Execution system and method of electricity measurement hybrid calculation task and electronic equipment

Abstract

The invention provides an execution system and method for a hybrid measurement and power calculation task and electronic equipment, and the system comprises a measurement and power calculation service platform which is used for analyzing the hybrid measurement and power calculation task to generate a sub-task set, and generating a task work queue corresponding to the sub-task set according to a dependency relationship between sub-tasks; and the scheduling adapter feeds back the task execution progress in the task work queue in real time, and the electricity measuring calculation service platform selects to issue the sub-tasks in the task queue to the calculation task scheduling execution system according to the task execution progress. The invention provides an execution system and method for a power measurement hybrid calculation task and electronic equipment, and the system employs a distributed task scheduling execution system, enables a power measurement calculation service platform to interact with the task scheduling execution system through employing a scheduling adapter as a middle layer, and distributes a calculation task to the corresponding task scheduling execution system. The resource allocation is more reasonable, the coupling degree of different task scheduling execution systems is reduced, and a new quantum computing cluster can be conveniently expanded.

Description

Technical Field

[0001] The present disclosure belongs to the technical field of computer networks, and particularly relates to an execution system, method, and electronic device for quantum-electric hybrid computing tasks. Background Art

[0002] The quantum-electric hybrid computing task scheduling and execution system realizes the scheduling and execution of computing tasks based on a classical computing cluster and a quantum computing cluster. Usually, a task includes quantum computing and classical computing. For the task execution system, it is necessary to coordinate the operation interaction between quantum tasks and classical tasks, realize the efficient connection of classical computing resources and quantum computing resources, and achieve efficient scheduling in the quantum-electric hybrid computing task environment.

[0003] In related technologies, for example, a method, medium, and system for model management and scheduling based on cloud-native deployment disclosed in CN202410962060.5 briefly describes the model management method and model scheduling method in a cloud-native environment, which is a distributed task scheduling method in a classical computing scenario. However, the heterogeneous resources it refers to are GPU computing resources, and it does not consider the extended access and scheduling optimization of quantum computing resources. When the quantum and classical computing clusters are integrated on the same platform, due to the heterogeneity of the quantum computing cluster, such as the different characteristics of superconducting, ion trap, and optical quantum computer clusters, the scheduling of the quantum computing task scheduling and execution system is not easy to be implemented in one system, which increases the coupling degree of the system.

[0004] For example, a task scheduling method and device for a quantum-electronic hybrid platform disclosed in CN202310571867.1 briefly describes a method for improving task scheduling efficiency in a quantum-electronic hybrid platform. For each node of the quantum-electronic hybrid platform, the expected time for the node to complete a virtual task is determined according to the current running state of the node, and this expected time is used as a reference to schedule real tasks. The above shows that the quantum-electronic hybrid platform only describes a part of the execution process of the hybrid computing task, without considering the way of directly submitting hybrid computing tasks and the task automatic parsing scheme, as well as the extended use of accessing a third-party platform, which is not convenient for the quantum-electronic hybrid platform to be accepted and used by users, and also does not consider the backup scheme when quantum resources fail.

[0005] For example, a method for optimizing the scheduling of hybrid computing power network resources and a follow-up control device disclosed in CN202310398653.9 briefly describe a target task scheduling method based on a quantum-electronic hybrid computing power network. This method needs to obtain the topological structure of the hybrid computing power network, layer and determine labels for each resource node based on the topological structure, use a classical operations research non-linear optimization global scheduling algorithm to determine the electronic nodes to be scheduled, use a quantum machine learning intelligent global scheduling algorithm to determine the quantum nodes to be scheduled, and finally determine the scheduling path based on the resource node labels to achieve target task scheduling. However, it does not consider the expansion of the computing power network for different types of quantum resource nodes and the upgrade and maintenance of different types of quantum computing task scheduling systems. At the same time, the computing power network defaults that the hybrid computing tasks have been split into quantum and electronic parts before submission, lacking a way to directly submit hybrid computing tasks. Summary of the Invention

[0006] The purpose of the present disclosure is to propose an execution system, method, and electronic device for quantum-electric hybrid computing tasks, which solves the problem that the scheduling of the quantum computing task scheduling execution system is not easy to be implemented in one system and improves the coupling degree of the system.

[0007] To this end, in the first aspect, the present disclosure provides an execution system for quantum-electric hybrid computing tasks, including:

[0008] A quantum-electric computing service platform, which is used to parse the quantum-electric hybrid computing tasks to generate a subtask set and generate a task work queue corresponding to the subtask set according to the dependency relationship between the subtasks;

[0009] A scheduling adapter, which is used to feedback the execution progress of the tasks in the task work queue in real time. The quantum-electric computing service platform selects to send the subtasks in the task queue to the computing task scheduling execution system according to the task execution progress. The subtasks include classical computing subtasks and quantum computing subtasks;

[0010] The computing task scheduling execution system is of a distributed architecture and is used to send the quantum computing subtasks to the corresponding quantum computing task scheduling execution system according to the different types of subtasks, and the service nodes in the quantum computing task scheduling execution system execute the quantum computing subtasks;

[0011] And send the classical computing subtasks to the classical computing task scheduling execution system, and the service nodes in the classical computing task scheduling execution system execute the classical computing subtasks.

[0012] Optionally, the power and electricity calculation service platform includes a task parsing module, a task classification module, and a task work queue module. The task parsing module is used to parse the power and electricity hybrid calculation task to obtain a number of subtasks; the task classification module is used to classify the subtasks and assign task tags; the task work queue module is used to arrange the subtasks into a task work queue according to the task tags.

[0013] Optionally, the quantum computing task scheduling and execution system includes at least one or more of a superconducting quantum computing task scheduling and execution system, an ion trap quantum computing task scheduling and execution system, and an optical quantum computing task scheduling and execution system.

[0014] Optionally, a classical computing partition is set in the classical computing task scheduling and execution system, a quantum computing partition is set in the quantum computing task scheduling and execution system, and a real machine computing partition and a simulation computing partition are set in the quantum computing partition.

[0015] In a second aspect, a task execution method for a power and electricity hybrid calculation task is provided, which is applied to the execution system of the power and electricity hybrid calculation task, and includes the following steps:

[0016] According to the type of subtasks, the power and electricity calculation service platform parses the power and electricity hybrid calculation task to generate a subtask set, and generates a task work queue corresponding to the subtask set according to the dependency relationship between the subtasks;

[0017] The scheduling adapter real-time feeds back the task execution progress in the task work queue, and the power and electricity calculation service platform selects to send the subtasks in the task queue to the computing task scheduling and execution system according to the task execution progress;

[0018] The quantum computing subtasks are sent to the quantum computing task scheduling and execution system, and the service nodes in the quantum computing task scheduling and execution system execute the quantum computing subtasks to obtain the calculation results of the quantum computing subtasks;

[0019] The classical computing subtasks are sent to the classical computing task scheduling and execution system, and the service nodes in the classical computing task scheduling and execution system execute the classical computing subtasks to obtain the calculation results of the classical computing subtasks.

[0020] Optionally, before the power - quantum computing service platform parses the power - quantum hybrid computing task to generate a subtask set and generates a task work queue corresponding to the subtask set according to the dependency relationship between subtasks, it further includes receiving a power - quantum hybrid computing task, where the power - quantum hybrid computing task is a tagged power - quantum hybrid computing task or has been disassembled into sequential hybrid computing subtasks. The tagged power - quantum hybrid computing task executes the step of the power - quantum computing service platform parsing the power - quantum hybrid computing task to generate several subtasks arranged in a task work queue, and the hybrid computing subtasks that have been disassembled into sequential ones execute the step of the scheduling adapter allocating the subtasks to the computing task scheduling and execution system according to the order of the task work queue and the task tags of the subtasks.

[0021] Optionally, when the power - quantum computing service platform parses the power - quantum hybrid computing task to generate a subtask set and generates a task work queue corresponding to the subtask set according to the dependency relationship between subtasks, it includes: disassembling the power - quantum hybrid computing task into several code blocks with serial numbers and task tags, constructing the code blocks into subtasks; arranging the subtasks into a task work queue according to the serial numbers and task tags of the subtasks.

[0022] Optionally, the task tags include power - quantum task information, serial - parallel information, and context information.

[0023] Optionally, the scheduling adapter provides real - time feedback on the progress of task execution in the task work queue, and when the power - quantum computing service platform selects to send the subtasks in the task queue to the computing task scheduling and execution system according to the task execution progress, it includes: the scheduling adapter polls the resource information of the computing task scheduling and execution system, clarifies the resource status of each computing task scheduling and execution system, and allocates the subtasks to the corresponding computing task scheduling and execution system according to the order of the task work queue and the power - quantum task information.

[0024] Optionally, when the service node in the quantum computing task scheduling and execution system executes the quantum computing subtask to obtain the calculation result of the quantum computing subtask, it includes: the quantum computing task scheduling and execution system re - orders the allocated quantum computing subtasks according to the scheduling policy, creates a scheduling queue, and sequentially allocates resources to the quantum computing subtasks according to the order of the scheduling queue to execute the quantum computing subtasks to obtain the calculation result of the quantum computing subtasks.

[0025] Optionally, when the quantum computing task scheduling and execution system executes the quantum computing subtask, it first executes through the real - machine computing partition. When the resource stability of the quantum computing subtask in the real - machine computing partition is lower than the set threshold, it switches to the simulation computing partition for execution.

[0026] In a third aspect, an electronic device is provided, which is characterized by including: a memory, a processor;

[0027] The memory stores computer-executable instructions;

[0028] The processor executes the computer-executable instructions stored in the memory, such that the processor executes the method described above.

[0029] Advantageous effects:

[0030] (1) The present disclosure provides an execution system, method, and electronic device for quantum-classical hybrid computing tasks. By adopting a distributed computing task scheduling and execution system, classical or different types of quantum computing tasks are asynchronously computed in different computing task scheduling and execution systems, and through a scheduling adapter as an intermediate layer, the quantum-classical computing service platform interacts with the computing task scheduling and execution system. Through the scheduling adapter, computing tasks can be assigned to corresponding computing task scheduling and execution systems, making resource allocation more reasonable, maximizing resource utilization, and the distributed computing task scheduling and execution system can reduce the coupling degree of different computing task scheduling and execution systems, facilitating the expansion of new quantum computing clusters and effectively reducing the system maintenance difficulty.

[0031] (2) The present disclosure automatically allocates tasks through the scheduling adapter to achieve seamless switching of computing resources. That is, when the quantum computing task scheduling and execution system executes a subtask, it first executes through the physical machine computing partition. When the resource stability of the subtask in the physical machine computing partition is lower than a set threshold, it switches to the simulation computing partition for execution. This can not only achieve reasonable allocation of task resources for different subtasks but also automatically adjust the resource nodes where the affected subtasks are located in case of resource anomalies, enabling the subtasks to execute smoothly. The resource switching is completed by the system itself without manual operation by the user, simplifying the user's operation and enhancing the user experience.

[0032] (3) In the present disclosure, through task parsing of quantum-classical hybrid computing tasks, the quantum-classical hybrid computing tasks are disassembled into multiple subtasks with task tags and arranged in a task work queue, providing the subtask allocation order for the scheduling adapter, separating quantum tasks and classical tasks, allocating the subtasks to suitable computing task scheduling and execution systems, and providing more degrees of freedom for task scheduling according to serial and parallel information.

[0033] It should be understood that the content described in this part is not intended to identify the key or important features of the embodiments of the present disclosure, nor is it used to limit the scope of the present disclosure. Other features of the present disclosure will become readily understood through the following description. Description of the Drawings

[0034] To more clearly illustrate the technical solutions in the embodiments of the present disclosure or the prior art, the following will briefly introduce the accompanying drawings required for the description of the embodiments or the prior art. Obviously, the accompanying drawings in the following description are only some embodiments of the present disclosure. For those of ordinary skill in the art, without creative efforts, other accompanying drawings can be obtained based on these drawings.

[0035] Figure 1 It is the system structure of an embodiment of an execution system for a quantum - electricity hybrid computing task in the present disclosure;

[0036] Figure 2 It is the method flow chart of an embodiment of an execution system for a quantum - electricity hybrid computing task in the present disclosure;

[0037] Figure 3 It is the method flow chart of an embodiment of a method for receiving a quantum - electricity hybrid computing task in an execution system for a quantum - electricity hybrid computing task in the present disclosure;

[0038] Figure 4 It is the schematic diagram of the work queue of sub - tasks in an embodiment of an execution system for a quantum - electricity hybrid computing task in the present disclosure;

[0039] Figure 5 It is the schematic structural diagram of an embodiment of an electronic device in the present disclosure.

[0040] In the figure, 100 - task execution system, 110 - quantum - electricity computing service platform, 111 - task parsing module, 112 - task classification module, 113 - task work queue module, 121 - classical computing task scheduling and execution system, 122 - classical computing partition, 123 - superconducting quantum computing task scheduling and execution system, 124 - superconducting quantum computing partition, 125 - ion - trap quantum computing task scheduling and execution system, 126 - ion - trap quantum computing partition, 127 - optical quantum computing task scheduling and execution system, 128 - optical quantum computing partition, 129 - other quantum computing task scheduling and execution system, 1210 - other quantum computing partition, 130 - scheduling adapter, 200 - electronic device, 201 - processor, 202 - memory, 203 - communication component, 204 - bus. Detailed implementation manners

[0041] To make the objectives, technical solutions, and advantages of the present application clearer, the following will clearly and completely describe the technical solutions in the present application with reference to the accompanying drawings in the present application. Obviously, the described embodiments are some, but not all, of the embodiments of the present application. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative efforts fall within the scope of protection of the present application.

[0042] In the description, claims and the above drawings of this application, the terms "first", "second", "third", "fourth", etc. are used to distinguish similar objects, and do not necessarily have to be used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged under appropriate circumstances. For example, without departing from the scope of this article, the first information can also be referred to as the second information, and similarly, the second information can also be referred to as the first information.

[0043] Depending on the context, the word "if" as used herein can be interpreted as "when" or "while" or "in response to determining".

[0044] Furthermore, as used herein, the singular forms "a", "an" and "the" are also intended to include the plural forms, unless the context indicates otherwise.

[0045] It should be further understood that the terms "comprising", "including" indicate the presence of features, steps, operations, elements, components, items, types, and / or groups, but do not exclude the presence, occurrence or addition of one or more other features, steps, operations, elements, components, items, types, and / or groups.

[0046] The terms "or" and "and / or" used herein are interpreted inclusively, or mean any one or any combination. Thus, "A, B or C" or "A, B and / or C" means "any one of the following: A; B; C; A and B; A and C; B and C; A, B and C". An exception to this definition only occurs when the combination of elements, functions, steps or operations are mutually exclusive in some way.

[0047] Classical computing cluster: A classical computing cluster is a collection of computing resources based on traditional computer architectures. It is composed of multiple physical servers or virtual machines that work together to process large-scale computing tasks. Based on the classical computing model (von Neumann architecture), using traditional hardware such as CPUs and GPUs, it is suitable for high-performance computing, big data processing, and distributed computing. Currently, it is mainly applied to supercomputer systems and cloud computing platforms.

[0048] Quantum computing cluster: A quantum computing cluster consists of multiple quantum processing units (QPUs) that utilize the properties of quantum mechanics (such as superposition and entanglement) to perform calculations. It performs calculations based on qubits, with processing speeds far exceeding classical computing, and requires combination with classical computers for control, correction, and result analysis. Currently, it is mainly applied to fields such as cryptography cracking and factorization.

[0049] Quantum-Classical Hybrid Computing: Quantum-classical hybrid computing is a hybrid computing model that combines quantum computing and classical computing, aiming to combine the advantages of quantum computing with the wide applicability of classical computing. Among them, classical computing handles general tasks, and quantum computing handles specific high-complexity problems. Currently, it is mainly applied in fields such as optimization algorithm acceleration and quantum-assisted AI training.

[0050] The task scheduling and execution system for quantum-classical hybrid computing realizes the scheduling and execution of computing tasks based on classical computing clusters and quantum computing clusters. Usually, a task includes quantum computing and classical computing. For the task execution system, it is necessary to coordinate the operation interaction between quantum tasks and classical tasks, achieve the efficient connection of classical computing resources and quantum computing resources, and realize efficient scheduling in the quantum-classical hybrid computing task environment.

[0051] In related technologies, quantum and classical computing clusters are integrated under the same platform. Due to the heterogeneity of quantum computing clusters, such as the different characteristics of superconducting, ion trap, and optical quantum computer clusters, the scheduling of the quantum computing task scheduling and execution system is not easy to be implemented in one system, which improves the coupling degree of the system, increases the maintenance difficulty of the system, and reduces the scalability of the system.

[0052] For this reason, on the one hand, the present disclosure provides an execution system 100 for a quantum-classical hybrid computing task as shown in Figure 1 and includes:;

[0053] A quantum-classical computing service platform 110, which is used to parse a quantum-classical hybrid computing task to generate a subtask set, and generate a task work queue corresponding to the subtask set according to the dependency relationship between the subtasks;

[0054] Among them, the quantum-classical computing service platform 110 includes a task parsing module 111, a task classification module 112, and a task work queue module 113. The task parsing module 111 is used to parse a quantum-classical hybrid computing task to obtain a number of subtasks; the task classification module 112 is used to classify the subtasks and assign task tags; the task work queue module 113 is used to arrange the subtasks into a task work queue according to the task tags.

[0055] When a quantum-classical hybrid computing task is uploaded to the quantum-classical computing service platform 110, the task parsing module 111 disassembles the quantum-classical hybrid computing task into multiple code blocks, the task classification module 112 assigns task tags to the code blocks, constructs these code blocks into executable files as subtasks, and the task work queue module 113 sorts the subtasks according to the task tags.

[0056] The computing task scheduling and execution system 120 has a distributed architecture. According to the different types of subtasks, the quantum computing subtasks are sent to the corresponding quantum computing task scheduling and execution systems, and the service nodes in the quantum computing task scheduling and execution systems execute the quantum computing subtasks; and the classical computing subtasks are sent to the classical computing task scheduling and execution system, and the service nodes in the classical computing task scheduling and execution system execute the classical computing subtasks. Among them, the computing task scheduling and execution system adopts a distributed architecture, that is, different task scheduling and execution systems are distributed on multiple computers or servers, and the architecture mode of communicating and collaborating through the network can improve the scalability, availability, performance and fault tolerance of the system.

[0057] In each classical computing task scheduling and execution system or quantum computing task scheduling and execution system, there are a task monitoring module, a task management module, a resource management module and a task queue management module. The task monitoring module is used to monitor the subtasks, the task management module is used to manage the subtasks, the resource management module is used to manage the resources, and the task queue management module is used to reorder the allocated subtasks to obtain a scheduling queue.

[0058] In one embodiment, at least one of the classical computing task scheduling and execution system 121, the superconducting quantum computing task scheduling and execution system 123, the ion trap quantum computing task scheduling and execution system 125, and the optical quantum computing task scheduling and execution system 127 is included in several computing task scheduling and execution systems 120. The classical computing task scheduling and execution system 121 is a system based on classical physics principles that uses binary data 0 and 1 for information storage and processing. The superconducting quantum computing task scheduling and execution system 123 uses qubits made of superconducting materials for computing and processes quantum subtasks. The ion trap quantum computing task scheduling and execution system 125 is a system that processes quantum subtasks based on ion trap technology. The optical quantum computing task scheduling and execution system 127 is a system that processes quantum subtasks through the quantum properties of photons. In addition, since the task scheduling and execution system adopts a distributed architecture, other quantum computing task scheduling and execution systems 129 can also be extended to enhance the quantum computing ability. The other quantum computing task scheduling and execution systems 129 can be optionally a nuclear magnetic resonance quantum computing task scheduling and execution system, a quantum dot computing quantum task scheduling and execution system, or a cold atom quantum computing task scheduling and execution system.

[0059] Each computing task scheduling and execution system 120 is correspondingly provided with a computing partition, and multiple service nodes are registered in each computing partition. For example, the classic computing task scheduling and execution system 121 is provided with a classic computing partition 122, and a general computing partition and a parallel computing partition are provided in the classic computing partition 122. The superconducting quantum computing task scheduling and execution system 123 is provided with a superconducting quantum computing partition 124, and the superconducting quantum computing partition includes a real machine computing partition and a simulation computing partition. The ion trap quantum computing task scheduling and execution system 125 is provided with an ion trap quantum computing partition 126, the optical quantum computing task scheduling and execution system 127 is provided with an optical quantum computing partition 128, and the other quantum computing task scheduling and execution system 129 is provided with another quantum computing partition 1210.

[0060] A two-level high-speed storage architecture is adopted within the computing task scheduling and execution system 120, where the first level is the cache and the second level is the database permanent storage. During the scheduling and execution process of the subtasks of the quantum-electric hybrid computing task, after each subtask ends, the result is preferentially stored in the cache and then stored in the database permanent storage. When the next subtask in the work queue of the quantum-electric hybrid computing task starts to execute, data will be preferentially retrieved from the cache. If the data is not found in the cache, the data stored permanently in the database will be retrieved and cached in the first-level cache.

[0061] The computing task scheduling and execution system 120 adopts an asynchronous event-driven architecture. Through the implementation of loosely coupled asynchronous communication, the asynchronous message passing of multiple tasks in the scheduling and execution system is completed. In the asynchronous event-driven architecture, after the system triggers an event, it will not wait for the event to be processed before continuing to execute other tasks. During the scheduling and execution process of the subtasks, each subtask will register a callback interface in the scheduling and execution system when it starts to execute, and is uniformly managed by the computing task scheduling and execution system 120. When the subtask is executed, the computing task scheduling and execution system 120 triggers the event of the subtask completion, actively calls the above-registered callback interface, and the computing task scheduling and execution system 120 updates the progress node of the work queue of the quantum-electric hybrid computing task. When the last subtask of a quantum-electric hybrid computing task is executed, the work queue completion event is triggered, and the computing task scheduling and execution system 120 asynchronously notifies the quantum-electric computing service platform 110. After receiving the notification, the quantum-electric computing service platform 110 actively sends a message notification to the user. After receiving the message, the user can go to the quantum-electric computing service platform 110 in time to view the task execution result.

[0062] The computing task scheduling and execution system 120 adopts object storage or distributed storage technology, taking into account storage efficiency and sharing characteristics, and realizes the function of sharing the execution results of quantum-electric hybrid computing tasks and subtasks among multiple clusters. During the scheduling and execution process of subtasks in classical computing and quantum computing, when a subtask is completed, the computing task scheduling and execution system 120 adds a shared storage address of the subtask result to the task information database, so that data can be obtained from the shared storage address in the database even if the next subtask and the current subtask are not in the same cluster. After all the subtasks belonging to the quantum-electric hybrid computing task are completed, the computing task scheduling and execution system 120 will first aggregate and process the subtask results, and the formed result is the final execution result of the quantum-electric hybrid computing task. The shared storage address of the final execution result is stored as a link in the quantum-electric computing service platform 110, and users can view the execution result of the quantum-electric hybrid computing task through the shared storage address link on the quantum-electric computing service platform 110.

[0063] The scheduling adapter 130 is communicatively connected to the quantum-electric computing service platform 110 and the computing task scheduling and execution system 120. The scheduling adapter 130 is used to provide real-time feedback on the progress of task execution in the task work queue. The quantum-electric computing service platform selects to send subtasks in the task queue to the computing task scheduling and execution system 120 according to the task execution progress. The subtasks include classical computing subtasks and quantum computing subtasks. The scheduling adapter 130 is a software module unit, which includes an interface for communicating with the quantum-electric computing service platform 110 and an interface for communicating with the computing task scheduling and execution system 120. For example, the interface for communicating with the quantum-electric computing service platform 110 is optionally a RESTful API interface for receiving task distribution instructions and sending task execution progress reports, and the interface for communicating with the computing task scheduling and execution system 120 is optionally a message queue-based communication interface such as a JMS interface, an AMQP interface or an MQTT interface for receiving task execution status updates. Through the scheduling adapter 130, computing tasks can be assigned to the corresponding task scheduling and execution system, making resource allocation more reasonable and maximizing resource utilization.

[0064] In a second aspect, there is provided a Figure 2 task execution method for a quantum-electric hybrid computing task as shown in

[0065] S1. The quantum-electric computing service platform receives a quantum-electric hybrid computing task;

[0066] The receipt of the quantum-electricity hybrid computing task can be achieved by entering command-line code scripts in the script box provided by the quantum-electricity computing service platform, or by using the code online editing software provided by the platform, such as Vscode, Jupyter Notebook, etc. The quantum-electricity computing service platform is the top layer of the overall system architecture. It can directly face user developers or connect to various third-party platforms through interfaces. Therefore, users can also use the mature quantum algorithm application services in third-party platforms. The back end of the third-party platform connects to the interface of the quantum-electricity computing service platform to submit the quantum-electricity hybrid computing task.

[0067] The quantum-electricity hybrid computing task includes a labeled quantum-electricity hybrid computing task and a hybrid computing subtask that has been disassembled into an ordered sequence. The labeled quantum-electricity hybrid computing task refers to marking the classical code part and the quantum code part. The hybrid computing subtask that has been disassembled into an ordered sequence requires the user to manually disassemble the quantum-electricity hybrid computing task and sort it.

[0068] As Figure 3 shown, after the user edits the quantum-electricity hybrid computing task, they can choose whether to submit it through a third-party platform. They can either submit it through a third-party platform or directly on the quantum-electricity computing service platform. After submitting the quantum-electricity hybrid computing task, the quantum-electricity computing service platform determines the type of the quantum-electricity hybrid computing task. For the labeled quantum-electricity hybrid computing task, S2 is executed to parse the task. For the hybrid computing subtask that has been disassembled into an ordered sequence, S2 is skipped and S3 is directly executed to queue the subtasks into the task work queue.

[0069] After the user submits the quantum-electricity hybrid computing task, the quantum-electricity computing service platform uses a combination of symmetric encryption technology (AES) and asymmetric encryption (RSA) technology to ensure that the user's code is encrypted during storage and transmission and cannot be directly viewed. After the user's code is submitted, the system generates a symmetric key to encrypt the user's code, and then uses the public key of the system to encrypt the symmetric key. The encrypted user's code and symmetric key are stored in the database. When executing the task code subsequently, the encrypted code and key are retrieved from the database, and the user's private key is used to decrypt the symmetric key. The decryption of the code by the symmetric key is completed in an isolated sandbox environment to prevent the decrypted code from being accessed externally. During runtime, the code running environment is isolated using containers or virtual machines. The symmetric encryption key is set to be valid for a short period and is immediately destroyed after use, and a hardware security module is set up to protect the key. In this step, the platform only generates a symmetric key to encrypt the user's code. The symmetric key is encrypted by the public key submitted by the user and stored in the database, and the encrypted code is also stored in the database. The quantum-electricity hybrid computing task submitted by the user is transmitted to the quantum-electricity computing service platform after being encrypted.

[0070] S2. The quantum-electricity computing service platform parses the quantum-electricity hybrid computing task to generate a subtask set;

[0071] After decrypting the quantum-electric hybrid computing task, the quantum-electric hybrid computing task is disassembled into multiple code blocks with serial numbers and task tags, and the code blocks are constructed into an executable file, which is the subtask. The task tags include quantum-electric task information, serial-parallel information, and context information. Among them, the quantum-electric task information is used to determine whether it is a quantum code or a classical code, the serial-parallel information is used to determine whether it is a serial task or a parallel task, and the context information is used to determine the subtasks that are contextually related.

[0072] In one embodiment, the quantum-electric hybrid computing task T submitted by the user is divided into subtasks numbered T1, T2, T3,..., Tn. Assign the E tag to the classical computing part and the Q tag to the quantum computing part according to the code attributes; assign the S tag to the serial computing part and the P tag to the parallel computing part according to the serial-parallel characteristics of the code. Assign context tags according to the order of the subtasks. For example, the subtask T2 is marked as (Q; S; above: T1, below: T3), indicating that this subtask is the second subtask in the quantum-electric hybrid computing task, belongs to the quantum code, the characteristic is serial computing, the above is the subtask T1, and the below is the subtask T3.

[0073] S3. Generate a task work queue corresponding to the subtask set according to the dependency relationship between the subtasks;

[0074] The quantum-electric computing service platform performs work queue scheduling on the subtask set of the quantum-electric hybrid computing task obtained by parsing the quantum-electric hybrid computing task, clarifies its context dependency relationship, and ensures that the subtasks of the quantum-electric hybrid computing task are driven to execute in a clear order.

[0075] In one embodiment, the quantum-electric hybrid computing task T is parsed into subtasks numbered T1, T2, T3, T4, T5, and the tags are T1(E; S; above: null, below: T2), T2(Q; S; above: T1, below: T3, T4), T3(E; P; above: T2, below: T5), T4(E; S; above: T2, below: T5), T5(E; S; above: T3, T4, below: null), as Figure 4 shown, the task work queue is T1 → T2 → T3, T4 → T5. Each subtask has a context dependency relationship. T2 must run based on the running result of T1. T3 and T4 must run based on the running result of T2, and T3 and T4 can run asynchronously. T5 must run based on the running results of T3 and T4.

[0076] T3 and T4 can be flexibly adjusted according to the occupancy of existing resources. For example, T3 and T4 can be adjusted to serial, and the task work queue is T1→T2→T3→T4→T5. Compared with the parallel T3 and T4 task queues, the usage of classic computing resources in the same time period is reduced, but the total running time of the hybrid computing task is also increased. The scheduling strategy of the task work queue can be adjusted according to the priority selected by the user when uploading the hybrid computing task, so as to improve the user's freedom of choice.

[0077] The task work queue module automates the process management of the work queue of the hybrid computing task, tracks the progress of the hybrid computing task work queue in real time, and displays it visually on the hybrid computing service platform.

[0078] S4, the scheduling adapter feeds back the task execution progress in the task work queue in real time, and the power calculation service platform selects to send the subtasks in the task queue to the computing task scheduling execution system according to the task execution progress;

[0079] The scheduling adapter polls the computing task scheduling execution system for resource information, clarifies the resource status of each computing task scheduling execution system, and allocates subtasks to the corresponding computing task scheduling execution system according to the order of the task work queue and the amount of power task information.

[0080] When a task is in an intermediate state of execution, the scheduling execution system must also continuously poll the task status and update the task status to the database.

[0081] At the same time, the quantitative and electrical computing service platform needs to synchronize the task status and resource status with the quantitative and electrical hybrid computing task scheduling and execution system at all times, and display the task information and resource information to the user in a visual form.

[0082] In one embodiment, for example, there are four hybrid computing task queues A, B, C, and D being executed, and the subtasks in each hybrid computing task queue are arranged in series. At this time, the newly created tasks in the scheduling execution system are A1 (E; S), B4 (Q; S), C3 (E; P), and D6 (Q; S), wherein omitting the context information does not affect the example expression. The scheduling adapter assigns A1 and C3 to the classical computing task scheduling execution system according to the task labels, and assigns B4 and D6 to the quantum task scheduling execution system.

[0083] S5. Send the subtask to the corresponding computing task scheduling and execution system according to the type of the subtask. The computing task scheduling and execution system allocates resources to the subtask to execute the subtask and obtain the computing result of the subtask.

[0084] These include:

[0085] S51. The quantum computing subtask is sent to the quantum computing task scheduling and execution system, and the service node in the quantum computing task scheduling and execution system executes the quantum computing subtask to obtain the calculation result of the quantum computing subtask.

[0086] S52. The classical computing subtask is sent to the classical computing task scheduling and execution system, and the service node in the classical computing task scheduling and execution system executes the classical computing subtask to obtain the calculation result of the classical computing subtask.

[0087] The computing task scheduling and execution system reorders the allocated subtasks according to the scheduling policy, creates a scheduling queue, and allocates resources to the subtasks in sequence according to the order of the scheduling queue to execute the subtasks, obtaining the calculation results of the subtasks.

[0088] When the quantum computing task scheduling and execution system executes the quantum computing subtask, it first executes through the real machine computing partition. When the resource stability of the quantum computing subtask in the real machine computing partition is lower than the set threshold, it switches to the simulation computing partition for execution.

[0089] A process engine sub-module is set in the computing task scheduling and execution system, and the scheduling queue is controlled by the process engine sub-module.

[0090] The scheduling policy of the subtasks adopts first-in-first-out. After being rearranged by the scheduling policy, for the subtasks C3(E;P) and A1(E;S) allocated to the classical computing task scheduling and execution system, and D6(Q;S) and B4(Q;S) allocated to the quantum task scheduling and execution system, at this time, each scheduling and execution system partition starts to issue tasks according to the order of the subtask scheduling queue. Assuming that the resources are sufficient after the C3 subtask is issued in the classical partition, the A1 subtask will continue to be issued. After the D6 subtask is issued in the quantum partition, the remaining quantum resources are not enough to execute the B4 subtask, and the B4 subtask will continue to wait and not be executed until enough quantum resources are released. When the stability of the quantum computing resources is poor, other quantum computing resources are used as substitutes, or the quantum simulator takes over small-scale quantum computing tasks and returns the task execution results.

[0091] During the scheduling and execution process of the quantum-electric hybrid computing task subtasks, after each subtask ends, the result is preferentially stored in the cache and then stored in the database for permanent storage. When the next subtask in the quantum-electric hybrid computing task work queue starts to execute, it will preferentially extract data from the cache. If the data is not found in the cache, it will enter the database to obtain the permanently stored data and cache the data in the first-level cache.

[0092] After all the subtasks belonging to the equivalent power hybrid calculation task are completed, the calculation task scheduling and execution system 120 will first aggregate and process the subtask results. The formed result is the final execution result of the equivalent power hybrid calculation task. The shared storage address of the final execution result is stored in the power calculation service platform in the form of a link. Users can view the execution result of the equivalent power hybrid calculation task on the power calculation service platform through the link of the shared storage address.

[0093] Thirdly, as Figure 5 shown, an electronic device is provided, which is characterized by including: a memory and a processor;

[0094] The memory stores computer execution instructions;

[0095] The processor executes the computer execution instructions stored in the memory, so that the processor executes the above method.

[0096] In one embodiment, the electronic device 200 includes: at least one processor 201 and a memory 202. Optionally, the electronic device 200 further includes a communication component 203. Among them, the processor 201, the memory 202, and the communication component 203 are connected through a bus 204.

[0097] In the specific implementation process, at least one processor 201 executes the computer execution instructions stored in the memory 202, so that at least one processor 201 executes the above method.

[0098] For the specific implementation process of the processor 201, reference can be made to the above method embodiment. The implementation principle and technical effect are similar, and will not be elaborated here in this embodiment.

[0099] In the above embodiment, it should be understood that the processor may be a central processing unit (English: Central Processing Unit, abbreviated as: CPU), and may also be other general-purpose processors, digital signal processors (English: Digital Signal Processor, abbreviated as: DSP), application specific integrated circuits (English: Application Specific Integrated Circuit, abbreviated as: ASIC), etc. The general-purpose processor may be a microprocessor or the processor may also be any conventional processor, etc. The steps of the method disclosed in combination with the invention can be directly embodied as being completed by the execution of the hardware processor, or by the combination of hardware and software modules in the processor.

[0100] The memory may include a high-speed memory (Random Access Memory, RAM), and may also include a non-volatile memory (Non-volatile Memory, NVM), such as at least one disk memory.

[0101] The bus can be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, an Extended Industry Standard Architecture (EISA) bus, etc. The bus can be divided into an address bus, a data bus, a control bus, etc. For the sake of convenience in representation, the buses in the attached drawings of this application are not limited to only one bus or one type of bus.

[0102] Finally, it should be noted that: After considering the specification and practicing the invention disclosed herein, those skilled in the art will readily think of other embodiments of the present disclosure. The present disclosure is intended to cover any variations, uses, or adaptations of the present disclosure, which follow the general principles of the present disclosure and include the common general knowledge or conventional technical means in the technical field not disclosed in the present disclosure. It is not limited to the exact structures described above and shown in the drawings, and various modifications and changes can be made without departing from its scope. The scope of the present disclosure is only limited by the appended claims.

Claims

1. A system for executing a hybrid computing task, characterized in that: include: The power quantity calculation service platform is used to parse the power quantity mixed calculation task to generate a subtask set, and generate a task work queue corresponding to the subtask set according to the dependency relationship between the subtasks; A scheduling adapter, used to provide real-time feedback on the progress of task execution in the task work queue. The quantum electrical computing service platform selects to send subtasks in the task queue to the computing task scheduling execution system according to the progress of task execution. The subtasks include classical computing subtasks and quantum computing subtasks. The computing task scheduling execution system is a distributed architecture, which is used to send the quantum computing subtask to the corresponding quantum computing task scheduling execution system according to different subtask types, and the service node in the quantum computing task scheduling execution system executes the quantum computing subtask; And the classical computing subtask is sent to the classical computing task scheduling and execution system, and the classical computing subtask is executed by the service node in the classical computing task scheduling and execution system.

2. The execution system of a hybrid quantitative and electrical computing task according to claim 1, characterized in that: The quantitative electricity computing service platform includes a task parsing module, a task classification module and a task work queue module. The task parsing module is used to parse the quantitative electricity mixed computing task to obtain a plurality of subtasks; the task classification module is used to classify the subtasks and add task labels; the task work queue module is used to arrange the subtasks into a task work queue according to the task labels.

3. The execution system of a hybrid quantitative and electrical computing task according to claim 1, characterized in that: The quantum computing task scheduling and execution system at least includes one or more of a superconducting quantum computing task scheduling and execution system, an ion trap quantum computing task scheduling and execution system, and an optical quantum computing task scheduling and execution system.

4. The execution system of a hybrid quantitative and electrical computing task according to claim 3, characterized in that: The classical computing task scheduling execution system is provided with a classical computing partition, the quantum computing task scheduling execution system is provided with a quantum computing partition, and the quantum computing partition is provided with a real machine computing partition and a simulation computing partition.

5. A method for executing a hybrid computing task, characterized in that: A system for executing a hybrid quantitative and electrical computing task as claimed in any one of claims 1 to 4, comprising the following steps: The quantitative and electrical computing service platform analyzes the quantitative and electrical hybrid computing task to generate a subtask set, and generates a task work queue corresponding to the subtask set according to the dependency relationship between the subtasks; The scheduling adapter feeds back the task execution progress in the task work queue in real time, and the power calculation service platform selects to send the subtasks in the task queue to the computing task scheduling execution system according to the task execution progress; According to the type of the subtask, the quantum computing subtask is sent to the quantum computing task scheduling and execution system, and the service node in the quantum computing task scheduling and execution system executes the quantum computing subtask to obtain the calculation result of the quantum computing subtask; The classical computing subtask is sent to the classical computing task scheduling and execution system, and the service node in the classical computing task scheduling and execution system executes the classical computing subtask to obtain the calculation result of the classical computing subtask.

6. The method for executing a hybrid quantitative and electrical computing task according to claim 5, characterized in that: Before the electrical quantity computing service platform parses the electrical quantity hybrid computing task to generate a subtask set, and generates a task work queue corresponding to the subtask set according to the dependency relationship between the subtasks, it also includes receiving the electrical quantity hybrid computing task, wherein the electrical quantity hybrid computing task is a marked electrical quantity hybrid computing task or has been disassembled into ordered hybrid computing subtasks, and the marked electrical quantity hybrid computing task executes the step of the electrical quantity computing service platform parsing the electrical quantity hybrid computing task to generate a plurality of subtasks arranged in a task work queue, and the step of the hybrid computing subtasks that have been disassembled into ordered executing the scheduling adapter to allocate the subtasks to the computing task scheduling execution system according to the order of the task work queue and the task tags of the subtasks.

7. The method for executing a hybrid quantitative and electrical computing task according to claim 5, characterized in that: The quantitative-electrical computing service platform parses the quantitative-electrical hybrid computing task to generate a subtask set, and generates a task work queue corresponding to the subtask set according to the dependency relationship between the subtasks, including: decomposing the quantitative-electrical hybrid computing task into a plurality of code blocks with sequence numbers and task labels, and constructing the code blocks into subtasks; and arranging the subtasks into a task work queue according to the sequence numbers and task labels of the subtasks.

8. The method for executing a hybrid quantitative and electrical computing task according to claim 7, characterized in that: The task tag includes power measurement task information, serial and parallel information, and context information.

9. The method for executing a hybrid quantitative and electrical computing task according to claim 7, characterized in that: The scheduling adapter provides real-time feedback on the progress of task execution in the task work queue, and the power measurement computing service platform selects to send the subtasks in the task queue to the computing task scheduling execution system according to the progress of task execution, including: the scheduling adapter polls the computing task scheduling execution system for resource information, clarifies the resource status of each computing task scheduling execution system, and allocates the subtasks to the corresponding computing task scheduling execution system according to the order of the task work queue and the power measurement task information.

10. The method for executing a hybrid quantitative and electrical computing task according to claim 8, characterized in that: The service node in the quantum computing task scheduling execution system executes the quantum computing subtask, and obtaining the calculation result of the quantum computing subtask includes: the quantum computing task scheduling execution system reorders the assigned quantum computing subtasks according to the scheduling strategy, creates a scheduling queue, and allocates resources to the quantum computing subtasks in sequence according to the order of the scheduling queue to execute the quantum computing subtasks and obtain the calculation result of the quantum computing subtask.

11. The method for executing a quantum-electric hybrid computing task according to claim 10, characterized in that: When the computing task scheduling execution system executes a quantum computing subtask, it first executes it through the real machine computing partition. When the resource stability of the quantum computing subtask in the real machine computing partition is lower than the set threshold, it switches to the simulated computing partition for execution.

12. An electronic device, characterized in that: include: Memory, processor; The memory stores computer-executable instructions; The processor executes the computer-executable instructions stored in the memory, so that the processor performs the method according to any one of claims 5 to 11.

Images