Quantum chip systems, quantum computers and their performance improvement methods, and related devices
By optimizing the quantum circuit design, the problem of complex and inefficient carry operations in quantum computing has been solved, achieving more efficient quantum computing. In particular, in complex financial calculations such as basket option pricing, the number of qubits and circuit depth have been reduced, improving computational efficiency and accuracy.
Patent Information
- Application Number
- CN202311248831.6
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2023-09-25
- Publication Date
- 2026-11-17
- Estimated Expiration
- 2043-09-25
AI Technical Summary
In quantum computing, carry operations are complex and inefficient, leading to an increase in the number of qubits and circuit depth, making it difficult to perform arithmetic operations efficiently, especially addition, which consumes too many resources, particularly in complex financial calculations such as basket option pricing.
By optimizing quantum circuit design, quantum states of variable bit groups are prepared using configuration information and statistical models of combined products. Summation bit groups are prepared by combining weights. Carry operations are performed using the target quantum circuit, reducing the number of quantum gate operands and working bits, and optimizing the number of qubits and circuit depth.
It enables faster and more efficient quantum computing, reduces quantum resource consumption, lowers computation time and noise errors, and improves computational efficiency.
Smart Images

Figure CN117494824B_ABST
Abstract
Description
Technical Field
[0001] This application relates to the field of quantum computing technology, and in particular to quantum chip systems, quantum computers, methods for improving the performance of quantum computers, and classical computers. Background Technology
[0002] A quantum computer is a physical device that performs high-speed mathematical and logical operations, stores and processes quantum information in accordance with the laws of quantum mechanics. When a device processes and calculates quantum information and runs quantum algorithms, it is a quantum computer. Because of its ability to process mathematical problems more efficiently than ordinary computers—for example, reducing the time to crack RSA keys from hundreds of years to hours—quantum computers have become a key technology under research.
[0003] In classical computing, carry-over can occur during binary addition. For example, 1 + 1 = 10, where 0 is the result bit and 1 is the carry bit. When calculating multi-digit numbers, this carry-over can occur consecutively, propagating from the least significant bit to the most significant bit. Implementing carry operations in quantum computing is relatively difficult because quantum operations need to be reversible and cannot destroy information in other qubits. Therefore, classical carry-over methods would be very cumbersome and inefficient in a quantum environment. Thus, a more optimized quantum circuit is needed that significantly reduces both the number of qubits and the circuit depth when performing carry operations.
[0004] Based on this, this application provides a quantum chip system, a quantum computer, a method for improving the performance of a quantum computer, and a classical computer, in order to improve related technologies. Summary of the Invention
[0005] This application provides multiple embodiments of a quantum chip system, a quantum computer, a method for improving the performance of a quantum computer, and a classical computer, which significantly reduce the number of qubits and the line depth when performing carry operations based on more optimized quantum circuits.
[0006] In a first aspect, one embodiment of this application provides a quantum chip system, including at least one quantum processor, which is used to execute quantum operations corresponding to a quantum program to achieve the following steps:
[0007] The quantum state of the variable bit group is prepared based on the joint probability distribution of the parameters of interest of the combined products; the joint probability distribution of the parameters of interest is obtained according to the configuration information of the combined products and the preset statistical model; the combined products include multiple products, and the configuration information includes the identifier and weight of each product.
[0008] Based on the quantum state of the variable bit group and the weight of each product, prepare the quantum state of the summation bit group;
[0009] The target function is achieved through a target quantum circuit; the target quantum circuit includes the summation bit group and the result bit group.
[0010] The target quantum circuit is executed at least once so that the measuring device performs at least one measurement on the resulting bit group, thereby calculating the expected value of the target function.
[0011] Secondly, one embodiment of this application provides a quantum computer, the quantum computer comprising:
[0012] The quantum chip system described above includes at least one quantum processor, which is used to execute quantum operations corresponding to a quantum program to process qubits;
[0013] The measurement and control system includes a control device and a measurement device. The control device is used to convert the quantum program into corresponding control signals and send them to the quantum processor. The measurement device is used to measure the qubits.
[0014] A support system is provided to provide the operating environment conditions to ensure the operation of the quantum chip system;
[0015] An operating system is provided to offer software that enables interaction between the user and the quantum chip system and the measurement and control system.
[0016] Thirdly, one embodiment of this application provides a method for improving the performance of a quantum computer, the method comprising:
[0017] The quantum state of the variable bit group is prepared based on the joint probability distribution of the parameters of interest of the combined products; the joint probability distribution of the parameters of interest is obtained according to the configuration information of the combined products and the preset statistical model; the combined products include multiple products, and the configuration information includes the identifier and weight of each product.
[0018] Based on the quantum state of the variable bit group and the weight of each product, prepare the quantum state of the summation bit group;
[0019] The target function is achieved through a target quantum circuit; the target quantum circuit includes the summation bit group and the result bit group.
[0020] The target quantum circuit is executed at least once so that the measuring device performs at least one measurement on the resulting bit group, thereby calculating the expected value of the target function.
[0021] Fourthly, one embodiment of this application provides a quantum memory that stores a quantum program, which, when executed by at least one quantum processor, implements the functions of any of the above-described quantum chip systems or the steps of any of the above-described methods.
[0022] Fifthly, one embodiment of this application provides a quantum program product, the quantum program product including a quantum program, which, when executed by at least one quantum processor, implements the functions of any of the above-described quantum chip systems or the steps of any of the above-described methods.
[0023] Sixthly, one embodiment of this application provides a classic computer, comprising:
[0024] Memory, used to store computer programs;
[0025] At least one processor is configured to execute the computer program to perform the following steps:
[0026] Based on the joint probability distribution of the parameters of interest of the combined products, a sampling interval is determined, and the sampling interval is uniformly sampled to obtain multiple combinations of values of the parameters of interest and the probability density value corresponding to each combination of values.
[0027] According to several embodiments provided in this application, firstly, the relevant information of the combined products is converted into a quantum state. The joint probability distribution of the parameters of interest of the combined products is calculated based on a classical pre-defined statistical model and configuration information. After obtaining this joint probability distribution information, a quantum state of a variable bit group corresponding to this probability distribution is prepared using quantum operations. That is, the quantum state of the variable bit group can be processed by specific quantum operations, combined with the weights of each product, to finally prepare a quantum state representing a summation bit group of weights. Next, a pre-designed target quantum circuit is used to evaluate the objective function. This target quantum circuit combines the summation bit group with other control bits and operations to prepare a quantum state of a result bit group, which contains the information needed to be obtained from the objective function. Then, the specific target quantum circuit is executed and the results are measured to estimate certain values. The target quantum circuit is executed multiple times, and after each execution, a measuring device measures the result bit group. These measurement results can be used to calculate the expected value of the objective function.
[0028] In quantum computing, arithmetic operations, especially addition, require a carry-over process. When the sum of two bits exceeds its maximum value (e.g., 1+1 in binary), this excess value needs to be carried over to the next higher-order bit. In classical computing, this operation is relatively straightforward, but in quantum computing, due to the superposition and entanglement properties of quantum mechanics, the carry-over process can involve complex quantum gate operations and additional working bits. This quantum chip system is optimized to more quickly and intuitively determine which bits require a carry-over operation. Specifically, it can predict and accurately identify which bits need a carry-over based on the weight of each product. This optimization in judgment greatly simplifies the carry-over operation, thereby reducing the number of related quantum gate operations and the required working bits. This design strategy significantly reduces the number of qubits and the depth of the quantum circuit when performing carry-over operations. In other words, this not only saves the quantum resources required for the algorithm but also shortens the computation time, making the entire quantum computing process more efficient. Attached Figure Description
[0029] This application will be further described below with reference to the accompanying drawings and specific embodiments.
[0030] Figure 1 This is a structural block diagram of a quantum chip system provided in one embodiment of this application.
[0031] Figure 2 This is a schematic diagram of the structure of a quantum weighted adder provided in one embodiment of this application.
[0032] Figure 3 This is a schematic diagram of another quantum weighted adder provided in one embodiment of this application.
[0033] Figure 4 This is a schematic diagram of the structure of a quantum computer provided in one embodiment of this application.
[0034] Figure 5 This is a flowchart illustrating a method for improving the performance of a quantum computer, provided in one embodiment of this application. Detailed Implementation
[0035] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.
[0036] In the description of the embodiments of this application, it should be understood that the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Therefore, features defined with "first" and "second" may explicitly or implicitly include one or more of the stated features. In the description of the embodiments of this application, "multiple" means two or more, unless otherwise explicitly specified.
[0037] Quantum computers have a hybrid architecture, comprising two main parts: a classical computing component responsible for performing classical calculations and control, and a quantum computing component responsible for running quantum programs to achieve quantum computation. A quantum program is a sequence of instructions written in a quantum language such as QRunes that can run on a quantum computer, supporting operations on quantum logic gates and ultimately enabling quantum computation. Specifically, a quantum program is a sequence of instructions that operates on quantum logic gates according to a specific timing order.
[0038] Quantum circuits, also known as quantum logic circuits, are a manifestation of quantum programming and are the most commonly used general-purpose quantum computing model. They represent circuits that operate on qubits under an abstract concept. They consist of qubits, circuits (timelines), and various quantum logic gates. Finally, the results are often read out through quantum measurement operations.
[0039] Unlike traditional circuits that use metal wires to transmit voltage or current signals, in quantum circuits, the circuits can be seen as being connected by time. That is, the state of a quantum bit evolves naturally over time, following the instructions of the Hamiltonian operator until it encounters a logic gate and is operated on.
[0040] A quantum program corresponds to a single quantum circuit, where the total number of qubits in the quantum circuit is the same as the total number of qubits in the quantum program. This can be understood as follows: a quantum program can consist of a quantum circuit, measurement operations on the qubits within that circuit, registers to store the measurement results, and control flow nodes (jump instructions). A single quantum circuit can contain dozens, hundreds, or even thousands of quantum logic gate operations. The execution of a quantum program is the process of executing all the quantum logic gates in a specific timing order. It should be noted that the timing order refers to the chronological sequence in which individual quantum logic gates are executed.
[0041] It's important to note that in classical computing, the most basic unit is the bit, and the most fundamental control mode is the logic gate. Circuit control can be achieved through combinations of logic gates. Similarly, the way to process qubits is through quantum logic gates. Quantum logic gates enable the evolution of quantum states and are the foundation of quantum circuits. Quantum logic gates include: single-qubit quantum logic gates, such as Hadamard gates (H-gates), Pauli-X gates (X-gates), Pauli-Y gates (Y-gates), Pauli-Z gates (Z-gates), RX gates, RY gates, RZ gates, etc.; and multi-qubit quantum logic gates, such as CNOT gates, CR gates, iSWAP gates, Tofoli gates, etc. Quantum logic gates are generally represented using unitary matrices, which are not only matrix forms but also operations and transformations. Generally, the effect of a quantum logic gate on a quantum state is calculated by left-multiplying the unitary matrix by the matrix corresponding to the right vector of the quantum state.
[0042] Option pricing is a common predictive activity in finance. Basket options, involving multiple underlying assets, are among the most common exotic options. Pricing multiple underlying assets is more complex than pricing a single-asset European option, requiring consideration of the weighted sum of the asset values under a joint probability distribution and the relationship between the weighted sum and the strike price. This cannot be solved using the Black-Scholes formula in classical algorithms and is generally solved using the Monte Carlo method. Basket option pricing requires significant computational resources in classical calculations, especially when multiple underlying assets are involved. As a multi-asset derivative, the pricing method for basket options is relatively complex. The payout of a basket option is based on the combined price of multiple underlying assets, meaning that the joint probability distribution of all underlying assets must be considered when pricing the option. This joint probability distribution is computationally intensive in classical calculations, and the classical Monte Carlo method is inefficient when dealing with large samples.
[0043] In situations where the Black-Sc formula cannot be used, quantum algorithms offer significant advantages over classical algorithms. A key operation in quantum computing is quantum amplitude encoding, which maps continuous probability distributions to qubits. While amplitude estimation based on phase estimation has been applied in quantum algorithms, its time and space complexity still requires further optimization. Furthermore, preliminary solutions have been proposed for efficiently calculating the sum of qubit weights, but these methods lack universality, require substantial quantum resources, and necessitate a large number of qubits and high circuit depth. Since the number of qubits and circuit depth directly affect computational efficiency and practicality, excessive qubits and circuit depth not only increase computational complexity but may also exceed available quantum hardware resources, leading to resource bottlenecks in practical computations on current quantum hardware. Moreover, as the number of qubits increases, noise and errors in quantum computing may also increase, affecting the accuracy of the results.
[0044] In other words, as quantum computing is applied to basket option pricing, the related quantum algorithm design methods may lead to unnecessary resource consumption, such as excessive bits and complex circuit structures, making these algorithms difficult to implement on current quantum computing hardware. Therefore, reducing the number of bits and circuit depth has become an issue that needs improvement in quantum computing for basket option pricing.
[0045] To address the aforementioned issues, this application provides quantum chip systems, quantum computers, methods for improving the performance of quantum computers, quantum memories, quantum program products, and classical computers to improve related technologies. Based on the classical Black-Scholes-Merton option pricing model, this application presents a quantum version of the Monte Carlo algorithm for option pricing. Given the parameters required for option pricing, the algorithm can obtain the corresponding option price, fully leveraging the advantages of quantum computing to achieve greater optimization in time and space complexity. It also provides a general solution, particularly for basket option pricing and other complex financial derivatives pricing problems. By optimizing the design of the quantum algorithm to address the characteristics of basket options, it reduces the required number of qubits and circuit depth, lowering noise and errors in quantum computing, thereby achieving a more efficient and practical quantum computing method to solve complex computational problems such as option pricing.
[0046] It should be noted that although this application uses the option pricing scenario in the financial field as an example, this application can be applied to other calculation scenarios in the financial field, such as risk exposure estimation of credit default swaps (CDS), cash flow simulation of complex structured mortgage-backed securities (MBS), value risk (VaR) calculation of portfolios, rating simulation of leveraged loan obligations (CLO), price path simulation of energy derivatives, such as oil and gas futures, and expected rate of return and risk exposure estimation of mixed funds or hedge funds.
[0047] In addition to the financial field, this application can also be applied to computing scenarios in other technical fields, such as optimizing multimodal transportation routes, simulating inventory levels, and forecasting demand in logistics and supply chain management; simulating protein structure folding, screening drug molecules, and analyzing gene sequences in biomedical research; simulating and predicting long-term climate change and predicting short-term extreme weather events in meteorology; simulating the properties of new materials and optimizing nanostructure design in materials science; analyzing power grid stability and optimizing power distribution in power grid management; and simulating traffic flow, optimizing routes, and predicting traffic accidents in transportation. This application does not set any limitations in these areas.
[0048] This application relates to quantum computers, whose system operating environments are suitable for high-performance computing centers, research laboratories, or specialized quantum technology companies.
[0049] This application does not limit the type of quantum computer, which may include, but is not limited to, the following types.
[0050] Quantum computers based on superconducting circuits: These computers use superconducting circuits to realize qubits, resulting in high integration and fast operation speed.
[0051] Ion trap-based quantum computers: These use ion traps to realize qubits, resulting in high fidelity and long coherence time.
[0052] Quantum computers based on optical systems: These computers use optical systems to implement qubits, resulting in high parallelism and fast operation speed.
[0053] Quantum computers based on topological systems: These computers use topological systems to realize qubits, which have higher fault tolerance and longer coherence time.
[0054] (Quantum chip system)
[0055] See Figure 1 , Figure 1 This is a structural block diagram of a quantum chip system provided in one embodiment of this application.
[0056] One embodiment of this application provides a quantum chip system, including at least one quantum processor, which is used to execute quantum operations corresponding to a quantum program to implement steps S101 to S104.
[0057] Step S101: Prepare the quantum state of the variable bit group based on the joint probability distribution of the parameters of interest of the combined product.
[0058] The joint probability distribution of the parameters of interest is obtained based on the configuration information of the combined products and a preset statistical model. The combined products include multiple products, and the configuration information includes the identifier and weight of each product. Specific quantum states are prepared based on the probability distribution. In this embodiment, uniformly distributed quantum states can be prepared using Hadamard gates, and quantum states with specific distributions can be prepared using specific quantum gates, etc.
[0059] Step S102: Prepare the quantum state of the summation bit group based on the quantum state of the variable bit group and the weight of each product.
[0060] By combining the quantum state with its corresponding weights, a new quantum state is prepared. In this embodiment, the weight information of the product is used to prepare a quantum state for the summing bit group. As an example, a controlled gate is used to perform weighted summation on two qubits.
[0061] Step S103: Implement the target function through the target quantum circuit; the target quantum circuit includes the summation bit group and the result bit group.
[0062] The target function is computed using a target quantum circuit. In this embodiment, the information of the summing bit group can be converted into the output of a target quantum circuit, which represents the value of the target function. As an example, a target quantum circuit for the target function is constructed using specific quantum gates such as Pauli-X gates, Pauli-Y gates, or Pauli-Z gates.
[0063] Step S104: Execute the target quantum circuit at least once so that the measuring device performs at least one measurement on the resulting bit group, thereby calculating the expected value of the target function.
[0064] By executing the target quantum circuit and performing measurements, the amplitude or probability of a certain quantum state in the resulting bit group is estimated. In this embodiment, the target quantum circuit is executed multiple times to measure and calculate the expected value of the objective function. As an example, the expected value is estimated using statistical methods by repeatedly executing the target quantum circuit and performing measurements.
[0065] A quantum chip system is a miniature device used to process quantum information and execute quantum operations corresponding to quantum programs. In this embodiment, the quantum chip system is used as the core computing component for option pricing, leveraging its parallel processing capabilities to bring efficiency to option pricing. As an example, the quantum chip system can be a superconducting quantum chip system, an ion trap quantum chip system, a photonic quantum chip system, etc., including a quantum processor and peripheral devices providing quantum chip packaging.
[0066] A portfolio product can refer to a collection or combination of multiple products. The products forming a portfolio product can share certain commonalities. In some cases, a portfolio product can include multiple products of the same category to achieve an optimal allocation of similar products. Of course, a portfolio product can also consist of products belonging to different categories to optimize the allocation of products from different categories within the portfolio product. For example, a portfolio product can be a stock investment portfolio containing stocks of companies A, B, and C. Alternatively, a portfolio product can be a basket of options, including stocks, stock indices, bonds, funds, commodities, foreign exchange (currency), or interest rates, etc. A product is a single component of a portfolio product.
[0067] A parameter of interest can be a measure or identifier of certain important or key properties of a portfolio product. For example, a parameter of interest could be the value of a basket of options, or the total return or total volatility of a stock portfolio.
[0068] A joint probability distribution is a probability distribution of multiple variables simultaneously satisfying a certain condition. It describes the relationship between two or more random variables, representing the probability that these random variables will simultaneously take different values. For two random variables X and Y, the joint probability distribution P(X=x, Y=y) represents the probability that X and Y will simultaneously take specific values x and y. This can be represented by a two-dimensional table or a function called the joint probability distribution function. For example, based on the joint probability distribution function, we can obtain the probability that the price of stock A is X and the price of fund B is Y.
[0069] A variable bit set is a combination of qubits used to represent a parameter of interest. For example, a variable bit set may include three qubits used to encode an octet probability distribution.
[0070] A quantum state refers to the physical state of a quantum system, which can be a pure state or a mixed state. For example, |0>, |1>, or a superposition state (α|0>+β|1>).
[0071] Configuration information describes the details or attributes of each component of the portfolio product. For example, a basket of options may include underlying assets such as stock A, stock index B, bond C, foreign exchange D, interest rate E, commodity F, and fund G. The weights of these underlying assets may be the same or different, meaning that the proportion of each underlying asset in the calculation of the option payout may be the same or different.
[0072] A statistical model is a mathematical model used to predict or explain the relationship between variables based on known data. Examples include the normal distribution and the log-normal distribution.
[0073] An identifier is information or a code that uniquely identifies a product. For example, stock code XXXX indicates that the stock is of company A.
[0074] Weight refers to the relative importance or share of a product in a portfolio. For example, stock A has a weight of 40% in the portfolio.
[0075] A summation bit set is a set of qubits used to store and represent the sum of weights or cumulative effects. For example, a set of qubits used to represent the returns of a basket of options.
[0076] A target quantum circuit can correspond to a series of quantum gate operations to implement a specific quantum algorithm or function. For example, a target quantum circuit for implementing the quantum Fourier transform. From a software perspective, a quantum circuit is an abstract model used to represent and design quantum algorithms, consisting of quantum gates, qubits, and measurement operations. These gates define various operations between qubits and within the qubits themselves. The quantum circuit model is used to describe, design, and analyze the behavior of quantum algorithms. From a hardware perspective, when a quantum algorithm needs to be executed on an actual quantum computer, this abstract quantum circuit must be implemented in physical hardware. This means that each operation of a quantum gate in the target quantum circuit corresponds to a certain actual operation in the hardware, such as a microwave pulse or a laser pulse, which changes the state of the qubit. "Executing a target quantum circuit" refers to performing the physical operation matching the description of the target quantum circuit on the actual quantum computer hardware. The quantum circuit is used to perform computations on the input qubit set and store the results in the output qubit set. In this embodiment, the input qubit set of the target quantum circuit includes a summation qubit set, and the output qubit set of the target quantum circuit is the result qubit set.
[0077] An objective function is a function used to describe or measure the optimization objective of a specific task or problem. Examples include the payoff function for a basket of options, or the function to maximize the total return of a stock portfolio.
[0078] The result bit set refers to the set of qubits that stores the result of the execution of the target quantum circuit. For example, qubits representing the payoff of a basket of options. In this embodiment, the result bit set can contain one or more bits.
[0079] A measuring device is a device used to read and interpret the state of a qubit. For example, a device used to read a qubit as |0> or |1>.
[0080] Expected value refers to the average or weighted average of the possible values of a random variable. For example, the expected value of the objective function is 0.43.
[0081] The purpose of preparation is to convert information into quantum states, enabling further processing and computation on a quantum computer. Preparation can be achieved by using specific quantum gates, such as Hadamard gates, to create uniform distributions, or by using complex quantum circuits to create specific distributions. Examples include Bell states and GHZ states.
[0082] Implementing an operation refers to performing a specific computation or operation on a quantum computer, which can be done using specific quantum gates, target quantum circuits, or combinations thereof. As an example, the quantum NOT operation is implemented using Pauli-X gates and Hadamard gates.
[0083] The purpose of executing a target quantum circuit is to obtain its output, which may include measuring the target quantum circuit or manipulating its quantum states. Examples include Shor's algorithm and Grover's search.
[0084] The purpose of measurement is to obtain classical information from a quantum state. Measurements can be performed on qubits using Z, X, or Y measurements, etc. As an example, measuring the Bell state determines the correlation between two qubits.
[0085] In the "joint probability distribution," "joint" describes the probability of two or more events occurring simultaneously, determining the correlation and common behavior of the parameters of interest in the combined product.
[0086] In the "pre-defined statistical model", "pre-defined" means that the statistical model is defined or set in advance, rather than dynamically generated, which ensures the consistency and reliability of the statistical model.
[0087] At least one execution means that the target quantum circuit can be executed once or multiple times, ensuring a sufficient number of executions and measurements to obtain more accurate results.
[0088] As an example, the portfolio product is a basket of options, with the value of the basket of options as the parameter of interest. The configuration information includes the weight of each product, initial asset value, volatility, and one or more of the following: time to maturity (or trading period), risk-free rate (i.e., risk-free interest rate), strike price, and correlation coefficient matrix. The preset statistical model is a log-normal distribution model, and the joint probability distribution of the basket of options is obtained based on the configuration information of the basket of options and the log-normal distribution model. In the case of the Black-Scholes-Merton model, the Brownian motion Wt follows a normal distribution, and the distribution of the asset value St of a single underlying asset is also a log-normal distribution. Volatility and the risk-free interest rate are obtained from historical market data. After providing parameters such as initial asset value, time to maturity (multiple underlying assets of the portfolio product correspond to the same or different time to maturity), volatility, and the risk-free interest rate, the distribution of the asset value St at the corresponding underlying asset's time to maturity is obtained.
[0089] Therefore, firstly, the relevant information of the combined products is transformed into a quantum state. The joint probability distribution of the parameters of interest of the combined products is calculated based on a classical pre-defined statistical model and configuration information. After obtaining this joint probability distribution information, quantum operations are used to prepare a quantum state of a variable bit group corresponding to this probability distribution. That is, through specific quantum operations, combined with the weights of each product, the quantum state of the variable bit group can be processed to finally prepare a quantum state representing the sum of weights. Next, a pre-designed target quantum circuit is used to evaluate the objective function. This target quantum circuit combines the sum of the bit group with other control bits and operations to prepare the quantum state of the result bit group, which contains the information needed to be obtained from the objective function. Then, the specific target quantum circuit is executed and the results are measured to estimate certain values. The target quantum circuit is executed multiple times, and after each execution, a measuring device measures the result bit group. These measurement results can be used to calculate the expected value of the objective function.
[0090] In quantum computing, arithmetic operations, especially addition, require a carry-over process. When the sum of two bits exceeds its maximum value (e.g., 1+1 in binary), this excess value needs to be carried over to the next higher-order bit. In classical computing, this operation is relatively straightforward, but in quantum computing, due to the superposition and entanglement properties of quantum mechanics, the carry-over process can involve complex quantum gate operations and additional working bits. This quantum chip system is optimized to more quickly and intuitively determine which bits require a carry-over operation. Specifically, it can predict and accurately identify which bits need a carry-over based on the weight of each product, rather than checking each bit individually. This optimization in judgment greatly simplifies the carry-over operation, thereby reducing the number of related quantum gate operations and the required working bits. This design strategy significantly reduces the number of qubits and the depth of the quantum circuit when performing carry-over operations. In other words, this not only saves the quantum resources required for the algorithm but also shortens the computation time, making the entire quantum computing process more efficient.
[0091] Furthermore, in quantum computing, simulating specific probability distributions often requires complex quantum circuits and a large number of qubits. This quantum chip system uses specially designed circuits and algorithms to more efficiently encode the joint probability distribution of the parameters of interest when preparing the quantum states of variable qubit groups, thus avoiding many unnecessary quantum operations and qubits. Subsequently, the quantum states of the summation qubit group are prepared directly based on the quantum states and the weights of each product, enabling the quantum chip system to perform fast weighted summation operations without redundant operations. This quantum chip system forms a highly optimized quantum computing flow, where the functional modules corresponding to each step are optimized for their specific tasks. When these modules work together, the overall efficiency far exceeds the efficiency of considering each module individually. Therefore, this quantum chip system improves the efficiency of quantum computing, achieving the goal of further reducing the number of qubits and circuit depth.
[0092] Compared to classical computational methods, this quantum chip system significantly improves computational efficiency through parallel processing and quantum superposition, especially when dealing with large-scale combinations of products and complex probability distributions. Utilizing quantum amplitude estimation and specific quantum circuit design, the quantum chip system can more accurately calculate the expected value of the objective function, thereby reducing errors. The design of this quantum chip system takes into account the needs of various combinations of products and can be adapted to different product combinations and statistical models through simple configuration information adjustments. Through optimized quantum circuit design, this quantum chip system requires less quantum resources and depth during computation, thereby reducing the impact of quantum noise and errors.
[0093] In some embodiments, the process of preparing the quantum state of the variable bit group includes: mapping multiple combinations of values of the parameter of interest to multiple eigenstates of the variable bit group, and performing amplitude encoding on the corresponding eigenstate according to the probability density value corresponding to each combination of values to prepare the quantum state of the variable bit group; wherein, the multiple combinations of values of the parameter of interest and the probability density value corresponding to each combination of values are determined according to the joint probability distribution of the parameter of interest, and the sampling interval is uniformly sampled.
[0094] In this embodiment, the classical computer determines the sampling interval of the interest parameters of each product based on the joint probability distribution of the interest parameters of the combined products, and uniformly samples the interest parameters of each product within the sampling interval corresponding to each product to obtain multiple values of the interest parameters of each product, thereby obtaining multiple combinations of values of the interest parameters of the combined products; and obtains the probability density value corresponding to each value combination based on the joint probability distribution of the interest parameters of the combined products.
[0095] Each value combination contains one value of the parameter of interest for each product. For example, if the product combination includes three products, and each product has 2, 4, and 8 values for its parameter of interest, respectively, then the number of value combinations for the parameter of interest in the product combination is 2 × 4 × 8 = 64. In one implementation, each value combination can be represented as a d-dimensional vector, where d is the number of products in the product combination.
[0096] To encode the distribution of the parameters of interest of the combined products into quantum states, discrete sampling is required within the sampling interval of the parameters of interest for each product. This discrete sampling point distribution can be used to represent the original continuous distribution; the more sampling points and the larger the sampling range, the better it represents the original distribution.
[0097] A sampling interval is a range within which samples are taken. The sampling intervals for the parameters of interest for each product can be the same or different; for example, the sampling interval could be between 10 yuan and 20 yuan.
[0098] Uniform sampling is a sampling method in which samples or data points are selected at equal intervals or with equal probability within a given sampling interval. The goal of this sampling method is to ensure that every possible data point or sample has an equal chance of being selected, thus maintaining fairness and representativeness in the sampling. In the one-dimensional case, uniform sampling means sampling data points at equal intervals within a continuous interval to ensure a uniform distribution of data points throughout the interval. The value of the parameter of interest for each product refers to the specific value obtained during the sampling process.
[0099] A probability density value is the value of a probability density function at a specific value (or a specific combination of values). For example, in a normal distribution, the probability density value is largest at the mean. A probability density function represents the probability density at each specific point within the possible range of values a random variable can take, that is, the probability density near that point.
[0100] The joint probability distribution of the parameters of interest can be a joint log-normal distribution, a joint normal distribution, etc., and this application does not set any restrictions on this.
[0101] The joint log-normal distribution is a statistical distribution typically used to model two or more random variables that all follow a log-normal distribution. A log-normal distribution is one that, when its logarithm is taken, follows a normal distribution. The probability density function of the joint log-normal distribution can be used to describe the association between two or more random variables, as well as the joint distribution of these random variables.
[0102] The probability density function of the joint log-normal distribution can be expressed as follows:
[0103]
[0104] Where s is the product's parameter of interest, μ is the mean parameter of the log-normal distribution, and σ is the standard deviation parameter of the log-normal distribution; the subscript i∈[1,d] refers to the corresponding parameter of the i-th product, d is the number of products in the product combination, and d is a positive integer.
[0105] An eigenstate is a state of a quantum system that is deterministic when measured, and the system remains in that state after the measurement. For example, a spin-1 / 2 particle can be in either an "up" or "down" eigenstate.
[0106] Amplitude encoding refers to the process of encoding information using the amplitude of quantum states. For example, the probability density value of each combination of values is normalized and then encoded into the amplitude of the corresponding eigenstate of the variable bit group, so that when a measurement is performed, the distribution of the result matches the distribution of the probability density value.
[0107] For example, when the portfolio product is a basket of options and the parameter of interest is price (or value), suppose we obtain 32 possible value combinations, and the variable bit set includes 5 variable bits. Prices = {10, 11, 12, ..., 41}
[0108] The mapping can then be:
[0109] |00000> corresponds to 10
[0110] |00001> corresponds to 11
[0111] |00010> corresponds to 12 ...
[0113] |11111> corresponds to 41
[0114] As an example, amplitude encoding of the corresponding eigenstate is performed based on the probability density value corresponding to each value combination. This includes: normalizing the probability density value corresponding to each value combination so that the sum of the normalized probability values is 1, and then encoding the corresponding eigenstate based on the normalized probability values. Quantum states satisfy the condition that the sum of the squares of their amplitudes is 1. Normalization ensures this, making each quantum state a legitimate and physically realizable state. Normalized data is easier to process and manipulate, which helps simplify algorithms and reduce potential errors in quantum computing. Unnormalized data may result in some probability values being too large or too small, affecting the stability of the algorithm. Normalization ensures that all data is within a reasonable range, thereby improving the robustness of the algorithm.
[0115] Therefore, firstly, a suitable sampling interval is determined based on the joint probability distribution of the parameters of interest. Once the sampling interval is determined, uniform sampling can be performed within this interval to obtain multiple combinations of values. Each combination of values has a corresponding probability density value. Each combination of values is mapped to a specific eigenstate of the variable bit group. Then, amplitude encoding is performed based on the probability density values. The significance of amplitude encoding is to adjust the amplitude of each eigenstate so that the square of its amplitude corresponds to the probability value of the combination of values. This means that when this quantum state is measured in subsequent quantum computing, the probability distribution of the result will reflect the probability distribution of the original parameters of interest.
[0116] The advantages of this approach are: it directly encodes the classical probability distribution onto the quantum state, avoiding extensive computation on a classical computer; quantum systems can perform computations simultaneously across multiple states, and this encoding method allows for the simultaneous processing of multiple value combinations, significantly improving computational efficiency; it is particularly suitable for complex optimization problems, such as combinatorial optimization problems, where the solution space can be directly encoded onto the quantum state, and efficient searching can be performed using the properties of quantum mechanics; and through reasonable sampling and amplitude encoding, it can ensure that the results of quantum computation are as close as possible to the true probability distribution, thereby improving computational accuracy.
[0117] In some implementations, the sampling interval is determined based on the mean and standard deviation of the joint probability distribution.
[0118] The sampling interval is determined based on the statistical properties of the joint probability distribution of the parameters of interest. Specifically, it is based on the mean and standard deviation of the probability distribution of the parameters of interest for each product corresponding to the joint probability distribution. The mean provides the center or "expected" location. In most cases, the mean provides an optimal "centralized" point. The standard deviation provides a measure of the dispersion or variability of the probability distribution. The larger the standard deviation, the wider the range of variation in the probability distribution; conversely, the smaller the standard deviation, the more concentrated the probability distribution. Based on these two parameters, a sampling interval can be determined for each probability distribution, such that the sampling interval includes most of the statistical results. Then, based on the probability distribution of the parameters of interest for all products, a uniform sampling interval can be determined, and the parameters of interest for all products can be uniformly sampled within this uniform sampling interval. Alternatively, different sampling intervals can be determined for the parameters of interest for each product, and the parameters of interest for the corresponding product can be uniformly sampled within the sampling interval for each product.
[0119] As an example, the sampling interval corresponding to the parameter of interest for each product can be set to [μ-3σ, μ+3σ].
[0120] As another example, the sampling interval corresponding to the parameter of interest for each product can be set to [S]. min S min +6σ], where S min =max{μ-3σ, 0}. While related quantum techniques can price options, their sampling intervals range from the mean minus three standard deviations (minimum 0) to the mean plus three standard deviations. This method significantly narrows the sampling interval when the standard deviation is very large, leading to increased estimation errors. Furthermore, this method is not precise enough in handling multivariate normal and joint log-normal distributions, making it unsuitable for practical financial models. Related techniques do not consider the impact of large standard deviations when setting the sampling interval, failing to adequately handle multivariate normal and joint log-normal distributions. Therefore, to more accurately simulate these distributions and improve pricing accuracy, the sampling interval setting in the quantum algorithm can be improved to prevent narrowing, thereby reducing estimation errors and addressing the inaccuracy of related techniques in option pricing under large standard deviations.
[0121] The advantage of this approach is that by using the mean and standard deviation to determine the sampling interval, the selection of the sampling interval is ensured to be scientific and reasonable. When preparing quantum states or performing other quantum operations, if the sampling interval is set too wide, quantum resources may be wasted on unlikely values; if it is set too narrow, important values may be missed. By selecting the sampling interval based on statistical properties, unnecessary waste can be avoided while retaining key information. Sampling intervals determined based on statistical properties are more likely to contain the most critical and probable values, thus ensuring the accuracy and reliability of the sampling.
[0122] In a specific application scenario, the portfolio product is a basket of options, each product in the portfolio product is an underlying asset, the parameter of interest is the value of the basket of options, and each combination of the value of the basket of options is represented by a d-dimensional vector (d is the number of underlying assets).
[0123] The process of uniformly sampling the joint log-normal distribution of the values of a basket of options includes: First, for S... i Determine the mean μ of its log-normal distribution. i Standard deviation σ i and sampling interval, parameter σ i The value is obtained, for example, from the volatility and correlation coefficient matrix of each underlying asset, and the distribution is prepared accordingly; for S i Perform uniform sampling to obtain There are 10 sampling points (values), all of which are located at S. i Within the corresponding sampling interval; S1, S2, S3, ..., Sd―1 S d Each corresponding combination of values is assembled into a d-dimensional vector (this d-dimensional vector is the joint sampling point), which serves as a combination of values for a basket of options, resulting in... A combination of values.
[0124] In obtaining After considering a number of value combinations, obtain the following based on the joint probability distribution: The probability density values of each combination of values are then normalized to obtain the normalized value. The probability values of each combination of values; and, Each value combination is mapped to Each variable bit Each eigenstate is encoded using quantum amplitude encoding, and the normalized state is... The probability values of each combination of values are prepared into Each variable bit corresponds to The amplitude of each eigenstate, to complete Preparation of each variable bit in N variable bits. Where i is a positive integer, N i It is a positive integer.
[0125] In other words, for d underlying assets, to obtain Each sampling point (corresponding to the i-th target asset) (sampling points), forming A possible combination of values. If The obtained sampling point is d·2 N One, requiring d·N variable bits for encoding.
[0126] Considering The sum of the probability density values corresponding to each combination of values is not 1, therefore... Normalize the discrete probability density values to obtain the probability values corresponding to each combination of values. The sum of these probability values is 1, which is used for amplitude coding.
[0127] As can be seen, the joint probability distribution of a basket of options is prepared to The quantum state of N variable bits, where the i-th target asset corresponds to N. i Each variable bit. That is, using Each qubit represents the joint probability distribution of a basket of options, achieving an identically distributed mapping for each underlying asset f:i→f(i), where i is the sampling point corresponding to the i-th underlying asset represented by the variable bit. f(i) represents the value of the corresponding underlying asset. As an example, if a basket of options has 3 underlying assets, and each underlying asset corresponds to 5 variable bits, then the value of the basket of options is prepared to 15 variable bits. The first underlying asset corresponds to the 1st to 5th variable bits, the second underlying asset corresponds to the 6th to 10th variable bits, and the third underlying asset corresponds to the 11th to 15th variable bits.
[0128] In a d-dimensional vector, each dimension corresponds to a target asset, and the number of variable bits used for each target asset can be the same or different. As an example, the numerical values represented by the variable bits for each dimension are equal. That is, the first register represents 0 to 127 (corresponding to actual values of 1 to 2), and the second register, if represented by 8 bits, represents 0 to 255 (corresponding to actual values of approximately 1 to 3). In other words, when adding the two registers (or the two dimensions of the vector), the representation of these registers must be consistent, or the actual values represented by these registers should be within the same numerical range for the addition to be meaningful.
[0129] In other words, you can't use the first register's 7 bits to represent the actual values 1 to 2, and the second register to represent the actual values 2 to 3. Such a representation has several problems.
[0130] First, in this representation, the range of the actual value 2 overlaps. In the first register, 2 corresponds to the number 127, while in the second register, 2 corresponds to the number 0. This means that the actual value 2 has different representations in the two registers. However, this overlapping representation may lead to conflicts or uncertainties during calculations.
[0131] Secondly, in the first register, the range from 1 to 2 is divided into 128 units, with a precision of 1 / 128; while in the second register, the range from 2 to 3 is divided into 256 units, with a precision of 1 / 256. This means that the same actual numerical change has different precisions in the two registers, which will lead to inconsistencies in data representation and calculation.
[0132] Third, if certain operations are performed on these two registers, such as summation or comparison, additional conversions and adjustments are required due to their different numerical mappings and precisions. Therefore, to simplify data processing and ensure computational consistency, a unified representation method is needed.
[0133] Fourth, increasing the number of qubits will increase the sampling range of the register exponentially by 2.
[0134] Therefore, for the application scenario of a basket of options, if the first variable takes the value 1 to 2, the second is 1.1 to 2.1, and the third is 1.2 to 2.4, we can uniformly use n variable bits to sample the range of 1 to 2.4, for a total of 3n variable bits, where n is a positive integer.
[0135] In this embodiment, the distribution preparation process can be implemented by the distribution preparation module packaged in the quantum chip system. The distribution preparation can be completed simply by inputting the corresponding set parameters (e.g., configuration information). Similarly, the weight summation process and the objective function implementation process can also be implemented by the packaged weight summation module and objective function module.
[0136] In some implementations, the process of preparing the quantum state of the summing bit group includes: applying at least one controlled NOT gate to the variable bit group, the summing bit group, and the carry bit group according to the weight of each product, in order to prepare the quantum state of the summing bit group.
[0137] The Controlled-Negative Gate (CNOT) is a two-bit quantum logic gate that performs a NOT operation on the second bit, but this operation is determined by the state of the first bit (the control bit). If the control bit is |1>, the target bit is flipped; if the control bit is |0>, the target bit remains unchanged.
[0138] When the control bit is |0> and the target bit is |0>, the result of the operation is |00>.
[0139] When the control bit is |0> and the target bit is |1>, the result of the operation is |01>.
[0140] When the control bit is |1> and the target bit is |0>, the result of the operation is that |10> becomes |11>.
[0141] When the control bit is |1> and the target bit is |1>, the result of the operation is that |11> becomes |10>.
[0142] In quantum mechanics and quantum information science, an "action" refers to the process by which a quantum gate or quantum operation alters the state of one or more qubits. This alteration can be linear, nonlinear, reversible, or irreversible, depending on the specific gate or operation used. Consider a simple case with two qubits, initially in the state |01>, where the first bit (control bit) is |0> and the second bit (target bit) is |1>. If a controlled NOT gate is applied to these two bits, the target bit remains unchanged since the control bit is |0>. Therefore, the resulting quantum state is still |01>.
[0143] Therefore, controlled NOT gates (also known as controlled NOT gates or CNOT gates) are used to implement the weight summation operation. In this embodiment, the variable bit group, summation bit group, and carry bit group are encoded in an appropriate manner according to the weight of each product. Then, controlled NOT gates are used to perform the weight summation operation based on this encoded information. This involves a series of controlled NOT gate operations, and the interaction and order between the quantum gate operations determine how the summation operation is performed according to the weight of each product, thereby obtaining the quantum state of the desired summation bit group.
[0144] The advantages of this approach are: using controlled-NOT gates for weighted summation directly leverages the parallelism of quantum systems, allowing multiple weights and values to be processed in a single operation, thus improving both computational speed and the accuracy of weighted summation; by employing controlled-NOT gates, the use of more complex quantum logic gates requiring more quantum resources is avoided, thereby improving the resource efficiency of quantum computing for a given task; different weight and variable configurations can be easily adapted by simply adjusting the control of the controlled-NOT gate and the selection of the target bit; and controlled-NOT gates are relatively simple gates and well-implemented in most quantum computing systems, resulting in a lower error rate and thus providing more robust performance.
[0145] In some implementations, the process of applying at least one controlled NOT gate to the variable bit group, the summation bit group, and the carry bit group includes: determining a corresponding binary matrix based on the weight of each product; and applying the corresponding controlled NOT gate to the variable bit group, the summation bit group, and the carry bit group based on the value and position of each binary bit in the binary matrix.
[0146] A binary matrix is a matrix whose elements contain only binary values 0 or 1.
[0147] A binary bit is the smallest unit that represents a binary value; it can only be 0 or 1.
[0148] The variable bit set includes variable bits corresponding to each product, where each product may correspond to one or more variable bits.
[0149] A carry bit group refers to a group of qubits used to process and store carry bits during binary addition.
[0150] For example, suppose there are two products with weights of 5 (101 in binary) and 3 (011 in binary). These weights can be represented as the following binary matrix: 1 0 1 0 1 1
[0153] That is, each row of the binary matrix represents the weight of a product. For example, when the portfolio is a basket of options, each row of the binary matrix can represent the weight of the value of an underlying asset.
[0154] In the example above, the quantum state of the variable bit group is prepared by the joint probability distribution, the summation bit group is used to store the weighted summation result, and the carry bit group is used to handle the carry in binary addition.
[0155] To calculate the weighted sum, a controlled NOT gate is used. Specifically, based on the value at each position in the binary matrix, a controlled NOT gate is applied to the variable bit group, the sum bit group, and the carry bit group.
[0156] The weight of the first row is 101, and controlled NOT gates are applied at positions 1 and 3.
[0157] The weight of the second row is 011, and a controlled NOT gate is applied at the 2nd and 3rd positions.
[0158] This method allows for the efficient summation of all product weights using quantum operations, with the result stored in a summation bit set.
[0159] Therefore, firstly, the weight of each product is converted into a corresponding binary matrix. In quantum computing, information is encoded as qubits, which are binary and represent states of |0> and |1> respectively. Then, based on the value (0 or 1) and position of each bit in these binary matrices, controlled NOT gates are applied to the variable bit group, the summing bit group, and the carry bit group. Specifically, if a bit in the binary matrix is 1, then the corresponding controlled NOT gate is used for the corresponding qubit, thereby performing a summation operation according to that weight. The weights are directly binary encoded, and then these encodings are used to control the operation of the controlled NOT gates, thus achieving the summation of weights at the quantum state level.
[0160] The advantages of this approach are as follows: by directly converting the weights into binary matrices, an intuitive and simplified encoding method is provided for quantum computing, which helps to simplify the design and implementation of algorithms; controlling the controlled NOT gates according to the binary matrix allows the quantum system to process multiple addition operations in parallel, improving computational efficiency, especially when dealing with a large number of weights; direct binary encoding and controlled NOT gate operations help reduce errors that may be introduced by complex encoding and multi-step operations; it is easy to extend to handle more complex or larger weight matrices, and can easily adapt to larger or more complex problems by simply adding more qubits and corresponding controlled NOT gate operations; compared with methods that require more complex quantum logic gates or multiple iterations, direct binary encoding and controlled NOT gate operations can make more efficient use of quantum resources.
[0161] In one embodiment, the controlled NOT gate includes a first Tooffoli gate and a first CNOT gate. The step of applying the corresponding controlled NOT gate to the variable bit group, the summation bit group, and the carry bit group according to the value and position of each binary bit in the binary matrix includes: if the value of the binary bit is 1 and it is the least significant bit, then applying the first Tooffoli gate to the variable bit, the summation bit corresponding to the least significant bit, and the carry bit corresponding to the least significant bit; and applying the first CNOT gate to the variable bit and the summation bit corresponding to the least significant bit. The control bit of the first Tooffoli gate is the summation bit corresponding to the least significant bit, and the control bit of the first CNOT gate is the variable bit.
[0162] In one embodiment, the controlled NOT gate further includes a second CNOT gate, and the process of applying the corresponding controlled NOT gate to the variable bit group, the summation bit group, and the carry bit group further includes: if the value of the binary bit is 1 and it is the most significant bit, then the second CNOT gate is applied to the variable bit and the summation bit corresponding to the most significant bit, wherein the control bit of the second CNOT gate is the variable bit.
[0163] In one embodiment, the controlled NOT gate further includes a third CNOT gate and a control virtual Tooffoli gate. The process of applying the corresponding controlled NOT gate to the variable bit group, the summation bit group, and the carry bit group further includes: if the value of the binary bit is 1 and it is the middle bit, then the third CNOT gate is applied to the variable bit and the carry bit corresponding to the middle bit, and the control virtual Tooffoli gate is applied to the variable bit, the summation bit corresponding to the middle bit, the carry bit corresponding to the middle bit and the carry bit corresponding to the middle bit, the control bit of the third CNOT gate is the variable bit, the control bit of the control virtual Tooffoli gate is the variable bit, and the virtual control bit of the control virtual Tooffoli gate is the carry bit corresponding to the middle bit and the carry bit corresponding to the middle bit and the previous carry bit.
[0164] In one embodiment, the step of applying the corresponding controlled NOT gate to the variable bit group, the summation bit group, and the carry bit group according to the value and position of each binary bit in the binary matrix includes: if the value of the binary bit is 0 and it is the least significant bit, then the controlled NOT gate operation is not performed on the variable bit, the summation bit, and the carry bit.
[0165] In one embodiment, the controlled NOT gate includes a fourth CNOT gate. The process of applying the corresponding controlled NOT gate to the variable bit group, the summation bit group, and the carry bit group further includes: if the value of the binary bit is 0 and it is the most significant bit, then the fourth CNOT gate is applied to the summation bit corresponding to the most significant bit and the carry bit preceding the carry bit corresponding to the most significant bit, wherein the control bit of the fourth CNOT gate is the carry bit preceding the carry bit corresponding to the most significant bit.
[0166] In one embodiment, the controlled NOT gate further includes a second Tooffoli gate and a fifth CNOT gate. The process of applying the corresponding controlled NOT gate to the variable bit group, the summation bit group, and the carry bit group further includes: if the value of the binary bit is 0 and it is the middle bit, then the second Tooffoli gate is applied to the summation bit corresponding to the middle bit, the carry bit preceding the carry bit corresponding to the middle bit, and the carry bit corresponding to the middle bit; and the fifth CNOT gate is applied to the carry bit preceding the summation bit corresponding to the middle bit and the carry bit corresponding to the middle bit. The control bit of the second Tooffoli gate is the carry bit preceding the summation bit corresponding to the middle bit and the carry bit corresponding to the middle bit, and the control bit of the fifth CNOT gate is the carry bit preceding the carry bit corresponding to the middle bit.
[0167] In one embodiment, the process of preparing the quantum state of the summation bit group after applying the corresponding controlled NOT gate to the variable bit group, the summation bit group, and the carry bit group according to the value and position of each binary bit in the binary matrix further includes: determining a target quantum logic gate based on the controlled NOT gate, the target quantum logic gate being used to reset the carry bit; and applying the target quantum logic gate to the carry bit.
[0168] For example, see Figure 2 , Figure 2 This is a schematic diagram of a quantum weighted summer (i.e., a weighted summing module) provided in one embodiment of this application. The value corresponding to a basket of options includes the weighted sum of the values of three underlying assets. The value of each underlying asset is encoded using one variable bit, where q0, q1, and q2 are variable bits with weights of 1, 2, and 4, respectively. The sum of the weights is 7, therefore, the number of summing bits required to represent this weighted sum is at least 3, where q3, q4, and q5 are summing bits. The number of carry bits is determined based on the number of summing bits, which is at least 2, where q6 and q7 are carry bits. Real control bits are represented by solid black dots, and virtual control bits are represented by hollow circles.
[0169] The weight of q0 is 1, its binary value is 001, and the least significant bit is 1. Therefore, the first Tooffoli gate is applied to the variable bit q0, the summation bit q3 corresponding to the least significant bit, and the carry bit q6 corresponding to the least significant bit (corresponding to timing 1). Then, the first CNOT gate is applied to the variable bit q0 and the summation bit q3 corresponding to the least significant bit (corresponding to timing 2). The control bits of the first Tooffoli gate are q0 and q3, and the control bit of the first CNOT gate is q0.
[0170] The weight of q0 is 1, its binary value is 001, and the middle bit is 0. The second Tooffoli gate is applied to the summation bit q4 corresponding to the middle bit, the carry bit q6 preceding the carry bit q7 corresponding to the middle bit, and the carry bit q7 corresponding to the middle bit (corresponding to timing 3). The fifth CNOT gate is applied to the summation bit q4 corresponding to the middle bit and the carry bit q6 preceding the carry bit q7 corresponding to the middle bit (corresponding to timing 4). The control bits of the second Tooffoli gate are q4 and q6, and the control bit of the fifth CNOT gate is q6.
[0171] For q0, the weight is 1, its binary value is 001, and the highest bit is 0. The fourth CNOT gate is applied to the summation bit q5 corresponding to the highest bit and the carry bit corresponding to the highest bit (q5). Figure 2 The carry bit q7 (not shown in the diagram) is the previous carry bit (corresponding to timing 5), and the control bit of the fourth CNOT gate is q7.
[0172] For q0, the weight is 1, and its binary value is 001. After applying the controlled NOT gates in time series 1-5 to the variable bit, summation bit, and carry bit based on the weights, the binary value is traversed in reverse order. The target quantum logic gate U1 (time series 6) for resetting the carry bit is determined according to the controlled NOT gates in time series 1-5, and U1 is applied to the carry bit. Since the controlled NOT gates that applied to the carry bits q6 and q7 have also applied to other bits, such as Tofoli (q4, q6, q7) in time series 3, CNOT (q6, q4) in time series 4, and CNOT (q7, q5) in time series 5, they may also apply to the variable bit and summation bit during the actual reset.
[0173] It should be noted that this operation is mainly to reset the carry bit, so that it can be reused in the next variable bit operation, thereby saving computing resources.
[0174] The weight of q1 is 2, its binary value is 010, and the least significant bit is 0. Controlled NOT gate operations are not performed on the variable bits, summation bits, and carry bits.
[0175] The weight of q1 is 2, its binary value is 010, and the middle bit is 1. The third CNOT gate is applied to the variable bit q1 and the carry bit q7 corresponding to the middle bit (corresponding to timing 7), and the control virtual Tooffoli gate is applied to the variable bit q1, the summation bit q4 corresponding to the middle bit, the carry bit q6 preceding the carry bit q7 corresponding to the middle bit, and the carry bit q7 corresponding to the middle bit (corresponding to timing 8). The control bit of the third CNOT gate is q1, the control bit of the control virtual Tooffoli gate is q1, and the virtual control bits of the control virtual Tooffoli gate are q4 and q6.
[0176] For q1, the weight is 2, its binary value is 010, and the highest bit is 1. The specific implementation of the controlled NOT gate operation (CNOT(q7, q5)) corresponding to the highest bit is given in the above embodiment; the reference is provided directly here. Figure 2 In time series 9, the specific implementation of the target quantum logic gate U2 corresponding to the controlled NOT gate with weight 2 is given in the above embodiment, and is directly provided here. Figure 2 The details of time series 10 will not be elaborated further.
[0177] The weight of q2 is 4, its binary value is 100, and its highest bit is 1. The second CNOT gate is applied to the variable bit q2 and the summation bit q5 corresponding to the highest bit. The control bit of the second CNOT gate is q2.
[0178] For q2, the weight is 4, its binary value is 100, and the least significant bit and the middle bit are 0. No controlled NOT gate operation is performed on the least significant bit. For the controlled NOT gate operations (Toffoli(q4, q6, q7) and CNOT(q7, q5)) corresponding to the middle bit, please refer to the above embodiment for specific implementation details; the reference is given directly here. Figure 2 For timing sequences 11 and 12, the specific implementation of the target quantum logic gate U3 corresponding to the controlled NOT gate with weight 4 is given in the above embodiment. Here, the reference is given directly. Figure 2 The details of time series 14 will not be elaborated further.
[0179] In some other possible implementations, the process of preparing the quantum state of the summing bit group includes: determining the rotation parameters of at least one controlled rotation logic gate according to the weight of each product; and applying the corresponding controlled rotation logic gate to the variable bit group according to the rotation parameters of the controlled rotation logic gate to prepare the quantum state of the summing bit group.
[0180] Controlled-rotation logic gates are a type of quantum logic gate that determines whether to perform a rotation operation on a target bit, depending on the state of one or more control bits. When the control bit is in a specific state (e.g., 1), the target bit will rotate according to the given rotation parameters; when the control bit is not in that state, the target bit is unaffected.
[0181] A rotation parameter is a parameter that determines the amount of rotation (usually expressed as an angle or radians). In quantum computing, the rotation parameter determines the degree of rotation of a qubit on the Bloch sphere. For example, a common controlled rotation logic gate is the Pauli-Y rotation gate, which can be described by an angle θ, indicating that the qubit should rotate by θ angles along the Y-axis. Assuming a product with a weight of 0.6, this weight can be mapped to a rotation parameter, for example: θ = π × 0.6 = 1.884 radians. Using a controlled rotation logic gate, when the control bit is 1, the target bit will rotate 1.884 radians along the Y-axis on the Bloch sphere; while when the control bit is 0, the target bit is unaffected.
[0182] To apply this rotation to the variable bit set and prepare the quantum state of the summing bit set, the following steps can be used: First, determine the rotation parameters according to the product weights, such as 1.884 radians in the example above. Next, use a controlled rotation logic gate to apply the rotation to the corresponding bits of the variable bit set. In this way, the qubits in the variable bit set will be rotated according to the product weights, thereby preparing the quantum state of the summing bit set.
[0183] Therefore, a method based on controlled rotation logic gates is adopted to handle the summation of weights. First, the weight of each product is used to determine the rotation parameters of at least one controlled rotation logic gate. The controlled rotation logic gate can rotate the state of the qubit at a specific angle according to the given rotation parameters. The calculation of these rotation parameters is directly related to the product weights and can therefore be regarded as a weighted operation of the quantum states. Next, according to the obtained rotation parameters, the corresponding controlled rotation logic gate is applied to the variable bit group. This operation changes the state of the qubit to reflect its corresponding weight. In other words, the summation operation of weights is mapped to the rotation operation of qubits, realizing the quantum expression of weights.
[0184] The advantages of this approach are that controlled rotation logic gates provide a continuous and accurate way to represent and process weights, reducing errors introduced by discretization and allowing for more precise reflection of small changes in weights; by directly rotating qubits to represent weights, the quantum computing process is simplified, avoiding the need for multiple quantum logic gates or complex quantum circuits; and controlled rotation logic gates provide an efficient method for quantum weighted summation, especially when a large number of weights are involved.
[0185] In some implementations, the process of determining the rotation parameters of the controlled rotating logic gate includes: determining the base angle of the controlled rotating logic gate; and determining the rotation parameters of the controlled rotating logic gate corresponding to each variable bit based on the product of the base angle and the weight of each product.
[0186] In quantum computing, a specific rotation operation can be described as a rotation on a Bloch sphere. The base angle can be viewed as a preset rotation angle, representing the base or unit rotation angle of a rotation operation. When multiplied by certain weights or other factors, this base angle can be used to determine the specific rotation angle or the rotation parameter of a controlled rotation logic gate. Assuming the base angle is: α = π / 10 = 0.314 radians. This means that when considering a unit weight, the corresponding rotation operation will rotate the Bloch sphere by 0.314 radians along an axis.
[0187] Therefore, firstly, the base angle of the controlled rotation logic gate is determined. The base angle can be considered as the basic rotation unit for each variable bit, providing a reference for subsequent rotation operations. Next, the determined base angle is multiplied by the weight of each product to obtain the rotation parameter of the controlled rotation logic gate corresponding to the variable bit of each product. Therefore, the rotation parameter is the product of the base angle and the weight, which means that this rotation parameter will provide the qubit with a rotation proportional to its weight.
[0188] The advantages of this approach are: it directly links the weights to the rotation parameters of the rotation logic gate, making the quantum implementation of the weight summation operation more intuitive and concise; by using a predetermined base angle multiplied by the weights, it ensures that the rotation operation accurately reflects the weight of each product, thereby reducing errors in weight representation; since each variable bit has its corresponding rotation parameter, it can flexibly handle different weights and variables, maintaining scalability when dealing with complex weight combinations or large-scale problems; and by directly using the base angle multiplied by the weights as the rotation parameter, it avoids situations requiring other complex calculations or additional quantum operations.
[0189] In one implementation, the controlled rotation logic gate can be a combinational logic gate of controlled RX gate, controlled RY gate, controlled RZ gate and controlled U1 gate.
[0190] For example, when the controlled rotary logic gate is a controlled RX gate or a controlled RY gate, its corresponding characteristic value is i is the imaginary sign, θ is the product of the base angle and the weight, and the base angle can be chosen to be either a controlled RY gate or a controlled RX gate. Where m is the number of summation bits in the summation bit group.
[0191] The step of determining the rotation parameters of the controlled rotating logic gate corresponding to each variable bit based on the product of the base angle and the weight of each product includes: constructing the rotation parameters of the controlled RY gate or controlled RX gate for each variable bit based on the base angle, with the rotation parameters being 2... m―1 θ, …2θ, θ. Where, w represents the weight.
[0192] For example, see Figure 3 , Figure 3 This is a schematic diagram of another quantum weight adder (i.e., weight summing module) provided in one embodiment of this application. It is assumed that there are 3 variable bits q0, q1, and q2 with weights of 1, 2, and 4 respectively, 3 summing bits q3, q4, and q5, and 1 auxiliary bit q6.
[0193] The controlled RY gates for the first variable bit are RY(θ) 11 ), RY(θ) 12 ), RY(θ) 13 ), where θ 11 =2 2 θ1, θ 12 =2θ1, θ 13 =θ1,
[0194] The controlled RY gates for the second variable bit are RY(θ) 21 ), RY(θ) 22 ), RY(θ) 23 ), where θ 21 =2 2 θ2, θ 22 =2θ2、θ 23 =θ2,
[0195] The controlled RY gates for the third variable bit are RY(θ) 31 ), RY(θ) 32 ), RY(θ) 33 ), where θ 31 =2 2 θ3, θ 32 =2θ3, θ 33 =θ3,
[0196] Figure 3 The roles of the H gate and the Fourier transpose conjugate operation are discussed in the context of phase estimation quantum circuits in related technologies. The H gate is used to create hyperstates for uniform phase estimation, while the Fourier transpose conjugate operation is used to convert the estimated phase information back to the computational basis to obtain the accurate phase value. These details will not be elaborated upon here.
[0197] For combinational logic gates that are controlled RZ gates and controlled U1 gates, except for the different bottom angles, the specific implementation process is the same as the specific implementation method of the controlled RY gates mentioned above, and will not be repeated here.
[0198] In one implementation, the portfolio product is a basket of options, and the objective function is the payoff function of the basket of options. The process of realizing the objective function through a target quantum circuit includes: constructing a target quantum circuit corresponding to a piecewise linear function of the payoff function based on a preset strike price and the payoff function of the basket of options; and mapping the function values of the piecewise linear function to the result bit group in the target quantum circuit based on the target quantum circuit.
[0199] In the financial field, the Black-Scholes-Merton model (option pricing model) can be used to determine the value of each underlying asset after time t. The specific calculation formula is as follows:
[0200]
[0201] t is the expiration time, S t Let S0 be the value of the asset at time t, S0 be the initial value, σ be the volatility parameter, and W be the initial value. t The motion is Brownian, and r is the risk-free interest rate.
[0202] A basket of options is an extension of a single-asset European option. Its return depends on the weighted sum of the d underlying assets. A basket of call options has the following payoff function:
[0203]
[0204] Where K is the strike price (i.e., the option's exercise price).
[0205] Since the payoff function F(S) is the same for both call and put options, basket The results are all piecewise linear based on the comparison with K. Therefore, the target quantum circuit can be constructed using a linear function amplitude preparation algorithm to obtain F(S). basket The expected value of the linear function amplitude preparation algorithm can be obtained by calling the encapsulated target function module. Only the corresponding parameters (e.g., configuration information) need to be input to realize a piecewise linear function using the target quantum circuit.
[0206] In one embodiment, the process of executing the target quantum circuit at least once includes: executing the target quantum circuit multiple times so that a measuring device performs multiple measurements on the resulting bit set to calculate the expected value of the target function.
[0207] This application does not limit the number of measurements, which may be, for example, once or multiple times.
[0208] Specifically, the amplitude value of the resulting bit group is measured, and the expected value of the piecewise linear function is determined based on the amplitude value of the resulting bit group.
[0209] In one implementation, the option premium for the basket of options is calculated based on the expected value and a preset conversion formula.
[0210] In this embodiment, the payoff of a basket of options is "the value of the basket of options after the expiration time". The maximum value between the difference between the value S and the strike price K, and "0", is defined as follows: when the value S > the strike price K (positive payout), f(S) = max{0, S - K}; when the value S < the strike price K (negative payout), f(S) = max{S - K, 0}. Here, f(S) is the payout function for a basket of options (this payout function is a piecewise linear function).
[0211] Next, a target quantum circuit corresponding to the piecewise linear function is constructed. Then, the function value of the piecewise linear function is mapped to a preset result bit group. The expected value of the piecewise linear function can be determined based on the amplitude corresponding to the result bit group. This allows the expected value of the piecewise linear function to be determined based on the target quantum circuit. Finally, the option of the target object is calculated based on the expected value of the piecewise linear function, completing the option estimation of the target object. The construction of the target quantum circuit can be seen in, for example, in Chinese patent document application number 202110129188.X, entitled "Option Estimation Method, Apparatus, Medium and Electronic Device Based on Quantum Circuits," and will not be elaborated upon here.
[0212] It should be noted that, based on the preset exercise price and the prepared d·N qubits, the target quantum circuit corresponding to the payoff function of a basket of options is constructed. In essence, this includes the construction of a comparison quantum circuit for comparing the value S with the exercise price K.
[0213] The method for calculating the amplitude of qubits (i.e., the resulting bit set) used in the above steps determines the amplification parameters of the target quantum circuit in each iterative step. This iterative process ensures that the difference between the upper and lower bounds of the amplitude angle remains within a precision threshold, avoiding the problem of amplitude convergence and improving the accuracy of the amplitude value. The underlying principles and more detailed implementation process can be found in Chinese patent document application number 202011591351.6, entitled "Amplitude Estimation Method, Apparatus, Storage Medium, and Electronic Device for Quantum Circuits," which will not be elaborated upon here.
[0214] One embodiment of this application also provides a quantum chip system, including at least one quantum processor, which is used to execute quantum operations corresponding to a quantum program to achieve the following steps:
[0215] Multiple combinations of values for the parameter of interest are mapped to multiple eigenstates of the variable bit group, and amplitude encoding is performed on the corresponding eigenstates according to the probability density value corresponding to each value combination to prepare the quantum state of the variable bit group; wherein, the multiple combinations of values for the parameter of interest and the probability density value corresponding to each value combination are determined according to the joint probability distribution of the parameter of interest, and the sampling interval is obtained by uniform sampling; the joint probability distribution of the parameter of interest is obtained according to the configuration information of the combined product and a preset statistical model, the combined product includes multiple products, and the configuration information includes the identifier and weight of each product; the sampling interval is determined according to the mean and standard deviation of the joint probability distribution;
[0216] Based on the quantum state of the variable bit group and the weight of each product, prepare the quantum state of the summation bit group;
[0217] The target function is achieved through a target quantum circuit; the target quantum circuit includes the summation bit group and the result bit group.
[0218] The target quantum circuit is executed at least once so that the measuring device performs at least one measurement on the resulting bit group, thereby calculating the expected value of the target function.
[0219] The quantum state of the summing bit group is prepared using any of the following methods.
[0220] The first method is to determine the corresponding binary matrix based on the weight of each product; and to apply the corresponding controlled NOT gate to the variable bit group, the summation bit group, and the carry bit group based on the value and position of each binary bit in the binary matrix, so as to prepare the quantum state of the summation bit group.
[0221] The second method involves: determining the base angle of the controlled rotation logic gate; determining the rotation parameter of the controlled rotation logic gate corresponding to each variable bit based on the product of the base angle and the weight of each product; and applying the corresponding controlled rotation logic gate to the variable bit group based on the rotation parameter of the controlled rotation logic gate to prepare the quantum state of the summing bit group.
[0222] In some implementations, since real-world quantum computers are still affected by noise and errors, noise-tolerant techniques or quantum error-correcting codes can be integrated into the quantum chip system to improve computational accuracy. Some complex problems may require a hybrid of quantum and classical computational methods. The quantum chip system can be designed to support hybrid computing, where some tasks are performed in the classical computation part and others in the quantum part. Module parameters can be dynamically adjusted according to different application scenarios. During design, the parameterization and configurability of each module can be fully utilized to make the quantum chip system more adaptable. Considering the potential of quantum chip systems in quantum simulation, quantum simulation modules can be added to the chip to support the simulation of specific physical systems or chemical processes. Since quantum computing has potential applications in machine learning, quantum machine learning modules can be added to implement commonly used quantum learning algorithms, such as quantum support vector machines and quantum neural networks. Furthermore, utilizing quantum parallelism, multiple tasks or multiple circuits can be executed simultaneously on the chip. During design, this characteristic can be effectively utilized to improve the chip's computational efficiency. To ensure efficient data processing, data flow and communication between modules can be further studied and optimized. There are various physical implementations of quantum computing, such as ion traps, superconducting qubits, and topological qubits. By designing the architecture of a quantum chip system to adapt to different physical implementation methods, its versatility can be improved.
[0223] (Quantum computer)
[0224] One embodiment of this application also provides a quantum computer, the specific implementation of which is consistent with the implementation method and the technical effect achieved in the above-described quantum chip system implementation method, and some contents will not be repeated.
[0225] See Figure 4 , Figure 4 This is a schematic diagram of the structure of a quantum computer provided in one embodiment of this application.
[0226] The quantum computer includes any one of the aforementioned quantum chip system 1, measurement and control system 2, support system 3, and operating system 4. For example, a superconducting quantum computer includes components such as a 24-qubit superconducting quantum chip system 1, a quantum computing measurement and control system 2, a quantum computer operating system 3, and a quantum computing environment support system 4. Figure 4 As shown.
[0227] The quantum chip system 1 includes at least one quantum processor, which is used to execute quantum operations corresponding to quantum programs to process qubits. For example, in a superconducting quantum computer, the superconducting quantum chip system 1 based on a superconducting quantum processor is the computing core of the quantum computer, capable of executing quantum programs.
[0228] The measurement and control system 2 includes a control device and a measurement device. The control device converts the quantum program into corresponding control signals and sends them to the quantum processor. The measurement device measures the qubits. The quantum computing measurement and control system 2 is the control system of the quantum computer, used to control the operation of the quantum chip system. The quantum chip system 1 receives control signals from the measurement and control system 2, executes the quantum operations corresponding to the quantum program to process the qubits, thereby realizing the execution of the quantum program.
[0229] This application does not limit the measuring device, which may be any one or a combination of the following devices.
[0230] Superconducting measurement device: This device utilizes the properties of superconducting materials at low temperatures to measure the state of qubits. It can accurately detect changes in the quantum state in a very short time.
[0231] Optical interferometer: It measures quantum states by interfering with light, and is especially suitable for photonic qubits.
[0232] Ion trap detector: For ion trap qubit systems, this device can detect the quantum state of ions by measuring their electromagnetic frequency.
[0233] Magnetic resonance measurement device: Using magnetic resonance technology, it is possible to measure the state of specific qubits, such as spin-based qubits.
[0234] Charge measurement device: This device can detect the charge state in quantum dots or other tiny structures, thereby determining their quantum state.
[0235] Quantum dot measurement device: For quantum dot-based quantum computing systems, this device is able to measure the electronic states in the quantum dots.
[0236] Quantum state discriminator: It can identify and classify a given quantum state, thereby determining the specific state.
[0237] Environmental monitor: Since quantum computing requires specific environmental conditions (such as low temperature and low noise), this device can monitor and control the computing environment in real time to ensure the accuracy of measurements.
[0238] Error correction module: Since errors may occur in quantum computing, the error correction module can correct and optimize the measurement results.
[0239] The measurement devices listed above can be used independently or in combination with other devices to meet specific quantum computing needs.
[0240] The support system 3 provides the necessary working environment conditions to ensure the operation of the quantum chip system 1. The support system 3 may include an ultra-low temperature refrigeration system and a host active vibration damping system, providing a stable working environment for the quantum computer.
[0241] The operating system 4 provides a software system to enable interaction between the user and the quantum chip system 1 and the measurement and control system 2. The quantum computer operating system 4 can, for example, employ the Sinan system, providing a quantum computing software framework for the quantum computer, and possessing functions such as parallel execution of multiple quantum computing tasks, automatic calibration of the quantum chip, and efficient management of quantum resources.
[0242] (Methods to improve the performance of quantum computers)
[0243] One embodiment of this application also provides a method for improving the performance of a quantum computer. The specific implementation method and the technical effects achieved are consistent with those described in the above-described quantum chip system implementation method, and some details will not be repeated here.
[0244] See Figure 5 , Figure 5 This is a flowchart illustrating a method for improving the performance of a quantum computer, provided in one embodiment of this application.
[0245] The method is used to improve the performance of quantum computers, and the method includes steps S101 to S104.
[0246] Step S101: Prepare the quantum state of the variable bit group according to the joint probability distribution of the parameters of interest of the combined products; the joint probability distribution of the parameters of interest is obtained according to the configuration information of the combined products and the preset statistical model, the combined products include multiple products, and the configuration information includes the identifier and weight of each product.
[0247] Step S102: Prepare the quantum state of the summation bit group based on the quantum state of the variable bit group and the weight of each product.
[0248] Step S103: Implement the target function through the target quantum circuit; the target quantum circuit includes the summation bit group and the result bit group.
[0249] Step S104: Execute the target quantum circuit at least once so that the measuring device performs at least one measurement on the resulting bit group, thereby calculating the expected value of the target function.
[0250] This method for improving the performance of quantum computers utilizes specific parameters of the combined products to optimize and tune the quantum computing process in order to reduce the number of bits and the line depth. By preparing quantum states based on the joint probability distribution of the parameters of interest for combined products, quantum computing tasks can be made more direct and efficient, thereby reducing resource waste. Preparing quantum states of summation qubit groups based on the quantum states of the weights and variable qubit groups of each product allows for more accurate simulation of actual financial models, resulting in computational results closer to reality. This method provides a standardized computational process for various combined products, easily applicable to different financial scenarios and needs. Because the method clarifies the operation of each step, quantum computing resources can be better allocated and utilized, improving overall computational efficiency. This technical solution can not only be used for combined products in the financial field but also extended to other fields such as supply chain management, project management, and resource scheduling, achieving cross-domain quantum computing optimization. This method allows for seamless integration between quantum computing and classical statistical models, better utilizing the advantages of both. Through rapid evaluation of the objective function, decision-makers can obtain feedback on the returns of combined products more quickly, accelerating the decision-making process. Because the method optimizes both quantum state preparation and the execution of the target quantum circuit, it effectively reduces computational errors and improves the reliability of results. For financial institutions and investors, the ability to quickly and accurately evaluate the returns of combined products is crucial, and this method meets this need, thereby improving user satisfaction.
[0251] In one embodiment, the step of preparing the quantum state of the variable bit group based on the joint probability distribution of the parameters of interest of the combined products (i.e., step S101) includes: mapping multiple combinations of values of the parameters of interest to multiple eigenstates of the variable bit group, and performing amplitude encoding on the corresponding eigenstates according to the probability density value corresponding to each combination of values, so as to prepare the quantum state of the variable bit group.
[0252] In one implementation, the sampling interval is determined based on the mean and standard deviation of the joint probability distribution.
[0253] In one embodiment, the step of preparing the quantum state of the summation bit group based on the quantum state of the variable bit group and the weight of each product (i.e., step S102) includes: applying at least one controlled NOT gate to the variable bit group, the summation bit group, and the carry bit group according to the weight of each product to prepare the quantum state of the summation bit group.
[0254] In one embodiment, the step of applying at least one controlled NOT gate to the variable bit group, the summing bit group, and the carry bit group according to the weight of each product to prepare the quantum state of the summing bit group includes: determining a corresponding binary matrix according to the weight of each product; and applying a corresponding controlled NOT gate to the variable bit group, the summing bit group, and the carry bit group according to the value and position of each binary bit in the binary matrix.
[0255] In one embodiment, the step of preparing the quantum state of the summation bit group based on the quantum state of the variable bit group and the weight of each product (i.e., step S102) includes: determining the rotation parameters of at least one controlled rotation logic gate based on the weight of each product; and applying the corresponding controlled rotation logic gate to the variable bit group based on the rotation parameters of the controlled rotation logic gate to prepare the quantum state of the summation bit group.
[0256] In one implementation, determining the rotation parameters of at least one controlled rotating logic gate based on the weight of each product includes: determining the base angle of the controlled rotating logic gate; and determining the rotation parameters of the controlled rotating logic gate corresponding to each variable bit based on the product of the base angle and the weight of each product.
[0257] (Quantum memory)
[0258] One embodiment of this application also provides a quantum memory, the specific implementation of which is consistent with the implementation method and the technical effect achieved in the above-described quantum chip system implementation method, and some contents will not be repeated.
[0259] The quantum memory stores a quantum program, which, when executed by at least one quantum processor, implements the functions of any of the aforementioned quantum chip systems or the steps of any of the aforementioned methods.
[0260] By introducing quantum memories, quantum computers can store a complex number of quantum programs, which can then be executed by quantum processors, thus greatly expanding the functionality and application scope of quantum computing. Compared to classical memories, quantum memories can store and process vast amounts of quantum state information, meaning quantum computers can execute more complex and advanced quantum algorithms and tasks. Due to the nature of quantum programs, quantum processors can handle multiple computational tasks simultaneously, significantly accelerating computation compared to classical computing. Quantum memories can store quantum keys and other information related to quantum communication and quantum cryptography, providing a higher level of security than classical technologies. The design of quantum memories allows for dynamic expansion of memory capacity to adapt to ever-increasing computational demands. Quantum memories can work alongside traditional classical memories and other computing resources, enabling collaborative work between quantum and classical computers. Because quantum memories can store multiple quantum states at superlocations, data redundancy can be effectively reduced, and storage efficiency improved. By storing quantum programs specifically designed for quantum processors, the potential of quantum computing, such as quantum machine learning and quantum simulation, can be better utilized. Users can load and execute different quantum programs according to their needs, enabling quantum computers to perform a wide variety of tasks. Efficient management and allocation of quantum storage resources ensures the efficient operation of quantum computers. The introduction of quantum memories creates conditions for further research and application of quantum computing, quantum communication, and other quantum technologies, thereby driving technological progress in the entire field of quantum information science.
[0261] (Quantum program products)
[0262] One embodiment of this application also provides a quantum program product, the specific implementation of which is consistent with the implementation method and the technical effect achieved in the above-described quantum chip system implementation method, and some contents will not be repeated.
[0263] The quantum program product includes a quantum program that, when executed by at least one quantum processor, implements the functions of any of the aforementioned quantum chip systems or the steps of any of the aforementioned methods.
[0264] Quantum program products provide specific operating and functional guidelines for quantum computers, allowing quantum processors to efficiently perform complex tasks. This enables a variety of quantum computing tasks to be performed in a modular and repeatable manner. Similar to classical software applications, quantum programs can be shared and deployed by multiple users or devices, thus expanding their application scope. Optimized quantum programs can improve the computational efficiency of quantum processors, reduce error rates, and ensure accurate computational results. Quantum program products can include complex algorithms and processes, supporting various advanced quantum computing tasks such as optimization, simulation, encryption, and machine learning. Users can customize or adjust quantum programs to meet specific computing tasks or business needs. Quantum programs can work in conjunction with classical computing programs, allowing for seamless integration between quantum and classical computers. For non-specialist users, quantum program products provide a simple and easy-to-use interface, enabling users to easily leverage the power of quantum computing without needing to delve into its complex underlying principles. Quantum program products can include various testing and verification tools to help users ensure the accuracy and reliability of quantum computing. As quantum computing technology develops, more and more quantum program products are entering the market, providing users with various functions and services, thereby driving the commercialization process of the entire quantum computing field. Quantum program products can also be used as educational and training tools to help students and researchers better understand and master the principles and applications of quantum computing.
[0265] This application does not limit the users. Users can be providers of portfolio products, such as banks, investment institutions, securities companies, insurance companies, fund management companies, private equity funds, hedge funds, fintech companies, financial consulting firms, etc.; users of portfolio products, such as individual investors, corporate finance departments, retirement funds, university endowments, non-governmental organizations, etc.; financial research and education institutions, such as universities, research institutes, financial training centers, economics schools, etc.; or regulatory agencies, rating agencies, auditing firms, etc. related to the financial market.
[0266] These users can utilize the aforementioned quantum computing technologies to perform various financial calculations and analyses, such as portfolio optimization, risk assessment, pricing models, investment strategy analysis, financial simulation, and time series analysis.
[0267] Banks and investment institutions can use quantum computing to optimize their portfolios, assess the risks and returns of different investment strategies, or conduct complex financial simulations.
[0268] Securities firms and fintech companies can leverage quantum computing to develop new trading strategies, pricing models, or high-frequency trading algorithms.
[0269] Insurance companies: Quantum computing can be used to assess the risks of various insurance products, or to conduct long-term financial planning and forecasting.
[0270] Individual investors and corporate finance departments: Quantum computing can be used for more accurate investment analysis and decision-making.
[0271] Financial research and education institutions: Quantum computing can be used for cutting-edge financial research or to provide advanced financial education and training for students.
[0272] Regulators and rating agencies: Quantum computing can be used to more accurately assess market stability and risk, or to conduct more complex financial audits and regulations.
[0273] In short, both financial institutions and individual investors can leverage the power of quantum computing to improve the quality and efficiency of their financial decisions.
[0274] (Classic Computer)
[0275] One embodiment of this application also provides a classic computer, including a memory and at least one processor.
[0276] Memory is used to store computer programs.
[0277] At least one processor is used to execute the computer program to perform the following steps: determining a sampling interval based on the joint probability distribution of the parameters of interest of the combined products, and uniformly sampling the sampling interval to obtain multiple combinations of values of the parameters of interest and the probability density value corresponding to each combination of values.
[0278] Therefore, by uniformly sampling the joint probability distribution of the parameters of interest in the combined product, the optimal or near-optimal parameter combination can be effectively determined. Uniform sampling can provide decision-makers with a more comprehensive view of the parameter space, thereby assisting them in making more accurate decisions. Compared with quantum computers, classical computers may offer faster processing speeds for certain tasks, especially when the task can be processed in parallel or does not require complex quantum operations. Furthermore, because classical computer technology is quite mature, it can be more easily integrated with existing systems and technologies, is easily scalable, and is stable and reliable.
[0279] It is understood that the specific examples in this document are only intended to help those skilled in the art better understand the implementation methods of this application, and are not intended to limit the scope of this application.
[0280] It is understood that in the various embodiments of this application, the sequence number of each process does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application.
[0281] It is understood that the various implementation methods described in this application can be implemented individually or in combination, and the implementation methods in this application are not limited in this respect.
[0282] Unless otherwise stated, all technical and scientific terms used in the embodiments of this application have the same meaning as commonly understood by one of ordinary skill in the art. The terminology used in this application is for the purpose of describing particular embodiments only and is not intended to limit the scope of this application. The term "and / or" as used in this application includes any and all combinations of one or more of the associated listed items. The singular forms "a," "the," and "the" as used in the embodiments of this application and the appended claims are also intended to include the plural forms unless the context clearly indicates otherwise.
[0283] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.
[0284] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the quantum computer, performance improvement method, quantum memory, and quantum program product described above can be referred to the corresponding processes in the aforementioned quantum chip system implementation methods, and will not be repeated here.
[0285] In the several embodiments provided in this application, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between devices or units may be electrical, mechanical, or other forms.
[0286] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment, depending on actual needs.
[0287] In addition, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.
[0288] If the aforementioned functions are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium (e.g., a classical computer or a quantum computer). Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.
[0289] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.
Claims
1. A quantum chip system, characterized by, Includes at least one quantum processor, said at least one quantum processor being used to execute quantum operations corresponding to a quantum program to achieve the following steps: The quantum state of the variable bit group is prepared based on the joint probability distribution of the parameters of interest of the combined products; the joint probability distribution of the parameters of interest is obtained according to the configuration information of the combined products and the preset statistical model; the combined products include multiple products, and the configuration information includes the identifier and weight of each product. Based on the quantum state of the variable bit group and the weight of each product, prepare the quantum state of the summation bit group, including: determining the corresponding binary matrix according to the weight of each product; applying the corresponding controlled NOT gate to the variable bit group, the summation bit group, and the carry bit group according to the value and position of each binary bit in the binary matrix; the controlled NOT gate includes a first Tooffoli gate, a second Tooffoli gate, a first CNOT gate, a second CNOT gate, a third CNOT gate, a fourth CNOT gate, a fifth CNOT gate, and a controlled virtual Tooffoli gate; The target function is achieved through a target quantum circuit; the target quantum circuit includes the summation bit group and the result bit group. The target quantum circuit is executed at least once so that the measuring device performs at least one measurement on the resulting bit group, thereby calculating the expected value of the target function; The step of applying the corresponding controlled NOT gate to the variable bit group, the summation bit group, and the carry bit group according to the value and position of each binary bit in the binary matrix includes: If the value of the binary bit is 1 and it is the least significant bit, then the first Toffoli gate is applied to the variable bit, the summation bit corresponding to the least significant bit, and the carry bit corresponding to the least significant bit, and the first CNOT gate is applied to the variable bit and the summation bit corresponding to the least significant bit. The control bit of the first Toffoli gate is the summation bit corresponding to the variable bit and the least significant bit, and the control bit of the first CNOT gate is the variable bit. If the value of the binary bit is 1 and it is the most significant bit, then the second CNOT gate is applied to the variable bit and the summation bit corresponding to the most significant bit, and the control bit of the second CNOT gate is the variable bit; If the value of the binary bit is 1 and it is the middle bit, then the third CNOT gate is applied to the variable bit and the carry bit corresponding to the middle bit, and the control virtual control Tooffoli gate is applied to the variable bit, the summation bit corresponding to the middle bit, the carry bit before the carry bit corresponding to the middle bit and the carry bit corresponding to the middle bit. The control bit of the third CNOT gate is the variable bit, the control bit of the control virtual control Tooffoli gate is the variable bit, and the virtual control bit of the control virtual control Tooffoli gate is the summation bit corresponding to the middle bit and the carry bit before the carry bit corresponding to the middle bit. If the value of the binary bit is 0 and it is the least significant bit, then no controlled NOT gate operation is performed on the variable bit, the summation bit, and the carry bit. If the value of the binary bit is 0 and it is the most significant bit, then the fourth CNOT gate is applied to the summation bit corresponding to the most significant bit and the carry bit preceding the carry bit corresponding to the most significant bit. The control bit of the fourth CNOT gate is the carry bit preceding the carry bit corresponding to the most significant bit. If the value of the binary bit is 0 and it is the middle bit, then the second Toffoli gate is applied to the sum bit corresponding to the middle bit, the carry bit preceding the carry bit corresponding to the middle bit, and the carry bit corresponding to the middle bit. The fifth CNOT gate is applied to the sum bit corresponding to the middle bit and the carry bit preceding the carry bit corresponding to the middle bit. The control bit of the second Toffoli gate is the carry bit preceding the sum bit corresponding to the middle bit and the carry bit corresponding to the middle bit. The control bit of the fifth CNOT gate is the carry bit preceding the carry bit corresponding to the middle bit.
2. The quantum chip system of claim 1, wherein, The process of preparing a quantum state for a variable bit set includes: Multiple combinations of values of the parameter of interest are mapped to multiple eigenstates of the variable bit group, and the corresponding eigenstates are amplitude encoded according to the probability density value corresponding to each value combination to prepare the quantum state of the variable bit group; wherein, the multiple combinations of values of the parameter of interest and the probability density value corresponding to each value combination are determined according to the joint probability distribution of the parameter of interest, and the sampling interval is uniformly sampled to obtain the sampling interval.
3. The quantum chip system of claim 2, wherein, The sampling interval is determined based on the mean and standard deviation of the joint probability distribution.
4. The quantum chip system of claim 1, wherein, The process of preparing the quantum state of the summing bit group includes: Based on the weight of each product, determine the rotation parameters of at least one controlled rotary logic gate; Based on the rotation parameters of the controlled rotation logic gate, the corresponding controlled rotation logic gate is applied to the variable bit group to prepare the quantum state of the summing bit group.
5. The quantum chip system of claim 4, wherein, The process of determining the rotation parameters of the controlled rotation logic gate includes: Determine the bottom angle of the controlled rotary logic gate; The rotation parameters of the controlled rotation logic gate corresponding to each variable bit are determined based on the product of the base angle and the weight of each product.
6. A quantum computer, characterized by The quantum computer includes: The quantum chip system according to any one of claims 1-5, the quantum chip system comprising at least one quantum processor, the at least one quantum processor being configured to execute quantum operations corresponding to a quantum program to process qubits; A measurement and control system, comprising a control device and a measurement device, wherein the control device is used to convert the quantum program into corresponding control signals and send them to the quantum processor, and the measurement device is used to measure the qubits; A support system is provided to provide the operating environment conditions to ensure the operation of the quantum chip system; An operating system is provided to offer software that enables interaction between the user and the quantum chip system and the measurement and control system.
7. A method for improving performance of a quantum computer, characterized by, The method for improving the performance of a quantum computer includes: The quantum state of the variable bit group is prepared based on the joint probability distribution of the parameters of interest of the combined products; the joint probability distribution of the parameters of interest is obtained according to the configuration information of the combined products and the preset statistical model; the combined products include multiple products, and the configuration information includes the identifier and weight of each product. Based on the quantum state of the variable bit group and the weight of each product, prepare the quantum state of the summation bit group, including: determining the corresponding binary matrix according to the weight of each product; applying the corresponding controlled NOT gate to the variable bit group, the summation bit group, and the carry bit group according to the value and position of each binary bit in the binary matrix; the controlled NOT gate includes a first Tooffoli gate, a second Tooffoli gate, a first CNOT gate, a second CNOT gate, a third CNOT gate, a fourth CNOT gate, a fifth CNOT gate, and a controlled virtual Tooffoli gate; The target function is achieved through a target quantum circuit; the target quantum circuit includes the summation bit group and the result bit group. The target quantum circuit is executed at least once so that the measuring device performs at least one measurement on the resulting bit group, thereby calculating the expected value of the target function; The step of applying the corresponding controlled NOT gate to the variable bit group, the summation bit group, and the carry bit group according to the value and position of each binary bit in the binary matrix includes: If the value of the binary bit is 1 and it is the least significant bit, then the first Toffoli gate is applied to the variable bit, the summation bit corresponding to the least significant bit, and the carry bit corresponding to the least significant bit, and the first CNOT gate is applied to the variable bit and the summation bit corresponding to the least significant bit. The control bit of the first Toffoli gate is the summation bit corresponding to the variable bit and the least significant bit, and the control bit of the first CNOT gate is the variable bit. If the value of the binary bit is 1 and it is the most significant bit, then the second CNOT gate is applied to the variable bit and the summation bit corresponding to the most significant bit, and the control bit of the second CNOT gate is the variable bit; If the value of the binary bit is 1 and it is the middle bit, then the third CNOT gate is applied to the variable bit and the carry bit corresponding to the middle bit, and the control virtual control Tooffoli gate is applied to the variable bit, the summation bit corresponding to the middle bit, the carry bit before the carry bit corresponding to the middle bit and the carry bit corresponding to the middle bit. The control bit of the third CNOT gate is the variable bit, the control bit of the control virtual control Tooffoli gate is the variable bit, and the virtual control bit of the control virtual control Tooffoli gate is the summation bit corresponding to the middle bit and the carry bit before the carry bit corresponding to the middle bit. If the value of the binary bit is 0 and it is the least significant bit, then no controlled NOT gate operation is performed on the variable bit, the summation bit, and the carry bit. If the value of the binary bit is 0 and it is the most significant bit, then the fourth CNOT gate is applied to the summation bit corresponding to the most significant bit and the carry bit preceding the carry bit corresponding to the most significant bit. The control bit of the fourth CNOT gate is the carry bit preceding the carry bit corresponding to the most significant bit. If the value of the binary bit is 0 and it is the middle bit, then the second Toffoli gate is applied to the sum bit corresponding to the middle bit, the carry bit preceding the carry bit corresponding to the middle bit, and the carry bit corresponding to the middle bit. The fifth CNOT gate is applied to the sum bit corresponding to the middle bit and the carry bit preceding the carry bit corresponding to the middle bit. The control bit of the second Toffoli gate is the carry bit preceding the sum bit corresponding to the middle bit and the carry bit corresponding to the middle bit. The control bit of the fifth CNOT gate is the carry bit preceding the carry bit corresponding to the middle bit.
Citation Information
Patent Citations
Amplitude estimation method and device of quantum circuit, storage medium and electronic device
CN114757357A
Data processing method and device, equipment and storage medium
CN114723317A
Quantum circuit-based option estimation method and device, medium and electronic device
CN114819170A
