Quantum computer and super computer connection fusion method

Abstract

The invention discloses a quantum computer and super computer connection fusion method, which comprises the following steps of: constructing a quantum computer architecture and a super computer architecture, and realizing connection between the quantum computer architecture and the super computer architecture through a hardware layer; inputting the calculation data into a super computer, and performing task classification on the calculation data by utilizing an innovation fusion module to obtain a plurality of calculation tasks; utilizing an AI model to predict execution time and resource consumption of the plurality of calculation tasks on the quantum computer and the super computer respectively, and adjusting the distribution proportion of the plurality of calculation tasks on the quantum computer and the super computer; and transmitting the calculation data to the super computer and the quantum computer for classical calculation and quantum calculation according to the distribution proportion, and transmitting a calculation result back to a central node of the super computer to complete a calculation task. The invention relates to the field of quantum computing and high-performance computing, and solves the technical problem that the computing efficiency is low when a quantum computer and a super computer are used for computing in the prior art.

Description

Technical Field

[0001] The present invention belongs to the intersection of quantum computing and high-performance computing, and relates to a quantum computer and supercomputer fusion architecture and collaborative computing technology, and specifically is a method for connecting and fusing a quantum computer and a supercomputer. Background Art

[0002] With the rapid development of information technology, quantum computing has attracted widespread attention as a computing paradigm with great potential. Quantum computers are expected to show performance advantages far exceeding traditional computers in specific computing fields due to the unique properties of quantum bits, such as quantum superposition and entanglement.

[0003] The common quantum computer architecture mainly includes three parts: quantum measurement and control system, environmental system and chip system. In the existing methods of integrating quantum computers and supercomputers for computing tasks, the task content is usually simply divided, and the quantum computer and supercomputer are respectively assigned to handle their respective parts, and the two interact through traditional information transmission methods. For example, when calculating the quantum variational algorithm, the supercomputer first prepares the input, builds the quantum computer input circuit and optimizes the parameters. After that, the quantum computer executes the quantum circuit and returns the results to the supercomputer to further optimize the circuit parameters, and repeats the iteration until convergence. However, this method has obvious defects.

[0004] On the one hand, quantum computers and supercomputers have not yet achieved deep connection and fusion. In essence, they are still two independent hardware devices that perform tasks independently and only merge the data results at the end. This serial workflow leads to extremely low computing efficiency, and a lot of time is spent on data transmission and task switching, which cannot give full play to the synergy advantages of quantum computers and supercomputers. For example, when dealing with large-scale and complex computing tasks, the overall computing speed is severely restricted due to the lack of close cooperation between the two.

[0005] On the other hand, as the number of quantum bits continues to grow, the quantum computer system itself faces limitations in the computing power required to manipulate quantum bits. The current relatively independent quantum measurement and control system that loads the computing power system not only increases the complexity of the system, but also makes the integration between quantum computers and supercomputers more difficult, which is not conducive to carrying out complex task operations. For example, in high-precision simulation calculations involving many quantum bits, independent quantum measurement and control systems are difficult to efficiently coordinate quantum computer and supercomputer resources, which can easily lead to waste of computing resources and difficulties in task execution. Summary of the invention

[0006] The present invention aims to solve at least one of the technical problems existing in the prior art; to this end, the present invention proposes a method for connecting and fusing a quantum computer and a supercomputer, which is used to solve the technical problems of low computing efficiency when using quantum computers and supercomputers for calculations and difficulty in fusing the two systems.

[0007] To achieve the above objectives, a first aspect of the present invention provides a method for connecting and fusing a quantum computer and a supercomputer, comprising:

[0008] S1, build quantum computer architecture and supercomputer architecture, and connect the quantum computer architecture and supercomputer architecture through the hardware layer;

[0009] S2, inputting the computing data into the supercomputer, using the supercomputer's innovative fusion module to classify the computing data into tasks, and obtaining a number of computing tasks;

[0010] S3, using the AI ​​model of the supercomputer to predict the execution time and resource consumption of several computing tasks on the quantum computer and the supercomputer respectively, and adjusting the allocation ratio of several computing tasks on the quantum computer and the supercomputer according to the execution time and resource consumption;

[0011] S4, according to the allocation ratio, transmits the calculation data to the supercomputer and quantum computer for classical calculation and quantum calculation, and transmits the calculation results of classical calculation and quantum calculation back to the central node of the supercomputer to complete the calculation task.

[0012] Based on the above technical steps, the present invention breaks the limitation of independent operation of quantum computers and supercomputers by constructing quantum computer architecture and supercomputer architecture and realizing hardware layer connection, and combines the advantages of both. It realizes the integration of quantum computing and classical computing resources, so that the system can make full use of the efficiency of quantum computing on specific problems and the powerful ability of supercomputers in classical computing when processing complex computing tasks, and improve the overall computing power.

[0013] Furthermore, the quantum computer architecture includes a quantum chip system and a classical control and support system; wherein,

[0014] The quantum chip system includes quantum bits, a quantum bit interconnection structure and a quantum gate operation module;

[0015] The classical control and support system includes a quantum measurement and control module, an environmental control module and a classical computing resource module.

[0016] Furthermore, the quantum bit is a physical carrier of quantum information;

[0017] The quantum bit interconnection structure is used for information exchange between quantum bits;

[0018] The quantum gate operation module is used to perform quantum logic gate operations on quantum bits.

[0019] Furthermore, the quantum measurement and control module is used to divide the programs of computing power circuit operation and quantum circuit operation;

[0020] The environmental control module is used to maintain the specific environment required by the quantum bit; wherein the specific environment includes ambient temperature, ambient air pressure and ambient light;

[0021] The classic computing resource module is used to monitor and adjust the environment control module using classic computing resources; wherein the classic computing resources include CPU and GPU.

[0022] Furthermore, the division of the programs for computing power circuit operation and quantum circuit operation includes:

[0023] Determine whether the implementation of the program is based on the direct principles of quantum mechanics; if so, divide the program into quantum circuits and execute it on a quantum computer; if not, mark the program as a computing power circuit and execute it on a supercomputer.

[0024] The quantum measurement and control module divides the programs reasonably and assigns them to quantum or supercomputers for execution based on whether the programs are based on the principles of quantum mechanics. This ensures the reasonable allocation of computing tasks between the two computing resources, allowing the quantum computer to focus on processing quantum-related computing tasks that it is good at, avoiding waste of resources. At the same time, it works in collaboration with supercomputers to achieve complementary advantages between quantum computing and classical computing, thereby improving overall computing efficiency.

[0025] Furthermore, the quantum measurement and control module also includes a classical control unit, a signal transceiver unit and a computing power processing control unit; wherein,

[0026] The classical control unit is connected to the input interface of the computing node in the supercomputer via a transmission interface;

[0027] The output end of the signal transceiver unit is connected to the network interface of the supercomputer, and the signal transceiver unit is used to transmit the control instructions and calculation data of the computing power processing unit, and output the microwave or laser pulse control signal to the computing power processing control unit for calculation;

[0028] The computing power processing control unit is used to complete the circuit compilation of the quantum computer input signal, instruction conversion, control of the signal transceiver unit, and processing and transmission feedback of the output signal.

[0029] Furthermore, the specific environment required to maintain the quantum bit includes:

[0030] A1, using sensors to collect ambient temperature, ambient air pressure and ambient light data to obtain monitoring data;

[0031] A2, real-time acquisition of the thermal excitation rate of quantum bits. When the thermal excitation rate is greater than the maximum preset threshold, the adjustment mechanism is activated to adjust the monitoring data until the thermal excitation rate is less than the minimum preset threshold.

[0032] The environmental control module collects environmental data through sensors, monitors parameters such as ambient temperature, air pressure, and light in real time, and adjusts the environment according to the thermal excitation rate of the quantum bit to ensure the stability of the quantum bit, reduce the risk of quantum bit decoherence, and extend the coherence time of the quantum bit, thereby improving the accuracy and reliability of quantum computing. A stable quantum bit state is the basis for achieving precise quantum computing operations and helps improve the overall performance of quantum computing.

[0033] Furthermore, the supercomputer architecture includes an innovative fusion module, a computing node, a central node, a heterogeneous operating system, a quantum task management module, a high-precision synchronization module, and a storage fusion module; wherein,

[0034] The innovative fusion module is used to perform task classification on computing data using a task classification model based on deep learning;

[0035] The computing node is used to receive and execute the assigned computing tasks;

[0036] The central node is used to receive the calculation results of several computing nodes of the supercomputer and the quantum computer;

[0037] The heterogeneous operating system is used to execute step S3, and includes a task scheduling module, a communication module, a monitoring module and a resource management module;

[0038] The quantum task management module is used to manage the data channel between the quantum chip and the computing node of the supercomputer, and monitor the data transmission status;

[0039] The high-precision synchronization module is used to synchronize the execution process of classical computing and quantum computing by using clock synchronization technology and timestamps;

[0040] The storage fusion module is used to map quantum state data into classical memory and utilize a distributed storage management mechanism to maintain data consistency between the calculation results of the supercomputer and the quantum computer.

[0041] The high-precision synchronization module uses clock synchronization technology and timestamps to accurately synchronize the execution process of classical computing and quantum computing, ensuring the accuracy and efficiency of the two working together, avoiding calculation errors and delays caused by synchronization problems, ensuring that quantum computing and classical computing are closely coordinated in time, and improving the overall efficiency of the computing process. The quantum task management module manages the data channel between the quantum chip and the supercomputer computing node, monitors the data transmission status, ensures fast and stable data transmission, avoids serial communication bottlenecks, improves data interaction efficiency, and accelerates the execution of computing tasks.

[0042] Furthermore, the task classification model and the AI ​​model are constructed based on a deep learning algorithm; wherein,

[0043] The input of the task classification model is the line depth and algorithm type of the calculation data, and the output is the task type;

[0044] The input of the AI ​​model is the resource consumption, execution time, decoherence time and thermal excitation rate of quantum bits during the calculation process of several historical computing tasks, and the output is the execution time and resource consumption of several computing tasks on quantum computers and supercomputers, as well as the decoherence time of quantum bits.

[0045] Furthermore, the adjusting the distribution ratio of a number of computing tasks on quantum computers and supercomputers according to execution time and resource consumption includes:

[0046] Setting the initial allocation ratio of several computing tasks on quantum computers and supercomputers based on execution time and resource consumption;

[0047] Several computing tasks input computing data into quantum computers and supercomputers for computing according to the initial allocation ratio, and use the monitoring module in the heterogeneous operating system to monitor the real-time performance and task load of quantum bits; the real-time performance includes quantum bit coherence time, gate fidelity and quantum bit error rate, and the task load includes computing task complexity, computing task data volume and resource utilization of quantum computers and supercomputers;

[0048] According to the real-time performance of quantum bits and task load, the distribution ratio of several computing tasks on quantum computers and supercomputers is dynamically adjusted.

[0049] With the help of the supercomputer's AI model, the execution time and resource consumption are predicted, and the computing task allocation ratio is adjusted accordingly, realizing the dynamic optimization configuration of resources. The computing tasks can be flexibly allocated according to the real-time performance of quantum computers and supercomputers and the specific requirements of the tasks. The system can be adaptively adjusted according to different computing tasks and hardware status, which improves the robustness of the system and can better meet the diverse computing needs.

[0050] Compared with the prior art, the present invention has the following beneficial effects:

[0051] By connecting the quantum computer architecture and the supercomputer architecture at the hardware level, the deep integration of the two is achieved, breaking the traditional independent working mode, enabling quantum computing and classical computing to be carried out in coordination, and improving the overall computing efficiency;

[0052] Through the built-in task classification model and AI model, the computing task type can be quickly identified and resource load estimation can be performed, realizing real-time adjustment and reasonable allocation of quantum and classical computing tasks. The computing tasks can be flexibly allocated according to the real-time performance of quantum computers and supercomputers and the specific requirements of the tasks.

[0053] The quantum task management module in the supercomputer architecture ensures high-speed data transmission between quantum chips and supercomputer computing nodes, avoiding communication bottlenecks; the storage fusion module realizes the effective mapping of quantum state data and classical memory and data consistency management, reduces storage conversion overhead and data access delay, improves the efficiency and reliability of data management, and helps to improve the performance of the entire computing system. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0055] Figure 1 A schematic diagram of a flow chart of a method for connecting and fusing a quantum computer and a supercomputer provided by the present invention;

[0056] Figure 2 A diagram of the quantum computer architecture provided by the present invention;

[0057] Figure 3 A schematic diagram of the framework of the quantum measurement and control module provided by the present invention;

[0058] Figure 4 A diagram of the supercomputer architecture provided by the present invention;

[0059] Figure 5 A schematic diagram of the framework of the heterogeneous operating system provided by the present invention. DETAILED DESCRIPTION

[0060] The technical solution of the present invention will be clearly and completely described below in conjunction with the embodiments. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0061] See also Figure 1-Figure 5 The first aspect of the present invention provides a method for connecting and fusing a quantum computer and a supercomputer, comprising:

[0062] S1, build quantum computer architecture and supercomputer architecture, and connect the quantum computer architecture and supercomputer architecture through the hardware layer;

[0063] S2, inputting the computing data into the supercomputer, using the supercomputer's innovative fusion module to classify the computing data into tasks, obtaining the task type of the computing data, and obtaining a number of computing tasks;

[0064] S3, using the AI ​​model of the supercomputer to predict the execution time and resource consumption of several computing tasks on the quantum computer and the supercomputer respectively, and adjusting the allocation ratio of several computing tasks on the quantum computer and the supercomputer according to the execution time and resource consumption;

[0065] S4, according to the allocation ratio, transmits the calculation data to the supercomputer and quantum computer for classical calculation and quantum calculation, and transmits the calculation results of classical calculation and quantum calculation back to the central node of the supercomputer to complete the calculation task.

[0066] The present invention starts from the perspective of deep integration of quantum computers and supercomputers, constructs quantum computer architecture and supercomputer architecture, and connects the two through the hardware layer; wherein the quantum computer architecture is divided into two main components: quantum chip system and classical control and support system;

[0067] The quantum chip system includes: quantum bits, quantum bit interconnection structure and quantum gate operation module; among them,

[0068] Quantum bits are the physical carriers of quantum information, carrying and storing quantum state information. Under the control of the quantum gate operation module, their quantum states evolve according to the principles of quantum mechanics to achieve the processing and calculation of quantum information.

[0069] The quantum bit interconnection structure is used for information exchange between quantum bits. Through specific physical connection methods (such as superconducting Josephson junctions, ion traps, etc.), quantum bits can work together to build complex quantum logic circuits to complete various quantum computing tasks.

[0070] The quantum gate operation module is used to perform quantum logic gate operations on quantum bits, including single-bit gate operations (such as Hadamard gates, Pauli gates, etc.) and multi-bit gate operations (such as CNOT gates, etc.). By precisely controlling microwave pulses or laser pulses, the quantum state of quantum bits can be precisely controlled to achieve various logical operations in quantum algorithms.

[0071] The classical control and support system provides the necessary control, support and operating environment for quantum computers, ensuring that the quantum chip system can operate stably and efficiently; it includes quantum measurement and control modules, environmental control modules and classical computing resource modules; among them,

[0072] Quantum measurement and control module: used to divide the programs of computing power circuit operation and quantum circuit operation, based on the differences in functions and physical implementation methods:

[0073] Specifically, the division is made by judging whether the implementation of the program is based on the direct principles of quantum mechanics; if the implementation of the program depends on the principles of quantum mechanics, such as programs involving quantum gate operations, quantum state measurements and other operations, it will be divided into quantum circuits and executed by quantum computers; on the contrary, if the implementation of the program does not involve the principles of quantum mechanics, such as classical optimization algorithms, conventional data processing and other tasks, it will be marked as a computing power circuit and executed by a supercomputer;

[0074] The classical control and support system is implemented using classical electronics and computer technology, and is responsible for the manipulation of quantum bits and the maintenance of the environment, requiring a large amount of classical computing resources; while the quantum chip system is based on the principles of quantum mechanics and uses the superposition and entanglement of quantum bits to process quantum information. The two interact through precisely calibrated microwave signals, ensuring that during the quantum computing process, the classical control and support system can transmit control instructions to the quantum chip system, while the quantum chip system can promptly feed back the results or status information of quantum computing to the classical control and support system, thereby achieving a close integration of quantum computing and classical computing, and building an efficient and stable quantum-classical hybrid computing platform.

[0075] The environment control module in the classical control and support system is used to maintain the specific environment required by the quantum bit; wherein the specific environment includes ambient temperature, ambient air pressure and ambient light;

[0076] Specifically, in the environmental control module, sensors are first used to collect ambient temperature, ambient air pressure, and ambient light data to obtain monitoring data;

[0077] Then, the thermal excitation rate of the quantum bit is obtained in real time. When the thermal excitation rate is greater than the maximum preset threshold, the adjustment mechanism is activated to adjust the monitoring data until the thermal excitation rate is less than the minimum preset threshold, so as to stabilize the environment of the quantum bit and provide a solid guarantee for the accuracy and reliability of quantum computing.

[0078] In the regulation mechanism, there are corresponding regulation strategies and technical means for different environmental parameters. Taking the ambient temperature as an example, when the temperature needs to be lowered, the environmental control module will accurately control the operating parameters of the refrigeration system, such as adjusting the compressor speed of the refrigerator, optimizing the flow distribution of the refrigerant, etc., based on the principle of heat conduction and the law of heat exchange, so that the heat is efficiently transferred from the environment where the quantum bit is located to achieve temperature reduction. For air pressure regulation, by controlling the opening of the gas valve and combining the principle of gas dynamics, the inlet and outlet flow of the gas is adjusted to stabilize the air pressure within a range suitable for the operation of the quantum bit. In terms of illumination regulation, optical control technology, such as variable aperture, optical power regulator and other equipment, is used to adjust the illumination intensity and spectral characteristics according to the specific requirements of the quantum bit for illumination, ensuring that the illumination environment meets the operating requirements of the quantum bit. The regulation process continues, and during the period, it is continuously adjusted according to the real-time feedback information of the quantum bit thermal excitation rate until the thermal excitation rate is less than the minimum preset threshold, so as to achieve the stabilization of the environment where the quantum bit is located, and provide a solid guarantee for the accuracy and reliability of quantum computing.

[0079] The classical computing resource module in the classical control and support system is used to utilize classical computing resources (such as CPU, GPU, etc.) to monitor and adjust the environmental control module; on the one hand, it runs the environmental control algorithm to continuously adjust the environmental parameters (such as dilution refrigerator temperature, magnetic field strength, etc.) according to the real-time status of the quantum bit (such as thermal excitation rate, coherence time, etc.); on the other hand, in the process of data transmission by the signal transceiver module of the quantum measurement and control system, it participates in the processing and conversion of classical digital signals, and provides computing power support for the stable operation of the entire system.

[0080] The quantum measurement and control module also includes a classical control unit, a signal transceiver unit and a computing power processing control unit;

[0081] Classic control unit: directly connected to the computing node (such as CPU or GPU) of the supercomputer through the high-speed PCIe interface or other high-speed interconnection technology, receiving the instructions of the supercomputer and passing them to the signal transceiver unit, and at the same time passing the data fed back by the signal transceiver unit to the supercomputer;

[0082] Signal transceiver module: When it is necessary to directly manipulate a single or a small number of quantum bits (such as preliminary verification of quantum algorithms, calibration of quantum gate operations, and other scenarios), control signals such as microwave / laser pulses are directly output to the control lines on the quantum chip. That is, it receives the classical digital signal sent by the classical computing resource module, first converts it into an analog signal, and the analog signal is converted into an actual voltage or current signal by a digital-to-analog converter (DAC), and then is converted into a microwave / laser pulse for controlling the quantum bit by a microwave / laser modulator and other equipment; when performing large-scale quantum simulations, quantum chemical calculations, and other tasks, it is mainly responsible for data transmission and status monitoring. That is, it receives the classical digital signal sent by the supercomputer (such as data representing the amplitude and frequency of microwave pulses), converts it into an analog signal, and then converts it into an actual control signal, and at the same time feeds back the status information of the quantum computer to the supercomputer;

[0083] Computing power processing control module (function replaced by supercomputer): The supercomputer directly takes over the line compilation, instruction conversion, control signal transceiver unit, output signal processing, transmission feedback and other tasks of the quantum computer input signal originally undertaken by the computing power processing control module.

[0084] It should be noted that the classic control unit is connected to the input interface of the computing node in the supercomputer through the transmission interface; the output end of the signal transceiver unit is connected to the network interface of the supercomputer.

[0085] The supercomputer architecture in this embodiment includes an innovative fusion module, a computing node, a central node, a heterogeneous operating system, a quantum task management module, a high-precision synchronization module, and a storage fusion module; wherein,

[0086] An innovative fusion module for task classification of computing data using a deep learning-based task classification model;

[0087] Computing nodes, used to receive and execute assigned computing tasks;

[0088] The central node is used to receive the calculation results of several computing nodes of the supercomputer and the quantum computer;

[0089] A heterogeneous operating system, used to execute step S3, and comprising a task scheduling module, a communication module, a monitoring module and a resource management module;

[0090] The quantum task management module is used to manage the data channel between the quantum chip and the computing nodes of the supercomputer and monitor the data transmission status;

[0091] High-precision synchronization module, used to synchronize the execution of classical and quantum computing processes using clock synchronization technology and timestamps;

[0092] The storage fusion module is used to map quantum state data into classical memory and use a distributed storage management mechanism to maintain data consistency between the calculation results of supercomputers and quantum computers.

[0093] In the supercomputer architecture, both the task classification model and the AI ​​model are built based on deep learning algorithms; the input of the task classification model is the line depth and algorithm type of the computing data, and the output is the task type; the input of the AI ​​model is the resource consumption, execution time, and decoherence time and thermal excitation rate of several historical computing tasks during the computing process, and the output is the execution time and resource consumption of several computing tasks on quantum computers and supercomputers, as well as the decoherence time of quantum bits.

[0094] In supercomputers, first of all, at the algorithm level, supercomputers use pre-trained task classification models to classify computing tasks according to the characteristics of input data, and obtain categories such as quantum Monte Carlo simulation, quantum chemistry simulation, and quantum machine learning. At the same time, with the help of historical data and real-time monitoring data, AI models are used to predict the resource consumption of tasks. Taking quantum Monte Carlo simulation as an example, the model can estimate the required computing resources and time based on the circuit depth and the number of quantum bits, so as to quickly identify the type of computing tasks and resource consumption;

[0095] Then, for different computing tasks, the heterogeneous operating system will select the appropriate hybrid training model (a specialized model trained for specific problems, such as a model based on a variational quantum eigensolver), and use the pre-compiled processing module (belonging to the classical control and support system, responsible for optimizing quantum circuits before the execution of quantum computing tasks, including gate decomposition, quantum bit mapping and other operations) to optimize the quantum circuit, such as implementing gate decomposition and optimizing quantum bit mapping. Then, based on the real-time performance of the hardware (such as the coherence time of the quantum bit, the gate fidelity) and the task load, the ratio of the quantum computing part and the classical computing part is dynamically adjusted to achieve adaptive task division between the two. For example, when the coherence time of the quantum bit is short, more computing tasks are assigned to the classical computer. During the execution process, when the quantum computing part is implemented, the classical computing part algorithm is executed synchronously with the assistance of the software and hardware layers, achieving heterogeneous computing parallel pipelining.

[0096] As an integrated component of the supercomputer, the quantum task management module is a high-speed data channel between the quantum chip and the supercomputer node, which enables the quantum data to be directly written into the supercomputer memory, effectively avoiding the bottleneck problem that may arise from the traditional serial communication method, improving the data transmission efficiency, and ensuring that the data can quickly and smoothly interact between the quantum chip and the supercomputer memory during the quantum computing process, providing timely and accurate data support for subsequent computing operations. At the same time, the module uses a high-precision synchronization module to ensure the temporal synergy between quantum computing and classical computing, so that the two can cooperate efficiently during the same task execution process, give full play to their respective advantages, and improve overall computing performance;

[0097] In order to reduce the storage transfer overhead, a storage fusion module is specially designed in the supercomputer architecture. It uses an innovative data mapping method to directly map quantum state data to classical memory, so that the supercomputer system can directly operate on quantum state data through direct memory access (DMA), reducing the time consumption and resource overhead caused by the transfer of data between different storage media, and improving the data processing speed. In the quantum computing process, the accuracy and integrity of the calculation results need to be reliably guaranteed. Therefore, the storage fusion module also uses a distributed storage management mechanism to avoid calculation errors or system failures caused by data inconsistency. At the same time, it reduces data access delays and optimizes the data storage and reading process, so that the system can respond more quickly to the data needs of computing tasks, improve the overall system operation efficiency, and ensure the stable and reliable operation of the system under complex computing tasks and high load conditions.

[0098] In a heterogeneous operating system, the ratio of quantum computing to classical computing is dynamically adjusted, including:

[0099] Setting the initial allocation ratio of several computing tasks on quantum computers and supercomputers based on execution time and resource consumption;

[0100] Several computing tasks input computing data into quantum computers and supercomputers for computing according to the initial allocation ratio, and use the monitoring module in the heterogeneous operating system to monitor the real-time performance and task load of quantum bits; the real-time performance includes quantum bit coherence time, gate fidelity and quantum bit error rate, and the task load includes computing task complexity, computing task data volume and resource utilization of quantum computers and supercomputers;

[0101] According to the real-time performance of quantum bits and task load, the distribution ratio of several computing tasks on quantum computers and supercomputers is dynamically adjusted.

[0102] The second embodiment of the present invention provides a computational process for completing molecular simulation using a quantum variational algorithm using the present method:

[0103] 1) Input molecular structure information and identify it as a quantum chemical calculation task through the task classification model. The supercomputer divides the molecules to be calculated into several groups (for example, each group contains 10 molecules). The grouping basis can be the similarity of molecular structure or the similarity of computational complexity. Then the properties that need to be calculated (such as ground state energy, dipole moment, excited state energy) are assigned to different computing resources. For example, the ground state energy is assigned to one group of quantum-classical computing resources, and the dipole moment is assigned to another group;

[0104] 2) Select the VQE hybrid training model and use the precompiled processing module to optimize the quantum circuit used to calculate the ground state energy of molecules. At the same time, the calculation tasks of each molecular group and the corresponding initial parameters are assigned to different computing nodes of the supercomputer. Each computing node is responsible for calculating the specific properties of a group of molecules;

[0105] 3) Each computing node sends the initial parameters of the assigned molecular group and the corresponding quantum circuit to the quantum measurement and control system. The quantum measurement and control system converts the classical control signal into a quantum control signal, and drives the quantum computer to execute the quantum circuit and measure the state of the quantum bit. The quantum computer returns the measurement result (quantum information) to the quantum measurement and control system. The quantum measurement and control system converts the quantum information into classical information and sends it back to the corresponding supercomputer computing node. While the quantum calculation is in progress, the classical optimizer on each computing node begins to execute the classical optimization algorithm (such as gradient descent, Nelder-Mead, etc.). The classical optimizer uses the quantum calculation results of the previous iteration (the initial parameters are the "results" of the first iteration) to update the variational parameters and generate new parameters;

[0106] 4) The AI ​​model predicts the convergence speed of the VQE algorithm and dynamically adjusts the parameters of the classic optimization part;

[0107] 5) After each computing node completes the property calculation of its assigned molecular group, it transmits the final calculation results (such as ground state energy, dipole moment, etc.) back to the central node of the supercomputer through a high-speed network. The central node summarizes and analyzes all calculation results and generates a final calculation report. The high-precision synchronization module ensures the synchronization of the classical optimization part and the quantum circuit execution part. The storage fusion module enables efficient access to quantum state data.

[0108] Part of the data in the above formula is calculated by removing the dimension and taking its numerical value. The formula is a formula closest to the actual situation obtained by software simulation of a large amount of collected data; the preset parameters and preset thresholds in the formula are set by technical personnel in this field according to actual conditions or obtained through simulation of a large amount of data.

[0109] Working principle of the present invention:

[0110] The present invention connects the quantum computer architecture and the supercomputer architecture through the hardware layer to achieve a deep integration of the two, breaking the traditional independent working mode and enabling quantum computing and classical computing to be carried out in collaboration; wherein, the innovative fusion module of the supercomputer uses a task classification model based on deep learning to classify the computing data, obtain the task type, and allocate it to the quantum computer and the supercomputer; then the AI ​​model of the supercomputer is used to predict the execution time and resource consumption of the computing tasks on the quantum computer and the supercomputer, and the task allocation ratio is dynamically adjusted according to these predictions; finally, according to the allocation ratio, the computing data is transmitted to the supercomputer and the quantum computer for classical computing and quantum computing, and the computing results are transmitted back to the central node of the supercomputer to complete the computing task.

[0111] At the same time, the environment control module of the quantum computer of the present invention is used to maintain the specific environment required by the quantum bit, such as temperature, air pressure and light, to ensure the stability of the quantum bit and the accuracy of the calculation;

[0112] The quantum task management module in the supercomputer of the present invention manages the data channel between the quantum chip and the supercomputer computing node, and monitors the data transmission status; the high-precision synchronization module uses clock synchronization technology and timestamps to synchronize the execution process of classical computing and quantum computing; the storage fusion module maps quantum state data to classical memory and maintains data consistency, reducing storage conversion overhead and data access delay.

[0113] The above embodiments are only used to illustrate the technical method of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical method of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical method of the present invention.

Claims

1. A method for connecting and fusing a quantum computer and a supercomputer, characterized in that: include: S1, build quantum computer architecture and supercomputer architecture, and connect the quantum computer architecture and supercomputer architecture through the hardware layer; S2, input the computing data into the supercomputer, and the task classification model of the innovative fusion module classifies the computing data into tasks to obtain several computing tasks; S3, using the AI ​​model of the supercomputer to predict the execution time and resource consumption of several computing tasks on the quantum computer and the supercomputer respectively, and adjusting the allocation ratio of several computing tasks on the quantum computer and the supercomputer according to the execution time and resource consumption; S4, according to the allocation ratio, transmits the calculation data to the supercomputer and quantum computer for classical calculation and quantum calculation, and transmits the calculation results of classical calculation and quantum calculation back to the central node of the supercomputer to complete the calculation task.

2. A method for connecting and fusing a quantum computer and a supercomputer according to claim 1, characterized in that: The quantum computer architecture includes a quantum chip system and a classical control and support system; wherein, The quantum chip system includes quantum bits, a quantum bit interconnection structure and a quantum gate operation module; The classical control and support system includes a quantum measurement and control module, an environmental control module and a classical computing resource module.

3. A method for connecting and fusing a quantum computer and a supercomputer according to claim 2, characterized in that: The qubit is the physical carrier of quantum information; The quantum bit interconnection structure is used for information exchange between quantum bits; The quantum gate operation module is used to perform quantum logic gate operations on quantum bits.

4. A method for connecting and fusing a quantum computer and a supercomputer according to claim 2, characterized in that: The quantum measurement and control module is used to divide the programs of computing power circuit operation and quantum circuit operation; The environmental control module is used to maintain the specific environment required by the quantum bit; wherein the specific environment includes ambient temperature, ambient air pressure and ambient light; The classic computing resource module is used to monitor and adjust the environment control module using classic computing resources; wherein the classic computing resources include CPU and GPU.

5. A method for connecting and fusing a quantum computer and a supercomputer according to claim 4, characterized in that: The division of the procedures for computing power circuit operation and quantum circuit operation includes: Determine whether the implementation of the program is based on the direct principles of quantum mechanics; if so, divide the program into quantum circuits and execute it on a quantum computer; if not, mark the program as a computing power circuit and execute it on a supercomputer.

6. A method for connecting and fusing a quantum computer and a supercomputer according to claim 2, characterized in that: The quantum measurement and control module also includes a classical control unit, a signal transceiver unit and a computing power processing control unit; wherein, The classical control unit is connected to the input interface of the computing node in the supercomputer via a transmission interface; The output end of the signal transceiver unit is connected to the network interface of the supercomputer, and the signal transceiver unit is used to transmit the control instructions and calculation data of the computing power processing unit, and output the microwave or laser pulse control signal to the computing power processing control unit for calculation; The computing power processing control unit is used to complete the circuit compilation of the quantum computer input signal, instruction conversion, control of the signal transceiver unit, and processing and transmission feedback of the output signal.

7. A method for connecting and fusing a quantum computer and a supercomputer according to claim 4, characterized in that: The specific environment required to maintain the quantum bit includes: A1, using sensors to collect ambient temperature, ambient air pressure and ambient light data to obtain monitoring data; A2, real-time acquisition of the thermal excitation rate of quantum bits. When the thermal excitation rate is greater than the maximum preset threshold, the adjustment mechanism is activated to adjust the monitoring data until the thermal excitation rate is less than the minimum preset threshold.

8. A method for connecting and fusing a quantum computer and a supercomputer according to claim 2, characterized in that: The supercomputer architecture includes an innovative fusion module, computing nodes, central nodes, heterogeneous operating systems, quantum task management modules, high-precision synchronization modules, and storage fusion modules; The innovative fusion module is used to perform task classification on computing data using a task classification model based on deep learning; The computing node is used to receive and execute the assigned computing tasks; The central node is used to receive the calculation results of several computing nodes of the supercomputer and the quantum computer; The heterogeneous operating system is used to execute step S3, and includes a task scheduling module, a communication module, a monitoring module and a resource management module; The quantum task management module is used to manage the data channel between the quantum chip and the computing node of the supercomputer, and monitor the data transmission status; The high-precision synchronization module is used to synchronize the execution process of classical computing and quantum computing by using clock synchronization technology and timestamps; The storage fusion module is used to map quantum state data into classical memory and utilize a distributed storage management mechanism to maintain data consistency between the calculation results of the supercomputer and the quantum computer.

9. A method for connecting and fusing a quantum computer and a supercomputer according to claim 1, characterized in that: The task classification model and the AI ​​model are constructed based on a deep learning algorithm; wherein, The input of the task classification model is the line depth and algorithm type of the calculation data, and the output is the task type; The input of the AI ​​model is the resource consumption, execution time, decoherence time and thermal excitation rate of quantum bits during the calculation process of several historical computing tasks, and the output is the execution time and resource consumption of several computing tasks on quantum computers and supercomputers, as well as the decoherence time of quantum bits.

10. A method for connecting and fusing a quantum computer and a supercomputer according to claim 1, characterized in that: The method of adjusting the distribution ratio of several computing tasks on quantum computers and supercomputers according to execution time and resource consumption includes: Setting the initial allocation ratio of several computing tasks on quantum computers and supercomputers based on execution time and resource consumption; Several computing tasks input computing data into quantum computers and supercomputers for computing according to the initial allocation ratio, and use the monitoring module in the heterogeneous operating system to monitor the real-time performance and task load of quantum bits; the real-time performance includes quantum bit coherence time, gate fidelity and quantum bit error rate, and the task load includes computing task complexity, computing task data volume and resource utilization of quantum computers and supercomputers; According to the real-time performance of quantum bits and task load, the distribution ratio of several computing tasks on quantum computers and supercomputers is dynamically adjusted.

Images