CN-121981291-A - Fusion computing method and related device of quantum classical mixed computing system

CN121981291ACN 121981291 ACN121981291 ACN 121981291ACN-121981291-A

Abstract

The application discloses a fusion computing method and a related device of a quantum classical mixed computing system, wherein the method comprises the steps of splitting task requests applied to the quantum classical mixed computing system to obtain computing subtasks independently completed by quantum computing, classical computing and intelligent computing; the method comprises the steps of responding to the type of a split computing subtask, configuring heterogeneous computing resources to corresponding computing subtasks, enabling the heterogeneous computing resources to comprise quantum computing resources, classical computing resources and intelligent computing resources, evaluating and analyzing the computing resources according to a resource evaluation and intelligent analysis module to obtain corresponding scheduling decision basis, inserting the computing subtasks into corresponding task queues, and scheduling the task queues based on the scheduling decision basis, a custom scheduling strategy and a first linkage relation of the computing subtasks.

Inventors

Dou Menghan
Request for anonymity
Request for anonymity
Request for anonymity

Assignees

本源量子计算科技（合肥）股份有限公司

Dates

Publication Date: 20260505
Application Date: 20241028

Claims (11)

1. A fusion computing method of a quantum classical hybrid computing system, the method comprising: Splitting task requests applied to a quantum classical mixed computing system to obtain computing subtasks independently completed by quantum computing, classical computing and intelligent computing, wherein the task requests are application programs with task coding characteristics based on quantum computing mixed programming, and a quantum computing part, a classical computing part and an intelligent computing part of the same application program establish a first linkage relation based on the task coding characteristics; responsive to the split computing subtask type, configuring heterogeneous computing resources to the corresponding computing subtasks, wherein the heterogeneous computing resources comprise quantum computing resources, classical computing resources and intelligent computing resources; According to the resource evaluation and the calculation resource evaluation analysis by the intelligent analysis module, a corresponding scheduling decision basis is obtained; Inserting the calculation subtasks into corresponding task queues, and scheduling the task queues based on the scheduling decision basis, the custom scheduling strategy and the first linkage relation of the calculation subtasks.
2. The method of claim 1, wherein splitting the task request for the quantum classical hybrid computing system application to obtain computing sub-tasks independently performed by quantum computing, classical computing, and intelligent computing comprises: responding to the user task requests of the mastered quantum computing part, the classical computing part and the intelligent computing part, programming components based on the quantum classical hybrid computing system so that a user determines the computing mode of a computing sub task; Or in response to a user task request without mastering the quantum computing part, the classical computing part and the intelligent computing part, determining the computing mode of the computing subtasks based on a unified programming interface and a cooperative scheduling module of the quantum classical hybrid computing system.
3. The method of claim 2, wherein the determining, in response to a user task request without grasping the quantum computing section, classical computing section, and intelligent computing section, a computing manner of the computing subtasks based on a unified programming interface and a co-scheduling module of the quantum classical hybrid computing system comprises: And acquiring a calculation task request submitted by a user, and transmitting calculation service information to a cooperative scheduling module through a unified programming interface of the quantum classical hybrid calculation system based on the calculation task request, wherein the cooperative scheduling module determines a calculation mode of the calculation task according to the calculation service information and calculation resources.
4. The method of claim 1, wherein the computation subtasks include quantum computation subtasks, classical computation subtasks, and intelligent computation subtasks.
5. The method of claim 1, wherein the scheduling the task queue based on the scheduling decision basis, a custom scheduling policy, and the first linkage relationship of the computing subtasks comprises: the method comprises the steps of scheduling a quantum task queue, a classical task queue and an intelligent task queue based on a quantum computing resource part, a first custom scheduling strategy and the first linkage relation of the quantum classical mixed computing system, or scheduling the classical task queue, the quantum task queue and the intelligent task queue based on a classical computing resource part, a second custom scheduling strategy and the first linkage relation of the quantum classical mixed computing system, or scheduling the intelligent task queue, the quantum task queue and the classical task queue based on an intelligent computing resource part, a third custom scheduling strategy and the first linkage relation of the quantum classical mixed computing system.
6. The method of claim 5, wherein the first custom scheduling policy, the second custom scheduling policy, and the third custom scheduling policy each comprise: Short job priority, first come first calculation, and task priority, or a combination thereof.
7. The method of claim 2, wherein the programming interface of the programming component of the quantum classical hybrid computing system comprises image recognition, unordered data search, network node ordering, sub-graph search, path search.
8. A fusion computing device of a quantum classical hybrid computing system, the device comprising: The system comprises a splitting unit, a first linkage relation and a second linkage relation, wherein the splitting unit is used for splitting task requests applied to a quantum classical mixed computing system to obtain computing subtasks independently completed by quantum computing, classical computing and intelligent computing; The system comprises a response unit, a response unit and a control unit, wherein the response unit is used for responding to the type of the split computing subtask and configuring heterogeneous computing resources to the corresponding computing subtask; the acquisition unit is used for carrying out evaluation and analysis on the computing resources according to the resource evaluation and the intelligent analysis module so as to acquire corresponding scheduling decision basis; The scheduling unit is used for inserting the calculation subtasks into corresponding task queues, and scheduling the task queues based on the scheduling decision basis, the self-defined scheduling strategy and the first linkage relation of the calculation subtasks.
9. An electronic device is characterized by comprising a processor and a memory; the processor is connected to a memory, wherein the memory is adapted to store a computer program, and the processor is adapted to invoke the computer program to perform the method according to claims 1-7.
10. A computer readable storage medium, characterized in that the computer readable storage medium stores a computer program comprising program instructions which, when executed by a processor, perform the method of claims 1-7.
11. A quantum computer operating system implementing a fusion calculation of a quantum classical hybrid computing system according to the method of any one of claims 1-7.

Description

Fusion computing method and related device of quantum classical mixed computing system Technical Field The invention belongs to the technical field of quantum computing, and particularly relates to a fusion computing method and a related device of a quantum classical mixed computing system. Background In the application of the current quantum classical mixed computing system, a large amount of classical computing power, quantum computing power and other multiple mixed computing power are generally required, in most application scenarios, one large computing task comprises a plurality of computing sub-tasks, dependency relations exist among the computing sub-tasks, dependency relations do not exist among the computing sub-tasks, and the computing power used by the computing sub-tasks and the optimal computing node number are different. Therefore, the application provides a fusion calculation method of a quantum classical mixed calculation system, which aims to reasonably allocate quantum classical mixed calculation resources and realize quantum classical efficient collaborative calculation. Disclosure of Invention The invention aims to provide a fusion computing method and a related device of a quantum classical mixed computing system, which aim to reasonably allocate quantum classical mixed computing resources and realize quantum classical efficient collaborative computing. One embodiment of the invention provides a fusion computing method of a quantum classical hybrid computing system, which comprises the following steps: Splitting task requests applied to a quantum classical mixed computing system to obtain computing subtasks independently completed by quantum computing, classical computing and intelligent computing, wherein the task requests are application programs with task coding characteristics based on quantum computing mixed programming, and a quantum computing part, a classical computing part and an intelligent computing part of the same application program establish a first linkage relation based on the task coding characteristics; responsive to the split computing subtask type, configuring heterogeneous computing resources to the corresponding computing subtasks, wherein the heterogeneous computing resources comprise quantum computing resources, classical computing resources and intelligent computing resources; According to the resource evaluation and the calculation resource evaluation analysis by the intelligent analysis module, a corresponding scheduling decision basis is obtained; Inserting the calculation subtasks into corresponding task queues, and scheduling the task queues based on the scheduling decision basis, the custom scheduling strategy and the first linkage relation of the calculation subtasks. Optionally, the splitting processing is performed on the task request applied by the quantum classical hybrid computing system to obtain computing sub-tasks independently completed by quantum computing, classical computing and intelligent computing, including: responding to the user task requests of the mastered quantum computing part, the classical computing part and the intelligent computing part, programming components based on the quantum classical hybrid computing system so that a user determines the computing mode of a computing sub task; Or in response to a user task request without mastering the quantum computing part, the classical computing part and the intelligent computing part, determining the computing mode of the computing subtasks based on a unified programming interface and a cooperative scheduling module of the quantum classical hybrid computing system. Optionally, the responding to the user task request without grasping the quantum computing part, the classical computing part and the intelligent computing part is based on a unified programming interface and a co-scheduling module of the quantum classical hybrid computing system, so as to determine the computing mode of the computing subtasks, including: And acquiring a calculation task request submitted by a user, and transmitting calculation service information to a cooperative scheduling module through a unified programming interface of the quantum classical hybrid calculation system based on the calculation task request, wherein the cooperative scheduling module determines a calculation mode of the calculation task according to the calculation service information and calculation resources. Optionally, the computation subtasks include a quantum computation subtask, a classical computation subtask, and an intelligent computation subtask. Optionally, the scheduling the task queue based on the scheduling decision basis, the custom scheduling policy, and the first linkage relation of the computing subtasks includes: the method comprises the steps of scheduling a quantum task queue, a classical task queue and an intelligent task queue based on a quantum computing resource part, a first custom scheduling strategy and the firs