Quantum computer

Abstract

Provided is a quantum computer capable of reducing the number of wires from a microwave control device compared to conventional systems. The quantum computer comprises: a quantum bit circuit 3 in which a plurality of quantum bits Q are arranged; a microwave control device 12 that outputs a plurality of frequency-multiplexed pulses obtained by multiplexing a plurality of microwave pulses having different frequencies; and a plurality of microwave transmission lines 16 for inputting the frequency multiplexed pulses to the quantum bits. The microwave transmission lines 16 include main lines 18 to which the frequency-multiplexed pulses are input from the microwave control device 12, and a plurality of branch lines 19 branched from each of the main lines 18. The main lines 18 are configured to be each shared by two or more quantum bits Q, wherein the plurality of branch lines 19 provided for each main line 18 are respectively connected to different quantum bits Q. The branch lines 19 from the respectively different main lines 18 are connected to adjacent quantum bits Q.

Description

quantum computer

[0001] The present invention relates to quantum computers.

[0002] In recent years, research into quantum computers, which perform calculations using quantum mechanical phenomena, has been attracting attention. Highly accurate control of quantum bits is essential for the realization of quantum computers, and various quantum computer-related technologies have been proposed (see, for example, Patent Document 1). For example, superconducting quantum bits, which are currently mainstream in quantum computers, are controlled by microwave pulses.

[0003] Transmon qubits are the main type of quantum computers using superconducting qubits. Known methods include a frequency-variable system, in which the resonant frequency (resonant frequency of the Transmon qubit) for microwaves is variable, and a frequency-fixed system, in which the resonant frequency for microwaves is fixed. The frequency-variable system requires a large number of wires per qubit, and generally tends to have a short coherence time. In contrast, the frequency-fixed system uses a single coaxial line for control per qubit, which generally results in a long coherence time and a simple configuration, making it suitable for large-scale applications.

[0004] Japanese Patent Application Laid-Open No. 2022-57269

[0005] However, in both frequency-variable and fixed-frequency quantum computers, as the number of quantum bits increases with the scale of the computer, there is a problem that the number of wirings, such as microwave transmission lines, that supply microwave pulses to the quantum bits also increases accordingly.

[0006] The present invention has been made in view of the above-mentioned problems, and has an object to provide a quantum computer that can reduce the number of wires from a microwave control device compared to conventional methods.

[0007] A quantum computer according to the present invention comprises a quantum bit circuit in which a plurality of quantum bits are arranged, a microwave control device that outputs a plurality of frequency-multiplexed pulses obtained by multiplexing a plurality of microwave pulses having different frequencies, and a plurality of microwave transmission lines that input the frequency-multiplexed pulses to the quantum bits, wherein the microwave transmission lines comprise a main line to which the frequency-multiplexed pulses are input from the microwave control device, and a plurality of branch lines branching off from the main line, wherein each of the main lines is shared by two or more of the quantum bits, and the plurality of branch lines provided for each main line are respectively connected to different quantum bits, and the branch lines from the different main lines are respectively connected to adjacent quantum bits.

[0008] According to the quantum computer of the present invention, by sharing one main line among a plurality of quantum bits, the number of wires from the microwave control device can be reduced accordingly compared to the conventional case.

[0009] 10 is a block diagram showing the configuration of a quantum computer. It is a schematic diagram showing a configuration in which a quantum bit circuit is connected to a microwave control device via a microwave transmission line. It is a schematic diagram showing an example of wiring assignment when using a triple-multiplexed frequency-multiplexed pulse in a quantum bit circuit in which quantum bits are arranged in an 8×8 square lattice. 1001 is a schematic diagram for explaining the alternating execution of an operation in which a one-qubit gate acts on a quantum bit and an operation in which a two-qubit gate acts on an adjacent quantum bit, 1002 is a schematic diagram for explaining a decoupling pulse, and 1003 is a diagram showing a generalized Hadamard matrix. It is a schematic diagram showing a state in which a one-qubit gate acts on each quantum bit. It is a schematic diagram showing a state in which a two-qubit gate acts on any adjacent quantum bit among a plurality of quantum bits. It is a schematic diagram showing a state in which a two-qubit gate acts on a specific adjacent quantum bit while canceling out interactions with other quantum bits in a region in which quantum bits are arranged in an 8×8 square lattice. Fig. 1 is a schematic diagram for explaining microwave pulses for frequency multiplexing; Fig. 2 is a block diagram showing the circuit configuration of a processing unit; Fig. 3 is a flowchart showing a wiring allocation process procedure for frequency multiplexing; Fig. 4 is a schematic diagram showing an example of wiring allocation when using frequency multiplexed pulses multiplexed by five in a quantum bit circuit in which quantum bits are arranged in a 16 × 16 square lattice; Fig. 5 is a graph showing the relationship between the lattice size of a quantum bit circuit in which quantum bits are arranged in a square lattice and the frequency multiplexing number.

[0010] DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings. In the following description, the same components are designated by the same reference numerals, and redundant description will be omitted.

[0011] <Overview of the quantum computer according to this embodiment> In the following embodiment, a superconducting quantum computer controlled by microwave pulses will be described as an example of a quantum computer, but the present invention can also be applied to various other quantum computers, such as semiconductor quantum computers.

[0012] 1 shows the configuration of a quantum computer 1 according to this embodiment. The quantum computer 1 includes a control device 2 and a quantum bit circuit 3 consisting of multiple quantum bits. The control device 2 includes a server 10, multiple microwave control devices 12, and a synchronization clock distribution unit 14.

[0013] The control device 2 is connected to the quantum bit circuit 3. The server 10 and the synchronization clock distribution unit 14 are connected to each microwave control device 12. The server 10 is provided with a dynamic decoupling processing unit 13, and each microwave control device 12 is provided with a signal processing unit 15. The synchronization clock distribution unit 14 distributes a common clock to each signal processing unit 15. Each signal processing unit 15 synchronizes with the other signal processing units 15 based on the clock. The quantum bit circuit 3 is provided in an extremely low temperature refrigerator.

[0014] Upon receiving instruction input from a user, the server 10 calculates waveform signals required for controlling and reading out the state of the quantum bits in the quantum bit circuit 3, and outputs them to the signal processing unit 15. The server 10 also reads out the input signals and analysis results from the signal processing unit 15 and performs predetermined processing.

[0015] Since each microwave control device 12 has the same configuration, the following description will focus on one microwave control device 12. Note that, as another embodiment, for example, a configuration in which only one microwave control device 12 is provided may be used. As shown in FIG. 2 , the microwave control device 12 includes a plurality of microwave sources 121-1, 121-2, .... The microwave control device 12 multiplexes microwave pulses of predetermined frequencies generated by each microwave source 121-1, 121-2, ... to generate a plurality of frequency-multiplexed pulses. The microwave control device 12 selects a frequency to be multiplexed for each frequency-multiplexed pulse, and inputs the generated plurality of frequency-multiplexed pulses to the quantum bit circuit 3 at a predetermined timing.

[0016] The quantum bit circuit 3 has a plurality of quantum bits arranged in, for example, a square lattice pattern. In this embodiment, which employs a superconducting quantum computer, a superconducting quantum bit having a Josephson junction element with an insulating layer between a pair of superconductors can be used as the quantum bit. The quantum bit according to this embodiment is preferably, for example, a frequency-locked quantum bit that resonates with a specific frequency and operates a one-qubit gate. Each quantum bit in the quantum bit circuit 3 is designed so that adjacent quantum bits have different resonant frequencies.

[0017] In the quantum bit circuit 3 according to this embodiment, there is a residual interaction between quantum bits even when no microwave pulse is input. This residual interaction, which always exists, is known as the permanent ZZ interaction. It is also known that inputting a microwave pulse of a predetermined non-resonant frequency to adjacent quantum bits having such a ZZ interaction can enhance the ZZ interaction in the adjacent quantum bits (see, for example, Non-Patent Document 1, "KX Wei et al., PRL 129, 060501 (2022); BK Mitchell et al., PRL 127, 200502 (2021)").

[0018] In the quantum bit circuit 3 according to this embodiment, when a microwave pulse of a predetermined non-resonant frequency is input as a non-resonant pulse to adjacent quantum bits among a plurality of quantum bits by a frequency multiplexed pulse, the interaction between the adjacent quantum bits to which the non-resonant pulse is input is enhanced, and a two-qubit gate is activated.

[0019] For example, if the adjacent quantum bits that operate the two-qubit gate have a non-resonant frequency of Ω1 cos(ω d t) microwave pulse and Ω2 cos(ω d When microwave pulses of ωt + φ are applied, the ZZ interaction rate ζ increases or decreases by the value shown in the following equation (1) (see Non-Patent Document 1). Note that Ω1 and Ω2 are values ​​obtained by converting the amplitude of each microwave pulse into the angular frequency of the resonant Rabi oscillation. dis the angular frequency of the microwave pulse. φ is the offset phase of the microwave pulse. In the following equation (1), α1 and α2 are the anharmonicities of the adjacent quantum bits Q, respectively. g is the coupling rate between the adjacent quantum bits. ω1 and ω2 are the resonant frequencies of the adjacent quantum bits Q, respectively.

[0020] In the quantum bit circuit 3, for example, when enhancing the interaction between adjacent quantum bits to operate a two-qubit gate, microwave pulses of a predetermined frequency are input as decoupling pulses from the microwave control device 12 to the other quantum bits, thereby canceling out unnecessary residual interactions in the other quantum bits. The microwave control device 12 generates decoupling pulses of a predetermined frequency at a predetermined timing based on a control signal received from the dynamic decoupling processing unit 13.

[0021] As described above, the microwave control device 12 according to this embodiment generates any one of a resonant pulse, a non-resonant pulse, and a decoupling pulse as a microwave pulse. A resonant pulse is a microwave pulse having an individual resonant frequency that causes a one-qubit gate to act on each quantum bit in the quantum bit circuit 3. A non-resonant pulse is a microwave pulse having a non-resonant frequency that causes a two-qubit gate to act on an adjacent quantum bit in the quantum bit circuit 3. A decoupling pulse is a microwave pulse having a frequency that cancels out interactions occurring between adjacent quantum bits in the quantum bit circuit 3.

[0022] In addition to this configuration, when the microwave control device 12 according to this embodiment causes a one-qubit gate to act on a quantum bit Q or when a two-qubit gate to act on an adjacent quantum bit Q, it generates one of the following four types of frequency-multiplexed pulses depending on the operation, and inputs the generated frequency-multiplexed pulse to a predetermined quantum bit Q. When causing a one-qubit gate to act on a quantum bit Q, the microwave control device 12 generates a frequency-multiplexed pulse by multiplexing a plurality of resonant pulses having different resonant frequencies, each of which causes a one-qubit gate to act on each quantum bit.

[0023] Furthermore, when a two-qubit gate is applied to adjacent quantum bit Q, microwave controller 12 generates, as a first frequency-multiplexed pulse, a frequency-multiplexed pulse obtained by multiplexing a plurality of non-resonant pulses of different non-resonant frequencies that enhance the interaction in adjacent quantum bits. Furthermore, when a two-qubit gate is applied to adjacent quantum bit Q, microwave controller 12 generates, as a second frequency-multiplexed pulse, a frequency-multiplexed pulse obtained by multiplexing one or more non-resonant pulses of non-resonant frequencies that enhance the interaction in adjacent quantum bits and one or more decoupling pulses of frequencies that cancel out the interaction occurring in adjacent quantum bits. Furthermore, when a two-qubit gate is applied to adjacent quantum bit Q, microwave controller 12 generates, as a third frequency-multiplexed pulse, a frequency-multiplexed pulse obtained by multiplexing a plurality of decoupling pulses of different frequencies that cancel out the interaction occurring in adjacent quantum bits.

[0024] In addition to this configuration, as shown in Fig. 2, a plurality of microwave transmission lines 16 that input frequency-multiplexed pulses to the quantum bits are connected to the microwave control device 12. Each of the microwave transmission lines 16 is connected to a plurality of quantum bits. Note that since each microwave transmission line 16 has the same configuration, the following description will focus on one microwave transmission line 16.

[0025] The microwave transmission line 16 includes a main line 18 to which frequency-multiplexed pulses are input from the microwave control device 12, and a plurality of branch lines 19 branching off from the main line 18. Each main line 18 is shared by two or more quantum bits Q. The plurality of branch lines 19 provided for each main line 18 are connected to different quantum bits Q. Adjacent quantum bits Q are connected to branch lines 19 from different main lines 18. Note that in FIG. 2 , diagrams of the branch lines 19 to the quantum bits Q are omitted to avoid complicating the drawing.

[0026] 3 is a schematic diagram showing an example of wiring assignment when using frequency-multiplexed pulses obtained by triply multiplexing three microwave pulses of different frequencies in a quantum bit circuit 3 in which quantum bits Q are arranged in an 8 × 8 square lattice. In this example, 22 microwave transmission lines 16 are provided for a total of 64 quantum bits Q arranged in an 8 × 8 square lattice.

[0027] In Figure 3, the numbers "0" to "21" arranged randomly in a square lattice pattern are label numbers i (in this case, i = any one of 0 to 21), and the same label number i is assigned to quantum bits Q connected to the same microwave transmission line 16. This label number i is an identifier for identifying quantum bits Q connected to the same microwave transmission line 16. When there is no need to distinguish between label numbers 0 to 21, they will hereinafter be simply referred to as label number i. Note that in Figure 3, each label number i is color-coded to enable visual identification.

[0028] Here, when using frequency-multiplexed pulses obtained by triply multiplexing three microwave pulses with different frequencies, the number of assigned identical label numbers i is three, and three quantum bits Q share one main line 18. However, in the example of Fig. 3 where a total of 64 quantum bits Q are provided, the last label number 21 is assigned only to the remaining one quantum bit Q. Therefore, a non-branched line extending from the main line 18, which is not branched from the main line 18, is connected to the quantum bit Q assigned the label number 21.

[0029] To explain this in detail using label number i, in Figure 3, there are three quantum bits Q assigned label number 0. These three quantum bits Q with label number 0 are connected to respective branch lines 19 branching off from one main line 18. Similarly, there are three quantum bits Q assigned label number 1. These three quantum bits Q with label number 1 are connected to respective branch lines 19 branching off from one main line 18 that is different from the main line 18 connected to the quantum bit Q with label number 0. In this way, different main lines 18 are connected to the quantum bit circuit 3 for each group of label numbers i with the same numerical value.

[0030] In addition to this configuration, in the quantum bit circuit 3, the label numbers i are arranged so that the numerical values ​​i of adjacent label numbers i are all different. Specifically, for example, four quantum bits Q101, Q102, Q103, and Q104 are arranged above, below, left, and right of quantum bit Q100, which is assigned label number 2. The four quantum bits Q101, Q102, Q103, and Q104 are assigned label numbers i that are different from the label number 2 of the central quantum bit Q100 and that are also different from each other.

[0031] 3 , quantum bit Q101 adjacent to the upper side of quantum bit Q100 is assigned label number 11, which is different from label number 2 of central quantum bit Q100 and label numbers 4, 7, and 13 of other quantum bits Q102, Q103, and Q104 adjacent to quantum bit Q100 to the lower, left, and right, respectively. As a result, quantum bit Q101 adjacent to the upper side of quantum bit Q100 is connected via branch line 19 to another main line 18 that is different from main lines 18 connected to central quantum bit Q100 and the other quantum bits Q102, Q103, and Q104 adjacent to quantum bit Q100 to the lower, left, and right, respectively.

[0032] 3, quantum bit Q102 adjacent to the lower side of quantum bit Q100 is also assigned label number 4, which is different from label number 2 of central quantum bit Q100 and label numbers 11, 7, and 13 of other quantum bits Q101, Q103, and Q104 adjacent to quantum bit Q100 above, to the left, and to the right, respectively. As a result, quantum bit Q102 adjacent to the lower side of quantum bit Q100 is connected via branch line 19 to another main line 18 that is different from main lines 18 connected to central quantum bit Q100 and the other quantum bits Q101, Q103, and Q104 adjacent to quantum bit Q100 above, to the left, and to the right, respectively.

[0033] Similarly, in Figure 3, quantum bit Q103 adjacent to quantum bit Q100 on the left side and quantum bit Q104 adjacent to quantum bit Q100 on the right side are each assigned a label number i that is different from the surrounding quantum bits Q, and the main lines 18 connected to each of them are also connected via branch lines 19 to main lines 18 that are different from the main lines 18 connected to the surrounding quantum bits Q.

[0034] In this way, in the quantum bit circuit 3, label numbers 0 to 21 are assigned so that a combination of adjacent label numbers i and the same combination of label numbers i do not exist anywhere else in the quantum bit circuit 3. Here, the same main line 18 is connected to each of the label numbers i with the same numerical value. In the quantum bit circuit 3, all of the combinations of adjacent label numbers i are different, and therefore all of the combinations of main lines 18 connected to adjacent quantum bits Q are different combinations.

[0035] In a quantum bit circuit 3 having such a configuration, it is possible to alternate between a time domain in which a one-qubit gate is applied to all quantum bits Q and a time domain in which a two-qubit gate is applied only to quantum bits Q that are adjacent to each other at a predetermined position among the quantum bits Q.

[0036] 4 is a schematic diagram illustrating the alternating execution of an operation of causing each quantum bit Q0 to Q7 to act on a one-qubit gate and an operation of causing adjacent quantum bits Q2, Q3, etc. to act on a two-qubit gate. For simplicity's sake, 1001 in FIG. 4 uses eight quantum bits Q0 to Q7 arranged in a row as an example. Note that when there is no need to particularly distinguish between the quantum bits Q0 to Q7, they are simply referred to as quantum bits Q. In practice, a frequency-multiplexed pulse in which multiple microwave pulses with different frequencies are multiplexed is input to each quantum bit Q0 to Q7. However, the following description focuses on the microwave pulse within the frequency-multiplexed pulse that operates each quantum bit Q0 to Q7.

[0037] In the example shown in 1001 of FIG. 4 , after a time region T1 in which a one-qubit gate (1Q gate) is applied to all of the quantum bits Q0 to Q7, there is a time region T2 in which a two-qubit gate (CZ (Controlled-Z) gate) is applied only to adjacent quantum bits Q2 and Q3 of the quantum bits Q0 to Q7. Then, there is a time region T3 in which a one-qubit gate is applied to all of the quantum bits Q0 to Q7 again, followed by a time region T4 in which a two-qubit gate is applied to adjacent quantum bits Q1 and Q2, adjacent quantum bits Q4 and Q5, and adjacent quantum bits Q6 and Q7 of the quantum bits Q0 to Q7. After that, there is a time region T5 in which a one-qubit gate is applied to all of the quantum bits Q0 to Q7 again. Similarly, thereafter, a time region in which a two-qubit gate is applied to the quantum bits Q0 to Q7 and a time region in which a one-qubit gate is applied to the quantum bits Q0 to Q7 are alternately executed. The following description will be focused on the time regions T1 to T5.

[0038] Here, different resonant frequencies for causing a one-qubit gate to operate are individually set in advance for the quantum bits Q0 to Q7, and therefore, in the time domains T1, T3, and T5 in which the one-qubit gate operates, microwave pulses of different resonant frequencies corresponding to the quantum bits Q0 to Q7 are input to each quantum bit Q0 to Q7 by frequency-multiplexed pulses.

[0039] Furthermore, non-resonant frequencies at which the two-qubit gate operates are individually set in advance for the quantum bits Q0 to Q7. Therefore, in the time domain T2 in which the two-qubit gate operates on the adjacent quantum bits Q2 and Q3, non-resonant pulses are input to the adjacent quantum bits Q2 and Q3 by frequency-multiplexed pulses. At this time, decoupling pulses of a predetermined frequency are input to the other quantum bits Q0, Q1, Q4 to Q7 in which the two-qubit gate is not operated by the frequency-multiplexed pulses, respectively, in order to cancel out unnecessary interactions (resident ZZ interactions) existing between the quantum bits.

[0040] In the figure, a quantum bit to which a decoupling pulse of a predetermined frequency is input to cancel out unwanted interactions is labeled A. A quantum bit to which a decoupling pulse of a different frequency than the decoupling pulse input to the quantum bit labeled A is input to cancel out unwanted interactions is labeled B. In the figure, a quantum bit to which a non-resonant pulse is input to operate a two-qubit gate is labeled C.

[0041] In the time domain T4 in which two-qubit gates are applied to adjacent quantum bits Q1 and Q2, adjacent quantum bits Q4 and Q5, and adjacent quantum bits Q6 and Q7, non-resonant pulses are input to these adjacent quantum bits Q1 and Q2, adjacent quantum bits Q4 and Q5, and adjacent quantum bits Q6 and Q7 by frequency-multiplexed pulses.

[0042] In this case, the pair of quantum bits Q4 and Q5 on which the two-qubit gate operates is adjacent to another pair of quantum bits Q6 and Q7 on which the two-qubit gate operates, and therefore, a non-resonant pulse of a different frequency from that of the adjacent pair of quantum bits Q6 and Q7 is input to the pair of quantum bits Q4 and Q5 by a frequency multiplexed pulse.

[0043] In the figure, quantum bits Q6 and Q7, which operate a two-qubit gate by inputting a non-resonant pulse different from the microwave pulse input to the pair of quantum bits Q4 and Q5 labeled C, are labeled D.

[0044] In addition, at this time, a decoupling pulse of a predetermined frequency is input to the other quantum bits Q0 and Q3, which do not have the two-qubit gate acted on, by a frequency-multiplexed pulse in order to cancel out unnecessary interactions (permanent ZZ interactions) that exist between the quantum bits.

[0045] To cancel unwanted interactions (permanent ZZ interactions) between quantum bits, decoupling pulses with different patterns are input for each quantum bit whose interactions are to be canceled. The decoupling pulse sequence has different sequence patterns depending on the label orders of labels A to D, as shown in 1002 of FIG. 4 . Such decoupling pulse sequences can be mechanically derived by arranging inverted pulses to realize the signs of the generalized Hadamard matrix of the label orders, as shown in 1003 of FIG. 4 (see, for example, D.W. Leung et al., Phys. A 61, 042310 (1999)).

[0046] <Example of Applying a One-Qubit Gate to All Quantum Bits> Next, an example of applying a one-qubit gate to all quantum bits Q will be described. For simplicity, the description will be given using nine quantum bits Q0 to Q8 arranged in a row, as shown in Fig. 5. Also, for simplicity, the description will be given using an example of using a frequency-multiplexed pulse in which two microwave pulses with different frequencies are double-multiplexed.

[0047] In this case, five microwave transmission lines 16-1 to 16-5 are provided for nine quantum bits Q0 to Q8. Microwave transmission line 16-1 has two branch lines 19-1 and 19-2 branching off from main line 18-1. One branch line 19-1 of microwave transmission line 16-1 is connected to quantum bit Q0, and the other branch line 19-2 is connected to another quantum bit Q6 that is not adjacent to quantum bit Q0.

[0048] Microwave transmission line 16-2 includes two branch lines 19-3 and 19-4 branching off from main line 18-2. One branch line 19-4 of microwave transmission line 16-2 is connected to quantum bit Q5, which is adjacent to quantum bit Q6 connected to microwave transmission line 16-1. Therefore, microwave transmission line 16-2 connects the other branch line 19-3 to quantum bit Q2, which is not adjacent to quantum bits Q0 and Q6 connected to microwave transmission line 16-1 and is not adjacent to quantum bit Q5 connected to one branch line 19-4.

[0049] Microwave transmission line 16-3 includes two branch lines 19-5 and 19-6 branching off from main line 18-3. One branch line 19-6 of microwave transmission line 16-3 is connected to quantum bit Q7, which is adjacent to quantum bit Q6 connected to microwave transmission line 16-1. Therefore, microwave transmission line 16-3 connects the other branch line 19-5 to quantum bit Q4, which is not adjacent to quantum bits Q0 and Q6 connected to microwave transmission line 16-1 and is not adjacent to quantum bit Q7 connected to one branch line 19-6.

[0050] Microwave transmission line 16-4 has two branch lines 19-7 and 19-8 branching off from main line 18-4. One branch line 19-7 of microwave transmission line 16-4 is connected to quantum bit Q1, which is adjacent to quantum bits Q0 and Q2 connected to microwave transmission lines 16-1 and 16-2, respectively. Therefore, microwave transmission line 16-4 connects the other branch line 19-8 to quantum bit Q8, which is not adjacent to quantum bits Q0, Q6, Q2, and Q5 connected to microwave transmission lines 16-1 and 16-2, respectively, and is not adjacent to quantum bit Q1 connected to one branch line 19-7. Microwave transmission line 16-5 has a non-branch line 20 of main line 18-5 connected to the remaining quantum bit Q3.

[0051] In this way, different main lines 18-1 and 18-4 are connected to adjacent quantum bits Q0 and Q1, respectively, and different main lines 18-4 and 18-2 are connected to adjacent quantum bits Q1 and Q2, respectively. Different main lines 18-2 and 18-5 are also connected to adjacent quantum bits Q2 and Q3, different main lines 18-5 and 18-3 are also connected to adjacent quantum bits Q3 and Q4, respectively, and different main lines 18-3 and 18-2 are also connected to adjacent quantum bits Q4 and Q5, respectively. Furthermore, different main lines 18-2 and 18-1 are also connected to adjacent quantum bits Q5 and Q6, respectively, different main lines 18-1 and 18-3 are also connected to adjacent quantum bits Q6 and Q7, respectively, and different main lines 18-3 and 18-4 are also connected to adjacent quantum bits Q7 and Q8, respectively.

[0052] The quantum bits Q0 to Q8 are set with different resonant frequencies from the adjacent quantum bits Q0, Q1 (Q1, Q2, ...). For example, a low frequency in a certain band is set as the resonant frequency for quantum bit Q0, and a one-qubit gate operates when a low-frequency microwave pulse PL is input. The quantum bit Q1 adjacent to quantum bit Q0 is set with a high frequency in a certain low band as the resonant frequency, and a one-qubit gate operates when a high-frequency microwave pulse PH is input. In this way, the other quantum bits Q2 to Q8 are also set with resonant frequencies different from the adjacent quantum bits Q3 to Q7.

[0053] Here, when a one-qubit gate is to act on quantum bits Q0 to Q8, a frequency-multiplexed pulse is generated so that a microwave pulse having a preset resonant frequency for each quantum bit Q0 to Q8 is input to each quantum bit Q0 to Q8. For example, a frequency-multiplexed pulse obtained by multiplexing a microwave pulse PL having a resonant frequency for quantum bit Q0 connected via branch line 19-1 and a microwave pulse PM having a resonant frequency for quantum bit Q6 connected via branch line 19-2 is input from microwave control device 12 to main line 18-1.

[0054] Both microwave pulses PL and PM are input as frequency-multiplexed pulses to quantum bit Q0 from branch line 19-1 branching off from main line 18-1. Two microwave pulses PL and PM are input to quantum bit Q0, but one of the microwave pulses PM is not at the resonant frequency of quantum bit Q0, so nothing happens when that microwave pulse PM is input to quantum bit Q0. The other microwave pulse PL is at the resonant frequency of quantum bit Q0, so a one-qubit gate operates in quantum bit Q0 due to the other microwave pulse PL.

[0055] Like quantum bit Q0, quantum bit Q6 also receives both microwave pulses PL and PM as frequency-multiplexed pulses from branch line 19-2 branching off from main line 18-1. Similarly, quantum bit Q6 also receives two microwave pulses PL and PM, but one of the microwave pulses PL is not at the resonant frequency of quantum bit Q6, so nothing happens when this microwave pulse PL is input to quantum bit Q6. The other microwave pulse PM is at the resonant frequency of quantum bit Q6, so a one-qubit gate operates in quantum bit Q6 due to the other microwave pulse PM.

[0056] In this way, in microwave transmission line 16-1, frequency-multiplexed pulses of microwave pulses PL and PM are input to main line 18-1, so that microwave pulses PL and PM having resonant frequencies corresponding to two quantum bits Q0 and Q6 can be simultaneously input via one main line 18-1. Therefore, one qubit gate can be simultaneously applied to two quantum bits Q0 and Q6 via one main line 18-1.

[0057] In addition, a frequency-multiplexed pulse that is a multiplexed combination of a microwave pulse PL, which is the resonant frequency of quantum bit Q2 connected via branch line 19-3, and a microwave pulse PH, which is the resonant frequency of quantum bit Q5 connected via branch line 19-4, is input from microwave control device 12 to main line 18-2.

[0058] Both microwave pulses PL and PH are input to quantum bit Q2 as frequency-multiplexed pulses from branch line 19-3 branching off from main line 18-2. Two microwave pulses PL and PH are input to quantum bit Q2, but one of the microwave pulses PH is not at the resonant frequency of quantum bit Q2, so nothing happens when this microwave pulse PH is input to quantum bit Q2. The other microwave pulse PL is at the resonant frequency of quantum bit Q2, so a one-qubit gate acts on quantum bit Q2 due to the other microwave pulse PL.

[0059] Like quantum bit Q2, quantum bit Q5 receives both microwave pulses PL and PH as frequency-multiplexed pulses from branch line 19-4 branching off from main line 18-2. Similarly, quantum bit Q5 receives two microwave pulses PL and PH, but one of the microwave pulses PL is not at the resonant frequency of quantum bit Q5, so nothing happens when this microwave pulse PL is input to quantum bit Q5. The other microwave pulse PH is at the resonant frequency of quantum bit Q5, so a one-qubit gate operates on quantum bit Q5 due to the other microwave pulse PH.

[0060] In this way, frequency-multiplexed pulses in which microwave pulses having resonant frequencies corresponding to the quantum bits Q4, Q7, Q1, and Q8 connected to the main lines 18-3 and 18-4 are multiplexed are input to the other microwave transmission lines 16-3 and 16-4 from microwave control device 12. In the example shown in Fig. 5, only quantum bit Q3 is connected to microwave transmission line 16-5, and therefore only microwave pulse PH having the resonant frequency of quantum bit Q3 is input.

[0061] <Example of Applying a Two-Qubit Gate to Adjacent Quantum Bits> Next, as shown in Figure 6, an example will be described in which a two-qubit gate is applied to adjacent quantum bits Q2 and Q3 and adjacent quantum bits Q4 and Q5, respectively, among quantum bits Q0 to Q8. In this case, a non-resonant pulse (a microwave pulse with a predetermined non-resonant frequency that enhances the interaction) Psi1 is input to adjacent quantum bits Q2 and Q3 to enhance the interaction between quantum bits Q2 and Q3. In addition, another non-resonant pulse Psi2 is input to adjacent quantum bits Q4 and Q5 to enhance the interaction between quantum bits Q4 and Q5.

[0062] In this case, a frequency-multiplexed pulse, which is a multiplexed combination of the non-resonant pulse Psi1 of quantum bit Q2 connected via branch line 19-3 and the non-resonant pulse Psi2 of quantum bit Q5 connected via branch line 19-4, is input from microwave control device 12 to main line 18-2.

[0063] Furthermore, a non-resonant pulse Psi1 of quantum bit Q3 connected via non-branch line 20 is input from microwave control device 12 to main line 18-5. Furthermore, a frequency-multiplexed pulse obtained by multiplexing a non-resonant pulse Psi2 of quantum bit Q4 connected via branch line 19-5 and a decoupling pulse (described later) of a predetermined frequency that cancels out interactions between quantum bit Q7 connected via branch line 19-6 and the surrounding quantum bits Q6 and Q8 is input from microwave control device 12 to main line 18-3.

[0064] Decoupling pulses of a predetermined frequency that cancel out interactions with adjacent quantum bits Q1, Q6, etc. are also input to the other quantum bits Q0, Q1, Q6, and Q8 from the corresponding main lines 16-1 and 16-4.

[0065] In this case, both non-resonant pulses Psi1 and Psi2 are input to quantum bit Q2 as frequency-multiplexed pulses from branch line 19-3 branching off from main line 18-2. Two different non-resonant pulses Psi1 and Psi2 are input to quantum bit Q2, but one of the non-resonant pulses, Psi2, is not at the resonant frequency of quantum bit Q2 or the non-resonant frequency that activates the two-qubit gate, so nothing happens when non-resonant pulse Psi2 is input to quantum bit Q2.

[0066] The other non-resonant pulse Psi1 has a non-resonant frequency that causes a two-qubit gate to act on the quantum bit Q2. Therefore, when the same non-resonant pulse Psi1 is input from the main line 18-5 to the quantum bit Q3 adjacent to the quantum bit Q2, the interaction between the adjacent quantum bits Q2 and Q3 is enhanced. This causes a two-qubit gate to act on the quantum bits Q2 and Q3.

[0067] Similar to quantum bit Q2, quantum bit Q5 also receives both non-resonant pulses Psi1 and Psi2 as frequency-multiplexed pulses from branch line 19-4 branching off from main line 18-2. Similarly, quantum bit Q5 also receives two different non-resonant pulses Psi1 and Psi2, but one of the non-resonant pulses, Psi1, is not at the resonant frequency of quantum bit Q5 or the non-resonant frequency that causes a two-qubit gate to act on quantum bit Q5, so nothing happens when non-resonant pulse Psi1 is input to quantum bit Q5.

[0068] The other non-resonant pulse Psi2 has a non-resonant frequency that causes a two-qubit gate to act on the quantum bit Q5. Therefore, when the same non-resonant pulse Psi2 is input from the main line 18-3 to the quantum bit Q4 adjacent to the quantum bit Q5, the interaction between the adjacent quantum bits Q4 and Q5 is enhanced. This causes a two-qubit gate to act on the quantum bits Q4 and Q5.

[0069] 7 is a schematic diagram showing an example of grouping of adjacent quantum bits Q arranged at predetermined positions in a quantum bit circuit 3 in which 64 quantum bits Q are arranged in an 8 × 8 square lattice, in a time domain in which a two-qubit gate is applied to adjacent quantum bits Q. In FIG. 7, the connection between adjacent quantum bits Q is represented by an edge Ed.

[0070] 7, quantum bits that cancel interactions are represented by labels A and B, and quantum bits that enhance interactions are represented by labels C and D. In the quantum bit circuit 3, as shown in the above-mentioned FIG. 3, label numbers 0 to 21 are assigned to each quantum bit Q, and different main lines 18 are connected to each of the label numbers 0 to 21 via branch lines 19.

[0071] In the quantum bit circuit 3, frequency-multiplexed pulses including decoupling pulses with different frequencies for the labels A and B are input from main lines 18 assigned to the positions of the labels A and B according to the label numbers 0 to 21 to the quantum bits Q at the positions of the labels A and B via branch lines 19. In the quantum bit circuit 3, different labels A and B are assigned to adjacent quantum bits Q that cancel out the interaction, and therefore the labels A and B are arranged in a checkerboard pattern.

[0072] Furthermore, in the quantum bit circuit 3, frequency-multiplexed pulses including a different non-resonant pulse for each of the labels C and D are input from main lines 18 assigned to the positions of the labels C and D according to the label numbers 0 to 21, respectively, to the quantum bits at the positions of the labels C and D via branch lines 19. When label C is assigned to adjacent quantum bits that cause a two-qubit gate to operate, a label D different from label C is assigned to another adjacent quantum bit that is arranged adjacent to the label C and causes a two-qubit gate to operate.

[0073] <Example of microwave pulse frequencies when frequency multiplexing> Next, the frequencies of microwave pulses when frequency multiplexing will be described. FIG. 8 shows an example in which six frequencies are multiplexed within 1330 MHz. The g frequency (denoted as "g" in FIG. 8) is the resonance frequency between the ground state (g level) and the first excited state (e state) of transmons. The ef frequency (denoted as "ef" in FIG. 8) is the resonance frequency between the first excited state (e level) and the second excited state (f level) of transmons. The gf frequency (denoted as "gf" in FIG. 8) is the resonance frequency between the ground state (g level) and the second excited state (f level) of transmons.

[0074] The frequency of the microwave pulse input to each quantum bit to cause a one-qubit gate to act on the quantum bit is one of multiple Ge frequencies. When this microwave pulse is input to the quantum bit, it is desirable to prevent any influence on resonant frequencies other than the target Ge frequency (suppress non-resonant excitation). To achieve this, it is desirable to make the frequency difference between the Ge frequency and other resonant frequencies as large as possible. On the other hand, in some cases, it is also possible to multiplex microwave pulses of many different frequencies into one frequency-multiplexed pulse.

[0075] In this embodiment, an example is shown in which six sets are provided, each consisting of three frequencies, g e , g f , and ef . By arranging these sets of g e , g f , and ef frequencies apart from each other, non-resonant excitation can be suppressed. In this case, in order to increase the number of multiplexed microwave pulses in one frequency-multiplexed pulse, it is desirable to set the resonant frequencies of multiple quantum bits so that the ef frequency is located between adjacent g e frequencies.

[0076] For example, if the ef frequency located midway between adjacent g e frequencies is set to be detuned from the g e frequency by 130 MHz to 150 MHz, the g e frequency (= the driving microwave frequency for a 1-qubit gate) can be positioned at an appropriate distance from any resonant frequency.

[0077] In the example of Fig. 8, adjacent g e frequencies of the microwave pulses to be multiplexed are detuned by 266 MHz. Also, in the example of Fig. 8, the anharmonicity is set to -400 MHz, and the ef frequency, which is intermediate between the g e and g e frequencies of the microwave pulse, is detuned by 133 MHz from the g e frequency ((400 MHz / 3) ≒ 133 MHz). Also, the g f frequency, which is intermediate between the g e and ef frequencies of the microwave pulse, is detuned by 66 MHz from the ef frequency.

[0078] In this way, the frequencies of the multiple microwave pulses supplied to the main line 18 as frequency-multiplexed pulses are detuned so as to avoid frequency collisions between adjacent or nearby quantum bits Q. The frequencies of the microwave pulses to be multiplexed as frequency-multiplexed pulses are desirably selected from within the frequency band of 1330 MHz.

[0079] <Wiring Allocation Method for Allocating Microwave Transmission Lines to Multiple Quantum Bits> Next, a wiring allocation method for allocating different microwave transmission lines 16 to adjacent quantum bits Q in a quantum bit circuit 3 having multiple quantum bits Q as shown in Fig. 3 will be described. Fig. 9 is a block diagram showing the circuit configuration of a processing device 200 that presents the calculation result of allocating different microwave transmission lines 16 to adjacent quantum bits Q.

[0080] The arithmetic processing device 200 includes a control unit 41 configured as a microcomputer including a CPU (Central Processing Unit), RAM (Random Access Memory), ROM (Read Only Memory), etc. The arithmetic processing device 200 has a configuration in which an input unit 42, a display unit 43, a memory unit 44, a label placement unit 45, a rearrangement count determination unit 46, and a redo count determination unit 47 are connected to the control unit 41 via a bus B. The control unit 41 comprehensively controls various functions of the arithmetic processing device 200 by loading various programs, such as a basic program and a wiring assignment processing program, stored in the memory unit 44 into the RAM and starting them up.

[0081] The arithmetic processing device 200 executes frequency multiplexing wiring assignment processing based on information input by the user via the input unit 42, and displays the obtained calculation results on the display unit 43. Based on the calculation results displayed on the display unit 43, the user can design a configuration in which a predetermined number of wirings (cable_num) of microwave transmission lines 16 are connected to each quantum bit Q of the quantum bit circuit 3. The user can manufacture the quantum computer 1 according to this embodiment based on the obtained design.

[0082] Next, the procedure for the frequency multiplexing wiring assignment process will be described below with reference to the flowchart shown in Fig. 10. When the control unit 41 starts the wiring assignment process as shown in Fig. 10, the process proceeds from the start step to step S1. In step S1, when the user operates the input unit 42 to input a designation of a lattice that determines the arrangement of the quantum bits Q, the arithmetic processing device 200 proceeds to the next step S2.

[0083] Here, the specification of the lattice refers to the specification of a square lattice in which to arrange the quantum bits Q, such as an 8×8 square lattice as shown in Fig. 3 or a 16×16 square lattice as shown in Fig. 11. For example, the user inputs a total number of nodes Bn, with the quantum bits to be arranged as nodes, from the input unit 42. The label arrangement unit 45 defines the wiring connecting adjacent quantum bits Q as edges Ed from the total number of nodes Bn input from the input unit 42.

[0084] In step S2, when the user inputs a frequency multiplexing number (mux_num) that determines the number of multiplexed frequencies through an input operation of the input unit 42, the arithmetic processing device 200 proceeds to the next step S3. In the example shown in Fig. 3, triple multiplexing, in which three frequencies are multiplexed, is specified.

[0085] In step S3, the label placement unit 45 defines the number of wirings, which is the number of microwave transmission lines 16, as (cable_num), and defines the number of wirings (cable_num) as the smallest integer not less than the total number of nodes Bn divided by the frequency multiplexing number (mux_num) based on the total number of nodes (total number of quantum bits Q) Bn of the lattice input in step S1 and the frequency multiplexing number (mux_num) input in step S2, and then proceeds to the next step S4.

[0086] In step S4, the retry count determination unit 47 sets the retry count (counter_1) to 0 (counter_1=0), and proceeds to the next step S5. In step S5, the label placement unit 45 sets the label number i to 0 (i=0), and proceeds to the next step S6. In step S6, the rearrangement count determination unit 46 sets the random count (counter_2) to 0 (counter_2=0), and proceeds to the next step S7.

[0087] In step S7, the label placement unit 45 randomly places the label number i of the frequency multiplexing number (mux_num) on the nodes in the lattice, and then proceeds to the next step S8. Note that when the label placement unit 45 randomly places the label number i on the nodes in the lattice, it does so in a way that the same label number i is not placed on the nodes at both ends of each edge Ed of the lattice. Specifically, it places the labels so that label number 0 is not both placed on the nodes at both ends of each edge Ed of the lattice (i.e., nodes with label number 0 are not adjacent).

[0088] In step S8, the label placement unit 45 determines whether the combinations of the two label numbers i assigned to the nodes at both ends of each edge Ed of the lattice are all different. Note that initially, since the label number i is set to 0 in step S5, there are no other label numbers i other than label number 0 in the lattice, and therefore a positive result is obtained in step S8.

[0089] If a positive result is obtained in step S8, this means that the combinations of the numerical values ​​of the two label numbers i assigned to the nodes at both ends of each edge Ed of the lattice are all different, and in this case the label placement unit 45 proceeds to the next step S9.

[0090] In step S9, the label placement unit 45 determines whether the label number i is equal to or greater than the number of wirings (cable_num). If a negative result is obtained in step S9, this indicates that the label number i is equal to or less than the number of wirings (cable_num), that is, that all of the label numbers i (label numbers 0 to 21 in the example shown in FIG. 3) corresponding to the total number of microwave transmission lines 16 to be assigned have not been assigned to nodes, and in this case the label placement unit 45 proceeds to the next step S10.

[0091] In step S10, the label placement unit 45 increments the value of the label number i by 1, returns to step S6, sets the random number (counter_2) to 0 (counter_2=0), and proceeds to the next step S7.

[0092] In step S7, the label placement unit 45 randomly places the label number i of the frequency multiplexing number (mux_num) (here, label number 1, which was incremented by 1 from label number 0 in step S10) on the nodes in the lattice, and then proceeds to the next step S8. When randomly placing label number 1 on the nodes in the lattice, the label placement unit 45 does so in a manner similar to that for label number 0, such that the same label number 1 is not placed on the nodes at both ends of each edge Ed of the lattice. As a result, in the example shown in FIG. 3, since the frequency multiplexing number (mux_num) is 3, three each of label numbers 0 and 1 are randomly placed in the lattice.

[0093] In step S8, the label placement unit 45 determines whether the combinations of the two label numbers i (label numbers 0 and 1) assigned to the nodes at both ends of each edge Ed of the lattice are all different. Specifically, when label number 1 is placed on a node adjacent to label number 0 already placed in the lattice, the label placement unit 45 determines whether any other nodes with adjacent label numbers 0 and 1 have been placed.

[0094] If a positive result is obtained in step S8, this indicates that the combinations of the values ​​of the two label numbers i assigned to the nodes at both ends of each edge of the lattice are all different, and in this case the label placement unit 45 proceeds to the next step S9. Specifically, if a positive result is obtained in step S8, this indicates that when label number 1 is placed on a node adjacent to label number 0 that has already been placed in the lattice, there are no other nodes placed with label numbers 0 and 1 adjacent to each other, and in this case the label placement unit 45 proceeds to the next step S9.

[0095] In step S9, the label placement unit 45 determines whether the label number i is equal to or greater than the number of wirings (cable_num). If a negative result is obtained in step S9, this indicates that the label number i is equal to or less than the number of wirings (cable_num), that is, that all of the label numbers i (label numbers 0 to 21) corresponding to the total number of microwave transmission lines 16 to be assigned have not been assigned to nodes, and in this case the label placement unit 45 proceeds to the next step S10.

[0096] Thereafter, if a positive result continues to be obtained in step S8 and a negative result continues to be obtained in step S9, the above-described process is repeated until a positive result is obtained in step S9. If a positive result is obtained in step S9, it indicates that the label number i is the number of wires (cable_num), that is, that all label numbers i (label numbers 0 to 21) corresponding to the total number of microwave transmission lines 16 to be assigned have been assigned to nodes, and at this time the arithmetic processing device 200 determines that the desired wiring assignment (i.e., frequency multiplexing assignment) has been found for all nodes, and terminates the wiring assignment process.

[0097] Next, a case where a negative result is obtained in step S8 described above will be described. If a negative result is obtained in step S8, this indicates that there is a location where the combination of the numerical values ​​of the two label numbers i assigned to the nodes at both ends of each edge Ed of the lattice is the same, and in this case, the label placement unit 45 proceeds to the next step S11. For example, when label numbers 0 and 1 are randomly placed in the lattice, if a negative result is obtained in step S8, this indicates that there are multiple nodes in the lattice where label numbers 0 and 1 are adjacent, and in this case, the label placement unit 45 proceeds to the next step S11.

[0098] In step S11, the rearrangement count determination unit 46 determines whether the random count (counter_2) is equal to or greater than N. Here, N indicates the upper limit number of times that only the label number i randomly placed in the most recent step S7 is erased and the erased label number i is randomly rearranged again in the grid. The upper limit number N is an integer preset in the rearrangement count determination unit 46, and is preferably several times the total number of nodes.

[0099] If a negative result is obtained in step S11, this indicates that the random count (counter_2) is less than the upper limit count N, and in this case the rearrangement count determination unit 46 proceeds to the next step S12. In step S12, the rearrangement count determination unit 46 increments the random count (counter_2) by 1 and proceeds again to step S7. In step S7, the label placement unit 45 again randomly places the label number i (here, label number 1) of the frequency multiplexing number (mux_num) on the nodes in the lattice, and repeats the above-described processing until a positive result is obtained in the next step S8.

[0100] On the other hand, if a positive result is obtained in step S11, this indicates that the random number (counter_2) has reached the upper limit number N, and at this time the rearrangement number determination unit 46 proceeds to the next step S13.

[0101] In step S13, the retry count determination unit 47 determines whether the number of retry attempts (counter_1) is equal to or greater than M. Here, M indicates the upper limit of the number of retry attempts for erasing all label numbers i that have been placed in the lattice nodes up to that point and randomly re-arranging label numbers i to the lattice again, starting with label number 0. The upper limit of the number of retry attempts M is an arbitrary integer that is preset in the retry count determination unit 47.

[0102] If a negative result is obtained in step S13, this indicates that the number of retry attempts (counter_1) is less than the upper limit number of retry attempts M, and in this case the number of retry attempts determination unit 47 proceeds to the next step S14.

[0103] In step S14, the retry count determination unit 47 increments the retry count (counter_1) by 1, erases all label numbers i that have been placed on nodes in the lattice, and proceeds to step S5 again. In step S5, the label placement unit 45 sets the label number i to 0 (i=0), proceeds to the next step S6, and repeats the above-mentioned process from the beginning.

[0104] On the other hand, if a positive result is obtained in step S13, this indicates that the number of retries (counter_1) has reached the upper limit number of retries M, and at this time, the arithmetic processing unit 200 determines that the desired frequency multiplexing allocation has not been made to all nodes and the wiring allocation has failed, and terminates the wiring allocation process.

[0105] Regarding how to arrange the resonant frequencies of each quantum bit Q in the quantum bit circuit 3, under the constraints that the resonant frequencies of adjacent quantum bits Q must not collide with each other and that microwave pulses of the same resonant frequency must not be input to one main line 18, how to arrange the resonant frequencies of each quantum bit Q can be calculated by arithmetic processing.

[0106] As described above, in the quantum computer 1 according to this embodiment, when the identifier for identifying quantum bits Q connected to the same main line 18 is defined as label number i (i is an integer), the combinations of the values ​​of the two label numbers i assigned to the quantum bits Q at both ends of the edge Ed representing the connection between adjacent quantum bits Q are all different, and the relationship between the label number i and the number of wires (cable_num) of the main line 18 satisfies the relationship label number i ≧ number of wires (cable_num).

[0107] <Simulation> A simulation was performed to determine whether wiring assignment could be performed by changing the frequency multiplexing number and the total number of nodes Bn in the square lattice according to the wiring assignment processing procedure described above. As a result, the results shown in Figure 12 were obtained. In Figure 12, the horizontal axis shows the number of nodes arranged in one direction (either the vertical or horizontal direction) out of the total number of nodes (total number of quantum bits) Bn arranged in a square lattice, and the vertical axis shows the frequency multiplexing number (mux_num).

[0108] In the simulation, the number of nodes in one direction was set to 4 to 16, and the frequency multiplexing number (mux_num) was set to 2 to 5. The number of wirings (cable_num) was defined as the smallest integer not less than the total number of nodes Bn ÷ the frequency multiplexing number (mux_num). If the solution of Bn ÷ (mux_num) is not an integer solution, one of the main lines 18 of the number of wirings (cable_num) is connected to the quantum bit via a non-branch line. The upper limit number N of the random number (counter_2) was set to 1000 times. The upper limit number M of the number of retries (counter_1) was set to 1000 times.

[0109] As a result of the simulation, as shown in Fig. 12, in configurations in which nodes are arranged in a 4x4, 5x5, and 6x6 square lattice pattern, label number i is optimally assigned if the number of frequency multiplexing is up to 2. Similarly, in configurations in which nodes are arranged in a 14x14, 15x15, and 16x16 square lattice pattern, label number i is optimally assigned if the number of frequency multiplexing is up to 5.

[0110] <Actions and Effects> As described above, the quantum computer 1 according to this embodiment includes a quantum bit circuit 3 in which a plurality of quantum bits Q having fixed resonant frequencies are arranged, a microwave control device 12 that outputs a plurality of frequency-multiplexed pulses obtained by multiplexing a plurality of microwave pulses having different frequencies, and a plurality of microwave transmission lines 16 that input the frequency-multiplexed pulses to the quantum bits. The microwave transmission line 16 includes a main line 18 to which the frequency-multiplexed pulses are input from the microwave control device 12, and a plurality of branch lines 19 branching from the main line 18.

[0111] In addition, each main line 18 is shared by two or more quantum bits Q, and a plurality of branch lines 19 provided for each main line 18 are connected to different quantum bits Q. Adjacent quantum bits Q are connected to branch lines 19 from different main lines 18.

[0112] In this way, in the quantum computer 1, by having multiple quantum bits Q share one main line 18, it is possible to reduce the number of microwave transmission lines 16 provided in the microwave control device 12 compared to conventional methods, and it is also possible to reduce the number of wires from the microwave control device 12 compared to conventional methods. Furthermore, because the number of wires from the microwave control device 12 can be reduced, the burden on the microwave control device 12 can be reduced accordingly compared to conventional methods.

[0113] Furthermore, when applied to a superconducting quantum computer, one main line 18 can be shared by multiple quantum bits Q, thereby reducing the number of wires (cable_num) of the main line 18 to about "one-fifth of the multiplexing number," thereby resolving the lack of space within the dilution refrigerator and mitigating the problem of heat inflow.

[0114] Here, when microwave control device 12 causes a one-qubit gate to act on quantum bit Q, it generates, as a first frequency-multiplexed pulse, a frequency-multiplexed pulse that multiplexes a plurality of resonance pulses having different resonance frequencies, which cause a one-qubit gate to act on each quantum bit Q. Then, by having the above-mentioned wiring configuration, quantum computer 1 can input a unique resonance pulse to each quantum bit Q by a frequency-multiplexed pulse from main line 18 via each branch line 19, and cause a one-qubit gate to act on each quantum bit.

[0115] Furthermore, when the microwave control device 12 operates a two-qubit gate on adjacent quantum bits Q, it generates one of the following: (i) a frequency-multiplexed pulse obtained by multiplexing a plurality of non-resonant pulses of different non-resonant frequencies that enhance the interaction in adjacent quantum bits; (ii) a frequency-multiplexed pulse obtained by multiplexing one or more non-resonant pulses of non-resonant frequencies that enhance the interaction in adjacent quantum bits and one or more decoupling pulses of frequencies that cancel out the interaction occurring in adjacent quantum bits; or (iii) a frequency-multiplexed pulse obtained by multiplexing a plurality of decoupling pulses of different frequencies that cancel out the interaction occurring in adjacent quantum bits.

[0116] In the quantum computer 1, frequency-multiplexed pulses including non-resonant pulses that enhance the interaction are input to, for example, two different main lines 18 connected to a pair of adjacent quantum bits Q whose interaction is to be enhanced, from among the multiple main lines 18. This enables the quantum computer 1 to enhance the interaction between the pair of adjacent quantum bits and operate a two-qubit gate.

[0117] In this case, the quantum computer 1 inputs frequency-multiplexed pulses including decoupling pulses that cancel out the interaction to the main lines 18 connected to the other quantum bits whose interactions are desired to be canceled, other than the adjacent quantum bit whose interaction has been enhanced. This allows the quantum computer 1 to enhance the interaction only of the quantum bit Q at the desired position where the interaction is desired to be enhanced, and to cancel out other unwanted interactions.

[0118] In the field of quantum computing, a method using a cross-resonant gate to operate a two-qubit gate is generally known. In a typical cross-resonant gate method, for example, to avoid frequency collisions where four qubits are arranged in a square lattice, the frequency of the microwave pulse between adjacent qubits must be detuned by approximately 850 MHz. In this case, the cross-resonant gate requires microwave power of 20 to 30 MHz higher than when a one-qubit gate is operated, which places a greater burden on the microwave control device. The requirement for such high output power from the microwave control device simultaneously requires conflicting requirements: a large dynamic range, low crosstalk, and low noise floor, resulting in increased packaging volume and power consumption for the quantum computer.

[0119] In the quantum computer 1 according to this embodiment, the two-qubit gate acts on adjacent qubits by enhancing the permanent ZZ interaction without using a cross-resonant gate that has a wide influence, and therefore the conditions for frequency collision can be alleviated compared to a cross-resonant gate. Therefore, even in places where qubits are arranged in a square lattice, a decrease in yield can be suppressed, and it is expected that the yield of large-scale qubit chips exceeding 1000 qubits will be improved.

[0120] In the quantum computer 1 according to this embodiment, in order to prevent a resonant pulse input to a certain quantum bit from affecting other quantum bits using frequency-multiplexed pulses, microwave pulses of different frequencies detuned by, for example, 266 MHz or more are multiplexed onto one main line 18. Therefore, the quantum computer 1 does not require the large microwave power required for a cross resonant gate, which can contribute to the miniaturization and power saving of the microwave control device 12. Furthermore, since a cross resonant gate is not used, the influence of quantum crosstalk caused by a strong non-resonant pulse in the cross resonant gate can also be reduced.

[0121] According to the above configuration, in the quantum computer 1 of this embodiment, the number of wires in the dilution refrigerator can be reduced, the burden on the microwave control device 12 can be alleviated, and the probability of frequency collisions can be reduced, thereby enabling larger-scale integration of quantum bits than before. Furthermore, in the quantum computer 1 of this embodiment, the burden on the microwave control device 12 can be reduced, so that an increase in the capacity of the microwave control device 12 can be suppressed and the power consumption of the microwave control device 12 can be reduced compared to before.

[0122] Other Embodiments The present invention is not limited to the above-described embodiments. Various modifications are possible within the spirit and scope of the present invention. Other embodiments and modifications made by those skilled in the art are also included in the present invention. In the above-described embodiments, the arrangement of the multiple quantum bits Q has been described as an arrangement in which the multiple quantum bits Q are arranged in a square lattice pattern. However, the present invention is not limited to this. Other arrangements of the quantum bits may include, for example, an arrangement in which the multiple quantum bits Q are arranged in a line, or an arrangement in which the multiple quantum bits Q are arranged in a honeycomb pattern (e.g., IBM's "IBM_Sherbrooke"). Even in such an arrangement, wiring can be assigned according to the above-described wiring assignment process ( FIG. 10 ).

[0123] Furthermore, for example, in addition to superconducting quantum computers, the present invention can be applied to various quantum computers, such as semiconductor quantum dot quantum computers that can be controlled with microwave pulses, quantum computers using cooled atomic gas systems that can be controlled with microwave pulses, and quantum computers that arrange electron spins using molecular technology or nanotechnology.

[0124] Furthermore, in the above-described embodiment, a case has been described in which, when enhancing the interaction between adjacent quantum bits Q, microwave pulses of the same non-resonant frequency are input as non-resonant pulses to the adjacent quantum bits Q, but the present invention is not limited to this. As the non-resonant pulse, microwave pulses of different non-resonant frequencies may be input as non-resonant pulses to the adjacent quantum bits Q, as long as the interaction between the adjacent quantum bits Q can be enhanced.

[0125] Furthermore, in the above-described embodiment, a case has been described in which a fixed-frequency quantum bit (a quantum bit with a fixed resonant frequency) in which the frequency of the microwave pulse is fixed is applied, but the present invention is not limited to this, and a variable-frequency quantum bit (a quantum bit with a variable resonant frequency) in which the resonant frequency of the microwave pulse is variable may also be applied. In a quantum bit circuit in which such variable-frequency quantum bits are arranged, similar to the above-described embodiment, by having multiple quantum bits Q share one main line 18, it is possible to reduce the number of microwave transmission lines 16 provided in microwave control device 12 compared to the conventional case, and accordingly the burden on microwave control device 12 can be reduced compared to the conventional case.

[0126] In the above-described embodiment, the connection portion where the plurality of branch lines 19 branching from the main line 18 are connected to the main line 18 may be provided inside a cryogenic refrigerator provided with the quantum bit circuit 3, or the connection portion may be provided outside the refrigerator. In a configuration where the connection portion where the plurality of branch lines 19 branching from the main line 18 are connected to the main line 18 is provided inside a cryogenic refrigerator provided with the quantum bit circuit 3, the wiring connecting the control device 2 to the refrigerator becomes the main line 18, and the number of wirings can be reduced. On the other hand, in a configuration where the connection portion where the plurality of branch lines 19 branching from the main line 18 are connected to the main line 18 is provided outside a cryogenic refrigerator provided with the quantum bit circuit 3, an existing refrigerator provided with the quantum bit circuit 3 inside can be reused, and the plurality of wirings from the quantum bit circuit 3 drawn out to the outside of the refrigerator can be used as branch lines 19.

[0127] 1 quantum computer 3 quantum bit circuit 12 microwave control device 16 microwave transmission line 18 main line 19 branch line Q quantum bit

Claims

1. A quantum computer comprising: a quantum bit circuit having a plurality of quantum bits; a microwave control device that outputs a plurality of frequency-multiplexed pulses obtained by multiplexing a plurality of microwave pulses having different frequencies; and a plurality of microwave transmission lines that input the frequency-multiplexed pulses to the quantum bits, wherein the microwave transmission lines comprise: a main line to which the frequency-multiplexed pulses are input from the microwave control device; and a plurality of branch lines branching off from the main line, wherein each of the main lines is shared by two or more of the quantum bits, and the plurality of branch lines provided for each main line are each connected to a different quantum bit, and the branch lines from the different main lines are each connected to adjacent quantum bits.

2. The quantum computer according to claim 1, wherein one of the main lines is shared by a number of the plurality of quantum bits equal to the number of frequency multiplexing of the microwave pulses to be multiplexed.

3. The quantum computer of claim 1, wherein the number of wirings (cable_num) of the main lines is the smallest integer not less than Bn÷(mux_num), where Bn is the total number of quantum bits and (mux_num) is the frequency multiplexing number, and when the solution of Bn÷(mux_num) is not an integer solution, at least one of the main lines of the number of wirings (cable_num) is connected to the quantum bit via a non-branch line.

4. The quantum computer described in claim 1, wherein the main lines connected to the quantum bits are each input with a frequency-multiplexed pulse including, as a resonant pulse, a microwave pulse having a resonant frequency that causes a one-qubit gate to act on the quantum bit.

5. The quantum computer of claim 1, wherein the frequency-multiplexed pulses including the microwave pulses of a non-resonant frequency that enhances the interaction as a non-resonant pulse are input to the different main lines connected to adjacent quantum bits for which it is desired to enhance the interaction.

6. The quantum computer of claim 1, wherein the frequency-multiplexed pulse including the microwave pulse of a frequency that cancels the interaction as a decoupling pulse is input to each of the main lines connected to the quantum bits for which it is desired to cancel the interaction.

7. A quantum computer as described in claim 1, wherein the frequency-multiplexed pulse including, as a non-resonant pulse, the microwave pulse of a non-resonant frequency that enhances the interaction is input to different main lines respectively connected to adjacent quantum bits whose interaction is to be enhanced, and the frequency-multiplexed pulse including, as a decoupling pulse, the microwave pulse of a frequency that cancels the interaction is input to main lines respectively connected to other quantum bits whose interaction is to be canceled other than the quantum bit whose interaction is to be enhanced.

8. The quantum computer of claim 1, wherein the microwave control device generates any of a frequency-multiplexed pulse obtained by multiplexing a plurality of non-resonant pulses of different non-resonant frequencies that enhance the interaction in the adjacent quantum bits, a frequency-multiplexed pulse obtained by multiplexing one or more non-resonant pulses of non-resonant frequencies that enhance the interaction in the adjacent quantum bits and one or more decoupling pulses of frequencies that cancel out the interaction occurring in the adjacent quantum bits, and a frequency-multiplexed pulse obtained by multiplexing a plurality of decoupling pulses of different frequencies that cancel out the interaction occurring in the adjacent quantum bits, and the microwave transmission line inputs the frequency-multiplexed pulse including the non-resonant pulse to the adjacent quantum bit whose interaction is to be enhanced, and inputs the frequency-multiplexed pulse including the decoupling pulse to the other quantum bit whose interaction is to be canceled.

9. The quantum computer of claim 1, wherein the multiple microwave pulses supplied to the main line as the frequency-multiplexed pulses are detuned in frequency so as to avoid frequency collisions between adjacent or nearby quantum bits.

10. The quantum computer according to claim 9, wherein the frequency of the microwave pulses to be multiplexed into the frequency-multiplexed pulse is selected from a frequency band of 1330 MHz.

11. A quantum computer as described in claim 1, wherein the frequency of the microwave pulse to be multiplexed into the frequency-multiplexed pulse is selected based on the constraint that the resonant frequencies of adjacent quantum bits must not collide with each other, and the constraint that microwave pulses of the same resonant frequency must not be input to one of the main lines.

12. A quantum computer as described in claim 1 or 8, wherein the quantum bit circuit has a time domain in which a one-qubit gate is applied to the quantum bit based on the frequency-multiplexed pulse input from the microwave transmission line, and a time domain in which a two-qubit gate is applied only to the quantum bit adjacent to the quantum bit at a predetermined position.

13. The quantum computer of claim 1, wherein when an identifier for identifying the quantum bits connected to the same main line is a label number i (i is an integer), the combinations of the values ​​of the two label numbers i assigned to the quantum bits at both ends of an edge representing a connection between adjacent quantum bits are all different, and the relationship between the label number i and the number of wirings (cable_num) of the main line satisfies the relationship: label number i ≧ number of wirings (cable_num).

Images