Calculation power scheduling method and system based on classical computer and superconducting quantum computer

Abstract

The invention discloses a computing power scheduling method and system based on a classical computer and a superconducting quantum computer. The computing power scheduling method comprises the steps of obtaining a target task; dividing the target task into a plurality of calculation sub-tasks; determining the calculation type of the current calculation subtask, wherein the calculation type comprises a quantum calculation subtask and a classic calculation subtask; and distributing the current calculation subtask to the target computer for processing according to the determined calculation type. According to the computing power scheduling method, the target task is divided into the computing sub-tasks, the computing types of the computing sub-tasks are judged, and the computing sub-tasks are allocated to the corresponding computers for processing according to the computing types, so that effective cooperation of classical and quantum computers is realized, the computing efficiency is greatly improved, and the cost of computing resources is reduced.

Description

Technical Field

[0001] The present application relates to the field of quantum computing technology, and more specifically, to a computing power scheduling method and system based on classical computers and superconducting quantum computers. Background Art

[0002] Quantum computing is an application that combines the basic principles of traditional quantum mechanics with modern computer technology. It is a new computing method that compiles, controls quantum information units, and performs calculations according to the basic principles of quantum mechanics. Superconducting quantum computing is a quantum computing solution based on superconducting circuits. Superconducting quantum circuits are composed of one or several different types of superconducting quantum bits and some necessary supporting quantum components.

[0003] In related technologies, pure superconducting quantum computers have problems with insufficient stability and high error rates, and are unable to solve all computing tasks efficiently and accurately. Summary of the invention

[0004] The embodiments of the present application provide a computing power scheduling method and system based on classical computers and superconducting quantum computers.

[0005] The present application embodiment provides a computing power scheduling method based on a classical computer and a superconducting quantum computer, and the computing power scheduling method includes:

[0006] Get the target task;

[0007] Dividing the target task into a plurality of computing subtasks;

[0008] Determine a computing type of a current computing subtask, where the computing type includes a quantum computing subtask and a classical computing subtask;

[0009] The current computing subtask is allocated to a target computer for processing according to the determined computing type.

[0010] In this way, the computing power scheduling method, computer device and computer-readable storage medium of the implementation mode of the present application divide the target task into computing subtasks, and determine the computing type of the computing subtasks, and assign the computing subtasks to the corresponding computers for processing according to the computing type, so as to realize the effective coordination of classical and quantum computers, take into account the advantages of quantum computing and classical computing technologies, not only can it greatly save resources, improve computing efficiency, and reduce computing costs, but also can give full play to the respective advantages of quantum computers and supercomputers to solve complex computing tasks in different fields, break through computing bottlenecks, and realize the engineering of quantum computing.

[0011] In some embodiments, dividing the target task into a plurality of computing subtasks includes:

[0012] Determine task information of the computing subtask according to the target task, wherein the task information includes input parameters and a type identifier;

[0013] The task information of the computing subtask is stored in the sub-computing unit of the computing temporary storage unit.

[0014] In this way, by determining the task information of the computing subtask according to the target task, the target task can be divided into a plurality of computing subtasks, and the task information of the computing subtasks can be stored in the sub-computing unit of the computing temporary storage unit.

[0015] In some implementations, determining the task information of the computing subtask according to the target task includes:

[0016] The task information of the computing subtask is determined according to the API method route called when the target task is written.

[0017] In this way, by calling the API method routing, the task information of the computing subtasks included in the target task can be determined, thereby dividing the target task into multiple computing subtasks.

[0018] In some implementations, determining the computing type of the current computing subtask includes:

[0019] The computing type of the current computing subtask is determined according to the type identifier of the current computing subtask.

[0020] In this way, the computing type of the current computing subtask can be determined according to the type representation, thereby facilitating accurate allocation of computing resources.

[0021] In some embodiments, allocating the current computing subtask to a corresponding target computer for processing according to the determined computing type includes:

[0022] If the computing type of the computing subtask is the classical computing type, assigning the computing subtask to the classical computer for processing;

[0023] If the computing type of the computing subtask is the quantum computing type, the computing subtask is assigned to the superconducting quantum computer for processing.

[0024] In this way, computing subtasks are assigned to classical computers or superconducting quantum computers for processing according to their computing types, achieving effective coordination between classical computing and superconducting quantum computing, greatly improving computing efficiency, and reducing resource costs.

[0025] In some embodiments, the computing power scheduling method further includes:

[0026] Obtaining a calculation result obtained by the target computer processing the current calculation subtask;

[0027] The target calculation result is packaged and transmitted to the cloud platform, wherein the target calculation result includes the calculation results of all the sub-computing tasks.

[0028] In this way, by encapsulating the target calculation results and transmitting them to the cloud platform, users can view and download the calculation results of the target tasks.

[0029] In some embodiments, after obtaining the calculation result obtained by the target computer processing the current calculation subtask, the computing power scheduling method further includes:

[0030] determining whether all of the computing subtasks are processed;

[0031] If all the computing subtasks are processed, the step of packaging the target computing results and transmitting them to the cloud platform is performed.

[0032] In this way, if all computing subtasks are processed, the computing task has been processed, and the target computing results are packaged and transmitted to the cloud platform for users to review.

[0033] In some embodiments, determining whether all computing subtasks of the sub-computing units are processed includes:

[0034] A termination function is executed, wherein the termination function is used to determine whether all the computing subtasks are processed.

[0035] In this way, by calling the termination function, it can be determined whether all computing subtasks have been processed and completed. If they have been completed, the target task results are packaged and transmitted to the cloud platform.

[0036] In some embodiments, the classical computer comprises a supercomputer.

[0037] In this way, by scheduling the computing power of supercomputers and superconducting quantum computers, supercomputers and superconducting quantum computers can work together to perform tasks that they are good at. This computing model fully utilizes the superiority of superconducting quantum computers in dealing with specific problems, while also relying on supercomputers to complete functions such as control, data processing, and error correction. In this way, the optimal configuration of quantum resources and classical resources can be effectively achieved, and the performance and efficiency of computers can be improved.

[0038] In some embodiments, the computing power scheduling method further includes:

[0039] If there is an unprocessed computing subtask, the next computing subtask is used as the current computing subtask, and the step of determining the computing type of the current computing subtask is performed.

[0040] In this way, if there is a computing subtask that has not been processed, the computing resource scheduling of the next computing subtask will continue.

[0041] The present application embodiment provides a computing power scheduling system, the computing power scheduling system comprising a scheduler, a superconducting quantum computer and a classical computer, the scheduler is configured as follows:

[0042] Get the target task;

[0043] Initializing a temporary computing unit according to the target task, the temporary computing unit comprising a plurality of sequentially arranged sub-computing units, the target task comprising a plurality of computing sub-tasks, and the sub-computing unit being used to store the computing sub-tasks;

[0044] Determine a computing type of a current computing subtask in a current computing sub-unit, wherein the computing type includes a quantum computing subtask and a classical computing subtask;

[0045] The current computing subtask is allocated to a target computer for processing according to the determined computing type.

[0046] In this way, by decomposing, scheduling and distributing the user's computing tasks between superconducting quantum computers and classical computers, efficient collaboration between superconducting quantum computers and classical computers can be achieved.

[0047] Additional aspects and advantages of the present application will be given in part in the description below, and in part will become apparent from the description below, or will be learned through the practice of the present application. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] The above and / or additional aspects and advantages of the present application will become apparent and easily understood from the description of the embodiments in conjunction with the following drawings, in which:

[0049] Figure 1 It is a flowchart of a computing power scheduling method according to an implementation mode of the present application;

[0050] Figure 2 It is a flowchart of a computing power scheduling method according to an implementation mode of the present application;

[0051] Figure 3 is a schematic diagram of a calculation temporary storage unit according to an embodiment of the present application;

[0052] Figure 4 It is a flowchart of a computing power scheduling method according to an implementation mode of the present application;

[0053] Figure 5 It is a flowchart of a computing power scheduling method according to an implementation mode of the present application;

[0054] Figure 6 It is a flowchart of a computing power scheduling method according to an implementation mode of the present application;

[0055] Figure 7 It is a flowchart of a computing power scheduling method according to an implementation mode of the present application;

[0056] Figure 8 It is a flowchart of a computing power scheduling method according to an implementation mode of the present application;

[0057] Fig. 9 It is a flowchart of a computing power scheduling method according to an implementation mode of the present application;

[0058] Fig.10 It is a schematic diagram of a computing power scheduling system according to an implementation mode of the present application. DETAILED DESCRIPTION

[0059] The embodiments of the present application are described in detail below, and examples of the embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present application, and cannot be understood as limiting the present application.

[0060] Quantum computing is an application that combines the basic principles of traditional quantum mechanics with modern computer technology. It is a new computing method that compiles, controls quantum information units, and performs calculations according to the basic principles of quantum mechanics. The quantum bit group composed of microscopic particles is the basic unit, and it also has the characteristics of quantum superposition and entanglement. Due to the controlled evolution of quantum states, quantum computing can realize a large amount of information encoding and computational storage, and therefore has a huge information carrying capacity that is incomparable in classical computing, and a huge information processing capability that far exceeds parallel computing.

[0061] Superconducting quantum computing is a quantum computing solution based on superconducting circuits. Superconducting quantum circuits are composed of one or several different types of superconducting quantum bits and some necessary supporting quantum components. Superconducting quantum computers have extremely demanding physical environment requirements and require low temperature and low noise environments to ensure operation. At the same time, the manipulation and reading of superconducting quantum bits generally require the application of highly synchronized radio frequency pulse signals. However, superconducting quantum circuits can use macroscopic quantum states to encode information, and are more easily strongly coupled with control and reading circuits to achieve high-speed quantum information processing. They have great potential for scalability, and superconducting quantum circuits can be prepared through micro-nano processing technology and integrated into planar circuits, with good scalability. Due to the above advantages, superconducting quantum computing has become one of the most promising quantum computing technology routes.

[0062] Among related technologies, pure superconducting quantum computers have problems with insufficient stability and high error rates, and cannot solve all computing tasks efficiently and accurately. In addition, since technical difficulties such as decoherence, scalability, and error correction capabilities need to be overcome, achieving engineering applications is still a challenge currently faced.

[0063] At the same time, technologies such as classical computing and artificial intelligence are developing rapidly, but the growing scale of data is gradually approaching the limit of classical computing. As the speed of processors increases, their power consumption and heat dissipation problems are becoming increasingly serious, which limits the further increase of processor speed. At the same time, the speed of the storage system does not match the speed of the processor, forming the so-called "storage wall", which has brought huge challenges to fields such as artificial intelligence.

[0064] Based on the above problems that may be encountered, please refer to Figure 1 The embodiment of the present application provides a computing power scheduling method based on a superconducting quantum computer 100 and a classical computer 200, and the computing power scheduling method includes:

[0065] 01: Get the target task;

[0066] 02: Divide the target task into multiple computing subtasks;

[0067] 03: Determine the computing type of the current computing subtask, which includes quantum computing subtask and classical computing subtask;

[0068] 04: Allocate the current computing subtask to the target computer for processing according to the determined computing type.

[0069] The embodiment of the present application provides a computing power scheduling device based on a superconducting quantum computer 100 and a classical computer 200, and the computing power scheduling device includes a first acquisition module, a division module, a first determination module, and an allocation module. The first acquisition module can be used to acquire a target task; the division module can be used to divide the target task into multiple computing subtasks; the first determination module can be used to determine the computing type of the current computing subtask, and the computing type includes a quantum computing subtask and a classical computing subtask; the allocation module can be used to allocate the current computing subtask to the target computer for processing according to the determined computing type.

[0070] An embodiment of the present application provides a computer device, which includes one or more processors and a memory, wherein the memory stores a computer program, and when the computer program is executed by the processor, the steps of the computing power scheduling method of any of the above embodiments are implemented.

[0071] Specifically, the computing power scheduling method can be implemented by the scheduler 300. Users can develop on the cloud platform based on online programming tools and programming SDK (Software Development Kit), and after selecting corresponding classical computing resources and quantum computing resources, submit the target task. The target task enters the task queue, and the cloud platform prompts the scheduler 300 that there is a new task, and the scheduler 300 obtains the target task from the cloud platform.

[0072] The target task can be developed by the user under a preset editing framework. That is, the structure of the target task follows the structural logic of the editing framework. When the user writes a computing task, the API method of the computing resource is called and written into a record.

[0073] After obtaining the target task, the scheduler 300 divides the target task into multiple computing subtasks. In some embodiments, the scheduler 300 can divide the computing subtasks according to the division logic preset in the structure of the editing framework, that is, the scheduler 300 can divide the target task into multiple computing subtasks according to its structure, for example, according to the API (Application Programming Interface) method routing called when the user writes the computing task, the number, type and task input parameter packaging of the computing subtasks.

[0074] After dividing the computing subtasks, the scheduler 300 determines the computing type of the current computing subtask to determine the computing type of the current computing subtask, and allocates it to the corresponding target computer for execution processing according to the determined computing type. In some embodiments, when the computing type is determined to be a quantum computing type, the current computing subtask is allocated to the superconducting quantum computer 100 for processing.

[0075] The target computer includes a superconducting quantum computer 100 and a classical computer 200. When the target computer is a superconducting quantum computer 100, the superconducting quantum computer 100 performs quantum circuit compilation according to the computing subtask, and delivers the computing subtask to the proxy layer (proxy layer) of the superconducting quantum computer 100, and then executes the computing subtask. When the target computer is a classical computer 200, the classical computer 200 applies for classical computing resources, creates a container object, and then executes the computing subtask.

[0076] In this way, by dividing the target task into computing subtasks and determining the computing type of the computing subtasks, the computing subtasks are assigned to the corresponding computers for processing according to the computing type, and effective collaboration between classical and quantum computers is achieved. Taking into account the advantages of quantum computing and classical computing technologies, it can not only greatly save resources, improve computing efficiency, and reduce computing costs, but also give full play to the respective advantages of quantum computers and supercomputers to solve complex computing tasks in different fields, break through computing bottlenecks, and realize the engineering of quantum computing.

[0077] See also Figure 2 and Figure 3 In some embodiments, step 02 (dividing the target task into multiple computing subtasks) includes:

[0078] 021: Determine the task information of the computing subtask according to the target task, and the task information includes input parameters and type identification;

[0079] 022: The task information of the computing subtask is stored in the sub-computing unit 11 of the computing temporary storage unit 10, and the sub-computing units 11 are arranged in sequence.

[0080] In some embodiments, the processor can be used to determine task information of the computing subtask based on the target task, the task information including input parameters and type identification; and to store the task information of the computing subtask in the sub-computing unit 11 of the computing temporary storage unit 10, and the sub-computing units 11 are arranged sequentially.

[0081] In some embodiments, the partitioning module includes a first determination submodule and a storage submodule. The first determination submodule can be used to determine the task information of the computing subtask according to the target task, and the task information includes input parameters and type identification; the storage submodule can be used to store the task information of the computing subtask in the sub-computing unit 11 of the computing temporary storage unit 10, and the sub-computing units 11 are arranged in sequence.

[0082] Specifically, the target task can be developed and designed according to the structure of the editing framework of the cloud platform, so the computing subtasks can be divided according to the structure of the target task, and the task information of the computing subtasks can be extracted. The scheduler 300 stores the task information of the computing subtasks in the sub-computing unit 11.

[0083] The task information of the computing subtask includes input parameters, type representation, and other computing parameters.

[0084] Furthermore, the scheduler 300 can initialize the computing temporary storage unit 10 according to the target task. The computing temporary storage unit 10 includes a plurality of sub-computing units 11, such as Figure 3In the example, the computing temporary storage unit 10 includes n sub-computing units 11. The scheduler 300 stores the multiple computing sub-tasks obtained by division in the multiple sub-computing units 11, and each sub-computing unit 11 stores one computing sub-task, so as to realize the division and storage of computing sub-tasks.

[0085] The multiple sub-computing units 11 may be arranged in sequence, and the scheduler 300 performs computing resource allocation scheduling for each computing sub-task in sequence, that is, implements serial computing power scheduling.

[0086] In addition, after the target computer executes the calculation subtask and obtains the calculation result, the calculation result can also be stored in the corresponding sub-computing unit 11.

[0087] In this way, by determining the task information of the computing subtask according to the target task, the target task can be divided into a plurality of computing subtasks, and the task information of the computing subtasks can be stored in the sub-computing unit 11 of the computing temporary storage unit 10 .

[0088] See also Figure 4 In some implementations, step 021 (determining task information of computing subtasks according to the target task) includes:

[0089] 0211: Determine the task information of the computing subtask according to the API method route called when the target task is written.

[0090] In some embodiments, the processor may be configured to determine the task information of the computing subtask according to an API method route called when the target task is written.

[0091] In some embodiments, the first determination submodule includes a determination unit, which can be used to determine the task information of the computing subtask according to the API method route called when the target task is written.

[0092] Specifically, when the user writes the target task, the API is called according to the computing resources that need to be scheduled. The scheduler can determine the task information of the computing subtask according to the API method routing called by the target task, for example, determine the number of computing subtasks, computing resource types, input parameters, etc., thereby dividing the target task into multiple computing subtasks.

[0093] In this way, by calling the API method routing, the task information of the computing subtasks included in the target task can be determined, thereby dividing the target task into multiple computing subtasks.

[0094] See also Figure 5 In some implementations, step 03 (determining the computing type of the current computing subtask) includes:

[0095] 031: Determine the calculation type of the current calculation subtask according to the type identifier of the current calculation subtask.

[0096] In some implementations, the processor may be configured to determine the computing type of the current computing subtask according to the type identifier of the current computing subtask.

[0097] In some embodiments, the first determination module includes a second determination submodule, and the second determination submodule can be used to determine the computing type of the current computing subtask according to the type identifier of the current computing subtask.

[0098] Specifically, when dividing the target task to obtain computing subtasks, different modules of the target task are identified, and the computing subtasks that require quantum computing resources are calibrated as quantum computing types, and the computing subtasks that require classical computing resources are calibrated as classical computing types. When determining the computing type of the current computing subtask of the current subcomputing unit 11, the computing type is determined according to the type identifier of the current computing subtask.

[0099] The calibration process of the type identifier is completed when dividing the computing subtasks. When the scheduler 300 schedules computing resources, the type identifier can be used to determine the computing type of the computing subtask.

[0100] In one embodiment, the type of the current computing subtask in the current sub-computing unit 11 is identified as a quantum computing type. When the scheduler 300 identifies the quantum computing type, it determines that the current computing subtask needs to allocate quantum computing resources for processing.

[0101] In this way, the computing type of the current computing subtask can be determined according to the type representation, thereby facilitating accurate allocation of computing resources.

[0102] See also Figure 6 In some embodiments, step 04 (assigning the current computing subtask to the target computer for processing according to the determined computing type) includes:

[0103] 041: If the computing type of the computing subtask is a classical computing type, the computing subtask is assigned to the classical computer 200 for processing;

[0104] 042: If the computing type of the computing subtask is a quantum computing type, the computing subtask is assigned to the superconducting quantum computer 100 for processing.

[0105] In some embodiments, the processor may be used to assign the computing subtask to the classical computer 200 for processing if the computing type of the computing subtask is a classical computing type; and to assign the computing subtask to the superconducting quantum computer 100 for processing if the computing type of the computing subtask is a quantum computing type.

[0106] In some embodiments, the allocation module includes a first allocation submodule and a second allocation submodule. The first allocation submodule can be used to allocate the computing subtask to the classical computer 200 for processing if the computing type of the computing subtask is a classical computing type; the second allocation submodule can be used to allocate the computing subtask to the superconducting quantum computer 100 for processing if the computing type of the computing subtask is a quantum computing type.

[0107] Specifically, when determining that the computing type of the computing subtask is a classical computing type, the scheduler 300 allocates the computing subtask to the classical computer 200 for processing. The classical computer 200 allocates resources and processes the computing subtask, and writes the computing result back to the register after the processing is completed.

[0108] In some embodiments, data processing and user interface related computing subtasks may be handled by the classical computer 200, while the superconducting quantum computer 100 may focus on executing a specific quantum algorithm.

[0109] In other embodiments, the superconducting quantum computer 100 is a superconducting quantum computer (QPU service resource), and the classical computer 200 is a supercomputer (HPC service resource). When the scheduler 300 determines that the current computing subtask of the current subcomputing unit 11 is a quantum computing type, the current computing subtask is submitted to the superconducting quantum computer side for calculation. When the scheduler 300 determines that the current computing subtask of the current subcomputing unit 11 is a classical computing type, the current computing subtask is submitted to the supercomputer side for calculation.

[0110] When it is determined that the computing type of the computing subtask is a quantum computing type, the scheduler 300 assigns the computing subtask to the superconducting quantum computer 100 for processing. The superconducting quantum computer 100 performs quantum circuit compilation on the computing subtask and delivers it to the superconducting quantum computer proxy layer, executes the quantum computing task to obtain the computing result, and then writes the computing result back to the register.

[0111] In this way, computing subtasks are assigned to the classical computer 200 or the superconducting quantum computer 100 for processing according to the computing type of the computing subtask, thereby achieving effective coordination between classical computing and superconducting quantum computing, greatly improving computing efficiency, and reducing resource costs.

[0112] See also Figure 7 In some implementations, the computing power scheduling method further includes:

[0113] 05: Get the calculation result obtained by the target computer processing the current calculation subtask;

[0114] 06: Encapsulate the target calculation results and transmit them to the cloud platform. The target calculation results include the calculation results of all sub-computing tasks.

[0115] In some embodiments, the processor can be used to obtain the calculation results obtained by the target computer processing the current calculation subtask; and to encapsulate the target calculation results and transmit them to the cloud platform, and the target calculation results include the calculation results of all sub-computing tasks.

[0116] In some embodiments, the computing power scheduling device further includes a second acquisition module and a first encapsulation module. The second acquisition module can be used to obtain the calculation result obtained by the target computer processing the current calculation subtask; the first encapsulation module can be used to encapsulate the target calculation result and transmit it to the cloud platform, and the target calculation result includes the calculation results of all sub-computing tasks.

[0117] Specifically, after the classical computer 200 or the superconducting quantum computer 100 processes and obtains the calculation result, the calculation result is packaged and stored in the register. The scheduler 300 uniformly packages the calculation results of all sub-computation tasks to obtain the total calculation result, and uploads the total calculation result to the task queue of the cloud platform, so that the user can view or download the corresponding target task and the total calculation result.

[0118] In this way, by encapsulating the target calculation results and transmitting them to the cloud platform, users can view and download the calculation results of the target tasks.

[0119] See also Figure 8 In some embodiments, after step 05 (obtaining the calculation result obtained by the target computer processing the current calculation subtask), the computing power scheduling method further includes:

[0120] 07: Determine whether all computing subtasks have been processed;

[0121] 08: If all computing subtasks are processed, execute step 04 (encapsulate the target computing results and transmit them to the cloud platform);

[0122] In some embodiments, the processor may be used to determine whether all computing subtasks are processed; and if all computing subtasks are processed, execute the steps of packaging the target computing results and transmitting them to the cloud platform.

[0123] In some embodiments, the computing power scheduling device also includes a second determination module and a second encapsulation module. The second determination module can be used to determine whether all computing subtasks are processed; the second encapsulation module can be used to execute the steps of encapsulating the target computing results and transmitting them to the cloud platform if all computing subtasks are processed.

[0124] Specifically, each time computing resources are scheduled for a computing subtask and the computing results obtained by the target computer processing the computing subtask are obtained, it is determined whether all computing subtasks of the sub-computing units 11 are processed. If all computing subtasks are processed, the target computing results are packaged and transmitted to the cloud platform, so that users can view or download the computing tasks and computing results through the cloud platform.

[0125] In this way, if all computing subtasks are processed, the computing task has been processed, and the target computing results are packaged and transmitted to the cloud platform for users to review.

[0126] See also Fig. 9 In some embodiments, step 07 (determining whether all computing subtasks are processed) includes:

[0127] 071: Execute the termination function, which is used to determine whether all computing subtasks have been processed.

[0128] In some embodiments, the processor may be configured to execute a termination function, wherein the termination function is configured to determine whether all computing subtasks have been processed.

[0129] In some embodiments, the second determination module includes an execution submodule, and the execution submodule can be used to execute a termination function, and the termination function is used to determine whether all computing subtasks are processed.

[0130] Specifically, the termination function is usually used to terminate the computing task, that is, to terminate the computing task and return to the main process. The scheduler 300 can determine whether the computing task has reached the end stage, that is, to determine whether all computing subtasks have been processed, by executing the termination function.

[0131] If after executing the termination function, it is determined that there are computing subtasks that have not been processed, the scheduler 300 continues to process the computing subtasks in the next sub-computing unit 11, determines its computing type and allocates computing resources to it.

[0132] If it is determined that all computing subtasks have been processed after executing the termination function, the scheduler 300 will encapsulate all computing results in the register as target computing results and transmit them to the cloud platform for users to view or download.

[0133] In this way, by calling the termination function, it can be determined whether all computing subtasks have been processed and completed. If they have been completed, the target task results are packaged and transmitted to the cloud platform.

[0134] In some embodiments, classical computer 200 comprises a supercomputer.

[0135] Specifically, a superconducting quantum computer is a computer that uses quantum bits (qubits) made of superconducting materials to perform quantum computing. These qubits are implemented through superconducting circuits and can operate at extremely low temperatures to reduce energy loss and maintain the stability of the quantum state. The core of a superconducting quantum computer lies in superconducting qubits, which use the quantum properties of superconducting materials to store and process information. Unlike traditional classical bits, qubits can exist in multiple states at the same time, a phenomenon called quantum superposition. In addition, quantum entanglement allows the states of qubits to be correlated with each other, and even if they are far apart, changing the state of one qubit will immediately affect other bits entangled with it.

[0136] In addition, the classical computer 200 refers to a computer system that uses classical computing resources for computing, including supercomputers, classical computers, etc. Classical computing resources mainly refer to traditional computer hardware and software resources, including central processing units (CPUs), graphics processing units (GPUs), memory, storage devices, operating systems, and application software. The working principle of classical computing resources is based on a binary computing model, that is, all data and operations are represented and processed in the form of 0 and 1. This is different from the quantum bits of quantum computing, which use phenomena such as quantum superposition and entanglement to encode and process information.

[0137] In hybrid quantum computing, superconducting quantum computers are responsible for executing specific quantum algorithms, such as the preparation, manipulation, and measurement of quantum states, while classical computers are responsible for controlling the operation of superconducting quantum computers and analyzing and processing quantum computing results.

[0138] In this way, by scheduling the computing power of supercomputers and superconducting quantum computers, supercomputers and superconducting quantum computers can work together to perform tasks that they are good at. This computing model fully utilizes the superiority of superconducting quantum computers in dealing with specific problems, while also relying on supercomputers to complete functions such as control, data processing, and error correction. In this way, the optimal configuration of quantum resources and classical resources can be effectively achieved, and the performance and efficiency of computers can be improved.

[0139] See also Figure 8 In some implementations, the computing power scheduling method further includes:

[0140] 09: If there is an unprocessed computing subtask, the next computing subtask is used as the current computing subtask, and step 03 (determining the computing type of the current computing subtask) is executed.

[0141] In some embodiments, the processor may be configured to, if there is an unprocessed computing subtask, use the next computing subtask as the current computing subtask and execute step 03 (determine the computing type of the current computing subtask).

[0142] In some embodiments, the computing power scheduling device includes a judgment module, which can be used to judge the computing type of the next computing subtask if there are unprocessed computing subtasks.

[0143] Specifically, if there is an unprocessed computing subtask, the next computing subtask is used as the current computing subtask, and step 03 is repeatedly executed until all computing subtasks are processed.

[0144] In this way, if there is a computing subtask that has not been processed, the computing resource scheduling of the next computing subtask will continue.

[0145] See also Fig.10 The embodiment of the present application provides a computing power scheduling system 1000, which includes a scheduler 300, a superconducting quantum computer 100 and a classical computer 200. The scheduler 300 is configured to: obtain a target task; initialize a computing temporary storage unit 10 according to the target task, the computing temporary storage unit 10 includes a plurality of sequentially arranged sub-computing units 11, the target task includes a plurality of computing sub-tasks, and the sub-computing unit 11 is used to store computing sub-tasks; determine the computing type of the current computing sub-task in the current sub-computing unit 11, the computing type includes a quantum computing sub-task and a classical computing sub-task; allocate the current computing sub-task to a target computer for processing according to the determined computing type, and the target computer includes a superconducting quantum computer 100 and a classical computer 200.

[0146] Specifically, the scheduler 300 of the computing power scheduling system 1000 can manage the work of the computer and provide available interfaces for different user tasks, providing more convenient interface services for users to solve problems using superconducting quantum computers (superconducting quantum computers 100) and supercomputers (classical computers 200).

[0147] That is to say, the computing power scheduling system 1000 provides interface functions for packaging different problems or algorithms. After receiving the user's call, the scheduler 300 decomposes the target task, distributes the decomposed computing subtasks to the classical computing nodes (classical computer 200) or quantum computing nodes (superconducting quantum computer 100), and determines the total computing result after obtaining the feedback result, and finally returns the computing result of the target task to the user.

[0148] Among them, see Fig.10 The data module is used to temporarily store the calculation results obtained by processing the calculation subtask, and the calculation task module is used to store the calculation temporary storage unit 10. Since the input parameter of the calculation subtask can be the calculation result of the previous calculation subtask, the calculation result in the data module can be passed to the calculation task module. The database can be used to store the data in the data module, for example, to store or process the calculation results.

[0149] In this way, by decomposing, scheduling and allocating the user's computing tasks between the superconducting quantum computer 100 and the classical computer 200, efficient collaboration between the superconducting quantum computer 100 and the classical computer 200 is achieved.

[0150] An embodiment of the present application provides a computer-readable storage medium having a computer program stored thereon. When the program is executed by a processor, the steps of the computing power scheduling method of any of the above embodiments are implemented.

[0151] It is understood that a computer program includes computer program code. The computer program code may be in source code form, object code form, executable file or some intermediate form. Computer readable storage media may include: any entity or device capable of carrying computer program code, recording medium, USB flash drive, mobile hard disk, magnetic disk, optical disk, computer memory, read-only memory (ROM), random access memory (RAM), and software distribution medium.

[0152] In the description of this specification, the description with reference to the terms "one embodiment", "some embodiments", "illustrative embodiments", "examples", "specific examples" or "some examples" etc. means that the specific features, structures, materials or characteristics described in conjunction with the embodiments or examples are included in at least one embodiment or example of the present application. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. In addition, those skilled in the art may combine and combine different embodiments or examples described in this specification and the features of different embodiments or examples without contradiction.

[0153] In addition, the term "connection" should be understood in a broad sense, for example, it can include fixed connection, detachable connection, or integral connection; it can include direct connection, indirect connection through an intermediate medium, and internal communication between two elements. For ordinary technicians in this field, the specific meanings of the above terms in this application can be understood according to specific circumstances.

[0154] In addition, the terms "first" and "second" are used for descriptive purposes only and should not be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features. Therefore, the features defined as "first" and "second" may explicitly or implicitly include at least one of the features. In the description of this application, the meaning of "plurality" is at least two, such as two, three, etc., unless otherwise clearly and specifically defined.

[0155] Any process or method description in a flowchart or otherwise described herein may be understood to represent a module, fragment or portion of code that includes one or more executable instructions for implementing the steps of a specific logical function or process, and the scope of the preferred embodiments of the present application includes alternative implementations in which functions may not be performed in the order shown or discussed, including performing functions in a substantially simultaneous manner or in reverse order depending on the functions involved, which should be understood by technicians in the technical field to which the embodiments of the present application belong.

[0156] Although the embodiments of the present application have been shown and described above, it can be understood that the above embodiments are exemplary and cannot be understood as limitations on the present application. Ordinary technicians in this field can change, modify, replace and modify the above embodiments within the scope of the present application.

Claims

1. A computing power scheduling method based on classical computers and superconducting quantum computers, characterized in that: The computing power scheduling method includes: Get the target task; Dividing the target task into a plurality of computing subtasks; Determine a computing type of a current computing subtask, where the computing type includes a quantum computing subtask and a classical computing subtask; The current computing subtask is allocated to a target computer for processing according to the determined computing type.

2. The computing power scheduling method according to claim 1, characterized in that: The step of dividing the target task into a plurality of computing subtasks comprises: Determine task information of the computing subtask according to the target task, wherein the task information includes input parameters and a type identifier; The task information of the computing subtask is stored in the sub-computing unit of the computing temporary storage unit.

3. The computing power scheduling method according to claim 2, characterized in that: The determining the task information of the computing subtask according to the target task includes: The task information of the computing subtask is determined according to the API method route called when the target task is written.

4. The computing power scheduling method according to claim 2, characterized in that: The determining of the computing type of the current computing subtask includes: The computing type of the current computing subtask is determined according to the type identifier of the current computing subtask.

5. The computing power scheduling method according to claim 1, characterized in that: The step of allocating the current computing subtask to a target computer for processing according to the determined computing type includes: If the computing type of the computing subtask is the classical computing type, assigning the computing subtask to the classical computer for processing; If the computing type of the computing subtask is the quantum computing type, the computing subtask is assigned to the superconducting quantum computer for processing.

6. The computing power scheduling method according to claim 1, characterized in that: The computing power scheduling method further includes: Obtaining a calculation result obtained by the target computer processing the current calculation subtask; The target calculation result is packaged and transmitted to the cloud platform, wherein the target calculation result includes the calculation results of all the sub-computing tasks.

7. The computing power scheduling method according to claim 6, characterized in that: After obtaining the calculation result obtained by the target computer processing the current calculation subtask, the computing power scheduling method further includes: determining whether all of the computing subtasks are processed; If all the computing subtasks are processed, the step of packaging the target computing results and transmitting them to the cloud platform is performed.

8. The computing power scheduling method according to claim 7, characterized in that: The determining whether all the computing subtasks are processed comprises: A termination function is executed, wherein the termination function is used to determine whether all the computing subtasks are processed.

9. The computing power scheduling method according to claim 7, characterized in that: The computing power scheduling method further includes: If there is an unprocessed computing subtask, the next computing subtask is used as the current computing subtask, and the step of determining the computing type of the current computing subtask is performed.

10. A computing power scheduling system, comprising a scheduler, a superconducting quantum computer and a classical computer, characterized in that: The scheduler is configured as follows: Get the target task; Initializing a temporary computing unit according to the target task, the temporary computing unit comprising a plurality of sequentially arranged sub-computing units, the target task comprising a plurality of computing sub-tasks, and the sub-computing unit being used to store the computing sub-tasks; Determine a computing type of a current computing subtask in a current computing sub-unit, wherein the computing type includes a quantum computing subtask and a classical computing subtask; The current computing subtask is allocated to a target computer for processing according to the determined computing type.

Images