Computing system, artificial intelligence external device and operating method thereof

Abstract

This invention proposes a computing system, an artificial intelligence (AI) external device, and an operating method thereof. The computing system includes an electronic device and an AI external device. The electronic device includes an input unit, a display unit, a first transmission interface, and a first computing unit. The AI ​​external device is pluggably mounted on the electronic device. The AI ​​external device includes a second transmission interface and an AI model storage unit. The AI ​​model storage unit stores AI models. When the first computing unit receives an input command through the input unit, the AI ​​external device executes the AI ​​model and inputs a first tag command generated based on the input command into the AI ​​model to generate a second tag command. The first computing unit displays the response content through the display unit based on the output information generated based on the second tag command.

Description

Technical Field

[0001] This invention relates to a computing technology, and more particularly to a computing system, an artificial intelligence external device, and a method of operating the same. Prior Technology

[0002] Traditional artificial intelligence (AI) computing involves users connecting to a network via computer and then linking to an AI computing model hosted on a cloud server. Input data is transmitted to the AI ​​computing model over the network, and the results are received from the AI ​​computing model via the same network. Therefore, traditional AI computing scenarios are prone to high data latency and low privacy issues. Summary of the Invention

[0003] This invention provides a computing system, an artificial intelligence external device, and its operating method, which can realize convenient artificial intelligence computing functions.

[0004] The computing system of the present invention includes an electronic device and an artificial intelligence (AI) external device. The electronic device includes an input unit, a display unit, a first transmission interface, and a first computing unit. The first computing unit is coupled to the input unit, the display unit, and the first transmission interface. The AI ​​external device is pluggably disposed in the electronic device. The AI ​​external device includes a second transmission interface and an AI model storage unit. The AI ​​model storage unit is coupled to the second transmission interface and stores the AI ​​model. When the first computing unit receives an input command through the input unit, the AI ​​external device executes the AI ​​model and inputs a first tag command generated based on the input command into the AI ​​model to generate a second tag command. The first computing unit displays the response content through the display unit based on the output information generated based on the second tag command.

[0005] The artificial intelligence (AI) external device of the present invention is suitable for being installed in an electronic device in a pluggable manner. The electronic device includes an input unit, a display unit, a first transmission interface, and a first processing unit. The AI ​​external device includes a second transmission interface and an AI model storage unit. The AI ​​model storage unit is coupled to the second transmission interface and stores the AI ​​model. When the first processing unit receives an input command through the input unit, the AI ​​external device executes the AI ​​model and inputs a first tag command generated based on the input command into the AI ​​model to generate a second tag command. The first processing unit displays the response content through the display unit based on the output information generated based on the second tag command.

[0006] The operation method of the artificial intelligence external device of the present invention includes the following steps: when the first computing unit of the electronic device receives an input instruction through the input unit, the artificial intelligence model is executed through the artificial intelligence external device; a first tag instruction generated based on the input instruction is input into the artificial intelligence model to generate a second tag instruction; and the first computing unit of the electronic device displays the response content through the display unit according to the output information generated based on the second tag instruction.

[0007] Based on the above, the computing system, artificial intelligence external device and its operation method of the present invention can achieve highly secure artificial intelligence computing functions with low device cost through the artificial intelligence external device.

[0008] To make the above features and advantages of the present invention more apparent and understandable, specific embodiments are described below in conjunction with the accompanying drawings for detailed explanation. Simple Explanation of the Diagram

[0009] Figure 1 is a schematic diagram of a computing system according to an embodiment of the present invention. Figure 2 is a flowchart of an operation method according to an embodiment of the present invention. Figure 3 is a schematic diagram of a computing system according to an embodiment of the present invention. Figure 4 is a schematic diagram of an electronic device and an artificial intelligence external device according to an embodiment of the present invention. Figure 5 is a schematic diagram of the operation of a computing system according to an embodiment of the present invention. Figure 6 is a flowchart of an operation method according to an embodiment of the present invention. Figure 7 is a schematic diagram of a computing system according to an embodiment of the present invention. Figure 8 is a schematic diagram of an electronic device and an artificial intelligence external device according to an embodiment of the present invention. Figure 9 is a flowchart of an operation method according to an embodiment of the present invention. Figure 10 is a schematic diagram of a computing system according to an embodiment of the present invention. Figure 11 is a schematic diagram of an electronic device and an artificial intelligence external device according to an embodiment of the present invention. Figure 12 is a schematic diagram of a computing system according to an embodiment of the present invention. Figure 13 is a schematic diagram of an electronic device and an artificial intelligence external device according to an embodiment of the present invention. Implementation

[0010] To make the contents of this invention more readily apparent, the following specific embodiments are provided as examples on which this invention can indeed be implemented. Furthermore, wherever possible, elements / components / steps using the same reference numerals in the drawings and embodiments represent the same or similar parts.

[0011] Figure 1 is a schematic diagram of a computing system according to an embodiment of the present invention. Referring to Figure 1, the computing system 100 includes an electronic device 110 and an artificial intelligence (AI) external device 120. The electronic device 110 is coupled to the AI ​​external device 120. The electronic device 110 includes a first computing unit 111, a first transmission interface 112, an input unit 113, and a display unit 114. The first computing unit 111 is coupled to the first transmission interface 112, the input unit 113, and the display unit 114. The AI ​​external device 120 includes a second transmission interface 122 and an AI model storage unit 123. The second transmission interface 122 is coupled to the AI ​​model storage unit 123. In this embodiment, the AI ​​external device 120 can be pluggably disposed in the electronic device 110 to provide AI computing functions. In one embodiment, the AI ​​external device 120 can also be pluggably embedded in the electronic device 110.

[0012] In this embodiment, the first processing unit 111 may include, for example, a central processing unit (CPU), or other programmable general-purpose or special-purpose microprocessors, digital signal processors (DSPs), image processing units (IPUs), graphics processing units (GPUs), programmable controllers, application-specific integrated circuits (ASICs), programmable logic devices (PLDs), other similar processing devices, or combinations thereof.

[0013] In this embodiment, the first transmission interface 112 may include, for example, a female terminal, and the second transmission interface 122 may include, for example, a male terminal. When the AI ​​external device 120 is disposed in the electronic device 110, the AI ​​external device 120 can connect to the first transmission interface 112 through the second transmission interface 122 to form a communication path. The processing unit of the electronic device 110 can determine whether the second transmission interface 122 belongs to an external AI database. If so, the processing unit of the electronic device 110 can connect to the AI ​​external device 120. The electronic device 110 can transmit power to the AI ​​external device 120 through the first transmission interface 112 and the second transmission interface 122 to drive the relevant internal units of the AI ​​external device 120.

[0014] In this embodiment, the input unit 113 may include, for example, a mouse, keyboard, camera, touchpad, microphone, or communication module, but the invention is not limited thereto. In one embodiment, the input unit 113 may be, for example, a standalone touch panel or integrated with a display panel. In this embodiment, the display unit 114 may include a display panel and a display driver.

[0015] In this embodiment, the size of the artificial intelligence external device 120 is smaller than the size of the electronic device 110. The electronic device 110 may include a laptop, tablet, or smartphone. The artificial intelligence external device 120 may include a USB flash drive, a portable hard drive, or a portable solid-state drive (SSD). In this embodiment, the artificial intelligence external device 120 can be implemented at a lower cost while conforming to existing architectures. The size of the artificial intelligence model or language model stored in the artificial intelligence external device 120 may be related to the number of parameters it uses. Generally, the capacity of a small generative artificial intelligence model will not exceed 1TB, in line with the capacity size of existing portable storage devices.

[0016] In one embodiment, the artificial intelligence external device 120 can be a USB flash drive, which uses flash memory storage technology and has the advantages of being very small, typically only the size of a thumb, and with a capacity generally ranging from several GB to several hundred GB. The storage speed of a USB flash drive is typically between 10-100 MB / s. USB flash drives are suitable for storing and transferring small amounts of data such as files, photos, and music.

[0017] In one embodiment, the artificial intelligence external device 120 can be a portable hard drive, and uses hard disk drive (HDD) technology. The capacity of the portable hard drive can range from several hundred GB to several TB. The read and write speed of the portable hard drive is typically between 50-150 MB / s. The portable hard drive is suitable for backing up large amounts of data, such as videos, photos, or games.

[0018] In one embodiment, the artificial intelligence external device 120 can be a portable solid-state drive (SSD) using SSD technology, and its internal memory is also flash memory. The size of a portable SSD is typically about the size of a mobile phone. The capacity of a portable SSD can range from several hundred GB to several TB. The read and write speeds of a portable SSD are typically between 200-2000 MB / s.1 Portable SSDs are suitable for situations requiring fast access and transfer of large amounts of data, such as video editing or gaming.

[0019] In this embodiment, the artificial intelligence model storage unit 123 may include, for example, memory, such as Dynamic Random Access Memory (DRAM), Flash memory, or Non-Volatile Random Access Memory (NVRAM). In this embodiment, the artificial intelligence model storage unit 123 is used to store one or more artificial intelligence models, such as generative artificial intelligence (Generative AI) models, large language models (LM), generative adversarial networks, variational autoencoders, autoregressive models, or diffusion models. The artificial intelligence model storage unit 123 of the artificial intelligence external device 120 allows the language model to be separated from the electronic device 110. Therefore, users can easily select the type and size of the artificial intelligence model storage unit 123 according to the parameter requirements of different language models (e.g., legal / accounting / patent / customer data parameters). In this embodiment, the artificial intelligence external device 120 can independently store the language model in the artificial intelligence model storage unit 123, without sharing it with storage units storing other data, thus allowing users to conveniently update the data. The artificial intelligence models described in various embodiments of the present invention can, for example, combine a retrieval and generation model architecture to achieve retrieval-augmented generation (RAG) functionality. In this regard, when the artificial intelligence model receives a query, it uses advanced retrieval algorithms to search for query-related information from a large database, and the retrieved information is input into a generation model (such as GPT-3 or GPT-4) so ​​that the generation model can construct a response based on this information. Then, the artificial intelligence model can integrate retrieval and generation, and learn how to most effectively combine the retrieved information with its own generation capabilities to produce high-quality output. Therefore, the artificial intelligence model can cite specific and relevant information sources when answering queries, thereby improving the quality and accuracy of the answer.

[0020] Figure 2 is a flowchart of an operation method according to an embodiment of the present invention. Referring to Figures 1 and 2, the computing system 100 can execute the following steps S210 to S230. In step S210, when the first computing unit 111 of the electronic device 110 receives an input instruction through the input unit 113, the artificial intelligence external device 120 can execute the artificial intelligence model. In step S220, the electronic device 110 can input a first token generated based on the input instruction into the artificial intelligence model to generate a second token. In this embodiment, in step S230, the first computing unit 111 of the electronic device 110 can display the response content through the display unit 114 according to the output information generated based on the second token.

[0021] For example, the input instruction can be a prompt instruction. The input instruction can be voice or text input by the user with intent information, including the message "Help me organize our company's environmental strategy into 20 words." The first tag instruction can include multiple tags for multiple words broken down from the input instruction (one word as one tag), such as: "help," "me," "organize," "this," "company," "of," "environmental," "strategy," "to," "20," "words," and "within." The second tag instruction can be multiple tags for multiple other words generated by an artificial intelligence model based on the first tag instruction, such as: "this," "company," "of," "environmental," "strategy," "for," "sustainable," "development," "net zero," "carbon emissions," "recycling," "utilization," "green," "innovation," "enterprise," and "identification." The output information can be, for example, a combination of multiple words based on the second tag instruction, such as: "Our company's environmental strategy is 'sustainable development, net zero carbon emissions, recycling, green innovation, and corporate identification.'" Therefore, the display unit 114 can be used with the data dialog box 3141 and the above-mentioned output information to display a text response that corresponds to the user's voice or text input request (i.e., "help me organize our company's environmental protection strategy to within 20 words").

[0022] Accordingly, the computing system 100 of this embodiment can execute specific artificial intelligence models according to usage requirements. Since the artificial intelligence model is provided by the artificial intelligence external device 120, the electronic device 110 does not need additional hardware, and different artificial intelligence computing functions can be achieved by replacing different artificial intelligence external devices 120. The artificial intelligence model described in the various embodiments of the present invention can be an on-device AI. The artificial intelligence model can execute artificial intelligence tasks on the device itself, without relying on a remote server or cloud computing. Therefore, the computing system 100 and the artificial intelligence external device 120 of the present invention can have lower computing latency, better privacy protection, and can still execute artificial intelligence computing functions without a network connection. Furthermore, when the artificial intelligence external device 120 is computing, it can reduce the memory burden of the electronic device 110 because the language model and data can be loaded as needed; therefore, in the usage process, it is not necessary to load all the data into the memory of the electronic device 110. In one embodiment, the artificial intelligence model may be a language model, a neural network model, or an equivalent computational model. In another embodiment, the artificial intelligence model may be used to implement a text-to-text artificial intelligence model, but the present invention is not limited thereto. The artificial intelligence model described in the various embodiments of the present invention may also be other types of generative models.

[0023] Figure 3 is a schematic diagram of a computing system according to an embodiment of the present invention. Referring to Figure 3, the computing system 300 includes an electronic device 310 and an artificial intelligence (AI) external device 320. The electronic device 310 is coupled to the AI ​​external device 320. The electronic device 310 includes a first computing unit 311, a first transmission interface 312, an input unit 313, and a display unit 314. The first computing unit 311 is coupled to the first transmission interface 312, the input unit 313, and the display unit 314. The AI ​​external device 320 includes a second computing unit 321, a second transmission interface 322, an AI model storage unit 323, a biometric identification module 324, and an indicator unit 325. The second computing unit 321 is coupled to the second transmission interface 322, the AI ​​model storage unit 323, the biometric identification module 324, and the indicator unit 325. In this embodiment, the artificial intelligence external device 320 can be installed in the electronic device 310 in a pluggable manner to provide artificial intelligence computing functions.

[0024] In this embodiment, the second computing unit 321 may be, for example, a neural processing unit (NPU). In this embodiment, the biometric identification module 324 may include, for example, at least one of a fingerprint identification module, a voiceprint identification module, and a face identification module. In this embodiment, the indicating unit 325 may include a light-emitting unit, such as a light-emitting diode (LED), but the present invention is not limited thereto.

[0025] Figure 4 is a schematic diagram of an electronic device and an artificial intelligence external device according to an embodiment of the present invention. Referring to Figures 3 and 4, in this embodiment, the electronic device 310 may be, for example, a notebook computer, but the present invention is not limited thereto. In one embodiment, the electronic device 310 may be, for example, a tablet computer or a smartphone. In this embodiment, the electronic device 310 may include a first body 301 and a second body 302. The first body 301 and the second body 302 may be connected through a rotatable component. A display unit 314 may be disposed on a surface of the first body 301. An input unit 313 may be disposed on a surface of the second body 302. A first transmission interface 312 may be disposed in the side of the second body 302 and may be combined with a second transmission interface 322 of the artificial intelligence external device 320. A portion of the artificial intelligence external device 320 may be housed in the side of the second body 302, and a biometric identification module 324 may be exposed. The biometric identification module 324 may be, for example, a fingerprint identification module. When the artificial intelligence external device 320 is installed on the electronic device 310, the outer casing of the artificial intelligence external device 320 can be flush with the outer casing of the electronic device 310.

[0026] Figure 5 is a schematic diagram of the operation of a computing system according to an embodiment of the present invention. Figure 6 is a flowchart of the operation method according to an embodiment of the present invention. Referring to Figures 3, 5, and 6, the computing system 300 may, for example, operate the following steps S610 to S680. In step S610, the artificial intelligence external device 320 can determine whether the user's identity has been verified. The artificial intelligence external device 320 can determine whether the user's identity has authorization. If the user's identity has authorization, the artificial intelligence external device 320 can unlock to couple to the electronic device 310. As shown in Figure 5, the second computing unit 321 can provide an indicator light through the indicator unit 325 to prompt the user to press their finger 500 on the biometric identification module 324, so that the second computing unit 321 can obtain the user's fingerprint feature through the biometric identification module 324 to identify the user's identity. The second processing unit 321 can verify whether the user's identity is the pre-registered identity information, which can be configured or stored in the electronic device 310 or the artificial intelligence external device 320. If not, step S610 can be re-executed after a preset time interval. If yes, in step S620, the artificial intelligence external device 320 can unlock and access the artificial intelligence model storage unit 323, and output a start command to the electronic device 310. In this embodiment, before the user's identity is verified, the second processing unit 321 can temporarily withhold power to the artificial intelligence model storage unit 323 to ensure data security. Conversely, after the second processing unit 321 confirms the user's identity through the biometric identification module 324, the second processing unit 321 unlocks and accesses the artificial intelligence model storage unit 323, and can output a start command to the first transmission interface 312 through the second transmission interface 322.

[0027] In step S630, the electronic device 310 can display a data dialog box 3141. In step S640, the electronic device 310 can receive an input command. In step S650, the artificial intelligence external device 320 can generate a first tag command based on the input command. In this embodiment, the first processing unit 311 can execute an artificial intelligence assistant application according to a start command. As shown in FIG5, the first processing unit 311 can, for example, execute an artificial intelligence assistant application and can display the data dialog box 3141 through the display unit 314. The data dialog box 3141 may include an input box, an output box, and related dialogue information. The user can, for example, input input commands for related questions through the operation input unit 313 (keyboard). In this embodiment, the first processing unit 311 can output the input command to the second transmission interface 322 through the first transmission interface 312. The second processing unit 321 can obtain the input command from the second transmission interface 322 and analyze the input command to generate a first tag command.

[0028] In step S660, the artificial intelligence model storage unit 323 can generate a second tagging instruction according to the first tagging instruction. In this embodiment, the artificial intelligence model storage unit 323 can input the first tagging instruction into the artificial intelligence model, so that the artificial intelligence model outputs the second tagging instruction to the second processing unit 321. In step S670, the artificial intelligence external device 320 generates output information according to the second tagging instruction. The second processing unit 321 can generate output information according to the second tagging instruction and output the output information to the first transmission interface 312 through the second transmission interface 322.

[0029] In step S680, the electronic device 310 can display the response content according to the output information. In this embodiment, the electronic device 310 can display the corresponding response content through the data dialog box 3141 displayed by the display unit 314. Therefore, the computing system 300 and the operation method of this embodiment can realize effective artificial intelligence computing functions, and can provide the computing power and artificial intelligence model required for artificial intelligence computing through the artificial intelligence external device 320.

[0030] Figure 7 is a schematic diagram of a computing system according to an embodiment of the present invention. Referring to Figure 7, the computing system 700 includes an electronic device 710 and an artificial intelligence (AI) external device 720. The electronic device 710 is coupled to the AI ​​external device 720. The electronic device 710 includes a first computing unit 711, a first transmission interface 712, an input unit 713, a display unit 714, and a biometric identification module 715. The first computing unit 711 is coupled to the first transmission interface 712, the input unit 713, the display unit 714, and the biometric identification module 715. The AI ​​external device 720 includes a second transmission interface 722 and an AI model storage unit 723. The second transmission interface 722 is coupled to the AI ​​model storage unit 723. In this embodiment, the AI ​​external device 720 can be plugged into the electronic device 710 to provide AI computing functions. In this embodiment, the biometric identification module 715 may include, for example, at least one of a fingerprint identification module, a voiceprint identification module, and a face identification module.

[0031] Figure 8 is a schematic diagram of an electronic device and an artificial intelligence external device according to an embodiment of the present invention. Referring to Figures 7 and 8, in this embodiment, the electronic device 700 may be, for example, a notebook computer, but the present invention is not limited thereto. In one embodiment, the electronic device 700 may be, for example, a tablet computer or a smartphone. In this embodiment, the electronic device 700 may include a first body 701 and a second body 702. The first body 701 and the second body 702 may be connected through a rotatable component. A display unit 714 may be disposed on a surface of the first body 701. An input unit 713 may be disposed on a surface of the second body 702. A first transmission interface 712 may be disposed in the side of the second body 702 and may be combined with a second transmission interface 722 of the artificial intelligence external device 720. A portion of the artificial intelligence external device 720 may be housed in the side of the second body 702. In this embodiment, a biometric identification module 715 may be disposed in the side of the second body 702, but the present invention is not limited thereto. The biometric identification module 715 may be, for example, a voiceprint identification module. When the artificial intelligence external device 720 is installed on the electronic device 710, the device casing of the artificial intelligence external device 720 may be flush with the device casing of the electronic device 710.

[0032] Figure 9 is a flowchart of an operation method according to an embodiment of the present invention. Referring to Figures 7 to 9, the computing system 700 may, for example, operate the following steps S910 to S980. In step S910, the electronic device 710 may determine whether the user's identity has been verified. As shown in Figure 8, the first computing unit 711 may obtain the user's voiceprint features through the biometric identification module 715 to identify the user's identity. The first computing unit 711 may verify whether the user's identity is pre-registered identity information, wherein the pre-registered identity information may be configured or stored in the electronic device 710 or the artificial intelligence external device 720. If not, step S910 may be re-executed after a preset time interval. If yes, in step S920, the electronic device 710 may unlock the artificial intelligence external device 720 and access the artificial intelligence model storage unit 723. In this embodiment, before the user's identity has been verified, the first computing unit 711 may temporarily not supply power to the artificial intelligence external device 720 to ensure data security. Conversely, once the first computing unit 711 confirms the user's identity through the biometric identification module 715, the first computing unit 711 can unlock and access the artificial intelligence model storage unit 723 through the first transmission interface 712 and the second transmission interface 722.

[0033] In step S930, the electronic device 710 can display a data dialog box. In step S940, the electronic device 710 can receive an input command. In step S950, the electronic device 710 can generate a first tag command based on the input command. In this embodiment, the first processing unit 711 can, for example, execute an artificial intelligence assistant application and can display a data dialog box through the display unit 714. The user can, for example, input a relevant question through the operation input unit 713 (keyboard). In this embodiment, the first processing unit 711 can analyze the input command to generate the first tag command. The first processing unit 711 can provide the first tag command to the artificial intelligence model storage unit 723 through the first transmission interface 712 and the second transmission interface 722.

[0034] In step S960, the artificial intelligence model storage unit 723 can generate a second tagging instruction according to the first tagging instruction. In this embodiment, the artificial intelligence model storage unit 723 can input the first tagging instruction into the artificial intelligence model, so that the artificial intelligence model outputs the second tagging instruction to the second transmission interface 722. In step S970, the electronic device 710 generates output information according to the second tagging instruction. In this embodiment, the first arithmetic unit 711 can generate output information according to the second tagging instruction.

[0035] In step S980, the electronic device 710 can display the response content according to the output information. In this embodiment, the electronic device 710 can display the corresponding response content through the data dialog box displayed by the display unit 714. Therefore, the computing system 700 and the operation method of this embodiment can realize effective artificial intelligence computing functions, and can provide the computing power and artificial intelligence model required for artificial intelligence computing through the artificial intelligence external device 720.

[0036] Figure 10 is a schematic diagram of a computing system according to an embodiment of the present invention. Referring to Figure 10, the computing system 1000 includes an electronic device 1010 and an artificial intelligence (AI) external device 1020. The electronic device 1010 is coupled to the AI ​​external device 1020. The electronic device 1010 includes a first computing unit 1011, a first transmission interface 1012, an input unit 1013, and a display unit 1014. The first computing unit 1011 is coupled to the first transmission interface 1012, the input unit 1013, and the display unit 1014. The AI ​​external device 1020 includes a second transmission interface 1022, an AI model storage unit 1023, and a biometric identification module 1024. The second transmission interface 1022 is coupled to the AI ​​model storage unit 1023 and the biometric identification module 1024. In this embodiment, the artificial intelligence external device 1020 can be installed in the electronic device 1010 in a pluggable manner to provide artificial intelligence computing functions.

[0037] Figure 11 is a schematic diagram of an electronic device and an artificial intelligence external device according to an embodiment of the present invention. Referring to Figures 10 and 11, in this embodiment, the electronic device 1010 may be, for example, a notebook computer, but the present invention is not limited thereto. In one embodiment, the electronic device 1010 may be, for example, a tablet computer or a smartphone. In this embodiment, the electronic device 1010 may include a separate first body 1001. The display unit 1014 may include a display panel and is disposed on a surface of the first body 1001. The input unit 1013 may be a touch panel and is integrated with the display panel of the display unit 1014. A first transmission interface may be disposed within the side of the first body 1001 and may be combined with a second transmission interface of the artificial intelligence external device 1020. A portion of the artificial intelligence external device 1020 may be housed within the side of the first body 1001, and the biometric identification module 1024 may be exposed. The biometric identification module 1024 may be, for example, a face recognition module. When the artificial intelligence external device 1020 is installed on the electronic device 1010, the device casing of the artificial intelligence external device 1020 can be flush with the device casing of the electronic device 1010.

[0038] In this embodiment, compared to the embodiment in FIG7, the first computing unit 1011 can obtain the user's facial features through the biometric recognition module 1024 installed in the artificial intelligence external device 1020 to identify the user's identity. The first computing unit 1011 can verify whether the user's identity is pre-registered identity information, wherein the pre-registered identity information can be configured or stored in the electronic device 1010 or the artificial intelligence external device 1020. Furthermore, before the user's identity is verified, the first computing unit 1011 can temporarily withhold power to the artificial intelligence model storage unit 1023 to ensure data security. Conversely, after the first computing unit 1011 confirms the user's identity through the biometric recognition module 1024, the first computing unit 1011 can unlock and access the artificial intelligence model storage unit 1023 through the first transmission interface 1012 and the second transmission interface 1022.

[0039] In this embodiment, the display unit 1014 of the electronic device 1010 can display a data dialog box, and the user can input input commands via touch, so that the electronic device 1010 can generate a first tag command based on the input commands. In this embodiment, the first processing unit 1011 can, for example, execute an artificial intelligence assistant application, and can display a data dialog box through the display unit 1014. The user can, for example, input input commands for related questions through the operation input unit 1013 (touch panel). In this embodiment, the first processing unit 1011 can analyze the input commands to generate a first tag command. The first processing unit 1011 can provide the first tag command to the artificial intelligence model storage unit 1023 through the first transmission interface 1012 and the second transmission interface 1022.

[0040] In this embodiment, the artificial intelligence model storage unit 1023 can generate a second tagging instruction based on the first tagging instruction. In this embodiment, the artificial intelligence model storage unit 1023 can input the first tagging instruction into the artificial intelligence model, causing the artificial intelligence model to output the second tagging instruction to the second transmission interface 1022. The first processing unit 711 can receive the second tagging instruction through the first transmission interface 1012 and generate output information based on the second tagging instruction.

[0041] In this embodiment, the electronic device 1010 can display response content based on the output information. In this embodiment, the electronic device 1010 can display the corresponding response content through the data dialog box displayed by the display unit 1014. Therefore, the computing system 1000 and operating method of this embodiment can realize effective artificial intelligence computing functions, and can provide the computing power and artificial intelligence model required for artificial intelligence computing through the artificial intelligence external device 1020.

[0042] Figure 12 is a schematic diagram of a computing system according to an embodiment of the present invention. Referring to Figure 12, the computing system 1200 includes an electronic device 1210 and an artificial intelligence (AI) external device 1220. The electronic device 1210 is coupled to the AI ​​external device 1220. The electronic device 1210 includes a first computing unit 1211, a first transmission interface 1212, an input unit 1213, a display unit 1214, and a biometric identification module 1215. The first computing unit 1211 is coupled to the first transmission interface 1212, the input unit 1213, the display unit 1214, and the biometric identification module 1215. The AI ​​external device 1220 includes a second computing unit 1221, a second transmission interface 1222, an AI model storage unit 1223, and an indicator unit 1225. The second computing unit 1221 is coupled to the second transmission interface 1222, the AI ​​model storage unit 1223, and the indicator unit 1225. In this embodiment, the artificial intelligence external device 1220 can be installed in the electronic device 1210 in a pluggable manner to provide artificial intelligence computing functions.

[0043] Figure 13 is a schematic diagram of an electronic device and an artificial intelligence external device according to an embodiment of the present invention. Referring to Figures 12 and 13, in this embodiment, the electronic device 1210 may be, for example, a notebook computer, but the present invention is not limited thereto. In one embodiment, the electronic device 1210 may be, for example, a tablet computer or a smartphone. In this embodiment, the electronic device 1210 may include a first body 1201 and a second body 1202. The first body 1201 and the second body 1202 may be connected through a rotatable component. A display unit 1214 may be disposed on a surface of the first body 1201. An input unit 1213 may be disposed on a surface of the second body 1202. A first transmission interface 1212 may be disposed in the side of the second body 1202 and may be combined with the second transmission interface 1222 of the artificial intelligence external device 1220. A portion of the artificial intelligence external device 1220 can be housed within the side of the second body 1202, and the biometric identification module 1215 can be, for example, the camera lens of the electronic device 1210. When the artificial intelligence external device 1220 is installed on the electronic device 1210, the device housing of the artificial intelligence external device 1220 can be flush with the device housing of the electronic device 1210.

[0044] In this embodiment, compared to the embodiment in FIG3, the first computing unit 1211 can supply power to the second computing unit 1221 and the indicator unit 1225, and can provide indicator lights through the indicator unit 1225 to prompt the user to look at the camera lens of the electronic device 1210 (i.e., the biometric identification module 1215), so that the first computing unit 1211 can obtain the user's facial features through the biometric identification module 1215 to identify the user's identity. The first computing unit 1211 can verify whether the user's identity is the pre-registered identity information, so as to unlock and access the artificial intelligence model storage unit 1223 through the artificial intelligence external device 1220. The pre-registered identity information can be configured or stored in the electronic device 1210 or the artificial intelligence external device 1220. In this embodiment, before the user's identity is verified, the second computing unit 1221 can temporarily not supply power to the artificial intelligence model storage unit 1223 to ensure data security. Conversely, once the first computing unit 1211 confirms the user's identity through the biometric identification module 1215, the first computing unit 1211 can notify the second computing unit 1221 to unlock and access the artificial intelligence model storage unit 1223.

[0045] In this embodiment, the electronic device 1210 can display a data dialog box 1216. The first processing unit 1211 can execute an artificial intelligence assistant application according to a startup command. The first processing unit 1211 can, for example, execute an artificial intelligence assistant application and can display the data dialog box 1216 through the display unit 1214. The data dialog box 1216 may include an input box, an output box, and related dialogue information. The user can, for example, input input commands for related questions through the input unit 1213 (keyboard). In this embodiment, the first processing unit 1211 can output input commands to the second transmission interface 1222 through the first transmission interface 1212. The second processing unit 1221 can obtain input commands from the second transmission interface 1222 and analyze the input commands to generate a first tag command.

[0046] In this embodiment, the artificial intelligence model storage unit 1223 can generate a second tagging instruction based on the first tagging instruction. In this embodiment, the artificial intelligence model storage unit 1223 can input the first tagging instruction into the artificial intelligence model, causing the artificial intelligence model to output the second tagging instruction to the second processing unit 1221. The artificial intelligence external device 1220 can generate output information based on the second tagging instruction. The second processing unit 1221 can generate output information based on the second tagging instruction and output the output information to the first transmission interface 1212 through the second transmission interface 1222.

[0047] In this embodiment, the electronic device 1210 can display response content based on the output information. In this embodiment, the electronic device 1210 can display the corresponding response content through the data dialog box 1216 displayed by the display unit 1214. Therefore, the computing system 1200 and operating method of this embodiment can realize effective artificial intelligence computing functions, and can provide the computing power and artificial intelligence model required for artificial intelligence computing through the artificial intelligence external device 1220.

[0048] In summary, the computing system, artificial intelligence external device, and operating method of the present invention can achieve artificial intelligence computing functions even without a network connection through an external artificial intelligence computing device, and have low computing latency. Furthermore, the computing system of the present invention can be combined with a biometric identification module to provide better privacy protection.

[0049] Although the present invention has been disclosed above by way of embodiments, it is not intended to limit the present invention. Anyone skilled in the art can make some modifications and refinements without departing from the spirit and scope of the present invention. Therefore, the scope of protection of the present invention shall be determined by the appended claims.

[0050] 100, 300, 700, 1000, 1200: Computing System 110, 310, 710, 1010, 1210: Electronic devices 111, 311, 711, 1011, 1211: First arithmetic unit 112, 312, 712, 1012, 1212: First transmission interface 113, 313, 713, 1013, 1213: Input Units 114, 314, 714, 1014, 1214: Display Units 120, 320, 720, 1020, 1220: Artificial Intelligence External Devices 122, 322, 722, 1022, 1222: Second transmission interface 123, 323, 723, 1023, 1223: Artificial Intelligence Model Storage Units 301, 701, 1001, 1201: First Unit 302, 702, 1202: Second Unit 3141, 1015, 1216: Data dialog box 321, 1221: Second arithmetic unit 324, 715, 1024, 1215: Biometric identification modules 325, 1225: Indicator Unit 500: fingers S210~S230, S610~S680, S910~S980: Steps

Claims

1. A computing system, comprising: An electronic device includes: an input unit; a display unit; a first transmission interface; and a first processing unit coupled to the input unit, the display unit, and the first transmission interface; and an artificial intelligence external device disposed in the electronic device in a pluggable manner, wherein the artificial intelligence external device includes: a second transmission interface; and an artificial intelligence model storage unit coupled to the second transmission interface and storing an artificial intelligence model, wherein when the first processing unit receives an input command through the input unit, the artificial intelligence external device executes the artificial intelligence model and inputs a first tag command generated based on the input command into the artificial intelligence model to generate a second tag command, wherein the first processing unit displays a response content through the display unit based on output information generated based on the second tag command, wherein the first processing unit analyzes the input command to generate the first tag command.

2. The computing system as described in claim 1, wherein the artificial intelligence external device further includes: A second computing unit is coupled between the second transmission interface and the artificial intelligence model storage unit, wherein the first computing unit outputs the input instruction to the second transmission interface through the first transmission interface, and the second computing unit analyzes the input instruction to generate the first tag instruction.

3. The computing system as claimed in claim 2, wherein the second computing unit generates the output information according to the second flag instruction, and outputs the output information to the first transmission interface through the second transmission interface.

4. The computing system as described in claim 2, wherein the artificial intelligence external device further includes: A biometric identification module is coupled to the second computing unit. When the second computing unit confirms a user's identity through the biometric identification module, the second computing unit unlocks and accesses the artificial intelligence model storage unit, and outputs a start command to the first transmission interface through the second transmission interface, so that the electronic device displays a data dialog box through the display unit according to the start command.

5. The computing system as claimed in claim 2, wherein the electronic device further comprises: A biometric identification module is coupled to the first computing unit. When the first computing unit confirms the identity of a user through the biometric identification module, the first computing unit notifies the second computing unit through the first transmission interface and the second transmission interface, so that the second computing unit unlocks and accesses the artificial intelligence model storage unit, and the first computing unit displays a data dialog box through the display unit.

6. The computing system as claimed in claim 1, wherein the second computing unit outputs the second tag instruction to the first transmission interface through the second transmission interface, and the first computing unit generates the output information based on the second tag instruction.

7. The computing system as claimed in claim 1, wherein the electronic device further comprises: A biometric identification module is coupled to the first computing unit. When the first computing unit confirms a user's identity through the biometric identification module, the first computing unit unlocks and accesses the artificial intelligence model storage unit through the first transmission interface and the second transmission interface, and displays a data dialog box through the display unit.

8. The computing system as described in claim 1, wherein the artificial intelligence external device further includes: A biometric identification module is coupled to the second transmission interface. When the first computing unit confirms a user's identity through the biometric identification module, the first computing unit unlocks and accesses the artificial intelligence model storage unit through the first and second transmission interfaces, and displays a data dialog box through the display unit.

9. An artificial intelligence external device, suitable for being pluggably installed in an electronic device, the electronic device including an input unit, a display unit, a first transmission interface, and a first processing unit, wherein the artificial intelligence external device includes: A second transmission interface; and an artificial intelligence model storage unit, coupled to the second transmission interface, and storing an artificial intelligence model, wherein when the first computing unit receives an input instruction through the input unit, the artificial intelligence external device executes the artificial intelligence model and inputs a first tag instruction generated based on the input instruction into the artificial intelligence model to generate a second tag instruction, wherein the first computing unit displays a response content through the display unit based on output information generated based on the second tag instruction, wherein a second computing unit of the artificial intelligence external device generates the output information based on the second tag instruction and outputs the output information to the first transmission interface through the second transmission interface.

10. The artificial intelligence external device as claimed in claim 9, wherein the second computing unit is coupled between the second transmission interface and the artificial intelligence model storage unit, wherein the first computing unit outputs the input instruction to the second transmission interface through the first transmission interface, and the second computing unit analyzes the input instruction to generate the first tag instruction.

11. The artificial intelligence external device as claimed in claim 10, wherein the artificial intelligence external device further comprises: A biometric identification module is coupled to the second computing unit. When the second computing unit confirms a user's identity through the biometric identification module, the second computing unit unlocks and accesses the artificial intelligence model storage unit, and outputs a start command to the first transmission interface through the second transmission interface, so that the electronic device displays a data dialog box through the display unit according to the start command.

12. The artificial intelligence external device as claimed in claim 10, wherein the electronic device further comprises: A biometric identification module is coupled to the first computing unit. When the first computing unit confirms the identity of a user through the biometric identification module, the first computing unit notifies the second computing unit through the first transmission interface and the second transmission interface, so that the second computing unit unlocks and accesses the artificial intelligence model storage unit, and the first computing unit displays a data dialog box through the display unit.

13. The artificial intelligence external device as claimed in claim 9, wherein the first processing unit analyzes the input instruction to generate the first tag instruction.

14. The artificial intelligence external device as claimed in claim 13, wherein the second computing unit outputs the second tag instruction to the first transmission interface through the second transmission interface, and the first computing unit generates the output information based on the second tag instruction.

15. The artificial intelligence external device as claimed in claim 13, wherein the electronic device further comprises: A biometric identification module is coupled to the first computing unit. When the first computing unit confirms a user's identity through the biometric identification module, the first computing unit unlocks and accesses the artificial intelligence model storage unit through the first transmission interface and the second transmission interface, and displays a data dialog box through the display unit.

16. The artificial intelligence external device as claimed in claim 9, wherein the artificial intelligence external device further comprises: A biometric identification module is coupled to the second transmission interface. When the first computing unit confirms a user's identity through the biometric identification module, the first computing unit unlocks and accesses the artificial intelligence model storage unit through the first and second transmission interfaces, and displays a data dialog box through the display unit.

17. A method for operating an artificial intelligence external device, comprising: When a first processing unit of an electronic device receives an input instruction through an input unit, it executes an artificial intelligence model through an artificial intelligence external device; it inputs a first tag instruction generated based on the input instruction into the artificial intelligence model to generate a second tag instruction; and the first processing unit of the electronic device displays a response content through a display unit based on output information generated based on the second tag instruction, wherein a second processing unit of the artificial intelligence external device generates the output information based on the second tag instruction and outputs the output information to a first transmission interface of the electronic device through a second transmission interface of the artificial intelligence external device.

Images