Quantum circuit calculation method and quantum calculation platform

Abstract

The invention discloses a quantum circuit calculation method. The method comprises the steps that an original quantum circuit is decomposed, a target quantum circuit is generated, and the target quantum circuit only comprises a single-bit gate and a double-bit gate. Then, based on the quantum gate dependency relationship of the target quantum circuit, splitting the target quantum circuit to generate a first sub-circuit and a second sub-circuit; then, based on the quantum simulation machine, the first sub-line is executed, and a first density matrix is determined. Then, based on the quantum processor, the second sub-line is executed, and a second density matrix is determined. And finally, according to the first density matrix and the second density matrix, determining a target execution result corresponding to the original quantum circuit. Thus, through decomposition processing and splitting processing, the circuit structure of the original quantum circuit is simplified, and high resource consumption of a traditional full-circuit single processing mode is avoided.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of quantum computing, and more particularly, to a quantum circuit calculation method and a quantum computing platform. BACKGROUND

[0002] In related technologies, a quantum circuit is usually executed in a single processing mode of full-circuit classical simulation or full-circuit quantum true machine execution. However, a quantum circuit often includes Clifford gates and non-Clifford gates, and the single processing mode has poor utilization of computing resource, which affects the execution efficiency of the quantum circuit. SUMMARY

[0003] The present application provides a quantum circuit calculation method and a quantum computing platform.

[0004] The present application provides a quantum circuit calculation method, which comprises: decomposing an original quantum circuit to generate a target quantum circuit, wherein the target quantum circuit only includes single-bit gates and double-bit gates; splitting the target quantum circuit based on quantum gate dependency of the target quantum circuit to generate a first sub-circuit and a second sub-circuit; executing the first sub-circuit based on a quantum simulator to determine a first density matrix; executing the second sub-circuit based on a quantum processor to determine a second density matrix; determining a target execution result corresponding to the original quantum circuit according to the first density matrix and the second density matrix.

[0005] In this way, the original quantum circuit is decomposed to generate a target quantum circuit, and the target quantum circuit only includes single-bit gates and double-bit gates. Then, the target quantum circuit is split based on quantum gate dependency of the target quantum circuit to generate a first sub-circuit and a second sub-circuit. Subsequently, the first sub-circuit is executed based on a quantum simulator to determine a first density matrix. Then, the second sub-circuit is executed based on a quantum processor to determine a second density matrix. Finally, a target execution result corresponding to the original quantum circuit is determined according to the first density matrix and the second density matrix. In this way, the circuit structure of the original quantum circuit is simplified through decomposition and splitting, and the high resource consumption of the traditional full-circuit single processing mode is avoided. Moreover, the first sub-circuit suitable for classical simulation is diverted to a quantum simulator, which reduces the occupation of scarce quantum processor resources and improves the overall operation efficiency.

[0006] In some embodiments, the decomposing the original quantum circuit to generate the target quantum circuit comprises: decompose the complex quantum gate in the original quantum circuit into a preset standard gate set, the preset standard gate set only includes the single-bit gate and the double-bit gate; clean up redundant gates of the decomposed quantum circuit to obtain the target quantum circuit.

[0007] In this way, the complex quantum gate in the original quantum circuit is decomposed into a preset standard gate set, and the preset standard gate set only includes the single-bit gate and the double-bit gate. Then, the decomposed quantum circuit is cleaned up to obtain the target quantum circuit. In this way, the complex quantum gate in the original circuit is converted into the universal single-bit gate and the double-bit gate, which provides a clear circuit basis for subsequent quantum gate type marking and dependency analysis, and ensures the accuracy of the splitting process. Moreover, by deleting the redundant gates, the circuit structure can be simplified, thereby reducing the execution burden of the quantum simulator and the quantum processor, and improving the overall operation efficiency.

[0008] In some embodiments, the quantum gates in the first sub-circuit are all Clifford gates, and the second sub-circuit includes at least one non-Clifford gate, and the Clifford gate includes a Hadamard gate, a phase gate, a Pauli gate and / or a controlled NOT gate.

[0009] In this way, the quantum gates in the first sub-circuit are all Clifford gates, and the second sub-circuit includes at least one non-Clifford gate, and the Clifford gate includes a Hadamard gate, a phase gate, a Pauli gate and / or a controlled NOT gate. In this way, the first sub-circuit of the full Clifford gate is diverted to the quantum simulator, avoiding the occupation of quantum processor resources by the classically simulatable circuit, and improving the utilization efficiency of quantum hardware.

[0010] In some embodiments, based on the quantum gate dependency relationship of the target quantum circuit, the target quantum circuit is split to generate a first sub-circuit and a second sub-circuit, including: traverse the quantum gates in the target quantum circuit to mark the Clifford gate and the non-Clifford gate; determine a cutting position based on the quantum gate dependency relationship, wherein the cutting position is located at a logical boundary between the Clifford gate and the non-Clifford gate, and the first sub-circuit and the second sub-circuit are not coupled to each other at the cutting position; split the target quantum circuit based on the cutting position to obtain the first sub-circuit and the second sub-circuit.

[0011] Thus, quantum gates in the target quantum circuit are traversed to mark Clifford gates and non-Clifford gates. Then, based on quantum gate dependencies, a cutting position is determined, where the cutting position is located at a logical boundary between a Clifford gate and a non-Clifford gate, and the first sub-circuit and the second sub-circuit are not coupled to each other at the cutting position. Then, based on the cutting position, the target quantum circuit is split to obtain the first sub-circuit and the second sub-circuit. In this way, through dependency analysis and uncoupling constraints, execution failures of the split sub-circuits caused by broken dependencies or coupling associations are avoided, and each sub-circuit is ensured to be independently executable, so that valid results can be output independently.

[0012] In some embodiments, the total number of the first sub-circuits and the second sub-circuits is 4^N, where N is the number of cutting positions, and N is a natural number less than 3.

[0013] Thus, the total number of the first sub-circuits and the second sub-circuits is 4^N, where N is the number of cutting positions, and N is a natural number less than 3. In this way, the number of sub-circuits is limited to 1, 4, and 16 groups, avoiding exponential consumption of classical resources, reducing storage, calculation, and transmission costs, and adapting to the resource carrying capacity of existing quantum computing platforms.

[0014] In some embodiments, the method further comprises: If the cutting position is located at the beginning of the target quantum circuit, an initial quantum state is configured for the sub-circuit at the beginning; If the cutting position is located at the end of the target quantum circuit, a Pauli measurement basis is configured for the sub-circuit at the end.

[0015] Thus, if the cutting position is located at the beginning of the target quantum circuit, an initial quantum state is configured for the sub-circuit at the beginning. Then, if the cutting position is located at the end of the target quantum circuit, a Pauli measurement basis is configured for the sub-circuit at the end. In this way, start conditions are provided for the sub-circuit cut at the beginning, and output methods are provided for the sub-circuit cut at the end, solving the execution obstacles of special position splitting.

[0016] In some embodiments, the first sub-circuit is executed based on a quantum simulator to determine a first density matrix, comprising: The first sub-circuit is submitted to the quantum simulator; The first sub-circuit is simulated by the quantum simulator to output the first density matrix.

[0017] Thus, the first sub-circuit is submitted to the quantum simulator. Then, the first sub-circuit is simulated by the quantum simulator to output a first density matrix. In this way, the quantum simulator simulates the polynomial time complexity of the Clifford circuit, which is lower than that of the quantum processor, so as to improve the overall execution speed of the quantum circuit. Moreover, the classical computing characteristics of the quantum simulator are not affected by hardware noise and decoherence, and the output first density matrix is free of error accumulation, thereby guaranteeing the accuracy of the overall result.

[0018] In some embodiments, the quantum processor is configured to execute the second sub-circuit to determine a second density matrix, including: submitting the second sub-circuit to the quantum processor; executing the second sub-circuit by the quantum processor to output the second density matrix.

[0019] Thus, the second sub-circuit is submitted to the quantum processor. Then, the second sub-circuit is executed by the quantum processor to output a second density matrix. In this way, the quantum processor is used to execute the second sub-circuit, so as to fully release the potential performance of quantum computing and ensure that the operation logic of the original quantum circuit is correctly implemented.

[0020] In some embodiments, the quantum processor is configured to determine a target execution result corresponding to the original quantum circuit according to the first density matrix and the second density matrix, including: performing tensor product merging on the first density matrix and the second density matrix to obtain a target density matrix; taking the target density matrix as the target execution result.

[0021] Thus, the first density matrix and the second density matrix are subjected to tensor product merging to obtain a target density matrix. Then, the target density matrix is taken as a target execution result. In this way, the tensor product operation can accurately restore the complete quantum state of the original quantum circuit, avoids the loss of quantum state correlation caused by split execution, and ensures that the target execution result is consistent with the original circuit.

[0022] The embodiments of the present application also provide a quantum computing platform, which includes a decomposition module, a split module, a circuit execution module and a merging module. The decomposition module is configured to perform decomposition processing on an original quantum circuit to generate a target quantum circuit, wherein the target quantum circuit only includes single-bit gates and double-bit gates. The split module is configured to perform split processing on the target quantum circuit based on the quantum gate dependency relationship of the target quantum circuit to generate a first sub-circuit and a second sub-circuit. The circuit operation module is configured to execute the first sub-circuit based on a quantum simulator to determine a first density matrix; The circuit operation module is configured to execute the second sub-circuit based on a quantum processor to determine a second density matrix; The merging module is configured to determine a target execution result corresponding to the original quantum circuit according to the first density matrix and the second density matrix.

[0023] In this way, the decomposition module is configured to perform a decomposition process on the original quantum circuit to generate a target quantum circuit, which only includes single-bit gates and double-bit gates. Then, the splitting module is configured to perform a splitting process on the target quantum circuit based on the quantum gate dependency relationship of the target quantum circuit to generate the first sub-circuit and the second sub-circuit. Then, the circuit operation module is configured to execute the first sub-circuit based on a quantum simulator to determine a first density matrix. Subsequently, the circuit operation module is configured to execute the second sub-circuit based on a quantum processor to determine a second density matrix. Finally, the merging module is configured to determine a target execution result corresponding to the original quantum circuit according to the first density matrix and the second density matrix. In this way, through the decomposition process and the splitting process, the circuit structure of the original quantum circuit is simplified, and the high resource consumption of the traditional full-circuit single processing mode is avoided. Moreover, the first sub-circuit suitable for classical simulation is diverted to the quantum simulator, reducing the occupation of scarce quantum processor resources and improving the overall operation efficiency.

[0024] Additional aspects and advantages of the embodiments of the present application will be in part apparent and in part pointed out hereinafter. BRIEF DESCRIPTION OF DRAWINGS

[0025] The above and / or additional aspects and advantages of the present application will become apparent and be readily appreciated from the following description, including the appended drawings, wherein: Figure 1 is one of the flow diagrams of the quantum circuit calculation method of the embodiments of the present application; Figure 2 is the second flow diagram of the quantum circuit calculation method of the embodiments of the present application; Figure 3 is the third flow diagram of the quantum circuit calculation method of the embodiments of the present application; Figure 4 is a quantum circuit decomposition diagram of the embodiments of the present application; Figure 5 is the fourth flow diagram of the quantum circuit calculation method of the embodiments of the present application; Figure 6 is the fifth flow diagram of the quantum circuit calculation method of the embodiments of the present application; Figure 7 Fig. 6 is a flowchart of a sixth example of a quantum circuit calculation method according to embodiments of the present application; Figure 8 Fig. 7 is a flowchart of a seventh example of a quantum circuit calculation method according to embodiments of the present application. DETAILED DESCRIPTION

[0026] Embodiments of the present application are described in detail below with reference to the accompanying drawings. Examples of the embodiments are shown in the drawings, in which the same or similar reference numerals are used throughout the drawings to designate the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the drawings are examples only, and are merely intended to explain the embodiments of the present application, and cannot be understood as limiting the embodiments of the present application.

[0027] In the related art, the execution of a quantum circuit usually relies on two single processing modes: either numerical simulation of the entire circuit by a classical computer or direct submission of the entire circuit to a quantum processor for execution. However, such single processing mode is not good for the utilization of computing resources.

[0028] Specifically, almost all practical quantum circuits include two types of quantum gates with completely different properties: Clifford gates and non-Clifford gates. If the quantum circuit is executed by classical simulation of the entire circuit, the quantum circuit including only Clifford gates can be efficiently simulated by a classical computer, and the computational complexity increases polynomially with the number of qubits, and the resource consumption is controllable. However, once the non-Clifford gate is included in the circuit, the complexity of the classical simulation will jump to exponential level. For example, for a circuit including n qubits and non-Clifford gates, the classical simulation needs to store a quantum state vector of 2 n dimension, and when n exceeds 30, even a supercomputer is difficult to complete the calculation within a reasonable time, resulting in inefficient occupation of classical computing resources, and even interruption of execution due to insufficient resources. If the quantum circuit is executed by a quantum computer, the non-Clifford gate operation that cannot be efficiently simulated by a classical computer can be efficiently processed, but the execution of the Clifford gate will waste quantum computing resources, affecting the efficiency of the quantum circuit execution.

[0029] Based on the above problems, please refer to Figure 1 The embodiments of the present application provide a quantum circuit calculation method, the method comprising: 01: decomposing an original quantum circuit to generate a target quantum circuit; 02: based on the quantum gate dependency relationship of the target quantum circuit, splitting the target quantum circuit to generate a first sub-circuit and a second sub-circuit; 03: executing the first sub-circuit based on a quantum simulator to determine a first density matrix; 04: Based on the quantum processor, a second sub-circuit is executed to determine a second density matrix; 05: According to the first density matrix and the second density matrix, a target execution result corresponding to the original quantum circuit is determined.

[0030] Specifically, the original quantum circuit refers to the initial circuit architecture for implementing a specific quantum computing task without optimization processing, composed of a series of quantum gates in a specific order, which may include complex gate structures such as multi-bit gates.

[0031] The decomposition process refers to the operation of transforming and simplifying the quantum gates in the original quantum circuit, which can eliminate the interference of complex gate structures and make the circuit meet the standards for subsequent splitting and execution.

[0032] The target quantum circuit refers to the standardized circuit obtained after decomposition processing, which only includes single-bit gates and double-bit gates as basic operation units, without redundant or invalid gate structures, and is the basis for subsequent splitting and execution.

[0033] Single-bit gates refer to quantum gates that only act on a single quantum bit, such as Hadamard gates and phase gates, which are used to implement state transformations of a single quantum bit and are the basic building blocks of quantum circuits.

[0034] Double-bit gates refer to quantum gates that act on two quantum bits, such as controlled NOT gates and CZ gates, which are used to implement the correlation and interaction between two quantum bits and are the core units for building complex quantum logic.

[0035] Quantum gate dependency refers to the logical association between different quantum gates in a quantum circuit, i.e., whether the execution of a subsequent quantum gate depends on the output state of a previous quantum gate, which directly determines the feasibility of circuit splitting.

[0036] Splitting processing refers to the operation of splitting the target quantum circuit at a specific location into independent sub-circuits with no coupling relationship based on quantum gate dependency, ensuring that each sub-circuit can be executed independently.

[0037] The first sub-circuit and the second sub-circuit refer to the two independent circuit units obtained after splitting, which are logically complementary and can completely restore the computing logic of the original circuit when combined.

[0038] Quantum simulation machine refers to a quantum circuit simulation system based on classical computer hardware, which can simulate the running process of quantum circuits in a classical environment, especially suitable for circuits composed of specific types of quantum gates, with high efficiency and no noise.

[0039] A quantum processor is a hardware device with quantum computing capabilities. It performs quantum gate operations by manipulating qubits and can handle quantum computing tasks that are difficult to simulate using classical simulations. It is the core hardware carrier of quantum computing.

[0040] The density matrix is ​​a mathematical tool for describing the state of a quantum system. It can comprehensively characterize the probability distribution and coherence of quantum states. It is a standardized expression of the execution results of quantum circuits, which facilitates the fusion of results from different sub-circuits.

[0041] The target execution result refers to the final result obtained by merging the execution results of the first and second sub-circuits, which is consistent with the expected computational effect of the original quantum circuit, ensuring that the split execution does not lose computational accuracy and integrity.

[0042] First, the original quantum circuit, as input data, is decomposed and transformed into a target quantum circuit containing only single-qubit and double-qubit gates.

[0043] Subsequently, by analyzing the dependencies of quantum gates in the target circuit, it is split into two independent sub-circuit data streams, namely the first sub-circuit and the second sub-circuit, to ensure the integrity and independence of each data stream.

[0044] Next, the first sub-circuit data stream is input into the quantum simulator, and after simulation calculation, it outputs the first density matrix data stream. The second sub-circuit data stream is input into the quantum processor, and after running on the actual machine, it outputs the second density matrix data stream.

[0045] Finally, the two density matrix data streams are integrated through preset mathematical fusion rules (such as tensor product operation) to generate the target execution result data stream corresponding to the original quantum circuit, thus completing the data flow of the entire calculation process.

[0046] In summary, the quantum circuit computation method provided in this application decomposes the original quantum circuit to generate a target quantum circuit, which includes only single-qubit and two-qubit gates. Then, based on the quantum gate dependencies of the target quantum circuit, it is further split to generate a first sub-circuit and a second sub-circuit. Subsequently, the first sub-circuit is executed using a quantum simulator to determine a first density matrix. Then, the second sub-circuit is executed using a quantum processor to determine a second density matrix. Finally, based on the first and second density matrices, the target execution result corresponding to the original quantum circuit is determined. This decomposition and splitting process simplifies the circuit structure of the original quantum circuit, avoiding the high resource consumption of traditional single-processing modes. Furthermore, by offloading the first sub-circuit, suitable for classical simulation, to the quantum simulator, the occupation of scarce quantum processor resources is reduced, improving overall operating efficiency.

[0047] Please seeFigure 2 In some embodiments, step 01 (decomposing the original quantum circuit to generate the target quantum circuit) includes: 011: Decompose the complex quantum gates in the original quantum circuit into a predefined standard gate set; 012: Redundancy gate cleanup is performed on the decomposed quantum circuit to obtain the target quantum circuit.

[0048] Specifically, complex quantum gates refer to quantum gates that go beyond the scope of single-qubit and two-qubit gates, including multi-qubit gates and non-universal gates specific to certain quantum processors. Their operational logic and structure are relatively complex and difficult to be directly used for unified analysis, such as Tofoli gates and Fredkin gates.

[0049] The predefined standard gate set refers to a predefined set of universal gates that contains only single-bit and two-bit gates. It serves as the benchmark for standardizing quantum circuits and ensures that different original circuits have a unified gate type format after decomposition.

[0050] Redundant gate cleanup refers to the logical analysis of the decomposed quantum circuit to identify and delete quantum gate operations that have no practical computational significance and do not affect the final result, including repetition gates, cancellation gates, and no-operation gates.

[0051] For the complex quantum gates in the data stream, an equivalence transformation algorithm is used to replace each complex quantum gate with a combination of single-qubit and two-qubit gates from a pre-defined standard gate set. Logical analysis is then performed on the decomposed quantum circuits. Redundant gate data is identified and removed from the data stream based on the equivalence of the quantum gate effects, ultimately outputting the target quantum circuit.

[0052] In this way, the complex quantum gates in the original quantum circuit are decomposed into a predefined standard gate set, which includes only single-qubit and two-qubit gates. Then, redundant gates are removed from the decomposed quantum circuit to obtain the target quantum circuit. This transforms the messy complex quantum gates in the original circuit into general single-qubit and two-qubit gates, providing a clear circuit foundation for subsequent quantum gate type labeling and dependency analysis, ensuring the accuracy of the decomposition process. Furthermore, by removing redundant gates, the circuit structure can be simplified, thereby reducing the execution burden on the quantum simulator and quantum processor, and improving overall operating efficiency.

[0053] In some implementations, the quantum gates in the first sub-circuit are all Clifford gates, and the second sub-circuit includes at least one non-Clifford gate. Clifford gates include Hadamard gates, phase gates, Pauli gates, and / or controlled NOT gates.

[0054] Specifically, Clifford gates refer to a class of quantum gates in quantum computing that have specific operational rules. They can map Pauli operators to linear combinations of other Pauli operators through conjugate mappings, forming a closed Clifford group. It should be noted that circuits composed solely of Clifford gates can perform polynomial-time efficient and noise-free simulations on classical computers.

[0055] The Hadamard gate (H gate) is a typical single-bit Clifford gate that can convert the ground state of a qubit into a superposition state.

[0056] The phase gate (S-gate) is a single-bit Clifford gate that can apply a phase shift to a qubit, flipping the phase of a quantum state without changing the probability distribution of the quantum state, only affecting quantum coherence. It is often used for quantum phase manipulation.

[0057] Pauli gates include Pauli-X gates, Pauli-Y gates, and Pauli-Z gates, which can respectively realize quantum state flipping, complex plane rotation flipping, and phase flipping, and are the most basic logic operation gates in quantum computing.

[0058] The controlled-NOT gate (CNOT gate) is a two-bit Clifford gate, consisting of a control bit and a target bit. Its core logic is: when the control bit is in |1... When the control bit is in the |0 state, the target bit flips. When the target bit state remains unchanged, it is the core gate for constructing quantum entanglement and quantum logical connections.

[0059] Non-Clifford gates are quantum gates that do not satisfy the Clifford gate operation rules. They cannot be closed-looped through the conjugate mapping of Pauli operators, making them difficult to simulate efficiently on classical computers. They must rely on the quantum parallelism of quantum processors to complete the operation. Examples include T gates and S† gates.

[0060] Thus, all quantum gates in the first sub-circuit are Clifford gates, while the second sub-circuit includes at least one non-Clifford gate. Clifford gates include Hadamard gates, phase gates, Pauli gates, and / or controlled NOT gates. This offloads the all-Clifford-gate first sub-circuit to the quantum simulator, preventing quantum processor resources from being occupied by classically simulable circuits and improving the utilization efficiency of quantum hardware.

[0061] Please see Figure 3 In some implementations, step 02 (based on the quantum gate dependencies of the target quantum circuit, splitting the target quantum circuit to generate a first sub-circuit and a second sub-circuit) includes: 021: Traverse the quantum gates in the target quantum circuit, marking Clifford gates and non-Clifford gates; 022: Determine the cutting position based on quantum gate dependence; 023: Based on the cutting position, the target quantum circuit is split to obtain the first sub-circuit and the second sub-circuit.

[0062] Specifically, traversal refers to the process of scanning each gate operation one by one according to the arrangement order of quantum gates in the quantum circuit, ensuring that no type identification of any quantum gate is missed.

[0063] The labeling refers to assigning a specific type identifier to each quantum gate after traversal, based on the definition of Clifford gates and non-Clifford gates.

[0064] The cutting position refers to the specific node in the target quantum circuit used to divide it into two types of sub-circuits. It is the core control point for the splitting process and must simultaneously meet three conditions: gate type boundary, independent destruction, and no coupling association.

[0065] Quantum bit coupling correlation refers to the interaction between two sub-circuits at the cut-off point, such as the state correlation of shared qubits or the entanglement relationship that has not been cut off. This causes the execution result of one sub-circuit to affect the state of another sub-circuit, thus destroying the independence of the sub-circuits.

[0066] Please see Figure 4 , Figure 4 This is a schematic diagram of quantum circuit segmentation. Each quantum gate in the target quantum circuit is traversed sequentially, and its type is identified based on the characteristic parameters of Clifford gates and non-Clifford gates. A type label is then added to each quantum gate.

[0067] Extract the action bit correlation data of quantum gates in the data stream, determine whether there is state dependency between the preceding and following quantum gates, locate the logical boundary between Clifford gates and non-Clifford gates, and filter out the cutting positions that are neither dependently broken nor coupled.

[0068] Based on the cutting position coordinates, the target quantum circuit is segmented into independent first and second sub-circuits. The decoupling and complete dependency relationships between the qubits of the two data streams are verified, and the final outputs are the first and second sub-circuits that can be directly used for distributed execution.

[0069] Thus, the quantum gates in the target quantum circuit are traversed, and Clifford gates and non-Clifford gates are marked. Next, based on quantum gate dependencies, the cutting position is determined, where the cutting position is located at the logical boundary between Clifford and non-Clifford gates, and the first and second sub-circuits are not coupled to any qubit at the cutting position. Then, the target quantum circuit is split based on the cutting position to obtain the first and second sub-circuits. In this way, through dependency analysis and decoupling constraints, execution failures caused by dependency breaks or coupling associations in the split sub-circuits are avoided, ensuring that each sub-circuit is independently executable and can output valid results independently.

[0070] In some implementations, the total number of the first sub-line and the second sub-line is 4^N, where N is the number of cutting positions and N is a natural number less than 3.

[0071] Specifically, in quantum circuit decomposition, the number of cutting positions directly determines the number of sub-circuits. Without quantization constraints, this can lead to uncontrolled resource consumption and a surge in computational complexity. Specifically, too many cutting positions cause the number of sub-circuits to increase exponentially by 4^N. The execution, result storage, and merging of each sub-circuit require classical or quantum resources, ultimately causing the overall computational resource requirements to exceed the carrying capacity limit. An excessive number of sub-circuits causes the computational workload of density matrix merging to increase geometrically, not only prolonging computation time but also potentially amplifying errors due to the accumulation of intermediate data, leading to a surge in computational complexity.

[0072] Thus, the total number of the first and second sub-lines is 4^N, where N is the number of cutting positions and is a natural number less than 3. This limits the number of sub-lines to 1, 4, and 16 groups, avoiding the exponential consumption of classic resources, reducing storage, computing, and transmission costs, and adapting to the resource capacity of existing quantum computing platforms.

[0073] Please see Figure 5 In some implementations, the method further includes: 024: If the cutting position is located at the beginning of the target quantum circuit, configure the initial quantum state for the sub-circuit at the beginning; 025: If the cutting position is located at the end of the target quantum circuit, configure the Pauli measurement basis for the sub-circuit at the end.

[0074] Specifically, the beginning of the target quantum circuit refers to the starting point of the quantum gate arrangement in the target quantum circuit, which is the starting position of the original quantum circuit operation. The quantum gate at the beginning usually acts directly on the initial quantum state.

[0075] The sub-circuit at the beginning refers to the independent sub-circuit unit located at the starting side of the circuit after being split when the cutting position is at the beginning. Its original input end is exposed due to the cutting and lacks the support of the initial quantum state.

[0076] The initial quantum state refers to the starting state of quantum circuit operations. It is the basic state of a qubit and a prerequisite for a quantum gate to perform operations. Its configuration must match the initial logic of the original quantum circuit to ensure that the operation direction of the sub-circuit is consistent with that of the original circuit.

[0077] The end of the target quantum circuit refers to the termination of the quantum gate arrangement in the target quantum circuit. It is the end position of the original quantum circuit operation. The quantum state at the end needs to be converted into readable classical data through measurement.

[0078] The terminus of a sub-line refers to an independent sub-line unit located on the termination side of the line after being split when the cutting position is at the end. Its original output end is exposed due to the cutting and lacks standardized measurement procedures.

[0079] Pauli measurement bases refer to the measurement reference system built on Pauli operators. They are standardized quantum state measurement tools in quantum computing, including X measurement base, Y measurement base, and Z measurement base. They can convert quantum states into corresponding classical measurement results, ensuring that the quantum states of the end sub-circuits can be output in a unified format.

[0080] Thus, if the cutting position is located at the beginning of the target quantum circuit, an initial quantum state is configured for the sub-circuit at the beginning. Next, if the cutting position is located at the end of the target quantum circuit, a Pauli measurement basis is configured for the sub-circuit at the end. This provides the start-up conditions for the sub-circuit cut at the beginning and the output mode for the sub-circuit cut at the end, overcoming the execution obstacle of splitting at special locations.

[0081] Please see Figure 6 In some implementations, step 03 (based on a quantum simulator, executing the first sub-circuit to determine the first density matrix) includes: 031: Submit the first sub-circuit to the quantum simulator; 032: Simulate the first sub-circuit using a quantum simulator and output the first density matrix.

[0082] Specifically, simulation computing refers to a quantum simulator simulating the interaction of quantum gates with quantum states using efficient methods such as the stabilizer simulation algorithm, based on the received first sub-circuit data and following the principles of quantum mechanics and the operational rules of Clifford gates. This process does not require manipulating real qubits and completes the computation using only classical computing resources.

[0083] In some implementations, the data stream of the first sub-line is formatted according to the input protocol of the quantum simulator and transmitted to the quantum simulator through a secure transmission channel, while data verification is performed to ensure no packet loss or errors.

[0084] Subsequently, after receiving the data, the quantum simulator analyzes the gate sequence and related parameters, calls the stabilizer simulation algorithm, and simulates the sequential action of Clifford gates on the quantum state. The calculation process does not require the participation of real qubits, and the quantum state evolution simulation is completed only through classical computing resources.

[0085] Next, the quantum simulator converts the simulated quantum states into density matrix form, obtaining the first density matrix.

[0086] Thus, the first sub-circuit is submitted to the quantum simulator. The quantum simulator then performs simulation calculations on the first sub-circuit, outputting the first density matrix. In this way, the quantum simulator simulates the polynomial time complexity of the Clifford circuit, which is lower than that of a quantum processor, thereby improving the overall execution speed of the quantum circuit. Furthermore, the classical computational characteristics of the quantum simulator make it unaffected by hardware noise and decoherence, ensuring that the output first density matrix has no error accumulation and guaranteeing the accuracy of the overall result.

[0087] Please see Figure 7 In some implementations, step 04 (based on a quantum processor, executing a second sub-circuit to determine a second density matrix) includes: 041: Submit the second sub-circuit to the quantum processor; 042: The second sub-circuit is run by a quantum processor to output the second density matrix.

[0088] Specifically, operation refers to the quantum processor manipulating the qubits through a measurement and control system based on the received data from the second sub-circuit, sequentially executing Clifford and non-Clifford gate operations in the circuit to simulate the quantum state evolution process of the original circuit. This process needs to overcome interference such as hardware noise and decoherence, and ensure computational accuracy through methods such as optimizing pulse sequences and real-time measurement and control.

[0089] In some implementations, the data stream of the second sub-line is converted into a format according to the hardware protocol of the quantum processor and transmitted to the quantum processor through a dedicated interface. At the same time, data verification and integrity confirmation are performed to avoid interruption of operation due to data problems.

[0090] Subsequently, after receiving the data, the quantum processor analyzes parameters such as the gate sequence and the mapping relationship between qubits, and generates a suitable pulse sequence through the measurement and control system. It then sequentially manipulates the target qubits to perform Clifford gate and non-Clifford gate operations, while monitoring the state of the qubits in real time and compensating for deviations caused by noise and decoherence.

[0091] Next, after the computation is completed, the quantum processor measures the quantum state using a preset Pauli measurement basis and transforms the obtained classical measurement data into a second density matrix using a density matrix reconstruction algorithm.

[0092] Thus, the second sub-circuit is submitted to the quantum processor. The quantum processor then runs the second sub-circuit, outputting the second density matrix. In this way, running the second sub-circuit on a real quantum processor fully unleashes the potential performance of quantum computing and ensures the correct implementation of the original quantum circuit's computational logic.

[0093] Please see Figure 8 In some implementations, step 05 (determining the target execution result corresponding to the original quantum circuit based on the first density matrix and the second density matrix) includes: 051: Combine the first and second density matrices using tensor products to obtain the target density matrix; 052: Use the target density matrix as the target execution result.

[0094] Specifically, tensor product merging: a method for joint quantum state operations based on quantum mechanics principles, is a standard mathematical tool for describing the joint state of two independent quantum systems. For the density matrices ρ1 and ρ2 of two independent sub-circuits, the tensor product operation (ρ1...)... ρ2) can fully preserve the quantum state information of each individual circuit through the form of matrix Kronecker product, while reconstructing the overall quantum state correlation of the original circuit, thus ensuring the physical rationality of the merging result.

[0095] The target density matrix refers to the final quantum state data obtained by merging tensor products. It is a standardized representation of the complete quantum state of the original quantum circuit. It includes both the noiseless computational information of the Clifford gate part and the quantum parallel computation results of the non-Clifford gate part, satisfying the mathematical constraints of the quantum state.

[0096] In some implementations, the first density matrix data stream output by the quantum simulator and the second density matrix data stream output by the quantum processor are received, and the two data streams are format-checked and dimension-matched to ensure that they meet the requirements of tensor product operation.

[0097] Next, based on the mathematical principles of quantum mechanics, the Kronecker product operation is performed on the two density matrices to generate a preliminary merged matrix. Subsequently, the preliminary merged matrix undergoes quantum state normalization verification. Through algorithms such as trace normalization and semi-positive definite correction, it is ensured that it satisfies the mathematical constraints of the quantum states, yielding the standard target density matrix. Finally, depending on the specific quantum computing task requirements, the target density matrix can be further transformed into classical measurement results (e.g., reading the probability distribution through Pauli measurement bases) or other practical formats, and the final output serves as the target execution result, completing the entire data processing flow of the quantum circuit computation.

[0098] Thus, the first and second density matrices are combined using a tensor product to obtain the target density matrix. This target density matrix is ​​then used as the target execution result. In this way, the tensor product operation can accurately reconstruct the complete quantum state of the original quantum circuit, avoiding the loss of quantum state correlations caused by split execution, and ensuring that the target execution result is consistent with the expected result of the original circuit.

[0099] This application also provides a quantum computing platform, which includes a decomposition module, a splitting module, a circuit operation module, and a merging module. The decomposition module is configured to decompose the original quantum circuit to generate a target quantum circuit, which includes only single-qubit gates and two-qubit gates. The splitting module is configured to split the target quantum circuit based on the quantum gate dependency of the target quantum circuit, generating the first sub-circuit and the second sub-circuit. The circuit operation module is configured to execute the first sub-circuit based on a quantum simulator and determine the first density matrix; The circuit operation module is configured to be based on a quantum processor, execute the second sub-circuit, and determine the second density matrix; The merging module is configured to determine the target execution result corresponding to the original quantum circuit based on the first density matrix and the second density matrix.

[0100] Specifically, a quantum computing platform refers to an integrated quantum computing system that integrates functions such as circuit processing, splitting, execution, and result merging. It includes core capabilities such as hardware interface adaptation, software module collaboration, and data flow management and control, and serves as the physical carrier and software support for realizing distributed quantum circuit computing methods.

[0101] The decomposition module refers to the preprocessing core of the quantum computing platform, which is responsible for the standardized transformation of the original quantum circuits. It can eliminate complex gate structures and redundant operations, and provide circuit data in a unified format for subsequent decomposition.

[0102] The split module refers to the split unit of the quantum computing platform. Based on the quantum gate dependency relationship and gate type characteristics, the target quantum circuit is split into two independent sub-circuits, providing a basis for distributed execution.

[0103] The circuit operation module refers to the execution core of the quantum computing platform, which can call the quantum simulator and the quantum processor to execute two types of sub-circuits and output a standardized density matrix.

[0104] The merging module refers to the result closed-loop unit of the quantum computing platform. It integrates the first density matrix and the second density matrix through scientific computation to generate the target execution result corresponding to the original quantum circuit, thus completing the entire process closed loop.

[0105] Thus, the decomposition module is configured to decompose the original quantum circuit to generate the target quantum circuit, which consists only of single-qubit and two-qubit gates. Next, the splitting module is configured to split the target quantum circuit based on its quantum gate dependencies, generating a first sub-circuit and a second sub-circuit. Then, the circuit execution module is configured to execute the first sub-circuit using a quantum simulator to determine the first density matrix. Subsequently, the circuit execution module is configured to execute the second sub-circuit using a quantum processor to determine the second density matrix. Finally, the merging module is configured to determine the target execution result corresponding to the original quantum circuit based on the first and second density matrices. This simplifies the original quantum circuit structure through decomposition and splitting, avoiding the high resource consumption of traditional single-processing modes. Furthermore, offloading the first sub-circuit, suitable for classical simulation, to the quantum simulator reduces the consumption of scarce quantum processor resources and improves overall operating efficiency.

[0106] This application also provides a computer-readable storage medium containing a computer program. When the computer program is executed by one or more processors, it causes the one or more processors to perform the method of this application.

[0107] It is understood that a computer program includes computer program code. Computer program code can be in the form of source code, object code, executable files, or some intermediate form. Computer-readable storage media can include: any entity or device capable of carrying computer program code, recording media, USB flash drives, portable hard drives, magnetic disks, optical disks, computer memory, read-only memory (ROM), random access memory (RAM), and software distribution media, etc.

[0108] In this specification, the terms "specifically," "furthermore," "particularly," "understandably," etc., refer to specific features, structures, materials, or characteristics described in connection with embodiments or examples that are included in at least one embodiment or example of this application. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples. Moreover, without contradiction, those skilled in the art can combine and integrate the different embodiments or examples described in this specification, as well as the features of different embodiments or examples.

[0109] Any process or method described in the flowchart or otherwise herein can be understood as representing a module, segment, or portion of code comprising one or more executable instructions for implementing a particular logical function or process, and the scope of the preferred embodiments of this application includes additional implementations in which functions may be performed not in the order shown or discussed, including substantially simultaneously or in reverse order depending on the function involved, as will be understood by those skilled in the art to which embodiments of this application pertain.

[0110] Although embodiments of this application have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting this application. Those skilled in the art can make changes, modifications, substitutions and variations to the above embodiments within the scope of this application.

Claims

1. A quantum circuit computing method, characterized in that, The method includes: The original quantum circuit is decomposed to generate a target quantum circuit, which includes only single-qubit gates and two-qubit gates. Based on the quantum gate dependency of the target quantum circuit, the target quantum circuit is split to generate a first sub-circuit and a second sub-circuit. Based on a quantum simulator, the first sub-circuit is executed to determine the first density matrix; Based on the quantum processor, the second sub-circuit is executed to determine the second density matrix; Based on the first density matrix and the second density matrix, the target execution result corresponding to the original quantum circuit is determined.

2. The method according to claim 1, characterized in that, The process of decomposing the original quantum circuit to generate the target quantum circuit includes: The complex quantum gates in the original quantum circuit are decomposed into a preset standard gate set, which includes only the single-bit gate and the two-bit gate. Redundancy gate cleanup is performed on the decomposed quantum circuit to obtain the target quantum circuit.

3. The method according to claim 1, characterized in that, The quantum gates in the first sub-circuit are all Clifford gates, and the second sub-circuit includes at least one non-Clifford gate, wherein the Clifford gate includes a Hadamard gate, a phase gate, a Pauli gate, and / or a controlled NOT gate.

4. The method according to claim 3, characterized in that, The process of splitting the target quantum circuit based on its quantum gate dependencies to generate a first sub-circuit and a second sub-circuit includes: Traverse the quantum gates in the target quantum circuit and mark Clifford gates and non-Clifford gates; Based on the quantum gate dependency, the cutting position is determined, wherein the cutting position is located at the logical boundary between the Clifford gate and the non-Clifford gate, and the first sub-circuit and the second sub-circuit have no quantum bit coupling at the cutting position; The target quantum circuit is split based on the cutting position to obtain the first sub-circuit and the second sub-circuit.

5. The method according to claim 4, characterized in that, The total number of the first sub-line and the second sub-line is 4^N, where N is the number of cutting positions and N is a natural number less than 3.

6. The method according to claim 4, characterized in that, The method further includes: If the cutting position is located at the beginning of the target quantum circuit, configure the initial quantum state for the sub-circuit at the beginning; If the cutting position is located at the end of the target quantum circuit, a Pauli measurement basis is configured for the sub-circuit at the end.

7. The method according to claim 1, characterized in that, The process of executing the first sub-circuit based on a quantum simulator to determine the first density matrix includes: Submit the first sub-circuit to the quantum simulator; The first sub-circuit is simulated and calculated using the quantum simulator, and the first density matrix is ​​output.

8. The method according to claim 1, characterized in that, The step of executing the second sub-circuit based on the quantum processor to determine the second density matrix includes: Submit the second sub-circuit to the quantum processor; The second sub-circuit is run by the quantum processor, and the second density matrix is ​​output.

9. The method according to claim 1, characterized in that, The step of determining the target execution result corresponding to the original quantum circuit based on the first density matrix and the second density matrix includes: The first density matrix and the second density matrix are combined by tensor product to obtain the target density matrix; The target density matrix is ​​used as the target execution result.

10. A quantum computing platform, characterized in that, The quantum computing platform includes a decomposition module, a splitting module, a circuit operation module, and a merging module; The decomposition module is configured to decompose the original quantum circuit to generate a target quantum circuit, which includes only single-qubit gates and two-qubit gates. The splitting module is configured to split the target quantum circuit based on the quantum gate dependency of the target quantum circuit, generating a first sub-circuit and a second sub-circuit. The circuit operation module is configured to execute the first sub-circuit based on a quantum simulator and determine the first density matrix; The circuit operation module is configured to execute the second sub-circuit based on a quantum processor and determine the second density matrix; The merging module is configured to determine the target execution result corresponding to the original quantum circuit based on the first density matrix and the second density matrix.

Images