EP4666227A2 - Étalonnage de processeur quantique avec mémoire et rétroaction - Google Patents

Étalonnage de processeur quantique avec mémoire et rétroaction

Info

Publication number
EP4666227A2
EP4666227A2 EP23959057.3A EP23959057A EP4666227A2 EP 4666227 A2 EP4666227 A2 EP 4666227A2 EP 23959057 A EP23959057 A EP 23959057A EP 4666227 A2 EP4666227 A2 EP 4666227A2
Authority
EP
European Patent Office
Prior art keywords
calibration
modifications
quantum
determining
model
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
EP23959057.3A
Other languages
German (de)
English (en)
Inventor
Paul KLIMOV
Julian Shaw KELLY
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Google LLC
Original Assignee
Google LLC
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Google LLC filed Critical Google LLC
Publication of EP4666227A2 publication Critical patent/EP4666227A2/fr
Pending legal-status Critical Current

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N10/00—Quantum computing, i.e. information processing based on quantum-mechanical phenomena
    • G06N10/70—Quantum error correction, detection or prevention, e.g. surface codes or magic state distillation
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N10/00—Quantum computing, i.e. information processing based on quantum-mechanical phenomena
    • G06N10/40—Physical realisations or architectures of quantum processors or components for manipulating qubits, e.g. qubit coupling or qubit control
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/04—Architecture, e.g. interconnection topology
    • G06N3/044—Recurrent networks, e.g. Hopfield networks
    • G06N3/0442—Recurrent networks, e.g. Hopfield networks characterised by memory or gating, e.g. long short-term memory [LSTM] or gated recurrent units [GRU]

Definitions

  • This specification relates to quantum computing.
  • Quantum computations are physically realized through the time-evolution of quantum systems steered by analog control signals. As quantum information is stored in continuous amplitudes and phases, these control signals must be carefully chosen to achieve the desired result. Calibration is the process of performing a series of experiments on the quantum system to learn optimal control parameters.
  • This specification relates to calibration of quantum processor operating parameters using memory and feedback.
  • one innovative aspect of the subject matter described in this specification can be implemented in a computer implemented method that includes determining, by a feedback system and based on historical data generated by a calibration and quantum computing pipeline and stored by the feedback system, that a calibration modification event is required; in response, determining, by the feedback system and using the historical data, one or more modifications to components of the calibration and quantum computing pipeline; and applying the one or more modifications to the components of the calibration and quantum computing pipeline, wherein subsequent calibration procedures are performed on physical qubits included in the calibration and quantum computing pipeline using the modified components.
  • implementations of this aspect include corresponding classical and quantum computer systems, apparatus, and computer programs recorded on one or more computer storage devices, each configured to perform the actions of the methods.
  • a system of one or more classical and quantum computers can be configured to perform particular operations or actions by virtue of having software, firmware, hardware, or a combination thereof installed on the system that in operation causes or cause the system to perform the actions.
  • One or more computer programs can be configured to perform particular operations or actions by virtue of including instructions that, when executed by data processing apparatus, cause the apparatus to perform the actions.
  • the foregoing and other implementations can each optionally include one or more of the following features, alone or in combination.
  • the historical data comprises calibration and characterization data and processor status validation data obtained during past calibration procedures performed by the calibration and quantum computing pipeline.
  • the modifications to components of the calibration and quantum computing pipeline comprise modifications to a calibration scheduler that is configured to schedule taking of characterization data from qubits included in the calibration and quantum computing pipeline; and applying the modifications to the calibration scheduler comprises modifying one or more of: when characterization data is taken from the qubits or a type of characterization data taken from the qubits.
  • determining that the calibration event has occurred comprises one or more of: determining that one or more performance metrics of respective qubits have not been measured for a predefined period of time or determining that historical values of one or more performance metrics of respective qubits fluctuate more than a predefined expected fluctuation threshold; and applying one or more modifications to the calibration scheduler comprises causing the calibration scheduler to schedule experiments for determining current values of the performance metrics.
  • determining that the calibration event has occurred comprises determining that a duration of previous calibration procedures exceeds a predetermined acceptable threshold; and applying one or more modifications to the calibration scheduler comprises causing the calibration scheduler to take less characterization data during the subsequent calibration procedures.
  • determining that the calibration event has occurred comprises determining that a calibration model used by a calibration optimizer included in the calibration and quantum computing pipeline has not been updated for a predetermined amount of time; and applying one or more modifications to the calibration scheduler comprises causing the calibration scheduler to take more characterization data during the subsequent calibration procedures.
  • the modifications to components of the calibration and quantum computing pipeline comprise modifications to a calibration model that maps quantum processor operating parameters and characterization data onto respective performance metrics.
  • determining that the calibration event has occurred comprises one or more of: determining that the calibration model has not been updated for a predetermined amount of time or determining that the calibration model is converging on operating parameter configurations that introduce system errors; and applying the modifications to the calibration model comprises adding or removing one or more keep-out cost functions to the calibration model.
  • determining that the calibration event has occurred comprises determining that the calibration model has not been updated for a predetermined amount of time; and applying the modifications to the calibration model comprises adding random noise to the calibration model.
  • determining that the calibration event has occurred comprises determining that a predefined modification schedule indicates that calibration model parameters are required for a subsequent calibration procedure; applying the modifications to the calibration model comprises reducing relative amplitudes of low trust calibration model parameters in the calibration model.
  • determining that the calibration event has occurred comprises determining that the historical data includes a sufficient amount of data; determining modifications to the calibration model comprises: computing correlations between calibrated operating parameters and performance metrics; and determining statistical biases that bias values of the calibrated operating parameters towards historically reliable values.
  • determining that the calibration event has occurred comprises obtaining a model configured to predict a performance of calibrated operating parameters in previously unexplored regimes; and applying the modifications to the calibration model comprises adding the obtained model to the calibration model.
  • determining that the calibration event has occurred comprises obtaining a model configured to forecast future values of calibrated operating parameters; and applying the modifications to the calibration model comprises adding the obtained model to the calibration model.
  • determining that the calibration event has occurred comprises determining an operating parameter has failed calibration more than a predefined threshold number of times; the modifications to components of the calibration and quantum computing pipeline comprise modifications to how a calibration scheduler included in the calibration and quantum computing pipeline selects operating parameters for calibration; and applying the modifications comprises causing the calibration scheduler to select and calibrate additional interacting elements or operating parameters when the operating parameter that failed calibration is calibrated.
  • determining that the calibration event has occurred comprises determining that a dimension of a calibration model included in the calibration and quantum computing pipeline has changed; the modifications to components of the calibration and quantum computing pipeline comprise modifications to a calibration optimizer included in the calibration and quantum computing pipeline; and applying the modifications to the calibration model comprises updating an optimization algorithm used by the calibration optimizer.
  • determining that the calibration event has occurred comprises identifying one or more calibrated operating parameters that transitioned operating status between calibration iterations; the modifications to components of the calibration and quantum computing pipeline comprise modifications to a calibrations scheduler included in the calibration and quantum computing pipeline; and applying the modifications comprises causing the calibration scheduler to revisit the one or more calibrated operating parameters in subsequent calibration iterations.
  • the method further comprises receiving and storing data generated by the calibration and quantum computing pipeline during a previous number of calibration procedures.
  • a purpose of quantum processor calibration is to rapidly and reliably achieve (1) high system performance where all calibrations succeeded, all operating parameters are set within hardware specifications, and all computational elements perform well enough to execute quantum algorithms of interest with high performance, and (2) high stability, where high system performance is maintained over long periods of time, e g., where long is defined relative to the length of time it takes to calibrate a quantum processor and to execute quantum algorithms of interest.
  • each qubit needs to perform a number of independent operations which are independently calibrated, e.g., including single-qubit gates, two-qubit gates, and readout.
  • the presently described quantum processor calibration techniques leverage historical data to apply feedback during a qubit calibration procedure.
  • the feedback is used to modify the typical behavior of components of the calibration system implementing the calibration procedure in order to improve the quantum computing system's performance and/or stability.
  • feedback can be applied to modify how characterization data is taken during the calibration system.
  • Such modifications can improve the accuracy and stability of an underlying calibration model, since fluctuations in performance metrics can be better monitored/taken account of and additional characterization data can be taken if required. Further, such modifications can improve the computational runtime of the calibration system, e.g., by dynamically adjusting the amount of calibration data to be taken.
  • feedback can be applied to modify the calibration model used by the calibration system, e.g., to improve the model’s accuracy and stability and/or to bias control parameters into a target direction.
  • feedback can be applied to modify the selection of calibration targets, e.g., to account for complex operating parameter dependencies due to crosstalk and/or engineered interactions.
  • feedback can be applied to modify the optimization algorithm used by the calibration system based on the type of search problem being solved.
  • Such modifications can improve the computational runtime of the calibration system, e.g., by avoiding using computationally intense optimization algorithms such as simulated annealing if the search problem is a low dimensional or convex problem.
  • FIG. 1 is a block diagram of an example quantum computing sy stem that implements repetitive calibration with memory and feedback.
  • FIG. 2 is a block diagram of an example workflow for calibrating an calibration target of a quantum processor.
  • FIG. 3 is a flowchart of an example process for modifying components of a calibration and quantum computing pipeline.
  • FIG. 4 illustrates an example process for using historical data to add keepout cost functions to a calibration model.
  • FIG. 5 illustrates an example process for using historical data to implement trustbased temporal weighting of a calibration model.
  • FIG. 6 is a block diagram of an example process for determining correlation-based modifications to a calibration model.
  • FIG. 7 is a block diagram of an example process for determining a biasing model for a calibration model.
  • FIG. 8 shows an example process for modifying how calibration targets are selected.
  • FIG. 9 illustrates an example process for using historical data to apply keepout cost functions and select additional calibration targets.
  • FIG. 10 depicts an example quantum processor.
  • This specification describes techniques for leveraging historical data to apply feedback during a qubit calibration procedure.
  • more historical data about the quantum computing system being calibrated is accumulated.
  • the historical data can include, for example, characterization and calibration data, processor-status validation data, gate benchmark data, quantum algorithm output data, quantum algorithm performance data, electronics data, etc. This data is leveraged to modify the typical behavior of components of the calibration system implementing the calibration procedure in order to improve the quantum computing system’s performance and/or stability.
  • FIG. 1 is a block diagram of an example quantum computing system 100 that implements repetitive calibration with memory and feedback.
  • the system 100 is an example of a system implemented as computer programs on one or more classical and quantum computing devices in one or more locations, in which the systems, components, and techniques described in this specification can be implemented.
  • the system 100 includes a calibration and quantum computing pipeline 102 and a feedback system 104.
  • the calibration and quantum computing pipeline 102 includes a calibration system 106, a quantum processor 108, and a validation module 110.
  • the feedback system 104 includes a calibration feedback module 112 and a historical data memory device 114.
  • Components of the system 100 can be in data communication with each other, e.g., through a communication network such as a local area network or wide area network.
  • the quantum processor 108 is configured to perform quantum computations.
  • the quantum processor 108 includes classical and quantum computing elements.
  • the quantum processor 108 includes multiple physical qubits that interact via respective interactions.
  • the qubits can be used to perform algorithmic operations or quantum computations.
  • the specific realization of the one or more qubits and their interactions may depend on a variety of factors including the type of quantum computations that the quantum processor is performing.
  • the qubits may include qubits that are realized via atomic, molecular or solid-state quantum systems.
  • the qubits may include, but are not limited to, superconducting qubits or semi-conducting qubits.
  • the interacting qubits can be frequency tunable. That is, each qubit can have associated operating frequencies that can be adjusted, e.g., using control devices, through application of voltage pulses via a driveline coupled to the qubit.
  • Example operating frequencies include qubit idling frequencies, qubit interaction frequencies, and qubit readout frequencies. Different frequencies correspond to different operations that the qubit can perform.
  • the operating frequency can be set to a corresponding idling frequency may put the qubit into a state where it does not strongly interact with other qubits.
  • the operating frequency can be set to frequencies at which the qubit implements a single qubit gate, frequencies at which the qubit can be measured, and frequencies at which the qubit can be reset.
  • qubits when the qubits interact via couplers with fixed coupling, qubits can be configured to interact with one another by setting their respective operating frequencies at some gate-dependent frequency detuning from their common interaction frequency.
  • qubits when the qubits interact via tunable couplers, qubits can be configured to interact with one another by setting the parameters of their respective couplers to enable interactions between the qubits and then by setting the qubit’s respective operating frequencies at some gate-dependent frequency detuning from their common interaction frequency. Such interactions may be performed in order to perform tw o-qubit or many-qubit gates.
  • the quantum processor 108 also includes classical computing components.
  • the quantum processor 108 can include control devices that operate the multiple qubits, e.g., by tuning the qubit’s operating frequencies or tuning frequencies of couplers that couple the multiple qubits.
  • Example control devices include arbitrary waveform generators, control signal synthesizers, and readout resonators.
  • the type of control devices that the quantum processor 108 utilizes is dependent on the type of qubits the quantum processor 108 uses.
  • qubits that are realized via atomic, molecular or solid-state quantum systems typically have energy separation of the relevant qubit levels in the microwave or optical domain.
  • the states of such qubits may be manipulated and controlled using external fields, such as microwave or optical fields.
  • mode-locked lasers may serve as control electronics due to their broad-band optical spectra that feature both radio frequency and microwave structure.
  • the control devices could include a collection of individual qubit controllers realized by a radio frequency generator as well as one or a collection of global excitation controllers realized by a radio frequency or microwave generator. In both cases, the control devices can be operated manually or connected to a computer and controlled via suitable software allowing for specifying and automatically running the required qubit operations.
  • the calibration system 106 is configured to repetitively calibrate operating parameters of computing elements included in the quantum processor 108.
  • the calibration system 106 includes a calibration scheduler 118 and a calibration optimizer 120.
  • the calibration scheduler 118 and calibration optimizer 120 can be classical computing components that perform classical computations.
  • the calibration scheduler 118 is configured to schedule characterization data 124 to be taken from the quantum processor 108 and schedule the calibration of quantum processor operating parameters.
  • Characterization data 124 includes data that is used to calibrate one or more calibration targets.
  • a calibration target is an operating parameter of a computing element included in the quantum processor 108.
  • Example calibration targets include single qubit gate frequencies for performing single qubit gates using respective qubits or two-qubit gate frequency trajectories for implementing two-qubit gates using a pair of qubits.
  • the calibrated values of arbitrary operating parameters can depend on one another due to crosstalk and/or engineered interactions, e.g., in superconducting qubits unintentional short-range interactions (i.e.
  • crosstalk due to capacitive coupling between qubits in close spatial proximity to one another, unintentional long-range interactions (i.e. crosstalk) due to coupling between control lines at any point in the quantum computing system, for example in the chip packaging or chip wiring solution, or intentional short-range engineered interactions used to implement two-qubit gates, for example via a coupling element and/or direct capacitive and/or inductive coupling.
  • the characterization data 124 can include data obtained by performing experiments on qubits included in the quantum processor. Performing an experiment on a qubit can include applying a static control waveform to the qubit and measuring the qubit, where during the experiment no qubit operational parameters are varied.
  • the experiments can correspond to, for example, a gate sequence and measurement to determine the output probability distribution or a gate sequence and tomography to interrogate the state of the qubit.
  • Each experiment can be repeated a number of times to gather statistics, e.g., to obtain a measured qubit energy-relaxation rates versus frequency, qubit dephasing rates versus frequency, and qubit or qubit crosstalk error amplitudes.
  • the calibration scheduler 118 can be configured to implement a calibration scheduling strategy that formulates dependency relationships between operating parameters to be calibrated as a directed graph.
  • Each operating parameter to be calibrated is represented by a node in the graph and dependency relationships between operating parameters are represented by respective directed edges in the graph, where the direction of the edge denotes which operating parameter depends on the other.
  • the task of scheduling the operating parameters to be calibrated then becomes a graph traversal problem, where characterization data 124 can be taken at the traversed nodes by performing experiments, e.g., to extract current values of control and system parameters.
  • the type of experiments performed at each traversed node can vary and include coarse-grain experiments that have interplay between fundamental operations and elements such as single-qubit gates, readout, and couplers.
  • the experiments can also include fine-grain experiments that involve a more precise metrology for each qubit operation: single-qubit gates, two-qubit gates, and readout.
  • Example operations performed by the calibration scheduler 118 are described in US Patent No.: 9,940,212 titled “‘Automatic qubit calibration,” the contents of which are incorporated herein by reference.
  • the calibration optimizer 120 is configured to determine calibrated values of operating parameters 126 included in calibration schedules determined by the calibration scheduler 118.
  • the calibration optimizer 120 can be configured to construct or receive a calibration model 122 that maps operating parameters and characterization data onto one or more relevant metrics, e.g., metrics correlated with system error such as N-qubit randomized benchmarking error or error suppression factor of a quantum error correction algorithm.
  • the calibration optimizer 120 is configured to optimize the calibration model 122 with respect to one or more of the operating parameters to determine calibrated values of operating parameters 126, e.g., values that optimize the one or more relevant metrics.
  • the calibration system 102 can provide the calibrated values of the operating parameters 126 to the quantum processor 108 to calibrate the quantum processor 108.
  • the calibration model 122 can include multiple component models, where each component model corresponds to a respective and distinct error channel, e.g., error due to energy-relaxation, error due to dephasing, or error die to crosstalk.
  • Example calibration models are described in US Patent No.: 11,556,813 titled “Refining qubit calibration models using supervised learning,’’ the contents of which are incorporated herein by reference.
  • the calibration model 122 can be high-dimensional, high-constraint, and highly non-convex, e.g., in implementations where the operating parameters to be calibrated include interacting operating parameters such as logic gate frequencies. In these implementations, optimizing the calibration model 122 to obtain a globally optimal solution can be intractable. Therefore, in some implementations the calibration optimizer 120 can be configured to implement an optimization strategy that obtains locally optimal solutions, which have been empirically verified to be sufficient for state-of-the-art quantum computing applications.
  • Example operations performed by the calibration optimizer 120 are described in US Publication No.: US20200387822 titled “Calibration of quantum processor operator parameters” and US Patent No.: 11,361,241 titled “Optimizing qubit operating frequencies,” the contents of which are incorporated herein by reference.
  • the calibration optimizer 120 can be configured to determine calibrated values of operating parameters 126 included in calibration schedules determined by the calibration scheduler 118 without using a calibration model, e.g., by performing a trial and error strategy that varies operating parameters based on rules applied to a sequence of validation values in discrete steps.
  • the validation module 1 10 is configured to validate the status 128 of the quantum processor 108, e.g., check for calibration failures, software benchmarks and/or hardware benchmarks, and quantum algorithm metrics.
  • the validation module 110 can be configured to generate processor status validation data 130 that includes values of validation metrics for some or all of the qubits and control devices included in the quantum processor.
  • the processor status validation data 130 can include Boolean (e.g. GOOD or BAD) or continuous (e.g. float from 0 to 1) signals that can be used to trigger a new calibration iteration, or can include arbitrary metadata that can be used to inform downstream systems such as the feedback system 104.
  • the validation module 110 can be configured to validate the status 128 of the quantum processor 108 whilst a quantum algorithm is executing on the quantum processor 108. That is. the validation module 110 can be configured to perform an online validation process, e.g., where parity checks are continuously streamed from the quantum processor during the execution of a quantum error correction algorithm such as a surface code algorithm. Alternatively or in addition, the validation module 110 can be configured to perform an offline validation process, e.g., to obtain randomized benchmarking error rates for quantum logic gates.
  • the calibration system 106 is configured to repetitively/iteratively calibrate operating parameters of computing elements included in the quantum processor 108. At each repetition, characterization data 124 and processor status validation data 130 is generated. In addition, other data 134 relating to properties of the calibration system 106 is generated, e.g., data specifying the current calibration model 122 being used by the calibration optimizer 120 (such as whether the calibration model includes keepout cost functions, noise terms, low or high trust component models) and data specifying optimization algorithms used by the calibration optimizer 120 to optimize the calibration model. The other data can also include quantum algorithm output data, quantum algorithm performance data, and electronics data obtained from the calibration and quantum computing pipeline 102.
  • Example quantum algorithm output data includes measurement outputs from the quantum processor 108, which can be tested against expected values via some arbitrary statistical test.
  • Example quantum algorithm performance data includes randomized benchmarking errors, which can be derived from the quantum algorithm output data via an analysis routine within the validation module.
  • Example electronics data includes data representing a status of control electronics included in the quantum processor, which can be obtained from control electronics directly and/or the calibration system. The status can indicate whether the control electronics hardw are is operating in- or out-of-spec.
  • the calibration and quantum computing pipeline 102 is configured to provide this data to the feedback system 104.
  • the feedback system 104 is configured to store received characterization data 124, processor status validation data 130, and other data 134 in the memory device 114. Because the data stored in the memory device 114 includes characterization data 124, processor status validation data 130, and other data 134 from past and cunent iterations of the calibration and quantum computing pipeline 102, the data stored in the memory device 114 is referred to herein as historical data.
  • the calibration feedback module 112 is configured to access and use the historical data stored by the memory device 114 to determine modifications 132 of components of the calibration system that will improve the performance and stability 7 of components of the calibration and quantum computing pipeline 102. Example operations performed by the calibration feedback module 112 and example modifications generated by the calibration feedback module 112 are described in more detail below with reference to FIGS. 2 to 7.
  • the determined modifications 132 of the components of the calibration system are referred to herein as feedback.
  • the calibration feedback module 112 is configured to provide feedback to the calibration and quantum computing pipeline 102 to modify the components of the calibration system accordingly.
  • the calibration and quantum computing pipeline 102 can then perform subsequent calibration procedures and/or subsequent quantum computations using the modified components.
  • FIG. 2 is a block diagram 200 of an example workflow for calibrating an calibration target of a quantum processor.
  • the example workflow can be implemented by a calibration and quantum computing pipeline, e.g., the calibration and quantum computing pipeline 102 of FIG. 1.
  • a calibration target is selected.
  • the calibration target is a frequency f S q at which qubit q A implements single qubit (sq) gates.
  • the calibration target can be selected by a calibration scheduler, e.g., the calibration scheduler 118 of FIG. 1, during implementation of a calibration scheduling strategy.
  • characterization data is taken from qubit q A .
  • the calibration and quantum computing pipeline can perform experiments on qubit q A to generate characterization data that provides information about the current behavior of qubit q A .
  • three different types of experiments are performed on qubit q A - including experiments that characterize the qubif s energyrelaxation rate fy versus the selected operating parameter / ⁇ , the qubif s dephasing rate fy, versus the selected operating parameter f S q, and amplitudes of the qubit’s crosstalk error X AB with qubit q B versus the selected operating parameter f S q.
  • the generated characterization data can be stored in memory, e.g., the historical data memory device 114 of FIG. 1.
  • a calibration model is built using the characterization data generated at stage (B).
  • additional data can be used to generate the calibration model, e.g., historical characterization data.
  • the plotted line labelled “ACTUAL’” represents the real cost of a qubit operating somewhere. This is a quantity that is not usually known and can in some cases be prohibitively time-consuming to measure.
  • the lines “ACTUAL” via “MODEL” are estimated using a model and characterization data, which are nominally easier to measure.
  • the calibration model can be built (or otherwise obtained) by a calibration optimizer, e.g., the calibration optimizer 120 of FIG. 1.
  • the calibration model built during stage (C) is optimized with respect to the selected operating parameter f s ⁇ to determine a calibrated value of the selected operating parameter
  • the calibrated value of the selected operating parameter can be stored in memory, e.g., the historical data memory device 1 14 of FIG. 1.
  • the calibration model can be optimized by a calibration optimizer, e.g., the calibration optimizer 120 of FIG. 1.
  • stage (E) of the example workflow for calibrating an calibration target of a quantum processor the value of the quantum processor operating parameter f S g is set to the calibrated value determined during stage (D).
  • stage (F) of the example workflow for calibrating an calibration target of a quantum processor modifications to the calibration workflow are determined using historical data stored in memory, e.g., the memory device 1 14 of FIG. 1.
  • stage (F) first occurs after stages (A)-(E) have been performed.
  • stage (F) can be implemented before stage (A).
  • the modifications can be applied in various combinations according to an arbitrary schedule, e.g., depending on the calibration iteration (or repetition) number, the execution time, or other arbitrary’ metrics.
  • modifications can be applied to various stages of the calibration workflow.
  • historical data stored by the memory device can be used to modify how calibration targets of future calibration procedures are selected at stage (A) of the calibration workflow.
  • historical data stored by the memory device can be used to modify when and what type of characterization data should be taken at stage (B) of the calibration workflow.
  • historical data stored by the memory device can be used to determine modifications of the calibration model built and/or used at stages (C) and (D) of the calibration workflow.
  • historical data stored by the memory device can be used to modify how the calibration model is optimized to determine calibrated values of the operating parameters at stage (D) of the calibration workflow, e.g., to determine modifications to hyper parameters of the optimization strategy.
  • historical data stored by the memory device can be used to determine modifications to the model-free calibration strategies used to determine the calibrated values of the operating parameters.
  • historical data stored by the memory device can be used to determine modifications to calibrated parameter values.
  • the feedback system can identify that the optimization system has converged on a value that is known to be bad or invalid, e.g., a value that is outside of hardware specifications, previously found to be bad, or appears bad from the perspective of the model.
  • the feedback system can replace the value with a fallback value that is valid but may not be optimal.
  • FIG. 3 is a flow chart of an example process 300 for modifying components of a calibration and quantum computing pipeline.
  • the process 300 will be described as being performed by a system of one or more classical computing devices located in one or more locations.
  • a feedback system in data communication with a calibration and quantum computing pipeline e.g., the feedback system 104 in data communication with the calibration and quantum computing pipeline 102 of FIG. 1, appropriately programmed in accordance with this specification, can perform the process 300.
  • the system uses historical data generated by the calibration and quantum computing pipeline to determine that a calibration modification event is required (step 302).
  • the historical data can include historical characterization data obtained during past calibration procedures performed by the calibration and quantum computing pipeline.
  • characterization data obtained during a calibration procedure includes data that characterizes the behavior of qubits included in the calibration and quantum computing pipeline at the time the calibration procedure was performed.
  • the characterization data can include data obtained by performing experiments on the qubits, e.g., experiments that measure a qubit energyrelaxation rate versus frequency, qubit dephasing rate versus frequency, and qubit or qubit crosstalk error amplitude.
  • the historical data can also include historical processor status validation data obtained during past calibration procedures performed by the calibration and quantum computing pipeline.
  • processor status validation data obtained during a calibration procedure includes data that validates an operating status of a quantum processor included in the calibration and quantum computing pipeline at the time the calibration procedure was performed.
  • the processor status validation data can include data representing calibration failures, software and/or benchmarks, performance metrics of quantum algorithms performed by the quantum processor, and electronics data.
  • the system uses the historical data to determine one or more modifications to components of a calibration system included in the calibration and quantum computing pipeline (step 304). The system then causes the one or more modifications to be applied to the components of the calibration system (step 306). The calibration system can then perform subsequent calibration procedures on physical qubits included in the calibration and quantum computing pipeline using the modified components.
  • Example calibration modification events and modifications that can be applied to components of the calibration system are described in detail below.
  • the examples can be implemented as part of example process 300 in arbitrary combinations.
  • the system can use the historical data to determine modifications to a calibration scheduler included in the calibration system.
  • the modifications can include modifications that modify how the calibration scheduler takes characterization data from qubits included in the quantum processor, e.g., when characterization data is taken from the qubits and/or the type of characterization data that is taken from the qubits included in the calibration and quantum computing pipeline.
  • it can be beneficial from a modeling and stability’ perspective to take characterization data at all or some calibration iterations to detect fluctuations in performance metrics that may have occurred since the last time characterization data was taken.
  • energy-relaxation rates can fluctuate unpredictably on all timescales (e.g., ranging from seconds to months) due to fluctuations in two-level-system (TLS) defect frequencies. Therefore, it can be advantageous to retake the relaxation rate spectrum at some iterations.
  • TLS two-level-system
  • the system can use the historical data to determine that one or more performance metrics of respective qubits have not been measured for a predefined period of time.
  • the system can use the historical data to determine that historical values of one or more performance metrics of respective qubits unexpectedly fluctuate, e.g., more than a predefined expected fluctuation threshold.
  • the system can then generate and send instructions to the calibration and quantum computing pipeline that cause the calibration scheduler to schedule experiments that can be used to determine current values of the performance metrics.
  • the current values of the performance metrics can then be used by the calibration system, e.g., to adjust and improve subsequent calibration procedures, and stored as historical data for subsequent implementations of example process 300.
  • a keepout strategy can be used to add avoidance regions where TLS may have moved since the last time characterization data was taken.
  • the system can use the historical data to determine that calibration procedures are taking too long, e.g., that a duration of previous calibration procedures exceeds a predetermined acceptable threshold.
  • the system can then generate and send instructions to the calibration and quantum computing pipeline that cause the calibration scheduler to take less characterization data or no characterization data for one or more subsequent calibration iterations.
  • the system can use the historical data to determine that the calibration model used by the calibration optimizer has not been updated for a predetermined amount of time.
  • the system can then generate and send instructions to the calibration and quantum computing pipeline that cause the calibration scheduler to take more characterization data for one or more subsequent calibration iterations, so that the additional characterization data can be used to update the calibration model, e.g., improve its accuracy or increase its search space.
  • the system can use the historical data to determine modifications to a calibration model included in the calibration system, e g., modifications that improve the calibration model.
  • the modifications can include adding one or more keepout cost functions to prevent the calibration optimizer from converging on past solutions.
  • the keepout functions can include step functions with high cost amplitudes at respective past solutions, e.g., previously calibrated frequencies. Adding keepout functions to the calibration model therefore penalizes solutions to the optimization performed using the calibration model that converge on the past solutions.
  • Keepout cost functions can be thought of as ad hoc elements introduced into the calibration model to compensate for model imperfections that do not correctly identify “bad” operating parameter configurations (where a “bad” operating parameter configuration is understood to refer to an operating parameter configuration that is incorrect and/or causes system errors).
  • keepout cost functions can be added to avoid all previously visited calibrated operating parameters that were found to be bad.
  • keepout cost functions can be added to avoid some recently visited parameters that were found to be bad. Since operating parameters can depend on each other in complex ways, previously visited calibrated parameters that were bad can evolve into “good” parameters over multiple iterations, even in the case of zero calibration drift (where a “good” operating parameter configuration is understood to refer to an operating parameter configuration that is correct and does not introduce system errors or introduces an acceptable amount of errors). Therefore, it can be beneficial to remove old keepout cost functions that may be stale and may unnecessarily over-constraining the optimization search space.
  • keepout cost functions can be added to avoid some previously visited calibrated parameters that were found to be good.
  • Some operating parameters can have many locally optimal values that can be used for high performance quantum computing. For example, high fidelity single-qubit gates can often be executed at many different frequencies. By forcing one operating parameter to move from one good solution to another, one or more interacting operating parameters can be promoted to reach good/better solutions. Promoting such exploration can become increasingly important for operating parameters that can interact with many other operating parameters.
  • the system can use the historical data to determine that one or more keepout cost functions should be added to or removed from the calibration model. For example, the system can determine that one or more calibrated operating parameters are converging on bad solutions and therefore determine modifications to the calibration model that add keepout cost functions corresponding to the bad solutions. As another example, the system can determine that the calibration model has not been updated for a predetermined amount of time and that it might be beneficial to explore other operating parameter configurations to reach better operating parameter configurations. The system can therefore determine modifications to the calibration model that remove some previously added keepout cost functions. In either example, the system can send instructions to the calibration and quantum computing pipeline that cause the calibration system to update the calibration model accordingly.
  • FIG. 4 illustrates an example process 400 for using historical data to add keepout cost functions to a calibration model.
  • the calibration system included in the calibration and quantum computing pipeline uses a current calibration model 402 to calibrate an operating parameter f S g of a qubit.
  • the calibration converges on solution 404, represented by the vertical dashed line.
  • the calibrated value 404 can be stored as historical data in the feedback system, along with other data such as current processor status validation data for the qubit.
  • the feedback system can use the historical data, e.g., processor status validation data for the qubit, to determine that a current operating status of the qubit is bad, e.g., the qubit not operating as it should and introducing an unacceptable amount of system error.
  • the feedback system determines that a keepout cost function 406 should be added to the initial calibration model 402, where the keepout cost function 406 penalizes solutions to the optimization performed using the calibration model that converge on solution 404.
  • Stage (D) of the example process is similar to stage (A).
  • the calibration system uses the updated calibration model (that includes keepout cost function 406) to calibrate the operating parameter of a qubit.
  • the calibration converges on solution 408.
  • the calibrated value 408 can be stored as historical data in the feedback system, along with other data such as current processor status validation data for the qubit.
  • the feedback system can use the historical data to determine that a current operating status of the qubit is bad.
  • the feedback system determines that a second keepout cost function 410 should be added to the updated calibration model, where the keepout cost function 410 penalizes solutions to the optimization performed using the calibration model that converge on solution 408.
  • Stages (G), (H), and (I) are similar to stages (D), (E), and (F), respectively.
  • the calibration system uses the updated calibration model (that includes keepout cost functions 406 and 10) to calibrate the operating parameter of a qubit.
  • the calibration converges another solution 412.
  • the calibrated value 412 can be stored as historical data in the feedback system, along with other data such as current processor status validation data for the qubit.
  • the feedback system uses the historical data to determine that a current operating status of the qubit is bad.
  • the feedback system determines that a third keepout cost function should be added to the updated calibration model, where the keepout cost function penalizes solutions to the optimization performed using the calibration model that converge on solution 412.
  • the system can use the historical data to determine modifications to a calibration model included in the calibration system, e.g., modifications that improve the calibration model.
  • the modifications can also include adding randomness to the calibration model. Injecting randomness into the calibration system can be a powerful feedback mechanism for promoting exploration of low and high dimensional solution spaces and can facilitate the discovery of good/better calibrated parameters, particularly at system scale.
  • the system can use the historical data to determine that random noise, e.g., random Gaussian noise, should be added to the calibration model. For example, the system can determine that the calibration model has not been updated for a predetermined amount of time and that it might be beneficial to explore other operating parameter configurations to reach better operating parameter configurations. The system can then generate and send instructions to the calibration and quantum computing pipeline that cause the calibration system to add random noise to the calibration model.
  • random noise e.g., random Gaussian noise
  • Example model feedback trust-based temporal weighting
  • the system can use the historical data to determine modifications to a calibration model included in the calibration system, e.g., modifications that improve the calibration model.
  • the modifications can include modifying parameters of the calibration models based on a schedule that depends on the calibration iteration number, and/or the current calibration runtime, and/or some other arbitrary metric.
  • the calibration model can include multiple component models, where each component model corresponds to a respective and distinct error channel.
  • a temporal weighting strategy can be implemented to reduce the relative amplitudes of “low trust” component models over iterations/time (where a “low trust” model is understood to refer to a model with an uncertain reliability ).
  • Example low-trust models include heuristic models without substantial physics backing, models for physical processes that are not well understood, models for physical processes that are expected to vary' in time - for example relaxation rate models, which can vary unpredictably due to TLS fluctuations.
  • example high-trust models include models that are based only on well understood and/or accurately measurable and/or reasonably static parameters - for example, models based only on qubit circuit parameters such as capacitances and/or inductances and/or Josephson-junction resistances.
  • the system can use the historical data to determine that calibration model parameters should be adjusted.
  • the historical data can include a modification schedule (a temporal weighting strategy) that specifies when and which model parameters should be adjusted, e.g., at which calibration iterations.
  • the system can use the historical data to determine that the modification schedule indicates that calibration model parameters should be adjusted at the next calibration iteration.
  • the system can then use the historical data to identify which calibration model parameters should be adjusted, e.g., which component models are low trust.
  • the system can then adjust the identified calibration model parameters. For example, the system can determine modifications to the calibration model that reduce relative amplitudes of low trust component models.
  • the system can then send instructions to the calibration and quantum computing pipeline that cause the calibration system to update the calibration model accordingly.
  • FIG. 5 illustrates an example process 500 for using historical data to implement trust-based temporal weighting of a calibration model.
  • the calibration system included in the calibration and quantum computing pipeline uses a current calibration model 502 to calibrate an operating parameter f S q of a qubit.
  • the calibration converges on solution 504, represented by the vertical dashed line.
  • the calibrated value 504 can be stored as historical data in the feedback system, along with other data such as current processor status validation data for the qubit.
  • the feedback system can use the historical data to determine that the modification schedule specifies that parameters of the calibration model should be adjusted for a next calibration iteration.
  • the modifications schedule can specify that both the dephasing rate and cross talk for the qubit are high trust and therefore the dephasing and crosstalk components of the calibration model do not need to be modified.
  • the modification schedule can also specify that the relaxation rate for the qubit is expected to become less trustworthy over time, due to, for example TLS fluctuations. To reflect the reduction in trust over time, the system can determine that the relative amplitude of the relaxation rate component of the calibration model should be reduced (relative to the more trustworthy dephasing and crosstalk components).
  • the feedback system applies the modifications to the calibration model.
  • the modifications cause a reduction in the amplitude of the peak 506.
  • the feedback system only applies the modifications to the calibration model if a current status of the calibrated value of the operating parameter for the qubit is bad. Stages (A)-(C) can be repeated until a status of the calibrated value of the operating parameter for the qubit is good.
  • the system can use the historical data to determine modifications to a calibration model included in the calibration system, e.g., modifications that improve the calibration model.
  • the modifications can be determined based on correlations between calibrated operating parameters and relevant performance metrics, e.g., validation outputs.
  • the feedback system can determine that a sufficient amount of data has been accumulated over previous calibration iterations, e.g., an amount of data that exceeds a predetermined threshold.
  • the feedback system can then compute correlations between calibrated operating parameters and relevant performance metrics.
  • An example operating parameter includes qubit idle frequencies.
  • An example performance metric includes randomized benchmarking error. For example, for frequency-tunable qubits, it is often better to operate qubits near their respective maximum frequencies to reduce dephasing. This could surface in a correlation study e.g. smaller detunings from maximum frequencies may be correlated with lower errors and the feedback system would bias qubits towards maximum frequencies.
  • the system can use the historical data to determine modifications to a calibration model included in the calibration system, e.g., modifications that improve the calibration model.
  • the modifications can be determined based on interpolations or extrapolations.
  • the feedback system can generate or otherwise obtain models, e.g., a simple polynomial model or a more advanced model such as a trained neural network, to predict the performance of calibrated operating parameters in previously unexplored regimes.
  • the feedback system can determine modifications to the calibration by adding the model or some transformation of the model to the calibration model to bias the calibration system towards solutions that are predicted to be good. For example, if the amount of historical data is limited, e.g., below a predetermined threshold, the model can bias calibration via a transformation of a computed gradient and/or hessian and/or some other simple metric.
  • FIG. 7 is a block diagram of an example process 700 for determining a biasing model for a calibration model.
  • the feedback system uses historical data to generate a model that correlates an operating parameter f S g of a qubit with a performance metric.
  • the feedback system determines a bias to add to the calibration model to steer future calibrated values of the operating parameter of the qubit towards values that were historically found to be good.
  • the biasing model is a multi-dimensional polynomial bias.
  • Example model feedback forecast-based
  • the system can use the historical data to determine modifications to a calibration model included in the calibration system, e.g., modifications that improve the calibration model.
  • the modifications can be determined using forecasting models.
  • the feedback system can generate or otherwise obtain dynamic models that forecast where good/bad calibrated parameters may lie in the future.
  • the model can be a simple polynomial, e.g., one that extrapolates where TLS frequencies may be in the future, or a more advanced recurrent LSTM neural network.
  • the feedback system can determine modifications to the calibration by adding the dynamic model to the calibration model, e.g., as a biasing cost function.
  • Such feedback can be particularly relevant to calibrating parameters whose optimal values vary in time. For example, optimal gate frequencies fluctuate in time due to TLS resonance frequencies varying in time unpredictably on all timescales).
  • the system can use the historical data to determine modifications to a calibration scheduler included in the calibration system.
  • the modifications can include modifications that modify how the calibration scheduler selects operating parameters to be calibrated.
  • the system can use the historical data to identify operating parameters that failed a calibration a predefined number of times, e.g., a number of times that exceeds a predetermined acceptable threshold.
  • the system can determine modifications to the calibration scheduler that modify how the calibration scheduler selects operating parameters to be calibrated, e.g., causes the calibration scheduler to select and calibrate additional interacting elements or operating parameters when the identified operating parameters are calibrated.
  • the system can then send instructions to the calibration and quantum computing pipeline that cause the calibration system to update the calibration scheduler accordingly.
  • the modifications can cause the calibration scheduler to select and calibrate qubits that neighbor the qubit that has failed the calibration, e.g., qubits that are directly coupled to the qubit that failed the calibration.
  • the number of calibration targets can be expanded iteratively and/or recursively, e.g., increasing the size of the neighborhood, until the qubit passes the calibration. In extreme cases, this may lead to the simultaneous selection of all operating parameters of one type, e.g. all gate frequencies for all qubits. Selecting more parameters promotes the calibration/optimization system to explore a higher dimensional and larger search space for better solutions.
  • FIG. 8 shows an example process 800 for modifying how calibration targets are selected. After calibration iteration 820, it is determined that qubit 802 has failed a calibration more than a threshold number of times N thresh . The feedback system can then use this information to modify how the calibration scheduler selects operating parameters for calibration, e.g., by instructing the calibration scheduler to also calibrate qubits 804 and 806 when qubit 802 is calibrated since qubits 804 and 806 are directly coupled to qubit 802.
  • FIG. 9 illustrates an example process 900 for using historical data to apply keepout cost functions and select additional calibration targets.
  • the calibration system included in the calibration and quantum computing pipeline uses a current calibration model 902 to calibrate an operating parameter f S g of a qubit.
  • the calibrated value can be stored as historical data in the feedback system, along with other data such as current processor status validation data for the qubit.
  • the feedback system can use the historical data to determine that the qubit failed calibration.
  • the feedback system can use the historical data to determine whether the qubit has failed calibration a predetermined number of times.
  • the qubit has not failed calibration the predetermined number of times. Therefore, at stage (C), the feedback system modifies the calibration model to include a keepout cost function that penalizes future optimizations that converge to the failed calibrated value.
  • the calibration system uses the updated calibration model (that includes the keepout cost function) to calibrate the operating parameter.
  • the newly calibrated value can be stored as historical data in the feedback system, along with other data such as current processor status validation data for the qubit.
  • the feedback system can use the historical data to determine that the qubit failed calibration again. In response, the feedback system can use the historical data to determine whether the qubit has failed calibration a predetermined number of times. In this example, the qubit has failed calibration the predetermined number of times. Therefore, at stage (F), the feedback system modifies the calibration scheduler so that the calibration scheduler selects an operating parameter for a qubit B that interacts with qubit A and causes subsequent calibrations to also be performed on qubit B when qubit A is calibrated. In this example the feedback system also modifies the calibration model to include a second keepout cost function that penalizes future optimizations that converge to the second failed calibrated value.
  • the system can use the historical data to determine modifications to a calibration optimizer included in the calibration system, e.g., modifications that improve the calibration optimizer.
  • the modifications can include modifications that modify how the calibration optimizer optimizes the calibration model to determine calibrated values of operating parameters, e.g.. modify the optimization algorithm used to optimizes the calibration model.
  • Such modifications can be particularly beneficial when the calibration model changes (or is expected to change) significantly between iterations. For example, selecting additional operating parameters, e.g., in targeted feedback, or reducing the number of operating parameters to be calibrated can expand or contract the dimension of the calibration model. Therefore, different optimization algorithms can be applied to balance processing efficiency and the qualify of obtained solutions. For example, for a ID calibration model exhaustive search can be used. For larger dimension calibration models that are convex, gradient descent can be used. For larger dimension calibration models that are not convex, simulated annealing or differential evolution can be used.
  • the system can use the historical data to determine modifications to a calibration scheduler included in the calibration system.
  • the modifications can include modifications that cause the calibration scheduler to revisit calibrated operating parameters that previously transitioned from “good” to “bad”. This strategy can be beneficial for operating parameters that spontaneously change from “good” to “bad” or even from “bad” to good”. For example, when optimizing gate frequencies, TLS can move into and out of gate frequency trajectories on arbitrary time scales.
  • the feedback system can determine that the calibration event has occurred by using the historical data to identify one or more calibrated operating parameters that transitioned from a “good” to “bad” value between calibration iterations.
  • the system generate and send instructions to the calibration scheduler that cause the scheduler to revisit the identified operating parameters in subsequent calibration iterations.
  • the system can use the historical data to determine model-free modifications to the calibration and quantum computing pipeline.
  • feedback can be applied without an explicit calibration model - for example as a trial-and-error system that varies one or more operating parameters based on rules applied to a sequence of validation values in discrete steps.
  • Model-free feedback can run faster than model-based feedback, and in some implementations can reduce metrology and analysis overhead. Such feedback can therefore be particularly useful for online feedback, where validation values may be streamed directly from the quantum processor during algorithm runtime - for example real-time parity outputs of a quantum error correction algorithm such as a surface code.
  • one single-qubit gate frequency f s ⁇ t can be tuned while a corresponding validation metric V t is measured over an indefinite number of time steps t.
  • the feedback system can process historical pairs ar
  • Such strategies can be extended to tuning arbitrary sets of operating parameters.
  • the feedback system can determine that the calibration event has occurred by using the historical data to select one or more operating parameters to be calibrated, e.g., operating parameters that have not been calibrated in a previous predefined number of iterations.
  • the system can process the historical data to compare historical values of the operating parameters and corresponding values of validation metrics.
  • the system can determine how to tune the operating parameters, e.g., directions in which to tune, and send instructions to the calibration system that cause the scheduler to use the directions in subsequent calibration iterations.
  • FIG. 10 depicts an example quantum processor 1000 that can be included in the calibration and quantum computing pipeline described in this specification.
  • the example quantum processor 1000 includes an example quantum computing device 1002.
  • the quantum computing device 1002 is intended to represent various forms of quantum computing devices.
  • the components shown here, their connections and relationships, and their functions, are exemplary only, and do not limit implementations of the inventions described and/or claimed in this document.
  • the example quantum computing device 1002 includes a qubit assembly 1052 and a control and measurement system 1004.
  • the qubit assembly includes multiple physical qubits, e.g., qubit 1006, that are used to perform algorithmic operations or quantum computations. While the qubits shown in FIG. 10 are arranged in a rectangular array, this is a schematic depiction and is not intended to be limiting.
  • the qubit assembly 1052 also includes adjustable coupling elements, e.g., coupler 1008, that allow for interactions between coupled qubits. In the schematic depiction of FIG. 10, each qubit is adjustably coupled to each of its four adjacent qubits by means of respective coupling elements.
  • Each qubit can be a physical two-level quantum system or device having levels representing logical values of 0 and 1.
  • the specific physical realization of the multiple qubits and how they interact with one another is dependent on a variety of factors including the type of the quantum computing device 1002 included in the example computer 1000 or the type of quantum computations that the quantum computing device is performing.
  • the qubits may be realized via atomic, molecular or solid-state quantum systems, e.g., hyperfine atomic states.
  • the qubits may be realized via superconducting qubits or semi-conducting qubits, e.g., superconducting transmon states.
  • the qubits may be realized via nuclear spin states.
  • a quantum computation can proceed by loading qubits, e.g., from a quantum memory, and applying a sequence of un it ary operators to the qubits. Applying a unitary operator to the qubits can include applying a corresponding sequence of quantum logic gates to the qubits.
  • Example quantum logic gates include single-qubit gates, e.g., Pauli-X, Pauli-Y, Pauli-Z (also referred to as X, Y, Z), Hadamard gates, S gates, rotations, two-qubit gates, e.g., controlled-X, controlled-Y.
  • controlled-Z also referred to as CX, CY, CZ
  • controlled NOT gates also referred to as CNOT
  • iSWAP gates and gates involving three or more qubits, e.g., Toffoli gates.
  • the quantum logic gates can be implemented by applying control signals 1010 generated by the control and measurement system 1004 to the qubits and to the couplers.
  • the qubits in the qubit assembly 1052 can be frequency tunable.
  • each qubit can have associated operating frequencies that can be adjusted through application of voltage pulses via one or more drive-lines coupled to the qubit.
  • Example operating frequencies include qubit idling frequencies, qubit interaction frequencies, and qubit readout frequencies. Different frequencies correspond to different operations that the qubit can perform. For example, setting the operating frequency to a corresponding idling frequency may put the qubit into a state where it does not strongly interact with other qubits, and where it may be used to perform single-qubit gates.
  • qubits can be configured to interact with one another by setting their respective operating frequencies at some gate-dependent frequency detuning from their common interaction frequency.
  • qubits can be configured to interact with one another by setting the parameters of their respective couplers to enable interactions between the qubits and then by setting the qubit’s respective operating frequencies at some gate-dependent frequency detuning from their common interaction frequency. Such interactions may be performed in order to perform multi-qubit gates.
  • control signals 1010 depends on the physical realizations of the qubits.
  • the control signals may include RF or microwave pulses in an NMR or superconducting quantum computer system, or optical pulses in an atomic quantum computer system.
  • a quantum computation can be completed by measuring the states of the qubits, e.g., using a quantum observable such as X, Y, or Z, using respective control signals 1010.
  • the measurements cause readout signals 1012 representing measurement results to be communicated back to the measurement and control system 1004.
  • the readout signals 1012 may include RF, microwave, or optical signals depending on the physical scheme for the quantum computing device and/or the qubits.
  • the control signals 1010 and readout signals 1012 shown in FIG. 10 are depicted as addressing only selected elements of the qubit assembly (i.e. the top and bottom rows), but during operation the control signals 1010 and readout signals 1012 can address each element in the qubit assembly 1052.
  • the control and measurement system 1004 is an example of a classical computer system that can be used to perform various operations on the qubit assembly 1052, as described above, as well as other classical subroutines or computations.
  • the control and measurement system 1004 includes one or more classical processors, e.g., classical processor 1014, one or more memories, e.g., memory 1016, and one or more I/O units, e.g., I/O unit 1018, connected by one or more data buses.
  • the control and measurement system 1004 can be programmed to send sequences of control signals 1010 to the qubit assembly, e.g. to carry out a selected series of quantum gate operations, and to receive sequences of readout signals 1012 from the qubit assembly, e.g. as part of performing measurement operations and post processing measurement results.
  • the processor 1014 is configured to process instructions for execution within the control and measurement system 1004. In some implementations, the processor 1014 is a single-threaded processor. In other implementations, the processor 1014 is a multithreaded processor. The processor 1014 is capable of processing instructions stored in the memory 1016.
  • the memory 1016 stores information within the control and measurement system 1004.
  • the memory 1016 includes a computer-readable medium, a volatile memory unit, and/or a non-volatile memory unit.
  • the memory 1016 can include storage devices capable of providing mass storage for the system 1004, e.g. a hard disk device, an optical disk device, a storage device that is shared over a network by multiple computing devices (e.g., a cloud storage device), and/or some other large capacity storage device.
  • the input/ output device 1018 provides input/output operations for the control and measurement system 1004.
  • the input/output device 1018 can include D/A converters, A/D converters, and RF/microwave/optical signal generators, transmitters, and receivers, whereby to send control signals 1010 to and receive readout signals 1012 from the qubit assembly, as appropriate for the physical scheme for the quantum computer.
  • the input/output device 1018 can also include one or more network interface devices, e.g., an Ethernet card, a serial communication device, e.g., an RS-232 port, and/or a wireless interface device, e g., an 802.8 card.
  • the input/output device 1018 can include driver devices configured to receive input data and send output data to other external devices, e.g., keyboard, printer and display devices.
  • implementations of the subject matter and the functional operations described in this specification can be implemented in other types of digital electronic circuitry, or in computer software, firmware, or hardware, including the structures disclosed in this specification and their structural equivalents, or in combinations of one or more of them.
  • Implementations of the digital and/or quantum subject matter and the digital functional operations and quantum operations described in this specification can be implemented in digital electronic circuitry, suitable quantum circuitry or, more generally, quantum computational systems, in tangibly-embodied digital and/or quantum computer software or firmware, in digital and/or quantum computer hardware, including the structures disclosed in this specification and their structural equivalents, or in combinations of one or more of them.
  • quantum processors may include, but is not limited to, quantum computers, quantum information processing systems, quantum cryptography systems, or quantum simulators.
  • Implementations of the digital and/or quantum subject matter described in this specification can be implemented as one or more digital and/or quantum computer programs, i.e., one or more modules of digital and/or quantum computer program instructions encoded on a tangible non-transitory storage medium for execution by, or to control the operation of, data processing apparatus.
  • the digital and/or quantum computer storage medium can be a machine-readable storage device, a machine-readable storage substrate, a random or serial access memory device, one or more qubits, or a combination of one or more of them.
  • the program instructions can be encoded on an artificially-generated propagated signal that is capable of encoding digital and/or quantum information, e.g., a machine-generated electrical, optical, or electromagnetic signal, that is generated to encode digital and/or quantum information for transmission to suitable receiver apparatus for execution by a data processing apparatus.
  • digital and/or quantum information e.g., a machine-generated electrical, optical, or electromagnetic signal
  • quantum information and quantum data refer to information or data that is carried by, held or stored in quantum systems, where the smallest non-trivial system is a qubit, i.e., a system that defines the unit of quantum information.
  • qubit encompasses all quantum systems that may be suitably approximated as a two-level system in the corresponding context.
  • Such quantum systems may include multi-level systems, e.g., with two or more levels.
  • such systems can include atoms, electrons, photons, ions or superconducting qubits.
  • the computational basis states are identified with the ground and first excited states, however it is understood that other setups where the computational states are identified with higher level excited states are possible.
  • data processing apparatus refers to digital and/or quantum data processing hardware and encompasses all kinds of apparatus, devices, and machines for processing digital and/or quantum data, including by way of example a programmable digital processor, a programmable quantum processor, a digital computer, a quantum computer, multiple digital and quantum processors or computers, and combinations thereof.
  • the apparatus can also be, or further include, special purpose logic circuitry', e.g., an FPGA (field programmable gate array), an ASIC (application-specific integrated circuit), or a quantum simulator, i.e.. a quantum data processing apparatus that is designed to simulate or produce information about a specific quantum system.
  • a quantum simulator is a special purpose quantum computer that does not have the capability to perform universal quantum computation.
  • the apparatus can optionally include, in addition to hardware, code that creates an execution environment for digital and/or quantum computer programs, e.g., code that constitutes processor firmware, a protocol stack, a database management system, an operating system, or a combination of one or more of them.
  • code that creates an execution environment for digital and/or quantum computer programs e.g., code that constitutes processor firmware, a protocol stack, a database management system, an operating system, or a combination of one or more of them.
  • a digital computer program which may also be referred to or described as a program, software, a software application, a module, a software module, a script, or code, can be written in any form of programming language, including compiled or interpreted languages, or declarative or procedural languages, and it can be deployed in any form, including as a stand-alone program or as a module, component, subroutine, or other unit suitable for use in a digital computing environment.
  • a quantum computer program which may also be referred to or described as a program, software, a software application, a module, a software module, a script, or code, can be written in any form of programming language, including compiled or interpreted languages, or declarative or procedural languages, and translated into a suitable quantum programming language, or can be written in a quantum programming language, e.g., QCL or Quipper.
  • a digital and/or quantum computer program may, but need not, correspond to a file in a file system.
  • a program can be stored in a portion of a file that holds other programs or data, e.g.. one or more scripts stored in a markup language document, in a single file dedicated to the program in question, or in multiple coordinated files, e g., files that store one or more modules, sub-programs, or portions of code.
  • a digital and/or quantum computer program can be deployed to be executed on one digital or one quantum computer or on multiple digital and/or quantum computers that are located at one site or distributed across multiple sites and interconnected by a digital and/or quantum data communication network.
  • a quantum data communication network is understood to be a network that may transmit quantum data using quantum systems, e.g. qubits. Generally, a digital data communication network cannot transmit quantum data, however a quantum data communication network may transmit both quantum data and digital data.
  • the processes and logic flows described in this specification can be performed by one or more programmable digital and/or quantum computers, operating with one or more digital and/or quantum processors, as appropriate, executing one or more digital and/or quantum computer programs to perform functions by operating on input digital and quantum data and generating output.
  • the processes and logic flows can also be performed by, and apparatus can also be implemented as, special purpose logic circuitry, e.g., an FPGA or an ASIC, or a quantum simulator, or by a combination of special purpose logic circuitry or quantum simulators and one or more programmed digital and/or quantum computers.
  • a sy stem of one or more digital and/or quantum computers to be “configured to” perform particular operations or actions means that the system has installed on it software, firmware, hardware, or a combination of them that in operation cause the system to perform the operations or actions.
  • one or more digital and/or quantum computer programs to be configured to perform particular operations or actions means that the one or more programs include instructions that, when executed by digital and/or quantum data processing apparatus, cause the apparatus to perform the operations or actions.
  • a quantum computer may receive instructions from a digital computer that, when executed by the quantum computing apparatus, cause the apparatus to perform the operations or actions.
  • Digital and/or quantum computers suitable for the execution of a digital and/or quantum computer program can be based on general or special purpose digital and/or quantum processors or both, or any other kind of central digital and/or quantum processing unit.
  • a central digital and/or quantum processing unit will receive instructions and digital and/or quantum data from a read-only memory, a random access memory, or quantum systems suitable for transmitting quantum data, e.g. photons, or combinations thereof .
  • the essential elements of a digital and/or quantum computer are a central processing unit for performing or executing instructions and one or more memory devices for storing instructions and digital and/or quantum data.
  • the central processing unit and the memory can be supplemented by, or incorporated in, special purpose logic circuitry or quantum simulators.
  • a digital and/or quantum computer will also include, or be operatively coupled to receive digital and/or quantum data from or transfer digital and/or quantum data to, or both, one or more mass storage devices for storing digital and/or quantum data, e.g., magnetic, magneto-optical disks, optical disks, or quantum systems suitable for storing quantum information.
  • mass storage devices for storing digital and/or quantum data, e.g., magnetic, magneto-optical disks, optical disks, or quantum systems suitable for storing quantum information.
  • a digital and/or quantum computer need not have such devices.
  • Digital and/or quantum computer-readable media suitable for storing digital and/or quantum computer program instructions and digital and/or quantum data include all forms of non-volatile digital and/or quantum memory, media and memory' devices, including by way of example semiconductor memory devices, e.g., EPROM, EEPROM, and flash memory devices; magnetic disks, e.g., internal hard disks or removable disks; magneto-optical disks; CD-ROM and DVD-ROM disks; and quantum systems, e.g., trapped atoms or electrons.
  • semiconductor memory devices e.g., EPROM, EEPROM, and flash memory devices
  • magnetic disks e.g., internal hard disks or removable disks
  • magneto-optical disks CD-ROM and DVD-ROM disks
  • quantum systems e.g., trapped atoms or electrons.
  • quantum memories are devices that can store quantum data for a long time with high fidelity' and efficiency, e.g., light-matter interfaces where light is used for transmission and matter for storing and preserving the quantum features of quantum data such as superposition or quantum coherence.
  • Control of the various systems described in this specification, or portions of them, can be implemented in a digital and/or quantum computer program product that includes instructions that are stored on one or more non-transitory machine-readable storage media, and that are executable on one or more digital and/or quantum processing devices.
  • the systems described in this specification, or portions of them, can each be implemented as an apparatus, method, or system that may include one or more digital and/or quantum processing devices and memory to store executable instructions to perform the operations described in this specification.

Landscapes

  • Engineering & Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Mathematical Analysis (AREA)
  • Computing Systems (AREA)
  • Evolutionary Computation (AREA)
  • Condensed Matter Physics & Semiconductors (AREA)
  • Computational Mathematics (AREA)
  • Mathematical Optimization (AREA)
  • Pure & Applied Mathematics (AREA)
  • Data Mining & Analysis (AREA)
  • General Engineering & Computer Science (AREA)
  • Mathematical Physics (AREA)
  • Software Systems (AREA)
  • Artificial Intelligence (AREA)
  • Complex Calculations (AREA)
  • Measurement Of Radiation (AREA)

Abstract

L'invention concerne des procédés, des systèmes et un appareil d'étalonnage de processeur quantique avec mémoire et rétroaction. Selon un aspect, un procédé fait appel à la détermination, par un système de rétroaction et sur la base de données historiques générées par un pipeline d'étalonnage et de calcul quantique et stockées par le système de rétroaction, du fait qu'un événement de modification d'étalonnage est nécessaire ; en réponse, à la détermination, par le système de rétroaction et à l'aide des données historiques, d'une ou de plusieurs modifications apportées à des composants du pipeline d'étalonnage et de calcul quantique ; et à l'application de la ou des modifications aux composants du pipeline d'étalonnage et de calcul quantique, des procédures d'étalonnage ultérieures étant mises en oeuvre sur des bits quantiques physiques inclus dans le pipeline d'étalonnage et de calcul quantique à l'aide des composants modifiés.
EP23959057.3A 2023-04-07 2023-04-07 Étalonnage de processeur quantique avec mémoire et rétroaction Pending EP4666227A2 (fr)

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
PCT/US2023/017973 WO2025128070A2 (fr) 2023-04-07 2023-04-07 Étalonnage de processeur quantique avec mémoire et rétroaction

Publications (1)

Publication Number Publication Date
EP4666227A2 true EP4666227A2 (fr) 2025-12-24

Family

ID=95782142

Family Applications (1)

Application Number Title Priority Date Filing Date
EP23959057.3A Pending EP4666227A2 (fr) 2023-04-07 2023-04-07 Étalonnage de processeur quantique avec mémoire et rétroaction

Country Status (2)

Country Link
EP (1) EP4666227A2 (fr)
WO (1) WO2025128070A2 (fr)

Family Cites Families (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US9940212B2 (en) 2016-06-09 2018-04-10 Google Llc Automatic qubit calibration
CA3085717C (fr) * 2017-12-14 2023-04-18 Google Llc Etalonnage de bits quantiques
WO2019117955A1 (fr) * 2017-12-15 2019-06-20 Klimov Paul Affinage de modèles d'étalonnage de bits quantiques à l'aide d'un apprentissage supervisé
WO2019168544A1 (fr) 2018-03-02 2019-09-06 Google Llc Optimisation de fréquences de fonctionnement de bits quantiques
US11699088B2 (en) 2019-06-07 2023-07-11 Google Llc Calibration of quantum processor operator parameters

Also Published As

Publication number Publication date
WO2025128070A3 (fr) 2025-10-09
WO2025128070A2 (fr) 2025-06-19
WO2025128070A9 (fr) 2025-09-04

Similar Documents

Publication Publication Date Title
AU2024200550B2 (en) Verified quantum phase estimation
US12380355B2 (en) Automatic qubit calibration
US20250131310A1 (en) Qubit calibration
US20250278665A1 (en) Compensation pulses for qubit readout
AU2024201618B2 (en) Quantum computing with kernel methods for machine learning
EP3759655B1 (fr) Optimisation des fréquences de fonctionnement des qubits
US12579460B2 (en) Iterative supervised learning of quantum processor error models
US20220261679A1 (en) Probe of qubit-resonator dispersive shift using ac stark shift
US12204406B2 (en) Weighted alternating paths in graphs for quantum computing
US12572832B2 (en) Backtesting quantum device calibration
US12614101B2 (en) Preprocessing for correlated topological quantum error correction
EP4666227A2 (fr) Étalonnage de processeur quantique avec mémoire et rétroaction
EP4666222A1 (fr) Réétalonnage de processeur quantique basé sur une métrique
US20260099742A1 (en) Systems and Methods for Evaluating Scalability of Quantum Computing Systems
WO2026096898A1 (fr) Atténuation d'erreurs qubit par soustraction de signaux corrélés

Legal Events

Date Code Title Description
STAA Information on the status of an ep patent application or granted ep patent

Free format text: STATUS: UNKNOWN

STAA Information on the status of an ep patent application or granted ep patent

Free format text: STATUS: THE INTERNATIONAL PUBLICATION HAS BEEN MADE

PUAI Public reference made under article 153(3) epc to a published international application that has entered the european phase

Free format text: ORIGINAL CODE: 0009012

STAA Information on the status of an ep patent application or granted ep patent

Free format text: STATUS: REQUEST FOR EXAMINATION WAS MADE

17P Request for examination filed

Effective date: 20250919

AK Designated contracting states

Kind code of ref document: A2

Designated state(s): AL AT BE BG CH CY CZ DE DK EE ES FI FR GB GR HR HU IE IS IT LI LT LU LV MC ME MK MT NL NO PL PT RO RS SE SI SK SM TR