A method for automatically calibrating DC flux crosstalk based on superconducting quantum chips

Abstract

The present invention discloses a method for automatically calibrating DC magnetic flux crosstalk based on superconducting quantum chips, which specifically relates to the field of chip technology, and includes the following steps: initially setting the pre-crosstalk matrix as a unit matrix, denoted as P(0)=I, and marking the bits according to whether there is a read resonant cavity; scanning the modulation curves of all read resonant cavities with magnetic flux bias. The present invention analyzes in detail the influence of magnetic flux crosstalk on bit frequency, which can be used as a guide for calibrating magnetic flux crosstalk, so that calibration with arbitrary precision can be achieved in theory; multiple iterations are used to gradually approach the required precision, which has good robustness and applicability; the present invention analyzes for the first time the influence of the coupler not carrying a read resonant cavity on magnetic flux crosstalk calibration, and shows that our invention can handle this situation well.

Description

Technical Field

[0001] The present invention relates to the field of chip technology, and in particular to a method for automatically calibrating direct current magnetic flux crosstalk based on a superconducting quantum chip. Background Art

[0002] As the scale of superconducting quantum chips expands, there are more and more various lines on the chip. This includes drive lines, readout lines, and flux bias lines that usually carry AC fields. Because the line isolation technology is immature, the fields on each line will overlap with each other, forming the so-called classical crosstalk. This will destroy the addressability of the quantum bit, making it impossible for us to accurately and independently manipulate each bit, which is a severe challenge for large-scale quantum computing. For this reason, it is necessary to seek a technology that can suppress these classical crosstalks in manipulation. On the one hand, large-scale quantum bit manipulation can be achieved at the stage where the isolation technology between lines is immature, and on the other hand, the problems existing in the existing line isolation technology can be explored.

[0003] Among superconducting quantum bits, there is a type of bit whose frequency can be modulated by the magnetic field in the Josephson junction, called an adjustable bit, such as the double-junction Transmon, Fluxonium, etc. The external flux bias line is used to adjust the magnetic field in the Josephson junction, thereby changing the bit frequency. When there is DC flux crosstalk, the flux bias applied to a single bit will affect multiple bits, causing the frequencies of multiple bits to change. In order to calibrate the DC flux crosstalk between two bits, we first divide the bits into crosstalk bits and crosstalked bits. Then we change the flux bias of the crosstalk bit and read the frequency of the crosstalked bit. When crosstalk exists, the frequency of the crosstalked bit will change. Finally, we apply an additional flux bias to the crosstalked bit so that the frequency of the crosstalked bit is the same as at the beginning. By calculating the ratio of the flux bias change of the crosstalk bit and the additional flux bias of the crosstalked bit, we can obtain the pre-crosstalk matrix element. Repeating the above operations for all bits in the superconducting quantum chip in turn, we can obtain the pre-crosstalk matrix, thereby completing the calibration of DC flux crosstalk. The most critical step in the above process is to find the correct additional flux bias so that the frequency of the interfered bit is the same as the original one. Although this bias can be obtained by traversing all flux bias values, it takes a very long time to achieve the required accuracy. Therefore, the existing technology is to obtain the bit frequency change by trying several bias values, and use the relationship between bias and frequency to fit the appropriate bias value.

[0004] There are two problems with the existing methods. The first problem is that the fitting accuracy is constrained by the fitting formula and the number of bias points in the experiment. Due to the limitation of sample preparation accuracy, the relationship between bias and frequency may differ from the theoretical formula. At the same time, in order to accurately fit, we need enough experimental data, which will undoubtedly increase the experimental time. The second problem is that the coupling between bits will cause the bit frequency to move. When the bias of the crosstalk bit is changed, the frequency of the crosstalk bit will also change periodically. The change in the frequency of the crosstalk bit will cause the frequency of the crosstalked bit to move due to coupling. This causes the frequency of the crosstalked bit to be disturbed and no longer a single variable function of the flux crosstalk. Summary of the invention

[0005] To solve these problems, the improved solution is to make the initial bias of the crosstalked bit in the linear range. This can reduce the fitting parameters and thus reduce the amount of experimental data. At the same time, the error caused by the fitting formula can be alleviated by reducing the change in the crosstalk bit bias to ensure that the crosstalked bit is always in the linear range. Although this can solve the first problem, because the change in the crosstalk bit bias must be small enough, it is difficult to ensure that the crosstalk bit frequency can remain unchanged across cycles. Therefore, the second problem cannot be solved.

[0006] In order to calibrate the flux crosstalk quickly, robustly and with high precision, it is crucial to solve the above problems.

[0007] To this end, the present invention provides a method for automatically calibrating DC magnetic flux crosstalk based on a superconducting quantum chip to solve the problems raised in the background technology.

[0008] In order to achieve the above object, the present invention provides the following technical solution: a method for automatically calibrating DC magnetic flux crosstalk based on a superconducting quantum chip, comprising the following steps:

[0009] Step 1: Initially set the pre-crosstalk matrix to the unit matrix, denoted as P(0)=I, and mark the bits according to whether the resonant cavity is read or not;

[0010] Step 2: Scan all modulation curves of the reading resonant cavity with magnetic flux bias;

[0011] Step 3: Set the crosstalk unit q e and crosstalk unit q c , according to whether the crosstalked unit carries a read resonant cavity, select the next step, if it carries a read resonant cavity, jump to step 4, otherwise jump to step 13;

[0012] Step 4: Scan and read the modulation curve of the resonant cavity with the magnetic flux bias to find the measurement point;

[0013] Step 5: Use Newton iteration method to scan and read the resonant cavity frequency to predict the pre-crosstalk matrix element

[0014] Step 6: Determine whether the pre-crosstalk matrix element meets the preliminary calibration accuracy requirement. If so, proceed to step 7; otherwise, update the pre-crosstalk matrix element and return to step 4;

[0015] Step 7: Scan the relationship between the energy spectrum of the crosstalked unit and the magnetic flux bias;

[0016] Step 8: Same as step 4;

[0017] Step 9: Use Newton iteration method and adaptive method to read the energy spectrum of the crosstalk unit and predict the pre-crosstalk matrix element

[0018] Step 10: Determine whether the pre-crosstalk matrix element meets the fine calibration accuracy requirement, if so, proceed to step 11, otherwise update the pre-crosstalk matrix element and return to step 8;

[0019] Step 11: Determine whether all readable bits have been traversed, if so, proceed to step 12, otherwise return to step 3;

[0020] Step 12: Determine whether all crosstalk units have been traversed. If yes, exit the iteration to complete the crosstalk matrix calibration. Otherwise, wait until all readable bits have been traversed before proceeding to step 14.

[0021] Step 13: Scan the crosstalk unit flux bias to check all reading resonant cavities and determine whether there is crossover avoidance. If so, perform indirect reading using crossover avoidance and mark it as a readable bit and return to step 4. Otherwise, wait until all readable bits are traversed before proceeding to step 14.

[0022] Step 14: Prepare the 1 state of all readable bits one by one, then scan the magnetic flux bias of the crosstalk unit to check all reading resonant cavities to determine whether there is cross-free. If so, use cross-free to perform indirect reading and mark it as a readable bit and then return to step 8. Otherwise, the crosstalk unit cannot be calibrated and is directly set to 0 and deemed to be calibrated and then return to step 11.

[0023] Preferably, in step 2, the flux bias of all bits is set to 0, and the flux bias of the bits is changed one by one to obtain the resonant cavity frequency f read by all bits. c With the modulation curves of their respective flux biases, the modulation curves are finally compared and the line attenuation is adjusted so that their modulation periods differ within one cycle.

[0024] Preferably, in step 4, specifically, q e The flux bias is set at The larger the magnetic flux bias is, the Q c The flux bias is set near the lowest point, and the corresponding flux bias is recorded as The remaining bits are also set near the lowest point, and the whole is recorded as The lowest point refers to the magnetic flux bias corresponding to the lowest bit frequency when the bit reading resonant cavity frequency corresponds to the lowest bit frequency.

[0025] Preferably, step five specifically includes: according to the measurement points obtained in step four Measured For the first two iterations, you need to pre-set and Common settings The set value can be set according to the crosstalk level of the sample.

[0026] Preferably, in step six, the pre-crosstalk matrix element is updated according to the formula

[0027] Preferably, in step seven, the flux bias of all bits is set to 0, and the flux bias of the bits is changed one by one to obtain all bits f q Modulation curves with respective flux bias.

[0028] Preferably, in step nine, the measurement points obtained in step eight are Measured For the first two iterations, you need to pre-set and Common settings The set value can be set according to the crosstalk level of the sample, where n represents the number of iterations in step 6. According to the following formula, it is obtained by adaptively over a large range;

[0029]

[0030] in is the initial scanning range, usually set to 20MHz. is to expand the scanning range, generally set to 10MHz, n is the number of iterations, n max is the maximum number of iterations; the energy spectrum search is considered successful when the data in the signal can reach a given signal-to-noise ratio threshold; if the energy spectrum is still not found after exceeding the maximum number of iterations, it means that the readout measurement point is not selected properly, resulting in the energy spectrum signal-to-noise ratio not reaching the set threshold, and the measurement point needs to be re-optimized; it is generally believed that the maximum energy spectrum scanning range is ±100MHz.

[0031] Preferably, in step 10, the pre-crosstalk matrix element is updated according to the formula

[0032] Beneficial Effects

[0033] The present invention analyzes in detail the influence of flux crosstalk on the adjustable bit frequency without considering the specific adjustable bit type, and provides an automated method for calibrating the flux crosstalk to the desired accuracy. Theoretically, this calibration accuracy converges with the measurement point. The present invention uses a method of gradually iteratively updating the pre-crosstalk matrix to enhance the robustness of the automatic calibration and uses the Newton iteration method to find the pre-crosstalk matrix elements to enhance the applicability of the automatic calibration, thereby achieving high-precision, high-robustness and high-applicability automatic calibration of superconducting quantum bit flux crosstalk;

[0034] However, the current experimental technical solutions all give a quantitative description of the flux crosstalk calibration accuracy, and the calibration accuracy is obtained only based on the final measurement results, rather than from the principle;

[0035] Secondly, the current experimental technical solution needs to use accurate bit energy spectrum curves or derivative information to find the pre-crosstalk matrix elements. Although this is fast, it is greatly affected by noise and has poor robustness. Although the iterative method requires multiple measurements, it ensures robustness, which is very important in automatic calibration. Considering that the time complexity of scanning and reading the resonant cavity frequency is not high, the time of the method of the present invention will not be too long.

[0036] Finally, the current experimental technical solution does not analyze the situation when the coupler does not carry a read resonant cavity.

[0037] The present invention analyzes in detail the effect of flux crosstalk on bit frequency, which can serve as a guide for calibrating flux crosstalk, so that calibration with arbitrary accuracy can be achieved in theory; it uses multiple iterations to gradually approach the required accuracy, and has good robustness and applicability; the present invention analyzes for the first time the effect of the coupler not carrying a read resonant cavity on flux crosstalk calibration, and shows that our invention can handle this situation well. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] Figure 1 This is an automatic calibration flow chart of the method for automatically calibrating DC magnetic flux crosstalk based on a superconducting quantum chip provided by the present invention. DETAILED DESCRIPTION

[0039] The following is a description of the implementation of the present invention by specific embodiments. People familiar with the art can easily understand other advantages and effects of the present invention from the contents disclosed in this specification. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0040] The general relationship between the ideal qubit frequency and flux bias is:

[0041]

[0042] The actual relationship when subjected to flux crosstalk is:

[0043]

[0044] is the bit frequency vector excluding bit i, is the flux bias vector except bit i. In order to better show the influence of crosstalk flux bias on bit frequency, we assume that the flux bias of the remaining bits crosstalking to bit i is linearly related to the flux bias applied to the remaining bits, that is:

[0045]

[0046] in is the flux bias vector actually felt by the bit, is the flux bias vector actually applied to the bit, and M is the crosstalk matrix. For bit i, its frequency is:

[0047]

[0048] Now we use the pre-crosstalk matrix P such that:

[0049] PM=I (5)

[0050] where I is the identity matrix. So, when the magnetic flux bias felt by the desired bit is When

[0051]

[0052] From the above analysis, the necessary and sufficient condition for the pre-crosstalk matrix P to satisfy formula (5) is:

[0053]

[0054] Where i is not equal to j. Because during the experimental measurement, only Follow The change in And it can be written as:

[0055]

[0056] In order to eliminate the influence of other bit frequencies, it is necessary to measure The value is much smaller than Considering that the influence of the remaining bit frequencies on the bit i frequency is achieved through coupling between quantum bits, the greater the difference between the two quantum bits, the smaller the influence of the remaining bit frequencies on the bit i frequency. The quantitative relationship can be approximately expressed as:

[0057]

[0058] Where i is not equal to j, and when Take the - sign when it is, otherwise take the + sign. From formula (4), we can know that for any bit k:

[0059]

[0060] From (7) and (9), we can obtain:

[0061]

[0062] Generally speaking, The accuracy that needs to be calibrated is higher than So we can Treat it as a small quantity. Ignoring the higher-order small quantities, we have:

[0063]

[0064] According to We can get:

[0065]

[0066] The summation symbol "" indicates k≠i≠j. From formula (12), we get:

[0067]

[0068] According to formula (8), using the typical parameter g ij ≈0.1GHz,Δ ij =|ω i -ω j |≈2GHz, we can get The magnitude of is 0.25%. From formula (13), we can know that S ij The calibration accuracy is the same as In order to make S ij To achieve a calibration accuracy of one ten-thousandth, λ should be set ji Reaching the order of 1%. ij When the calibration accuracy is one part per thousand, the effect of frequency rejection can be ignored. So:

[0069]

[0070] From formula (14), we can know is a necessary and sufficient condition for calibration to be completed. The measurement accuracy is also limited by S ij Factors affecting calibration accuracy. In actual measurements, differentials are used instead of differentials:

[0071]

[0072] The subscripts m0 and m1 represent different measurement points. It can be seen that the measurement accuracy is actually limited by the energy spectrum resolution, i.e., Δf i The resolution is usually on the order of 0.2MHz. Therefore, in order to improve the calibration accuracy, the measurement point should be as high as possible. Size and Although the approximation of using differences instead of differentials fails, the convergence condition is Therefore, this will not usually have an effect on the actual calibration. However, it is worth noting that because f i is a periodic function, when Too large a value will cause f i When the cycle is crossed, the effect of the approximation failure of using the difference instead of the differential will be reflected and cause calibration errors. To prevent this error, use the step-by-step calibration method, first use a small After taking the value and calibrating it once, at this time, because S ij becomes smaller, from formula (14) we can know Also becomes smaller, combined with formula (15), we can know that maintaining the same Δf i , can be increased The required calibration accuracy can be achieved after gradual iteration.

[0073] After discussing the accuracy and requirements of calibration, we will use the Newton iteration method to find the convergence condition of a single iteration. That is, Δf i = 0. Combining formula (14) and (15), we have:

[0074]

[0075] Assume the pre-crosstalk matrix P is:

[0076]

[0077] From formula (5), we can get:

[0078] S ij =∑ k M in P nj =M ij +P ij (18)

[0079] The crosstalk matrix M is used ii = 1. We keep the measurement point unchanged and keep changing P ij Measuring Δf i , record the 0th measurement First measurement Then the nth measurement P ij (n) is:

[0080]

[0081] Continue iterating until Δf i (n) Once the required accuracy is achieved, the calibration of the crosstalk matrix element is completed.

[0082] The following is an explanation of the hardware requirements of the calibration scheme for superconducting quantum chips. Because the energy spectrum of each unit that needs to be calibrated (including quantum bits, couplers, etc.) must be read out, it is best to require them to carry a readout resonant cavity. However, for today's large-scale quantum bits, couplers usually do not carry a readout resonant cavity. In view of this situation, the method designed by the present invention can be calibrated to a certain extent. The present invention has a fast calibration speed, few chip parameter requirements, and high calibration accuracy, and can be applied to most superconducting quantum chips.

[0083] like Figure 1 As shown, the present invention discloses a method for automatically calibrating DC magnetic flux crosstalk based on a superconducting quantum chip, comprising the following steps:

[0084] Step 1: Initially set the pre-crosstalk matrix to the unit matrix, denoted as P(0)=I, and mark the bits according to whether the resonant cavity is read or not;

[0085] Step 2: Scan the modulation curves of all read resonant cavities with flux bias, specifically: set the flux bias of all bits to 0, change the flux bias of each bit one by one, and obtain the reading resonant cavity frequency f for all bits c Following the modulation curves of their respective flux biases, the individual modulation curves are finally compared and the line attenuation is adjusted so that their modulation periods differ within one cycle;

[0086] Step 3: Set the crosstalk unit q e and crosstalk unit q c , according to whether the crosstalked unit carries a read resonant cavity, select the next step, if it carries a read resonant cavity, jump to step 4, otherwise jump to step 13;

[0087] Step 4: Scan and read the modulation curve of the resonant cavity with the magnetic flux bias to find the measurement point; specifically: e The flux bias is set at The larger the magnetic flux bias is, the Q c The flux bias is set near the lowest point, and the corresponding flux bias is recorded as The remaining bits are also set near the lowest point, and the whole is recorded as The lowest point refers to the magnetic flux bias corresponding to the lowest bit frequency when the resonant cavity frequency of the bit is read;

[0088] Step 5: Use Newton iteration method to scan and read the resonant cavity frequency to predict the pre-crosstalk matrix element Specifically include: According to the measurement points obtained in step 4 Measured For the first two iterations, you need to pre-set and Common settings The set value can be set according to the crosstalk level of the sample;

[0089] Step 6: Determine whether the pre-crosstalk matrix element meets the preliminary calibration accuracy requirements. If so, proceed to step 7. Otherwise, according to the formula Update the pre-crosstalk matrix element and return to step 4;

[0090] Step 7: Scan the relationship between the energy spectrum of the crosstalked unit and the flux bias; specifically: set the flux bias of all bits to 0, change the flux bias of the bits one by one, and obtain the flux bias of all bits f q Modulation curves with respective flux bias;

[0091] Step 8: Same as step 4;

[0092] Step 9: Use Newton iteration method and adaptive method to read the energy spectrum of the crosstalk unit and predict the pre-crosstalk matrix element Specifically: According to the measurement points obtained in step eight Measured For the first two iterations, you need to pre-set and Common settings The set value can be set according to the crosstalk level of the sample, where n represents the number of iterations in step 6. According to the following formula, it is obtained by adaptively over a large range;

[0093]

[0094] in is the initial scanning range, usually set to 20MHz. is to expand the scanning range, generally set to 10MHz, n is the number of iterations, n maxis the maximum number of iterations. When the data in the signal can reach the given signal-to-noise ratio threshold, the energy spectrum search is considered successful. If the energy spectrum is still not found after exceeding the maximum number of iterations, it means that the readout measurement point is not selected properly, resulting in the energy spectrum signal-to-noise ratio not reaching the set threshold. The measurement point needs to be re-optimized. It is generally believed that the maximum energy spectrum scanning range is ±100MHz.

[0095] Step 10: Determine whether the pre-crosstalk matrix element meets the fine calibration accuracy requirements. If so, proceed to step 11. Otherwise, according to the formula Update the pre-crosstalk matrix element and return to step eight;

[0096] Step 11: Determine whether all readable bits have been traversed. If yes, proceed to step 12; otherwise, return to step 3. ;

[0097] Step 12: Determine whether all crosstalk units have been traversed. If yes, exit the iteration to complete the crosstalk matrix calibration. Otherwise, wait until all readable bits have been traversed before proceeding to step 14.

[0098] Step 13: Scan the crosstalk unit flux bias to check all reading resonant cavities and determine whether there is crossover avoidance. If so, perform indirect reading using crossover avoidance and mark it as a readable bit and return to step 4. Otherwise, wait until all readable bits are traversed before proceeding to step 14.

[0099] In the indirect reading, the present invention adjusts the flux bias of the coupler to make it coincide with a certain bit (denoted as q cp ) is at the maximum cross-free point, and the frequency of the reading resonant cavity corresponding to the maximum cross-free point is the frequency of the coupler here. When the frequency of the coupler moves due to magnetic flux crosstalk, this cross-free will decrease or disappear. The size of the maximum cross-free determines the frequency range of the coupler to be read. Therefore, when reading, the reading resonant cavity corresponding to the maximum cross-free should be selected as the auxiliary reading cavity.

[0100] Step 14: Prepare the 1 state of all readable bits one by one, then scan the magnetic flux bias of the crosstalk unit to check all reading resonant cavities to determine whether there is cross-free. If so, use cross-free to perform indirect reading and mark it as a readable bit and then return to step 8. Otherwise, the crosstalk unit cannot be calibrated and is directly set to 0 and deemed to be calibrated and then return to step 11.

[0101] Here, the indirect reading is firstly that the coupler is not connected to q cp When the cross-free cp Prepare to 1 state. Then sweep the frequency of the coupler. When the coupler frequency is near the crossover-free state, q cpThe population of will change and cause the corresponding voltage change of the reading resonant cavity, thereby reading the frequency of the coupler. When the frequency of the coupler moves due to flux crosstalk, q cp The voltage change of the reading resonant cavity will be reduced or disappear. The maximum cross-free value determines the frequency range of the coupler that we can read. Therefore, when reading, the bit corresponding to the maximum cross-free value should be selected as the auxiliary readout bit.

[0102] Using the above reading method, we can read the frequency of the coupler. The crosstalk calibration method of the coupler is basically the same as the method described above. Just pay attention to the following points:

[0103] (1) Because the coupler has a limited frequency range, the measurement point should be selected with care to ensure that the coupler frequency cannot be read.

[0104] (2) For the second case, since the coupler needs to be adjusted to avoid bit crossover, q cp The indirect crosstalk introduced (remember the ) is no longer a small amount. Unable to calibrate q cp If you need to calibrate this crosstalk, you should choose another bit q that can form a cross-free mode with the coupler. cp’ As auxiliary bit.

[0105] Although the present invention has been described in detail above by general description and specific embodiments, it is obvious to those skilled in the art that some modifications or improvements can be made to the present invention. Therefore, these modifications or improvements made without departing from the spirit of the present invention all belong to the scope of protection claimed by the present invention.

Claims

1. A method for automatically calibrating DC magnetic flux crosstalk based on a superconducting quantum chip, characterized in that: The following steps are involved: Step 1: Initially set the pre-crosstalk matrix to the unit matrix, denoted as P(0)=I, and mark the bits according to whether the resonant cavity is read or not; Step 2: Scan all modulation curves of the reading resonant cavity with magnetic flux bias; Step 3: Set the crosstalk unit q e and crosstalk unit q c , according to whether the crosstalked unit carries a read resonant cavity, select the next step, if it carries a read resonant cavity, jump to step 4, otherwise jump to step 13; Step 4: Scan and read the modulation curve of the resonant cavity with the magnetic flux bias to find the measurement point; Specifically, the crosstalk unit q e The flux bias is set at The larger the magnetic flux bias is, the The crosstalk unit q c The flux bias is set near the lowest point, and the corresponding flux bias is recorded as The remaining bits are also set near the lowest point, and the whole is recorded as The lowest point refers to the magnetic flux bias corresponding to the lowest bit frequency when the frequency of the resonant cavity corresponding to the bit is read; Step 5: Use Newton iteration method to scan and read the resonant cavity frequency to predict the pre-crosstalk matrix element Step 6: Determine whether the pre-crosstalk matrix element meets the preliminary calibration accuracy requirement. If so, proceed to step 7; otherwise, update the pre-crosstalk matrix element and return to step 4; Step 7: Scan the relationship between the energy spectrum of the crosstalked unit and the magnetic flux bias; Step 8: Same as step 4; Step 9: Use Newton iteration method and adaptive method to read the energy spectrum of the crosstalk unit and predict the pre-crosstalk matrix element Specifically, the measurement points obtained in step eight Measured For the first two iterations, you need to pre-set and The value set is set according to the crosstalk level of the sample, where n represents the number of iterations in step 6. According to the following formula, it is obtained by adaptively over a large range; in is the initial scan range, set to 20MHz, is to expand the scanning range, set to 10MHz, n is the number of iterations, n max is the maximum number of iterations; When the data in the signal can reach a given signal-to-noise ratio threshold, the energy spectrum is considered to be successfully found. If the energy spectrum is still not found after the maximum number of iterations, the measurement point needs to be re-optimized; Step 10: Determine whether the pre-crosstalk matrix element meets the fine calibration accuracy requirement, if so, proceed to step 11, otherwise update the pre-crosstalk matrix element and return to step 8; Step 11: Determine whether all readable bits have been traversed, if so, proceed to step 12, otherwise return to step 3; Step 12: Determine whether all crosstalk units have been traversed. If yes, exit the iteration to complete the crosstalk matrix calibration. Otherwise, wait until all readable bits have been traversed before proceeding to step 14. Step 13: Scan the crosstalk unit flux bias to check all reading resonant cavities and determine whether there is crossover avoidance. If so, perform indirect reading using crossover avoidance and mark it as a readable bit and return to step 4. Otherwise, wait until all readable bits are traversed before proceeding to step 14. Step 14: Prepare the 1 state of all readable bits one by one, then scan the magnetic flux bias of the crosstalk unit to check all reading resonant cavities to determine whether there is cross-free. If so, use cross-free to perform indirect reading and mark it as a readable bit and then return to step 8. Otherwise, the crosstalk unit cannot be calibrated and is directly set to 0 and deemed to be calibrated and then return to step 11.

2. The method for automatically calibrating DC magnetic flux crosstalk based on a superconducting quantum chip according to claim 1, characterized in that: In step 2, the flux bias of all bits is set to 0, and the flux bias of the bits is changed one by one to obtain the modulation curves of the resonant cavity frequency fc read by all bits with their respective flux biases. Finally, the modulation curves are compared and the line attenuation is adjusted so that their modulation periods differ within one period.

3. The method for automatically calibrating DC magnetic flux crosstalk based on a superconducting quantum chip according to claim 1, characterized in that: Step 5 specifically includes: according to the measurement points obtained in step 4 Measured For the first two iterations, you need to pre-set and The set value is set according to the crosstalk level of the sample.

4. The method for automatically calibrating DC magnetic flux crosstalk based on a superconducting quantum chip according to claim 1, characterized in that: In step 6, the pre-crosstalk matrix element is updated according to the formula 5. The method for automatically calibrating DC magnetic flux crosstalk based on a superconducting quantum chip according to claim 1, characterized in that: In step 7, the flux bias of all bits is set to 0, and the flux bias of the bits is changed one by one to obtain all bits f q Modulation curves with respective flux bias.

6. The method for automatically calibrating DC magnetic flux crosstalk based on a superconducting quantum chip according to claim 1, characterized in that: In step 10, the pre-crosstalk matrix element is updated according to the formula

Images