the key determined by the second target AI accelerator card is the same as the key determined by the first target AI accelerator card.

[0035] It should be noted that the key negotiation algorithm in this application embodiment can be set according to the specific actual situation, and this application embodiment does not limit the key negotiation algorithm.

[0036] S340: Determine the key between the first target AI accelerator card and any one of the second target AI accelerator cards using the target negotiation algorithm through the first target AI accelerator card.

[0037] In one possible implementation, after executing S340, the key is broadcast to the port of the first target AI accelerator card via the first target AI accelerator card.

[0038] In one possible implementation, key negotiation between multiple sets of AI accelerator cards is processed in parallel, wherein any set of AI accelerator cards includes a first target AI accelerator card and a second target AI accelerator card. This improves the efficiency of key negotiation.

[0039] Therefore, this application embodiment implements a key negotiation method between multiple AI accelerator cards, for example, such as Figure 4 The diagram shown illustrates key negotiation between multiple AI accelerator cards. Figure 4 As can be seen, key negotiation is performed between any two AI accelerator cards.

[0040] Based on the same inventive concept, this application also provides a key negotiation device. Figure 5 This is a schematic diagram of the key negotiation device. Figure 5 As shown, the device 500 includes: The first accelerator card determination module 510 is used to determine multiple second target AI accelerator cards based on the key negotiation instruction received by the first target AI accelerator card from the electronic device, for any one of the multiple AI accelerator cards as a first target AI accelerator card. The second target AI accelerator card is an AI accelerator card that needs to perform key negotiation with the first target AI accelerator card. The instruction sending module 520 is used to send key negotiation instructions to the plurality of second target AI acceleration cards respectively; The negotiation algorithm determination module 530 is used to determine the target negotiation algorithm it supports from the at least one negotiation algorithm after receiving at least one negotiation algorithm sent by any second target AI acceleration card through the first target AI acceleration card. The key determination module 540 is used to determine the key between the first target AI accelerator card and any one of the second target AI accelerator cards using the target negotiation algorithm.

[0041] In one possible embodiment, the device further includes: an identity authentication module 550, used for: Before sending the key negotiation instruction to the plurality of second target AI accelerator cards respectively, an authentication request is sent to the second target AI accelerator cards through the first target AI accelerator card; After receiving the digital certificate sent by the second target AI acceleration card through the first target AI acceleration card, identity authentication is performed based on the digital certificate; Once identity authentication is successful, a successful identity authentication command is sent from the first target AI acceleration card to the second target AI acceleration card. The first target AI accelerator card receives the identity authentication request sent by the second target AI accelerator card, and sends its own digital certificate to the second target AI accelerator card so that the second target AI accelerator card can authenticate the first target AI accelerator card based on the digital certificate; After receiving the authentication success instruction from the second target AI accelerator card through the first target AI accelerator card, the step of sending key negotiation instructions to the plurality of second target AI accelerator cards is executed.

[0042] In one possible embodiment, the device further includes: The broadcast module 560 is used to broadcast the key to the port of the first target AI accelerator card after determining the key between the first target AI accelerator card and the second target AI accelerator card using the target negotiation algorithm.

[0043] In one possible embodiment, key negotiation between multiple sets of AI accelerator cards is processed in parallel, wherein any set of AI accelerator cards includes a first target AI accelerator card and a second target AI accelerator card.

[0044] In one possible embodiment, the device further includes: The sending module 570 is configured to, after determining the target negotiation algorithm it supports from the at least one negotiation algorithm, send the target negotiation algorithm to the second target AI accelerator card, so that the second target AI accelerator card can use the target negotiation algorithm to determine the key between the first target AI accelerator card and the second target AI accelerator card, and broadcast the key to the port of the second target AI accelerator card, wherein the key determined by the second target AI accelerator card is the same as the key determined by the first target AI accelerator card.

[0045] Based on the same inventive concept, this application also provides another key negotiation device. Figure 6 This is a schematic diagram of the key negotiation device. Figure 6 As shown, the device 600 includes: The acquisition module 610 is used to acquire a pre-constructed network topology map, wherein the network topology map includes the connection relationship between multiple AI acceleration cards in the artificial intelligence AI acceleration device; The second accelerator card determination module 620 is used to determine, using the network topology map, a second target AI accelerator card corresponding to any one of the first target AI accelerator cards among a plurality of AI accelerator cards, wherein the second target AI accelerator card is an AI accelerator card that needs to perform key negotiation with the first target AI accelerator card. The instruction construction module 630 is used to construct a key negotiation instruction based on the first target AI accelerator card and the second target AI accelerator card, and send the key negotiation instruction to the first target AI accelerator card so that the first target AI accelerator card can perform key negotiation with the second target AI accelerator card based on the key negotiation instruction.

[0046] In one possible implementation, the second accelerator card determining module 620 is specifically used for: For any first target AI accelerator card, search the network topology map for other AI accelerator cards connected to the first target AI accelerator card; The other AI accelerator cards and the first target AI accelerator card are respectively formed into key negotiation sets; For any given key negotiation set, compare the stored key negotiation sets corresponding to the multiple AI accelerator cards with the given key negotiation set. If there exists a set in the stored key negotiation set that is identical to any of the key negotiation sets, then the key negotiation set is deleted. If there is no set in the stored key negotiation set that is the same as any of the key negotiation sets, then the other AI accelerator cards in the key negotiation set are determined as the second target AI accelerator card corresponding to the first target AI accelerator card.

[0047] The key negotiation device provided in the embodiments of the present invention can execute the collective communication method provided in any embodiment of the present invention, and has the corresponding functional modules and beneficial effects of the execution method.

[0048] The collection, storage, use, processing, transmission, provision, and disclosure of user personal information involved in the technical solution disclosed herein comply with the provisions of relevant laws and regulations and do not violate public order and good morals.

[0049] Figure 7 A schematic diagram of an electronic device 10, which can be used to implement embodiments of the present invention, is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital processors, cellular phones, smartphones, wearable devices (e.g., helmets, glasses, watches, etc.), and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely illustrative and are not intended to limit the implementation of the invention described and / or claimed herein.

[0050] like Figure 7As shown, the electronic device 10 includes at least one processor 11 and a memory, such as a read-only memory (ROM) 12 or a random access memory (RAM) 13, communicatively connected to the at least one processor 11. The memory stores computer programs executable by the at least one processor. The processor 11 can perform various appropriate actions and processes based on the computer program stored in the ROM 12 or loaded from storage unit 18 into the RAM 13. The RAM 13 can also store various programs and data required for the operation of the electronic device 10. The processor 11, ROM 12, and RAM 13 are interconnected via a bus 14. An input / output (I / O) interface 15 is also connected to the bus 14.

[0051] Multiple components in electronic device 10 are connected to I / O interface 15, including: input unit 16, such as keyboard, mouse, etc.; output unit 17, such as various types of displays, speakers, etc.; storage unit 18, such as disk, optical disk, etc.; and communication unit 19, such as network card, modem, wireless transceiver, etc. Communication unit 19 allows electronic device 10 to exchange information / data with other devices through computer networks such as the Internet and / or various telecommunications networks.

[0052] Processor 11 can be a variety of general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of processor 11 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special-purpose artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, a digital signal processor (DSP), and any suitable processor, controller, microcontroller, etc. Processor 11 performs the various methods and processes described above, such as key negotiation methods.

[0053] In some embodiments, the key negotiation method may be implemented as a computer program tangibly contained in a computer-readable storage medium, such as a storage unit. In some embodiments, part or all of the computer program may be loaded into and / or installed on an electronic device via ROM and / or a communication unit. When the computer program is loaded into RAM and executed by a processor, one or more steps of the key negotiation method may be performed. Alternatively, in other embodiments, the processor may be configured to perform the key negotiation method by any other suitable means (e.g., by means of firmware).

[0054] Various embodiments of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field-programmable gate arrays (FPGAs), application-specific integrated circuits (ASICs), application-specific standard products (ASSPs), systems-on-a-chip (SoCs), payload-programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various embodiments may include implementations in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which may be a dedicated or general-purpose programmable processor, capable of receiving data and instructions from a storage system, at least one input device, and at least one output device, and transmitting data and instructions to the storage system, the at least one input device, and the at least one output device.

[0055] Computer programs used to implement the methods of the present invention may be written in any combination of one or more programming languages. These computer programs may be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, such that when executed by the processor, the computer programs cause the functions / operations specified in the flowcharts and / or block diagrams to be performed. The computer programs may be executed entirely on a machine, partially on a machine, or as a standalone software package, partially on a machine and partially on a remote machine, or entirely on a remote machine or server.

[0056] In the context of this invention, a computer-readable storage medium can be a tangible medium that may contain or store a computer program for use by or in conjunction with an instruction execution system, apparatus, or device. A computer-readable storage medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination thereof. Alternatively, a computer-readable storage medium may be a machine-readable signal medium. More specific examples of machine-readable storage media include electrical connections based on one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fibers, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof.

[0057] To provide interaction with a user, the systems and techniques described herein can be implemented on an electronic device having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and pointing device (e.g., a mouse or trackball) through which the user provides input to the electronic device. Other types of devices can also be used to provide interaction with the user; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including sound input, voice input, or tactile input).

[0058] The systems and technologies described herein can be implemented in computing systems that include backend components (e.g., as data servers), or middleware components (e.g., application servers), or frontend components (e.g., user computers with graphical user interfaces or web browsers through which users can interact with implementations of the systems and technologies described herein), or any combination of such backend, middleware, or frontend components. The components of the system can be interconnected via digital data communication of any form or medium (e.g., communication networks). Examples of communication networks include local area networks (LANs), wide area networks (WANs), blockchain networks, and the Internet.

[0059] A computing system can include clients and servers. Clients and servers are generally located far apart and typically interact through communication networks. The client-server relationship is created by computer programs running on the respective computers and having a client-server relationship with each other. The server can be a cloud server, also known as a cloud computing server or cloud host, which is a hosting product within the cloud computing service system to address the shortcomings of traditional physical hosts and VPS services, such as high management difficulty and weak business scalability.

[0060] It should be understood that the various forms of processes shown above can be used, with steps reordered, added, or deleted. For example, the steps described in this invention can be executed in parallel, sequentially, or in different orders, as long as the desired result of the technical solution of this invention can be achieved, and this is not limited herein.

[0061] The specific embodiments described above do not constitute a limitation on the scope of protection of this invention. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this invention should be included within the scope of protection of this invention.

Claims

1. A key negotiation method, characterized in that, The method is applied in an artificial intelligence (AI) acceleration device, which includes multiple AI acceleration cards, and includes: For any one of the plurality of AI accelerator cards as a first target AI accelerator card, after receiving a key negotiation instruction sent by an electronic device through the first target AI accelerator card, a plurality of second target AI accelerator cards are determined based on the key negotiation instruction, wherein the second target AI accelerator card is an AI accelerator card that needs to perform key negotiation with the first target AI accelerator card; And send key negotiation instructions to the multiple second target AI acceleration cards respectively; After receiving at least one negotiation algorithm from any second target AI acceleration card through the first target AI acceleration card, the target negotiation algorithm supported by itself is determined from the at least one negotiation algorithm; The target negotiation algorithm is then used to determine the key between the first target AI accelerator card and any one of the second target AI accelerator cards.

2. The method according to claim 1, characterized in that, Before sending key negotiation instructions to the plurality of second target AI accelerator cards respectively, the method further includes: Send an authentication request to the second target AI accelerator card through the first target AI accelerator card; After receiving the digital certificate sent by the second target AI acceleration card through the first target AI acceleration card, identity authentication is performed based on the digital certificate; Once identity authentication is successful, a successful identity authentication command is sent from the first target AI acceleration card to the second target AI acceleration card. The first target AI accelerator card receives the identity authentication request sent by the second target AI accelerator card, and sends its own digital certificate to the second target AI accelerator card so that the second target AI accelerator card can authenticate the first target AI accelerator card based on the digital certificate; After receiving the authentication success instruction from the second target AI accelerator card through the first target AI accelerator card, the step of sending key negotiation instructions to the plurality of second target AI accelerator cards is executed.

3. The method according to claim 1, characterized in that, After determining the key between the first target AI accelerator card and the second target AI accelerator card using the target negotiation algorithm, the method further includes: The key is broadcast to the port of the first target AI accelerator card via the first target AI accelerator card.

4. The method according to claim 1, characterized in that, Key negotiation between multiple AI accelerator cards is processed in parallel, wherein any group of AI accelerator cards includes a first target AI accelerator card and a second target AI accelerator card.

5. The method according to claim 1, characterized in that, After determining the target negotiation algorithm it supports from the at least one negotiation algorithm, the method further includes: The target negotiation algorithm is sent to the second target AI accelerator card so that the second target AI accelerator card can use the target negotiation algorithm to determine the key between the first target AI accelerator card and the second target AI accelerator card, and broadcast the key to the port of the second target AI accelerator card, wherein the key determined by the second target AI accelerator card is the same as the key determined by the first target AI accelerator card.

6. A key negotiation method, characterized in that, When applied to electronic devices, the method includes: Obtain a pre-constructed network topology map, wherein the network topology map includes the connection relationship between multiple AI accelerator cards in the artificial intelligence (AI) acceleration device; For any one of the multiple AI accelerator cards as a first target AI accelerator card, the network topology map is used to determine the second target AI accelerator card corresponding to the first target AI accelerator card, wherein the second target AI accelerator card is the AI ​​accelerator card that needs to perform key negotiation with the first target AI accelerator card. Based on the first target AI accelerator card and the second target AI accelerator card, a key negotiation instruction is constructed and sent to the first target AI accelerator card so that the first target AI accelerator card can perform key negotiation with the second target AI accelerator card based on the key negotiation instruction.

7. The method according to claim 6, characterized in that, The step of using the network topology map to determine the second target AI accelerator card corresponding to the first target AI accelerator card includes: For any first target AI accelerator card, search the network topology map for other AI accelerator cards connected to the first target AI accelerator card. The other AI accelerator cards and the first target AI accelerator card are respectively formed into key negotiation sets; For any given key negotiation set, compare the stored key negotiation sets corresponding to the multiple AI accelerator cards with the given key negotiation set. If there exists a set in the stored key negotiation set that is identical to any of the key negotiation sets, then the key negotiation set is deleted. If there is no set in the stored key negotiation set that is the same as any of the key negotiation sets, then the other AI accelerator cards in the key negotiation set are determined as the second target AI accelerator card corresponding to the first target AI accelerator card.

8. A key negotiation device, characterized in that, Applied in artificial intelligence (AI) acceleration devices, the AI ​​acceleration devices include multiple AI acceleration cards, and the device includes: The first accelerator card determination module is used to determine multiple second target AI accelerator cards based on the key negotiation instruction received by the first target AI accelerator card from the electronic device, for any one of the multiple AI accelerator cards as a first target AI accelerator card. The second target AI accelerator cards are AI accelerator cards that need to perform key negotiation with the first target AI accelerator card. The instruction sending module is used to send key negotiation instructions to the plurality of second target AI accelerator cards respectively; The negotiation algorithm determination module is used to determine the target negotiation algorithm it supports from the at least one negotiation algorithm after receiving at least one negotiation algorithm sent by any second target AI acceleration card through the first target AI acceleration card; A key determination module is used to determine, using the target negotiation algorithm, the key between the first target AI accelerator card and any one of the second target AI accelerator cards.

9. A key negotiation device, characterized in that, The device, used in electronic devices, includes: The acquisition module is used to acquire a pre-built network topology map, wherein the network topology map includes the connection relationship between multiple AI acceleration cards in the artificial intelligence AI acceleration device; The second accelerator card determination module is used to determine, for any one of the multiple AI accelerator cards, the second target AI accelerator card corresponding to the first target AI accelerator card using the network topology map, wherein the second target AI accelerator card is an AI accelerator card that needs to perform key negotiation with the first target AI accelerator card. The instruction construction module is used to construct a key negotiation instruction based on the first target AI accelerator card and the second target AI accelerator card, and send the key negotiation instruction to the first target AI accelerator card so that the first target AI accelerator card can perform key negotiation with the second target AI accelerator card based on the key negotiation instruction.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions that cause a processor to execute the key negotiation method according to any one of claims 1-7.

Images