Public computing power-oriented intelligent computing center development state computing power scheduling method and device

Abstract

The invention provides an intelligent computing center development state computing power scheduling method and device oriented to common computing power, and relates to the technical field of computing power infrastructure, and the method comprises the steps: S1, under the condition that a development request sent by a user is received, a first container is created based on a virtual graphics processing unit (GPU) resource, and the first container is used for adjusting a model code; and S2, a debugging instruction is sent to a second container based on the first container, the debugging instruction is operated based on the second container, a debugging result is obtained, the debugging instruction comprises a GPU instruction corresponding to the operated and adjusted model code, and the second container is a container created based on computing power resources of an intelligent computing center. According to the method, the computing power resource utilization rate of the intelligent computing center can be greatly improved, and the economic cost of model development is greatly reduced.

Description

Technical Field

[0001] The present invention relates to the field of computing power infrastructure technology, and in particular to a method and device for developing state computing power scheduling for an inclusive computing power intelligent computing center. Background Art

[0002] With the rapid development of artificial intelligence technology, "intelligent computing centers" and "intelligent computing centers" have emerged.

[0003] An "Intelligent Computing Center" is a facility that utilizes large-scale heterogeneous computing resources, including general-purpose and intelligent computing power, to provide the computing power, data, and algorithms required for AI applications (such as AI deep learning model development, model training, and model inference). The Intelligent Computing Center encompasses facilities, hardware, and software, providing a full stack of capabilities, from bottom-level computing power to top-level application enablement.

[0004] “Intelligent Computing Center” includes but is not limited to “Smart Computing Center”.

[0005] "Intelligent Computing Center" refers to an artificial intelligence computing center. It is a type of computing power infrastructure that is based on artificial intelligence theory, adopts artificial intelligence computing architecture, and provides computing power services, data services, and algorithm services required for artificial intelligence applications.

[0006] "Computing power" is the core of "intelligent computing center" and "intelligent computing center". It is the ability of computer equipment or computing / data center to process information. It is the ability of computer hardware and software to work together to perform certain computing needs. It is the computing power to achieve target result output by processing information data. It is a new type of productivity that integrates information computing power, network carrying capacity, and data storage capacity. It mainly provides services to society through computing power infrastructure.

[0007] Currently, users develop models using computing resources provided by intelligent computing centers. This process requires constant modification and adjustment of model code, which doesn't require the computing resources of the intelligent computing center during the code modification phase. However, in existing technologies, since models are deployed in the intelligent computing center, the center's computing resources are still required during code modifications, resulting in long-term usage and low utilization of computing resources. Furthermore, users need to lease computing resources from the intelligent computing center for long periods of time, which in turn leads to high economic costs for model development and hinders the widespread application of affordable computing power.

[0008] It can be seen that the existing technology has the problems of low utilization of computing resources and high economic cost of model development. Summary of the Invention

[0009] The present invention provides a method and device for scheduling computing power in the development state of a universal computing power intelligent computing center, so as to solve the problems in the prior art of low utilization of computing power resources and high economic cost of model development.

[0010] To solve the above problems, the present invention is achieved as follows: In a first aspect, the present invention provides a method for scheduling computing power in a development state for an inclusive computing power intelligent computing center, comprising: Step S1: upon receiving a development request from a user, creating a first container based on a virtual graphics processor (GPU) resource, where the first container is used to adjust the model code; Step S2: Send debugging instructions to the second container based on the first container, and run the debugging instructions based on the second container to obtain debugging results. The debugging instructions include GPU instructions corresponding to running the adjusted model code. The second container is a container created based on the computing power resources of the intelligent computing center.

[0011] In one embodiment, step S2 includes: Step S21: If the second container does not exist in the intelligent computing center, call the container operation application program interface API to create a second container based on the computing power resources of the intelligent computing center; Step S22: establishing a communication channel between the first container and the second container; Step S23: Send the debugging instruction to the second container through the communication channel based on the first container; Step S24: Run the debugging instruction based on the second container to obtain a debugging result.

[0012] In one embodiment, step S2 includes: Step S25: If the second container is running in the intelligent computing center, send the debugging instruction to the second container through the communication channel based on the first container; Step S26: Run the debugging instruction based on the second container to obtain a debugging result.

[0013] In one embodiment, after step S2, the method further includes: Step S3: Send the debugging result to the first container through the communication channel based on the second container.

[0014] In one embodiment, after step S3, the method further includes: Step S4: Close the second container and release the computing resources occupied by the second container.

[0015] In one embodiment, step S4 includes at least one of the following: Step S41: If the accumulated time reaches a preset time threshold and the second container has not received any new debugging instructions, shut down the second container and release the computing resources occupied by the second container. The accumulated time is the time since the second container last received a debugging instruction. Step S42: When the accumulated time does not reach the preset time threshold, the second container recalculates the accumulated time when receiving a new debugging instruction.

[0016] In a second aspect, the present invention further provides a computing power scheduling device for developing a state-based computing power intelligent computing center, comprising: A creation module, configured to create a first container based on a virtual graphics processor (GPU) resource upon receiving a development request from a user, wherein the first container is used to adjust the model code; A debugging module is used to send debugging instructions to a second container based on the first container, and run the debugging instructions based on the second container to obtain debugging results, wherein the debugging instructions include GPU instructions corresponding to running the adjusted model code, and the second container is a container created based on the computing power resources of the intelligent computing center.

[0017] In a third aspect, the present invention also provides an electronic device comprising a processor, a memory, and a computer program stored in the memory and executable on the processor. When the computer program is executed by the processor, the steps in the method for developing state computing power scheduling for a universal computing power intelligent computing center as described in the first aspect above are implemented.

[0018] In a fourth aspect, the present invention further provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps in the method for developing state computing power scheduling for a universal computing power intelligent computing center as described in the first aspect above.

[0019] In a fifth aspect, the present invention also provides a computer program product comprising computer instructions, which, when executed by a processor, implement the steps in the method for developing state computing power scheduling for a universal computing power intelligent computing center as described in the first aspect above.

[0020] In the present invention, when a development request is received from a user, a first container is created based on the virtual graphics processor GPU resources, and the first container is used to adjust the model code; a debugging instruction is sent to the second container based on the first container, and the debugging instruction is run based on the second container to obtain a debugging result, wherein the debugging instruction includes the GPU instruction corresponding to running the adjusted model code, and the second container is a container created based on the computing power resources of the intelligent computing center. In this way, the model code is adjusted through the first container without occupying the computing power resources of the intelligent computing center. When debugging is required, the adjusted model code is run based on the computing power resources of the intelligent computing center through the second container, thereby achieving the goal of occupying the computing power resources of the intelligent computing center only when the adjusted model code needs to be run, greatly improving the utilization rate of the computing power resources of the intelligent computing center. Furthermore, since users do not need to rent the computing power services of the intelligent computing center for a long time, the economic cost of model development for users is effectively and significantly reduced, and the widespread application of inclusive computing power is achieved. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] In order to more clearly illustrate the technical solution of the present invention, the following is a brief introduction to the drawings required for the description of the present invention. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative labor.

[0022] Figure 1 This is a flow chart of a method for scheduling computing power in a development state for a universal computing power intelligent computing center provided by the present invention; Figure 2 is a schematic diagram of the interaction between the first container and the second container provided by the present invention; Figure 3 This is a structural diagram of a computing power scheduling device for a development state of a universal computing power intelligent computing center provided by the present invention; Figure 4 This is a structural diagram of an electronic device provided by the present invention. DETAILED DESCRIPTION

[0023] The following will clearly and completely describe the technical solutions of the present invention in conjunction with the accompanying drawings. Obviously, the embodiments described are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making any creative efforts shall fall within the scope of protection of the present invention.

[0024] The "computing power" mentioned in the present invention refers to: the ability of computer equipment or computing / data centers to process information, the ability of computer hardware and software to work together to execute certain computing requirements, and the computing power to achieve target result output by processing information data. It is a new type of productivity that integrates information computing power, network carrying capacity, and data storage capacity, and mainly provides services to society through computing power infrastructure.

[0025] The "computing power" (Computational Power, CP) mentioned in the present invention refers to: the ability of a data center server to process data and output results. It is a comprehensive indicator to measure the computing power of a data center, including general computing power, super computing power and intelligent computing power. The commonly used unit of measurement is the number of floating-point operations performed per second (FLOPS, 1EFLOPS=10^18FLOPS). The larger the value, the stronger the comprehensive computing power. According to calculations, 1 EFLOPS is approximately the computing power output of 5 Tianhe-2A or 500,000 mainstream server CPUs or 2 million mainstream laptops. The calculation formula is: CP=CP 通用 +CP 智能 +CP 超级 .

[0026] The "Network Power" (NP) mentioned in this invention refers to: it is a manifestation of the data transmission capability of computing power facilities, including comprehensive capabilities such as network architecture, network bandwidth, transmission latency, intelligent management and scheduling, etc. It involves network transmission within and between data centers, and is a comprehensive indicator for measuring network transmission scheduling capabilities.

[0027] "Storage Power" (SP) as used in this document refers to the comprehensive capabilities of a data center in four areas: data storage capacity, performance, security and reliability, and environmental friendliness and low-carbon development. It serves as a comprehensive indicator of a data center's data storage capabilities, encompassing both external storage devices like storage arrays and internal server storage. Storage capacity is commonly measured in exabytes (EB, 1EB = 2^60 bytes), while performance is commonly measured in IOPS / TB (Input / Output Operations Per Second). Disaster recovery ratio is a key indicator of security and reliability.

[0028] The "computing power infrastructure" mentioned in the present invention refers to a new type of information infrastructure that integrates information computing power, network carrying capacity, and data storage capacity, and can realize the centralized calculation, storage, transmission and application of information.

[0029] The "new information infrastructure" referred to in this document primarily includes network infrastructure such as 5G networks, fiber-optic broadband networks, backbone networks, international communications networks, and satellite internet; computing infrastructure such as data centers, general computing centers, intelligent computing centers, and supercomputing centers; and new technological facilities such as artificial intelligence, blockchain, and quantum computing. With the emergence and widespread application of new general technologies, the new information infrastructure will become even more diverse.

[0030] The "computing power" mentioned in the present invention includes: general computing power, intelligent computing power and super computing power.

[0031] The "general computing power" mentioned in the present invention refers to the computing power provided by servers based on CPU (Central Processing Unit) chips, which is used to support basic general computing such as cloud computing and edge computing.

[0032] The "intelligent computing power" mentioned in this invention refers to: a computing platform based on specialized chips such as GPU (Graphics Processing Unit), FPGA (Field Programmable Gate Array), and ASIC (Application Specific Integrated Circuit) for various innovative artificial intelligence applications, such as natural language processing and machine vision.

[0033] The "supercomputing power" mentioned in the present invention refers to the computing power provided by high-performance computing clusters such as supercomputers. It utilizes the centralized computing resources of multiple computer systems working in parallel and uses a dedicated operating system to handle extremely complex or data-intensive problems. It is mainly used for calculations in cutting-edge scientific fields, such as planetary simulation, drug molecule design, genetic analysis, etc.

[0034] The term "intelligent computing center" as used in this document refers to a facility that utilizes large-scale heterogeneous computing resources, including general-purpose computing power (CPUs) and intelligent computing power (GPUs, FPGAs, ASICs, etc.), primarily to provide the computing power, data, and algorithms required for AI applications (such as AI deep learning model development, model training, and model inference). An intelligent computing center encompasses facilities, hardware, and software, providing a full stack of capabilities, from bottom-level computing power to top-level application enablement.

[0035] The "intelligent computing center" mentioned in the present invention includes but is not limited to the "intelligent computing center".

[0036] The "intelligent computing center" mentioned in the present invention is an artificial intelligence computing center, which is a type of computing power infrastructure based on artificial intelligence theory, adopts artificial intelligence computing architecture, and provides computing power services, data services and algorithm services required for artificial intelligence applications.

[0037] The "computing power center" mentioned in the present invention refers to: a facility that is mainly composed of infrastructure such as wind, fire, water, electricity, and IT hardware and software equipment, and has computing power, transportation capacity, and storage capacity, including general data centers, intelligent computing centers, supercomputing centers, etc.

[0038] The "supercomputing center" mentioned in the present invention refers to: a supercomputing data center, which is a data center based on a supercomputer or a large-scale computing cluster, which can provide large-scale computing, storage and network services and other functions, and is widely used in application scenarios such as aerospace, national defense, oil exploration, climate modeling and genome sequencing.

[0039] The "computing resources" mentioned in the present invention refer to: technologies and facilities with information computing, transmission, storage and application capabilities required for the development of a digital society, including but not limited to computing resources such as CPUs and GPUs, network resources such as switches and routers, storage resources such as storage arrays and distributed storage, security resources such as firewalls and intrusion detection systems, and supporting and guarantee resources such as wind, fire, water and electricity.

[0040] The "development state" mentioned in the present invention refers to the state during the process of writing model code.

[0041] The “model” mentioned in the present invention includes but is not limited to a “large language model” and a “multimodal large model”.

[0042] The "large language model" mentioned in this invention refers to a large language model (LLM), which is a language model with a large parameter scale. It is designed to understand and generate human language. It is trained with a large amount of text data and can perform a wide range of tasks including text summarization, translation, sentiment analysis, etc.

[0043] The "Multimodal Large Models" mentioned in this invention refer to models that combine multimodal information such as text, images, video, and audio for training, including but not limited to multimodal large language models.

[0044] See Figure 1 , Figure 1 This is a flowchart of a method for scheduling computing power in a development state for a universal computing power intelligent computing center provided by the present invention. Figure 1 As shown, the following steps are included: Step S1: upon receiving a development request from a user, creating a first container based on a virtual graphics processing unit (GPU) resource, where the first container is used to adjust a model code.

[0045] The above-mentioned virtual GPU resources are a kind of virtual resources that simulate GPUs. The first container created based on the virtual GPU resources will not occupy the real GPU resources of the intelligent computing center, but can simulate the operating environment of the real GPU resources. Users can adjust the model code through the first container. During the entire adjustment process, there is no need to occupy the computing power resources of the intelligent computing center, avoiding waste of computing power.

[0046] Development requests are requests sent by users when they need to develop or adjust a model. They can be sent from the user's terminal. For example, if a user chooses to adjust a model in development mode, the terminal generates a development request and sends it to the intelligent computing center. Upon receiving the development request, the intelligent computing center creates a first container based on virtual GPU resources, allowing the user to adjust the model code through the first container.

[0047] Among them, the development request is used to represent the current user needs to develop and adjust the model code.

[0048] In some optional embodiments, the first container can be a lightweight development container (Pod) started based on Kubernetes (k8s), with a built-in integrated development environment (IDE) and debugging tools, and can simulate GPU resources through a resource fake mechanism to obtain virtual GPU resources.

[0049] Specifically, by modifying the Kubernetes Device Plugin interface, virtual GPU resources (for example, nvidia.com / fake-gpu:1) are reported to the scheduler, instructing the scheduler to allocate these "virtual resources" without actually occupying real GPU resources. During the creation of the first container, the container image is pre-installed with the same CUDA driver version and dependent libraries (such as cuDNN) as the real GPU container, ensuring seamless compatibility and consistency between the development and runtime environments.

[0050] Furthermore, users can select different modes based on their needs, and the terminal will send different requests based on the different modes. For example, if the user selects development mode, the terminal sends a development request to the intelligent computing center; if the user selects run mode (run mode indicates that the user needs to repeatedly run the model), the terminal sends a run request to the intelligent computing center. In this case, the intelligent computing center creates a container based on the actual computing resources, and uses the container to run the model based on the computing resources of the intelligent computing center.

[0051] Step S2: Send debugging instructions to the second container based on the first container, and run the debugging instructions based on the second container to obtain debugging results. The debugging instructions include GPU instructions corresponding to running the adjusted model code. The second container is a container created based on the computing power resources of the intelligent computing center.

[0052] The aforementioned second container is created based on the computing resources of the intelligent computing center. This second container allows the adjusted model code to be run using the computing resources of the intelligent computing center, thereby confirming the effectiveness of the model adjustment. The second container occupies the computing resources of the intelligent computing center when running, but does not occupy the computing resources of the intelligent computing center when not running. This avoids unused computing resources from being occupied by the intelligent computing center and improves the utilization rate of the computing resources of the intelligent computing center.

[0053] The above-mentioned debugging instructions are instructions sent by the user when the user needs to debug the adjusted model code after completing the adjustment of the model code. The debugging instructions are sent to the second container through the first container, so that the adjusted model code can be run through the second container without occupying the computing power resources of the intelligent computing center in the first container, thereby reducing the occupation of the computing power resources of the intelligent computing center.

[0054] Among them, the debugging instructions include GPU instructions corresponding to running the adjusted model code, which are specifically sent through the virtual GPU service of the first container. The second container executes the GPU instructions based on the computing power resources of the intelligent computing center to try to run the adjusted model code.

[0055] In some optional implementations, the debugging instruction may also include a modified model code. After receiving the debugging instruction, the second container may run the modified model code to obtain a debugging result.

[0056] The modified model code may be a full code or an incremental code.

[0057] In the present invention, when a development request is received from a user, a first container is created based on the virtual graphics processor GPU resources, and the first container is used to adjust the model code; a debugging instruction is sent to the second container based on the first container, and the debugging instruction is run based on the second container to obtain a debugging result, wherein the debugging instruction includes the GPU instruction corresponding to running the adjusted model code, and the second container is a container created based on the computing power resources of the intelligent computing center. In this way, the model code is adjusted through the first container without occupying the computing power resources of the intelligent computing center. When debugging is required, the adjusted model code is run based on the computing power resources of the intelligent computing center through the second container, thereby achieving the goal of occupying the computing power resources of the intelligent computing center only when the adjusted model code needs to be run, thereby improving the utilization rate of the computing power resources of the intelligent computing center. Furthermore, since users do not need to rent the computing power services of the intelligent computing center for a long time, the economic cost of model development for users is effectively and significantly reduced, and the widespread application of inclusive computing power is achieved.

[0058] It should be noted that there may be a running second container in the intelligent computing center, or there may not be a running second container. When there is a running second container, the second container can directly execute the receiving and running of debugging instructions. When there is no running second container, it is necessary to start the second container to implement the receiving and running of debugging instructions through the second container.

[0059] Specifically, step S2 includes: Step S21: If the second container does not exist in the intelligent computing center, call a container operation application programming interface (API) to create a second container based on the computing resources of the intelligent computing center; Step S22: establishing a communication channel between the first container and the second container; Step S23: Send the debugging instruction to the second container through the communication channel based on the first container; Step S24: Run the debugging instruction based on the second container to obtain a debugging result.

[0060] In this case, the intelligent computing center does not have a running second container. Therefore, the computing resources of the intelligent computing center are needed to execute and receive debug instructions through the second container. To enable data transmission between the first and second containers, a communication channel must be established between the first and second containers.

[0061] In some optional implementations, the second container includes a virtual Compute Unified Device Architecture (CUDA) interface, and the virtual CUDA interface is used to establish a communication channel between the first container and the second container.

[0062] It should be noted that the virtual CUDA interface is a functional component added when creating the second container. The virtual CUDA interface can be used to intercept debugging instructions and implement CUDA API calls in the code (such as cudaMalloc, cudaMemcpy, etc.).

[0063] Specifically, such as Figure 2 As shown, when there is no second container running in the intelligent computing center, the container operation application program interface API is called through the container operation instruction orchestration (Operator) of kubernetes (k8s), and a second container is created based on the computing power resources of the intelligent computing center. A communication channel is established between the second container and the first container, so that the second container can receive debugging instructions through the communication channel.

[0064] The second container includes a virtual CUDA (rCUDA) interface, which receives debugging instructions through the rCUDA interface. Furthermore, the second container also includes a CUDA layer and a driver layer to call CUDA functions and drivers.

[0065] In some optional implementations, the debugging instruction is specifically a remote instruction package (Remote Procedure Call Instruction, RPC Instruction). By sending the remote instruction package to the operation component of k8s, the creation of the second container can be achieved.

[0066] Furthermore, after the container operation instruction orchestration receives the debugging instruction, the container operation instruction orchestration selects the optimal GPU resource to start the second container based on the priority queue and resource pool status of the GPU.

[0067] Specifically, after the first container sends the debugging instruction, a GPU list of the intelligent computing center is obtained based on the container operation instruction arrangement, the GPU list includes multiple GPUs, and the multiple GPUs in the GPU list are arranged according to the priority of the debugging model; Obtaining a resource usage ratio of each GPU among the multiple GPUs; An evaluation score for each GPU is calculated based on the resource usage ratio and priority.

[0068] A target GPU is determined based on the resource usage status, where the target GPU is a GPU with a highest evaluation score among the multiple GPUs.

[0069] In some implementations, the evaluation score is calculated by weighting the resource usage ratio and priority.

[0070] In this way, by selecting the GPU with the highest evaluation score to create the second container, the second container can quickly execute debugging instructions.

[0071] In some implementations, different second containers in the intelligent computing center are bound to different GPUs to avoid interference between different tasks.

[0072] In some optional implementations, the debugging instructions use binary encoding (such as Protocol Buffers) to compress instruction data and reduce network transmission delay.

[0073] In the present invention, if the second container does not exist in the intelligent computing center, the container operation application program interface (API) is called to create the second container based on the computing power resources of the intelligent computing center; a communication channel is established between the first container and the second container; the debugging instructions are sent to the second container via the communication channel based on the first container; and the debugging instructions are executed based on the second container to obtain the debugging results. In this way, if the second container does not exist in the intelligent computing center, the second container is created by calling the container operation API interface, and a communication channel is established between the first and second containers, so that the second container can receive and execute the debugging instructions through the communication channel.

[0074] In some optional implementations, based on the computing resources of the intelligent computing center, calling a container operation application programming interface (API) to create a second container includes: Acquire historical load data, where the historical load data is used to characterize the load conditions of similar models during the debugging process; The second container is created based on the historical load data.

[0075] In this way, by creating a second container through historical load data, the computing power resources of the intelligent computing center occupied by the second container are close to the actual computing power resources required, further reducing the computing power resource consumption of the intelligent computing center and improving the utilization rate of the computing power resources of the intelligent computing center.

[0076] In one embodiment, step S2 includes: Step S25: If the second container is running in the intelligent computing center, send the debugging instruction to the second container through the communication channel based on the first container; Step S26: Run the debugging instruction based on the second container to obtain a debugging result.

[0077] In the present invention, if the second container is already running in the intelligent computing center, the debugging instructions are sent to the second container via a communication channel based on the first container; the debugging instructions are then executed in the second container to obtain the debugging results. In this way, if the second container is already running in the intelligent computing center, the debugging instructions can be directly received and executed through the existing second container, thus achieving low-latency model debugging.

[0078] In one embodiment, after step S2, the method further includes: Step S3: Send the debugging result to the first container through the communication channel based on the second container.

[0079] In the present invention, after the debugging instruction is executed in the second container and the debugging result is obtained, the debugging result is sent to the first container through the communication channel, so that the user can determine the effect of the model code adjustment through the first container, so that the user can further adjust the model code through the first container.

[0080] In one embodiment, after step S3, the method further includes: Step S4: Close the second container and release the computing resources occupied by the second container.

[0081] In the present invention, the second container needs to occupy the computing power resources of the intelligent computing center. In order to reduce the occupation of the computing power resources of the intelligent computing center, the second container is closed when the second container is not needed to run the code, and the computing power resources occupied by the second container are released to further improve the utilization rate of the computing power resources, thereby realizing the widespread application of inclusive computing power.

[0082] In some optional implementations, the CUDA API call chain is monitored through container operation instruction orchestration. When an end signal (such as cudaDeviceReset) is monitored, the second container is closed and GPU resources are released to reduce the computing power resources of the intelligent computing center occupied by the second container for a long time.

[0083] In some optional embodiments, a timeout threshold is set. When the container operation instruction orchestration monitors that the second container has not processed data within the timeout threshold, the second container is forcibly closed to avoid the abnormality causing the second container to occupy the computing power resources of the intelligent computing center for a long time.

[0084] In some embodiments, step S4 includes at least one of the following: Step S41: If the accumulated time reaches a preset time threshold and the second container has not received any new debugging instructions, shut down the second container and release the computing resources occupied by the second container. The accumulated time is the time since the second container last received a debugging instruction. Step S42: When the accumulated time does not reach the preset time threshold, the second container recalculates the accumulated time when receiving a new debugging instruction.

[0085] It's important to note that creating or starting a second container consumes a considerable amount of time and computing power. If users frequently need to run adjusted model code, repeatedly shutting down and starting the second container not only consumes computing power but also results in longer debugging delays. To address this issue, the present invention provides a preset time threshold to optimize the startup and shutdown of the second container.

[0086] Specifically, if the accumulated time reaches a preset time threshold and the second container has not received a new debug command, the second container is closed and the computing resources occupied by the second container are released. The accumulated time is the time since the second container last received a debug command. If the accumulated time does not reach the preset time threshold and the second container receives a new debug command, the accumulated time is recalculated. In this way, if the time between two consecutive debug commands is less than or equal to the preset time threshold, the second container will not be closed and the debug commands can be directly processed. If the time between two consecutive debug commands is greater than the preset time threshold, the second container will be closed to reduce the computing resources occupied by the second container in the intelligent computing center.

[0087] See Figure 3 , Figure 3 This is a structural diagram of a computing power scheduling device for developing a state computing power intelligent computing center provided by the present invention, such as Figure 3 As shown, the state computing power scheduling device 300 for the inclusive computing power intelligent computing center includes: A creation module 301 is configured to create a first container based on a virtual graphics processor (GPU) resource upon receiving a development request from a user, wherein the first container is used to adjust the model code; The debugging module 302 is used to send debugging instructions to the second container based on the first container, and run the debugging instructions based on the second container to obtain debugging results. The debugging instructions include GPU instructions corresponding to running the adjusted model code. The second container is a container created based on the computing power resources of the intelligent computing center.

[0088] In one embodiment, the debugging module 302 includes: A creating unit, configured to, if the second container does not exist in the intelligent computing center, call a container operation application program interface (API) to create the second container based on the computing power resources of the intelligent computing center; an establishing unit, configured to establish a communication channel between the first container and the second container; A first sending unit, configured to send the debugging instruction to the second container through the communication channel based on the first container; The first debugging unit is configured to execute the debugging instruction based on the second container to obtain a debugging result.

[0089] In one embodiment, the debugging module 302 includes: A second sending unit is configured to send the debugging instruction to the second container through a communication channel based on the first container when the second container is running in the intelligent computing center; The second debugging unit is configured to execute the debugging instruction based on the second container to obtain a debugging result.

[0090] In one embodiment, after the debugging module 302, the computing power scheduling device 300 for the inclusive computing power intelligent computing center development state further includes: A sending module is configured to send the debugging result to the first container through a communication channel based on the second container.

[0091] In one embodiment, after the sending module, the computing power scheduling device 300 for the development of a state of inclusive computing power intelligent computing center further includes: The closing module is used to close the second container and release the computing resources occupied by the second container.

[0092] In one embodiment, the shutdown module includes at least one of the following: a first shutting down unit, configured to shut down the second container and release the computing resources occupied by the second container if the accumulated time reaches a preset time threshold and the second container has not received any new debugging instructions, where the accumulated time is the time since the second container last received a debugging instruction; The second closing unit is configured to recalculate the accumulated time when the accumulated time does not reach the preset time threshold and the second container receives a new debugging instruction.

[0093] The computing power scheduling device for the development status of the universal computing power intelligent computing center provided by the present invention can realize the various processes of the various embodiments of the above-mentioned computing power scheduling method for the development status of the universal computing power intelligent computing center. The technical features correspond one to one and can achieve the same technical effects. To avoid repetition, they will not be described here.

[0094] It should be noted that the computing power scheduling device developed for the inclusive computing power intelligent computing center in the present invention can be a device, or a component, integrated circuit, or chip in an electronic device.

[0095] The present invention also provides an electronic device, see Figure 4 , Figure 4 The electronic device includes a memory 401, a processor 402, and a program or instruction stored in the memory 401 and executed by the processor 402. Figure 1 Any steps in the corresponding embodiment of the state computing power scheduling method for the inclusive computing power intelligent computing center development and the same beneficial effects are achieved will not be repeated here.

[0096] The processor 402 may be a CPU, an ASIC, an FPGA, or a GPU.

[0097] Those skilled in the art will understand that all or part of the steps of the above-mentioned embodiment of the method for implementing the development status computing power scheduling method for the inclusive computing power intelligent computing center can be completed through hardware related to program instructions, and the program can be stored in a readable medium.

[0098] The present invention also provides a readable storage medium, which stores a computer program. When the computer program is executed by a processor, the above Figure 1 Any steps in the corresponding embodiments of the method for developing state computing power scheduling for the inclusive computing power intelligent computing center can achieve the same technical effect and are not described here to avoid repetition. The storage medium is, for example, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk.

[0099] The present invention also provides a computer program product, comprising computer instructions, which, when executed by a processor, implement the above Figure 1 The corresponding processes of the implementation method of the state computing power scheduling method for the inclusive computing power intelligent computing center development can achieve the same technical effect. To avoid repetition, they will not be repeated here.

[0100] The terms "first", "second" and the like in the present invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequential order. In addition, the terms "comprise" and "have" and any of their variations are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device comprising a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or that are inherent to these processes, methods, products or devices. In addition, "and / or" is used in this application to represent at least one of the connected objects, for example A and / or B and / or C, which means comprising seven situations including single A, single B, single C, and both A and B exist, both B and C exist, both A and C exist, and both A, B and C exist.

[0101] It should be noted that, in this document, the terms "comprises," "includes," or any other variations thereof are intended to encompass non-exclusive inclusion, such that a process, method, article, or apparatus comprising a series of elements includes not only those elements but also other elements not explicitly listed, or elements inherent to such process, method, article, or apparatus. In the absence of further limitations, an element defined by the phrase "comprising a ..." does not exclude the presence of other identical elements in the process, method, article, or apparatus comprising the element.

[0102] Through the description of the above embodiments, those skilled in the art can clearly understand that the above-mentioned embodiment methods can be implemented by means of software plus the necessary general hardware platform. Of course, they can also be implemented by hardware, but in many cases the former is a better embodiment. Based on this understanding, the technical solution of this application, or the part that contributes to the existing technology, can be embodied in the form of a software product. This computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk), and includes a number of instructions for enabling a terminal (which can be a mobile phone, computer, server, air conditioner, or second terminal device, etc.) to execute the methods of each embodiment of this application.

[0103] The embodiments of the present application are described above in conjunction with the accompanying drawings, but the present application is not limited to the above-mentioned specific implementation methods. The above-mentioned specific implementation methods are merely illustrative and not restrictive. Under the guidance of this application, ordinary technicians in this field can also make many forms without departing from the purpose of this application and the scope of protection of the claims, all of which are within the protection of this application.

Claims

1. A method for scheduling computing power for intelligent computing centers with a state-based computing power development, characterized in that: include: Step S1: upon receiving a development request sent by a user, creating a first container based on a virtual graphics processor GPU resource, wherein the first container is used to adjust a model code; Step S2: Send debugging instructions to the second container based on the first container, and run the debugging instructions based on the second container to obtain debugging results, where the debugging instructions include GPU instructions corresponding to running the adjusted model code, and the second container is a container created based on the computing power resources of the intelligent computing center.

2. The method according to claim 1, characterized in that The step S2 comprises: Step S21: If the second container does not exist in the intelligent computing center, based on the computing resources of the intelligent computing center, call a container operation application program interface API to create a second container; Step S22: establishing a communication channel between the first container and the second container; Step S23: sending the debugging instruction to the second container through the communication channel based on the first container; Step S24: Run the debugging instruction based on the second container to obtain a debugging result.

3. The method according to claim 1, characterized in that The step S2 comprises: Step S25: When the second container is running in the intelligent computing center, send the debugging instruction to the second container through the communication channel based on the first container; Step S26: Run the debugging instruction based on the second container to obtain a debugging result.

4. The method according to claim 1, characterized in that After step S2, the method further includes: Step S3: Send the debugging result to the first container through the communication channel based on the second container.

5. The method according to claim 4, characterized in that After step S3, the method further includes: Step S4: close the second container and release the computing resources occupied by the second container.

6. The method according to claim 5, characterized in that The step S4 includes at least one of the following: Step S41: When the accumulated time reaches a preset time threshold and the second container has not received a new debugging instruction, the second container is closed and the computing resources occupied by the second container are released. The accumulated time is the time after the second container last received a debugging instruction. Step S42: When the accumulated time does not reach the preset time threshold, the second container recalculates the accumulated time when receiving a new debugging instruction.

7. A computing power scheduling device for the development of a universal computing power intelligent computing center, characterized in that: include: A creation module, configured to create a first container based on a virtual graphics processor GPU resource when a development request sent by a user is received, wherein the first container is used to adjust the model code; A debugging module is used to send debugging instructions to a second container based on the first container, and run the debugging instructions based on the second container to obtain debugging results, wherein the debugging instructions include GPU instructions corresponding to running the adjusted model code, and the second container is a container created based on the computing resources of an intelligent computing center.

8. An electronic device, characterized in that: include: A processor, a memory, and a program stored in the memory and executable on the processor, wherein when the program is executed by the processor, the steps of a method for developing state computing power scheduling for a universal computing power intelligent computing center are implemented as described in any one of claims 1 to 6.

9. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, which, when executed by a processor, implements the steps of a method for developing a state computing power scheduling method for a universal computing power intelligent computing center as described in any one of claims 1 to 6.

10. A computer program product, characterized in that It includes computer instructions, which, when executed by a processor, implement the steps of the method for developing a state computing power scheduling method for a universal computing power intelligent computing center as described in any one of claims 1 to 6.

Images