Systems and methods for superconducting circuits and superconducting devices used in scalable computing - Patents.com
The integration of QFP-based shift registers and non-dissipative PMM in superconducting circuits addresses scalability issues in quantum processors, enhancing data input/output speed and reducing thermal load, thus improving the performance of large-scale quantum processors.
Patent Information
- Application Number
- JP2023125941
- Authority / Receiving Office
- JP · JP
- Patent Type
- Patents
- Current Assignee / Owner
- Priority Date
- 2016-10-06
- Filing Date
- 2023-08-02
- Publication Date
- 2024-11-15
- Estimated Expiration
- 2037-05-03
AI Technical Summary
Existing superconducting quantum processors face scalability limitations due to complex qubit parameter control systems, inefficient digital-to-analog converters (DACs), and non-dissipative readout (NDRO) designs that do not scale well with increasing processor size, leading to thermal load and performance trade-offs.
Implementing quantum flux parametron (QFP)-based shift registers, multiplexers, and demultiplexers in superconducting circuits with adiabatic operation, utilizing XY or XYZ addressing schemes, and non-dissipative programmable magnetic memory (PMM) to enhance scalability and reduce thermal load, while using QFP-DACs without bias resistors for efficient signal conversion.
This approach allows for increased data input/output speed without adding input/output wires, improving processor performance by reducing thermal load and enhancing scalability, enabling larger quantum processors with faster programming and readout capabilities.
Smart Images

Figure 0007584589000004 
Figure 0007584589000005 
Figure 0007584589000006
Abstract
Description
[Technical field]
[0001] The present disclosure relates generally to superconducting circuits and superconducting devices used in scalable computing, such as quantum computing, and more particularly to the design of superconducting digital-to-analog converters (DACs) in quantum processors and programming or reading logic elements such as qubits in quantum processors. [Background technology]
[0002] Quantum Bits The fundamental unit of information in a quantum computer can be a qubit. A qubit can represent an actual physical device in which information is stored, or it can represent an abstraction of the physical device that is the unit of information itself. Examples of qubits include quantum particles, atoms, electrons, photons, and ions.
[0003] A quantum bit generalizes the concept of a classical digital bit. A quantum bit has two discrete physical states that can also be distinguished as "0" and "1". Physically, these two discrete states are represented by two different distinguishable physical states of a quantum information storage device, such as the direction or magnitude of a magnetic field, current or voltage, whose quantities encoding the bit state behave according to the laws of quantum physics. If the physical quantities storing these states behave quantum mechanically, the device can further be in a superposition of 0 and 1. That is, a quantum bit can exist in both "0" and "1" states at the same time, and therefore can perform calculations on both states at the same time. In general, N quantum bits can be in a superposition of 2N states.
[0004] In standard notation, the basis states of a qubit are referred to as the |0> and |1> states. During quantum computing, the state of a qubit typically consists of a superposition of basis states such that the qubit has a nonzero probability of occupying the |0> basis state and at the same time a nonzero probability of occupying the |1> basis state. Mathematically, a superposition of basis states means that the full state of the qubit, denoted |Ψ>, has the form |Ψ>=a|0>+b|1>, where a and b are coefficients corresponding to the probabilities |a|2 and |b|2, respectively. The coefficients a and b each have real and imaginary components that allow one to characterize the phase of the qubit. The quantum nature of a qubit comes largely from the ability of the qubit to exist in a coherent superposition of basis states, and that the state of the qubit has a phase. A qubit retains this ability to exist as a coherent superposition of basis states if the qubit is sufficiently isolated from sources of decoherence.
[0005] To complete a computation with a qubit, the state of the qubit is measured (i.e., read out). Typically, when a qubit is measured, the quantum nature of the qubit is temporarily lost and the superposition of basis states collapses to either a |0> basis state or a |1> basis state, resulting in similarity to a classical bit. The actual state of the qubit after it collapses depends on the probabilities |a|2 and |b|2 immediately prior to the read operation.
[0006] Superconducting qubits There are many different hardware and software approaches being considered for use in quantum computers. One hardware approach uses integrated circuits formed from superconducting materials such as aluminum or niobium.
[0007] A superconducting qubit is a type of superconducting device that can be included in a superconducting integrated circuit. Exemplary superconducting qubits have the advantage of, for example, scalability, and are generally classified according to the physical property used to encode information, such as charge and phase devices, phase or flux devices, and hybrid devices. Charge devices store and manipulate information in the charge state of the device, where the unit charge consists of an electron pair called a Cooper pair. A Cooper pair has a charge of 2e and consists of two electrons bound by, for example, phonon interactions. Flux devices store information in a variable related to the magnetic flux through a portion of the device. Phase devices store information in a variable related to the superconducting phase difference between two regions of the phase device. Recently, hybrid devices have been developed that use two or more of the charge, flux, and phase degrees of freedom.
[0008] Examples of flux qubits include radio frequency SQUIDs that contain a superconducting loop interleaved with one Josephson junction, or a compound Josephson junction (where the single Josephson junction is replaced by two parallel Josephson junctions), or persistent current qubits that contain a superconducting loop interleaved with three Josephson junctions, etc. An example of a qubit is the hybrid charge phase qubit.
[0009] The qubits may be coupled to corresponding local bias devices. The local bias devices may include metal loops in close proximity to the superconducting qubits that provide an external magnetic flux to the qubits. The local bias devices may also include multiple Josephson junctions. Each superconducting qubit in a quantum processor may have a corresponding local bias device, or there may be fewer local bias devices than qubits. Some examples use charge-based readout and local bias devices. A conventional readout device includes a DC SQUID magnetometer inductively coupled to each qubit in the topology. The readout device may provide a voltage or a current. The DC SQUID magnetometer is typically formed by a loop of superconducting material interposed with at least one Josephson junction.
[0010] Quantum Processor The computer processor may take the form of an analog processor, e.g., a quantum processor, such as a superconducting quantum processor. A superconducting quantum processor may include a plurality of qubits and coupled local bias devices, e.g., two or more superconducting qubits. Further details and embodiments of exemplary quantum processors are described in U.S. Patent Publication No. 2006-0225165, entitled "Systems, Devices and Methods for Analog Processing," filed November 8, 2007, U.S. Patent Application No. 12 / 013,192, and U.S. Provisional Patent Application No. 60 / 986,554.
[0011] A superconducting quantum processor may include a number of coupling devices operable to selectively couple each pair of qubits. Examples of superconducting coupling devices include radio frequency SQUIDs and DC SQUIDs that couple the qubits by magnetic flux. SQUIDs include a superconducting loop interleaved with one Josephson junction (radio frequency SQUID) or two Josephson junctions (DC SQUID). Coupling devices may allow both ferromagnetic and antiferromagnetic coupling depending on how the coupling devices are utilized in the interconnected topology. In the case of flux coupling, ferromagnetic coupling means that parallel flux is energetically favored, and antiferromagnetic coupling means that antiparallel flux is energetically favored. Alternatively, charge-based coupling devices may be used. Other coupling devices can be found, for example, in U.S. Patent Publication No. 2006-0147154 and U.S. Patent Application No. 12 / 017,995. For example, the coupling strength of each coupling device may be adjusted between 0 and a maximum value to provide ferromagnetic or antiferromagnetic coupling between the qubits.
[0012] Regardless of the specific hardware implementation, managing a single qubit requires control over multiple parameters. Traditionally, this requirement has required external communication with the individual qubits (i.e., communication from outside the processor structure). However, as the total processing power increases with the number of qubits in a system, high capacity processors that exceed the capabilities of traditional supercomputers must manage large numbers of qubits, and therefore traditional approaches that utilize external control over multiple parameters for individual qubits require complex systems to program the qubit parameters.
[0013] Therefore, the scalability of quantum processors is limited by the complexity of the quantum parameter control system, and there is a need in the art for devices that enable scalable qubit parameter control systems.
[0014] Digital-to-Analog Converter (DAC) A quantum processor provides a number of programmable devices that perform computations using quantum effects. The programmable devices include qubits, couplers (which programmably couple the qubits) and components thereof. The programmable devices are programmed by signals provided to affect the operation of the programmable devices, for example, a bias signal may be provided to a flux qubit to affect the flux of the flux qubit during a computation.
[0015] Such signals often require conversion and / or storage before being provided to a programmable device. For example, a classical computer can generate digital signals for a quantum processor, which can be converted to analog form by one or more digital-to-analog converters (DACs). The converted analog signals can be provided to a programmable device. As another example, a signal (which can be digital or analog) can be received by a quantum processor at a point before or during a computation and stored by a DAC until a later point in time when the signal is to be provided to a programmable device. DACs have many uses and can be used for one or more of these purposes (i.e., conversion and / or memory), and / or for other purposes. Examples of the use of DACs for these and other purposes are described in more detail, for example, in U.S. Pat. Nos. 7,876,248 and 8,098,179.
[0016] Superconducting quantum processors often include multiple DACs for these and other functions. Such DACs include superconducting DACs that store magnetic flux (sometimes referred to as Φ-DACs), which typically include a storage inductor (e.g., a superconducting magnet coil) and a programmable coupling element. Φ-DACs use the rate of change of magnetic flux in a circuit (e.g., a storage inductor) to store energy in a magnetic field within the circuit, thereby generating an effective inductance (sometimes referred to as a magnetic inductance).
[0017] Φ-DAC designs may impose various tradeoffs on the processor design. For example, magnetic storage inductors capable of storing sufficient magnetic flux for conventional designs are often relatively large (and may require several fabrication layers with current technologies), which may limit the space available for other components on the processor. Furthermore, the magnetic fields generated by Φ-DACs are strong and may require significant shielding. Even with shielding, Φ-DACs are prone to interference with other devices on the processor that are sensitive to magnetic flux. Furthermore, at least some Φ-DAC designs are particularly sensitive to manufacturing variations. Examples of Φ-DAC designs are described, for example, in Johnson et al., "A scalable control system for a superconducting adiabatic quantum optimization processor," arXiv:0907.3757, and Bunyk et al., "A scalable control system for a superconducting adiabatic quantum optimization processor," arXiv:0907.3757, and in ... This is described in more detail in "Architectural considerations in the design of a superconducting quantum annealing processor", arXiv:1401.5504, by L.
[0018] Therefore, there is a general need for a system and method for providing a superconducting DAC that ameliorates at least some of these shortcomings.
[0019] Quantum Flux Parametron Quantum Flux Parametrons (QFPs) are superconducting Josephson junction devices similar in some respects to composite radio frequency SQUIDs. QFP devices can be used to generate a specific potential energy curve, which may resemble a "W," where the height of the central peak or "barrier" as well as the depth of each of the two wells on either side of the central barrier are adjustable. Although the word "quantum" appears in the name of a QFP device, the device generally operates classically; that is, rapidly raising the central barrier is classically thought to be highly disruptive to the energy configuration of the system. Therefore, damping resistors are traditionally incorporated into the QFP circuit to facilitate the dissipation of energy and return the system to a stable energy configuration. These damping resistors dissipate excess energy in the form of heat, a process that can be particularly detrimental to any system that is sensitive to thermal noise. Therefore, conventional QFP circuits are generally not suitable for use in devices that are sensitive to thermal noise, such as elements of a superconducting quantum processor.
[0020] Scalability Non-dissipative readout (NDRO) data for superconducting processors The data rate (such as NDRO described in U.S. Patent No. 8,169,231) is constant regardless of processor size, so this approach does not scale to large processor sizes, e.g., quantum processors with large qubit counts, or classical superconducting processors with large device counts.
[0021] The data rate can be increased by adding more NDRO wiring and related hardware, but this approach is not scalable to reading large numbers of qubits.
[0022] The additional NDRO wiring increases the heat load on the refrigerator (i.e., fridge), raising the base temperature of the refrigerator. High power driven on-chip items can increase the heat load on the chip. There can be a trade-off between processor performance and read speed, since processor performance can depend on low chip temperature. Also, adding more wiring increases hardware costs.
[0023] The performance of a superconducting quantum processor may be limited by the number and bandwidth of the input wires. For example, in some existing implementations, a superconducting quantum processor can be accessed through approximately 200 wires, each with a bandwidth of 30 MHz. The number and bandwidth of the input wires may determine, at least in part, the speed at which the system can encode new problems on the quantum processor.
[0024] Increasing the number of wires and bandwidth is not an easily scalable approach. Adding more wires can place many demands on the system, including requiring a larger sample space and requiring more contact pads on the periphery of the processor chip. Increasing the number of wires can also increase the thermal load on the processor. Increasing the number of wires can also increase the sensitivity of the processor to more non-thermal photons. This could open up a route to
[0025] The foregoing examples of related art and their associated limitations are intended to be illustrative and not exhaustive. Other limitations of the related art will become apparent to those of ordinary skill in the art upon reading this specification and studying the drawings. Summary of the Invention
[0026] Described herein are a number of techniques that are useful for operating processors with larger numbers of logic devices (e.g., qubits) and for addressing those logic devices in a scalable manner. Many of the techniques described herein take advantage of QFPs to implement, for example, shift registers, multiplexers (i.e., MUX), demultiplexers (i.e., DEMUX), and permanent magnetic memories (i.e., PMMs). Many of the techniques described herein utilize XY or XYZ addressing schemes to provide scalability as the number of distinct logic devices on a processor chip, or an assembly of processor chips, grows. Many of these described techniques are particularly suited to performing inputs to and / or outputs (i.e., reads) from such processors.
[0027] In superconducting circuits, QFPs may be implemented as flux-based logic devices, and of course, QFPs may be used to implement superconducting shift registers, superconducting memory arrays, superconducting adders, superconducting flip-flops, and other logic-based circuits.
[0028] From a quantum mechanical standpoint, it is possible to tune the central barrier height slowly enough that the system is not excited beyond the desired ground state. This is an example of adiabatic evolution of a quantum system. Thus, whereas conventional damped QFP devices are designed for operation at high switching speeds (greater than about 20 GHz), the present systems, methods and apparatus provide undamped QFP-like devices for use in superconducting circuits where heat dissipation must be controlled. The operation of the present systems, methods and apparatus is generally controlled adiabatically. Throughout the remainder of this specification and the appended claims, the terms "adiabatically", "adiabatically" and "adiabatic frequency" are used in accordance with the following formula:
[0029]
number
[0030] Therefore, this disclosure describes input / output systems and methods that are capable of scalability to large processor sizes and provide improved performance by increasing the speed at which data can be input and / or output to a processor without increasing the number of input / output wires leading to the processor.
[0031] The superconducting readout system may further include a digital electronics subsystem for generating and detecting baseband signals, the digital electronics subsystem having a field programmable gate array (FPGA), at least one digital-to-analog converter (DAC) communicatively coupled to an output of the FPGA, and at least one analog-to-digital converter (ADC) communicatively coupled to an input of the FPGA.
[0032] The system may be summarized as comprising: a first digital-to-analog converter (DAC) set, each DAC of the first DAC set including a respective material loop and a respective Josephson junction pair interposed in the respective material loop and electrically coupled in parallel to each other at the respective material loop; a first quantum flux parametron (QFP)-based shift register including a series of a plurality of QFP-based shift register elements, the first QFP-based shift register being magnetically or galvanically coupleable to the DACs of the first DAC set; a power supply line coupling the DACs of the first DAC set in series; and a trigger line coupling the DACs of the first DAC set in series, each of the DACs of the first DAC set being independently addressable successive times to store a variable number of flux quanta by a triplet of three signals, the triplet of three signals including a first signal obtained via the QFP-based shift register elements of the first QFP-based shift register, a second signal obtained via the power supply line, and a third signal obtained via the trigger line.
[0033] The system may further include an inductor ladder circuit that undergoes superconductivity at a critical temperature and includes a plurality of inductors, and each of the QFP-based shift register elements of the first QFP-based shift register may be magnetically coupled to one of the DACs of the first DAC set via the inductor ladder circuit.
[0034] The system may further include a first set of QFP amplifiers galvanically coupled as individual flux amplifiers to individual DACs of the first set of DACs, each of the QFP amplifiers of the first set of QFP amplifiers may include an individual loop of material that exhibits superconductivity at a critical temperature and an individual pair of Josephson junctions interposed in the individual loop and electrically coupled in parallel to each other within the material loop.
[0035] The system may further include an inductor ladder circuit that undergoes superconductivity at a critical temperature and includes a plurality of inductors, and each of the QFP amplifiers of the first QFP amplifier set may be coupled between a respective one of the DACs of the first DAC set and a QFP-based shift register element of the first QFP-based shift register via the inductor ladder circuit. The first QFP-based shift register may include at least three QFP register elements for each DAC of the first DAC set. Each of the QFP register elements of the first QFP-based shift register may include a respective material loop that undergoes superconductivity at a critical temperature and a respective Josephson junction pair interposed in the respective loop and electrically coupled in parallel to each other within the material loop. Each of the QFP register elements of the first QFP-based shift register may include a respective galvanic interface to one of the plurality of address lines.
[0036] The system may further comprise a number of further DAC sets and a number of further QFP-based shift registers.
[0037] The system may further include a control circuit communicatively coupled to provide signals to the power, trigger and address lines, the control circuit sequentially providing signals to the power, trigger and address lines to load selected ones of the DACs with single flux quantum (SFQ).
[0038] The control circuit may reset all of the DACs in the first DAC set, provide a sustained clockwise current to a first plurality of QFP-based shift register elements coupled to the DACs to which the flux quanta are being summed, provide a sustained counterclockwise current to a second plurality of QFP-based shift register elements coupled to the DACs to which the flux quanta are not being loaded, and provide a signal on the trigger line a first number of times, the first number of times being at least proportional to a total number of flux quanta being summed. The number of 1's may be equal to the total number of flux quanta loaded.
[0039] The control circuit may sort all of the DACs based on the number of flux quanta loaded into the DACs, reset all of the DACs, provide a clockwise sustaining current to the plurality of QFP-based shift register elements coupled to the DAC stages loaded with at least one flux quantum, provide a counterclockwise sustaining current to the plurality of QFP-based shift register elements coupled to the DAC stages loaded with less than one flux quantum, provide a signal to a trigger line, provide a clockwise sustaining current to the plurality of QFP-based shift register elements coupled to the DACs loaded with at least two flux quanta, provide a counterclockwise sustaining current to the plurality of QFP-based shift register elements coupled to the DACs loaded with less than two flux quanta, and provide a signal to the trigger line.
[0040] The control circuit may sort all of the DAC stages based on the number of flux quanta loaded into the DACs, reset all of the DACs in all of the DAC stages, provide a clockwise sustaining current to the plurality of QFP-based shift register elements coupled to the DACs loaded with at least i flux quanta for a programmed integer number of times i ranging from 1 to a maximum number of flux quanta n, provide a counterclockwise sustaining current to the plurality of QFP-based shift register elements coupled to the DACs loaded with less than i flux quanta, and provide a signal to the trigger line. The maximum number of flux quanta n may be 18 to 22. To reset all of the DACs in the first DAC set, the control circuit may repeatedly provide a signal to the trigger line.
[0041] The computing may further include a second DAC set, each of the DACs in the second DAC set including a respective loop of material that undergoes superconductivity at a critical temperature and a respective Josephson junction pair interposed in the respective loop and electrically coupled in parallel to each other within the material loop, and a second QFP-based shift register having a plurality of QFP-based shift register element strings, the QFP-based shift register elements of the second QFP-based shift register being magnetically or galvanically coupleable to the DACs of the second DAC set, each of the DACs in the second DAC set being independently addressable successive times to store a variable number of magnetic flux quanta by a triplet of three signals, the triplet of three signals including the first, second and third signals received via the QFP-based shift register elements of the second QFP-based shift register. Each DAC has a separate loop of material that superconducts at least at a critical temperature, and all of the DACs and the QFP-based shift register may reside on a single chip.
[0042] The system may be summarized as comprising a plurality of quantum flux parametrons (QFPs), at least one quantum flux parametron digital-to-analog converter (QFP-DAC) pair, a current bias line that provides a current bias to both of the QFP-DACs of the at least one QFP-DAC pair without the use of a bias resistor, and, for each QFP of the plurality of QFPs, a respective first plurality of quantum flux parametron latches (QFP latches) selectively operable to communicatively couple magnetic flux between the QFP and a first QFP-DAC of the QFP-DAC of the at least one QFP-DAC pair, and a respective second plurality of QFP latches selectively operable to communicatively couple magnetic flux between the QFP and a second QFP-DAC of the QFP-DAC of the at least one QFP-DAC pair. Each of the QFPs includes a respective first material loop interleaved with a respective second material loop, the first material loop and the second material loop undergoing superconductivity at a critical temperature, and has respective first and second interfaces, the second material loop interleaved with two Josephson junctions, and the first inductive interface positioned to communicatively couple magnetic flux to a first QFP latch of the respective first plurality of QFP latches. and the second inductive interface may be disposed to communicatively couple magnetic flux to a first QFP latch of the respective second plurality of QFP latches. The second QFP latch of the first plurality of QFP latches and the second QFP latch of the second plurality of QFP latches may each include a respective Josephson junction material loop having two Josephson junctions interposed therebetween. Each of the QFP latches may include a respective loop comprising a first interface disposed to communicatively couple magnetic flux to a first QFP latch of the respective first plurality of QFP latches and a second interface disposed to communicatively couple magnetic flux to a first QFP latch of the respective second plurality of QFP latches.
[0043] Each first QFP latch of the first plurality of individual QFP latches may include an individual loop with an individual second interface arranged to communicatively couple magnetic flux to a second QFP latch of the first plurality of individual QFP latches, and each first QFP latch of the second plurality of individual QFP latches may include an individual loop with an individual second interface arranged to communicatively couple magnetic flux to a second QFP latch of the second plurality of individual QFP latches. The first QFP latch of each of the first plurality of QFP latches may be responsive to an individual first latch signal, and the first QFP latch of each of the second plurality of QFP latches may be responsive to an inverted logic of the individual first latch signal. The first QFP latch of each of the first plurality of QFP latches may be responsive to an individual first offset signal, and the first QFP latch of each of the second plurality of QFP latches may be responsive to an individual first offset signal. A first QFP latch of each of the first plurality of QFP latches may be responsive to a respective first address signal and a first QFP latch of each of the second plurality of QFP latches may be responsive to an inverted logic of the respective address signal. A first QFP latch of each of the first plurality of QFP latches may be responsive to a respective first latch signal and a first QFP latch of each of the second plurality of QFP latches may be responsive to a respective first latch signal. Each of the at least one QFP-DAC pair may be responsive to a respective second latch signal.
[0044] Each of the QFP-DACs may include a respective primary material loop, a latch material loop interleaved with a first pair of Josephson junctions and latch interfaces, and a reset material loop interleaved with a second pair of Josephson junctions and reset interfaces, the primary material loop having a latch material loop interleaved with a reset material loop, the primary material loop including a first interface arranged to be communicatively coupled to a QFP latch of a respective one of the first plurality of QFP latches or the second plurality of QFP latches, and at least a second interface to one of the input or output. Each of the QFP-DACs may be responsive to a respective second latch signal. Each of the QFP-DACs may be responsive to a respective first chip signal. Each of the QFP-DACs may be responsive to a respective reset signal. A first QFP of each of the QFPs may be responsive to a respective offset signal.
[0045] The device may further include a control circuit communicatively coupled to control provide signals to the QFP-DAC and the QFP latch. The control circuit may provide signals to the QFP latch to load information into the DAC. The control circuit may provide signals to the QFP latch to operate as a quantum flux parametron demultiplexer (QFP-Demux). The control circuit may provide signals to the QFP latch to load information from the DAC. The control circuit may provide signals to the QFP latch to operate as a quantum flux parametron multiplexer (QFP-Mux). The QFP, QFP-DAC and QFP latch may all form a non-dissipative programmable magnetic memory (PMM) present on a single chip.
[0046] There is a need to be able to process at least some problems that have a size and / or connectivity that is larger than (and / or at least not fully provided for by) the working graph (i.e., physical structure or topology) of an analog processor. Computational systems and methods are described that, at least in some aspects, enable computation of at least some problem graphs that have representations that do not fit within the working graph of an analog processor (e.g., because the problem graph requires more computing devices and / or more / other couplers than the processor provides).
[0047] Aspects of the present disclosure provide a quantum processor with a plurality of programmable devices. The programmable devices have superconducting qubits and one or more couplers that communicatively couple the qubits. The quantum processor includes a first digital-to-analog converter (DAC) operable to drive a current in a target device of the plurality of programmable devices. The first DAC is couplable to the target device and has a first superconducting loop. In operation, the first superconducting loop has a first charge carrier density. The DAC also has a first energy storage element interposed in the first superconducting loop. In operation, the energy storage element has a second charge carrier density less than the first charge carrier density to provide a first kinetic inductance.
[0048] In some embodiments, the first superconducting loop includes a first superconducting material having a first effective penetration depth and the first energy storage element includes a second superconducting material having a second effective penetration depth that is greater than the first effective penetration depth.
[0049] In some embodiments, the first energy storage element comprises a superconducting film of a second superconducting material, the superconducting film having a thickness less than a thickness of the first superconducting material of the superconducting loop, the thickness of the superconducting film may be up to three times and / or half the penetration depth of the second.
[0050] In some embodiments, the superconducting film is disposed entirely within a layer of the quantum processor such that the superconducting film is substantially planar. In some embodiments, the superconducting film snakes within a planar region having a region length in a first dimension and a region width in a second dimension. The superconducting film extends in a plurality of lengthwise portions that extend across the region length in the first dimension. Each lengthwise portion is spaced apart from an adjacent lengthwise portion in the second dimension by a distance less than the region width and is electrically connected to the adjacent lengthwise portion by a widthwise portion that extends in the second dimension.
[0051] In some embodiments, the second superconducting material is selected from the group consisting of NbN, NbTiN, TiN, and granular aluminum.
[0052] In some embodiments, the first energy storage element has a plurality of Josephson junctions serially interleaved in the first superconducting loop. The first DAC may be galvanically coupled to the target device by a galvanic coupling that includes a shared Josephson junction. The first DAC is magnetically coupled to the target device, and R for each of the plurality of Josephson junctions is N A is the target R N The deviation from the A value is less than the threshold amount, and R N may be the Josephson resistance per junction and A the area per junction.
[0053] In some embodiments, the quantum processor includes a second DAC having a second superconducting loop and a second energy storage element interposed in the second superconducting loop to provide a second dynamic inductance. The second superconducting loop is galvanically coupled to the first superconducting loop of the first DAC. The first superconducting loop and the second superconducting loop are coupled to each other. The loop has a shared portion common to the first superconducting loop and the second superconducting loop, the shared portion having a shared energy storage element that, in operation, provides a kinetic inductance proportional to at least one of the first kinetic inductance and the second kinetic inductance. In some embodiments, a width of the shared portion is greater than a width of the first energy storage element and the second energy storage element, the width being measured perpendicular to the direction of current flow and the direction of thickness.
[0054] In some embodiments, the first DAC has a coupling element interposed in the first superconducting loop, the first DAC being directly coupled to the target device via the coupling element, the second DAC being indirectly coupled to the target device via the first DAC, the first DAC operable to represent a most significant digit of the input string, and the second DAC operable to represent a less significant digit of the input string.
[0055] In some embodiments, the first energy storage element has a first plurality of Josephson junctions interposed in series with the first superconducting loop and the second energy storage element has a second plurality of Josephson junctions interposed in series with the second superconducting loop. The shared portion has a third plurality of Josephson junctions. At least one of the Josephson junctions of the third plurality of Josephson junctions has an area larger than each of the Josephson junctions of the first plurality of Josephson junctions and the second plurality of Josephson junctions. In some embodiments, at least one of the Josephson junctions of the third plurality of Josephson junctions has an area at least twice the area of each of the Josephson junctions of the first plurality of Josephson junctions and the at least one of the Josephson junctions of the second plurality of Josephson junctions.
[0056] In some embodiments, the first superconducting loop is galvanically coupled to a target device.
[0057] The quantum processor may be summarized as comprising a plurality of programmable devices having superconducting qubits and one or more couplers communicatively coupling the qubits, and a first digital-to-analog converter (DAC) operable to drive a current in a target device of the plurality of programmable devices, the first DAC comprising a first superconducting loop couplable to the target device and having a first charge carrier density in operation, and a first energy storage element interposed in the first superconducting loop, the first energy storage element having a second charge carrier density in operation less than the first charge carrier density, thereby providing a first dynamic inductance. The first superconducting loop may include a first superconducting material having a first effective penetration depth, and the first energy storage element may include a second superconducting material having a second effective penetration depth greater than the first effective penetration depth. The first energy storage element may comprise a superconducting film of a second superconducting material, the superconducting film may have a thickness less than a thickness of the first superconducting material of the superconducting loop. The thickness of the superconducting film may be at most three times the second penetration depth. The thickness of the superconducting film may be at most one-half the second penetration depth. The superconducting film may be disposed entirely within a layer of the quantum processor such that the superconducting film is substantially planar.
[0058] The superconducting film may be serpentine in a planar region having a region length in a first dimension and a region width in a second dimension, the superconducting film extending in a plurality of lengthwise portions extending across the region length in the first dimension, each lengthwise portion being spaced apart from an adjacent lengthwise portion in the second dimension by a distance less than the region width and electrically connected to the adjacent lengthwise portion by a widthwise portion extending in the second dimension. The second superconducting material may be selected from the group consisting of NbN, NbTiN, TiN and granular aluminum. The first energy storage element may include a plurality of Josephson junctions interposed in series in the first superconducting loop. The first DAC is galvanically coupled to the target device by a galvanic coupling, the galvanic coupling may include a covalent Josephson junction. The first DAC may be coupled to the target device by a galvanic coupling, the galvanic coupling may include a covalent Josephson junction. a Josephson junction for each of the plurality of Josephson junctions. N A is the target R N A value may deviate by a threshold amount or less, and R N is the Josephson resistance per junction and A is the area per junction.
[0059] The quantum processor may include a second DAC having a second superconducting loop and a second energy storage element interposed in the second superconducting loop to provide a second kinetic inductance, the second superconducting loop being galvanically coupled to the first superconducting loop of the first DAC, the first superconducting loop and the second superconducting loop having a shared portion common to the first superconducting loop and the second superconducting loop, the shared portion having a shared energy storage element that provides a shared kinetic inductance proportional to at least one of the first kinetic inductance and the second kinetic inductance in operation. A width of the shared portion may be greater than a width of the first energy storage element and the second energy storage element, the width being measured perpendicular to the direction of current flow and the direction of thickness. A first region of the shared portion may be coupled to one or more superconductors having a smaller kinetic inductance that is smaller than the first kinetic inductance, and an uncoupled region of the shared portion may be defined at least in part by the superconductor to provide the shared kinetic inductance. The shared energy storage element may provide a shared dynamic inductance, in operation, proportional to the first dynamic inductance, and the first DAC may represent one or more orders of magnitude more significant than the second DAC, in operation.
[0060] The first DAC may include a coupling element interposed in the first superconducting loop, the first DAC being directly coupled to the target device via the coupling element and the second DAC being indirectly coupled to the target device via the first DAC, the first DAC operable to represent a most significant digit of the input string and the second DAC operable to represent a less significant digit of the input string.
[0061] The first energy storage element may include a first plurality of Josephson junctions interposed in series with the first superconducting loop, the second energy storage element may include a second plurality of Josephson junctions interposed in series with the second superconducting loop, and the shared portion may include a third plurality of Josephson junctions, and at least one of the Josephson junctions of the third plurality of Josephson junctions may have a larger area than each of the Josephson junctions of the first plurality of Josephson junctions and the second plurality of Josephson junctions. At least one of the Josephson junctions of the third plurality of Josephson junctions may have an area at least twice as large as at least one of the Josephson junctions of the first plurality of Josephson junctions and the second plurality of Josephson junctions. The first superconducting loop may be galvanically coupled to the target device. [Brief description of the drawings]
[0062] In the drawings, identical reference numbers identify similar elements or acts. Sizes and relative positions of elements in the drawings are not necessarily drawn to scale. For example, shapes and angles of various elements are not necessarily drawn to scale, and some of these elements may be arbitrarily enlarged and positioned to improve drawing legibility. Additionally, the particular shapes of drawn elements are not necessarily intended to convey any information regarding the actual shape of the particular elements, and may be selected solely for ease of recognition in the drawings. [Figure 1] FIG. 1 is a schematic diagram of a hybrid computing system, including a digital computer and a quantum computer, that may incorporate various logic devices and / or perform various addressing techniques described herein, according to at least one illustrated embodiment. [Diagram 2] FIG. 1 is a schematic diagram illustrating a quantum flux parametron (QFP)-based shift register communicatively coupled to a set of digital-to-analog converters (DACs) according to at least one illustrated embodiment. [Diagram 3]FIG. 1 is a schematic diagram illustrating a quantum flux parametron (QFP)-based shift register communicatively coupled to a set of digital-to-analog converters (DACs) via multiple intermediate QFPs, according to at least one illustrated embodiment. [Figure 4] FIG. 4 is a flow diagram illustrating a method of operation of a circuit utilizing the QFP-based shift register of FIG. 2 or FIG. 3 in accordance with at least one illustrated embodiment. [Diagram 5] 5 illustrates a method of operation of a circuit utilizing a QFP-based shift register, which may be a specific embodiment of the method of FIG. 4, in accordance with at least one illustrated aspect. [Figure 6] 5 illustrates a method of operation of a circuit utilizing a QFP-based shift register, which may be a generalization of the method of FIG. 4, according to at least one illustrated embodiment. [Figure 7A] 1 illustrates a latch-controlled quantum flux parametron demultiplexer (QFP-Demux) circuit coupled to a digital-to-analog converter (DAC) in accordance with at least one illustrated embodiment. [Figure 7B] 1 illustrates a body flux (address) controlled quantum flux parametron demultiplexer (QFP-Demux) circuit coupled to a digital-to-analog converter (DAC) in accordance with at least one illustrated embodiment. [Figure 8] 7B is a graph of a plot of signals associated with the latch-controlled QFP-Demux circuit of FIG. 7A in accordance with at least one illustrated embodiment. [Figure 9] 1 illustrates a three-junction, two-loop flux pump addressing a digital-to-analog converter (DAC) in accordance with at least one illustrated embodiment. [Figure 10] 10 is a graph illustrating a plot of applied flux waveforms versus Josephson phase of an end-most junction for a programming / deprogramming pattern for the three-junction, two-loop flux pump of FIG. 9 according to at least one illustrated embodiment. [Figure 11] 1 is a schematic diagram of a four-junction, three-loop flux pump circuit addressing a digital-to-analog converter (DAC) according to at least one illustrated embodiment. [Figure 12A] 12 is a graph illustrating a plot of applied flux waveforms versus Josephson phases of end-most junctions for a programming / deprogramming pattern for the four-junction, two-loop flux pump of FIG. 11 according to at least one illustrated embodiment. [Figure 12B] 1 is a schematic diagram of a direct galvanic connection between two flux DACs in accordance with at least one illustrated embodiment. [Figure 13] FIG. 1 is a schematic diagram of a set of eight four-junction, three-loop flux pump circuits addressing individual digital-to-analog converters (DACs) via multiple address lines according to at least one illustrated embodiment. [Figure 14] 12 is a graph illustrating a plot of applied flux waveforms versus Josephson phases of endmost Josephson junctions for a programming / deprogramming pattern for the four-junction, two-loop flux pump of FIG. 11 according to at least one illustrated embodiment. [Figure 15] FIG. 1 is a flow diagram of a method of operating a multiple junction superconducting quantum interface device (SQUID) flux pump according to at least one illustrated embodiment. [Figure 16A] FIG. 1 is a flow diagram of a method of operating a multiple junction superconducting quantum interface device (SQUID) flux pump in accordance with at least one illustrated embodiment. [Figure 16B] 12 is a graph illustrating a plot of an applied flux waveform versus Josephson phase of an endmost Josephson junction for resetting the four-junction, two-loop flux pump of FIG. 11 according to at least one illustrated embodiment. [Figure 17A] FIG. 1 is a schematic diagram of a set of four digital-to-analog converters (DACs), a number of signal lines arranged to individually address each of the DACs, and control circuitry communicatively coupled to provide signals to selected ones of the DACs by individual triplets of the signal lines, according to at least one illustrated embodiment. [Figure 17B]17B is a schematic diagram of the four digital-to-analog converter (DAC) sets and control lines of FIG. 17A, further showing a number of diagonal lines representing the direction of the braid control lines between the DAC sets. [Figure 18] 1 is a schematic diagram of a quartet of digital-to-analog converters (DACs), a power line, and three control lines in accordance with at least one illustrated embodiment. [Figure 19] FIG. 13 is a flow diagram of a method of operation utilizing blade-arranged control lines to individually address DACs in a two-dimensional set of DACs (eg, an ordered array) according to at least one illustrated embodiment. [Figure 20] FIG. 1 is a schematic diagram of a portion of an exemplary superconducting quantum processor designed for quantum annealing (and / or adiabatic quantum computing) components that may be used to implement the present systems and devices. [Figure 21] FIG. 1 is a schematic diagram of an exemplary single-loop DAC with a dynamic inductance energy storage element. [Figure 22] FIG. 1 is a schematic diagram of two exemplary cascaded DACs with dynamic inductance energy storage elements. [Diagram 23] FIG. 1 is a schematic diagram of an exemplary embodiment of a dynamic inductance energy storage element using thin film superconductors. [Figure 24] FIG. 4 is a schematic diagram of an exemplary embodiment of a multiple dynamic inductance energy storage element including a shared portion of a cascaded DAC, such as the DAC shown in FIG. 3, using thin film superconductors. [Figure 25A] FIG. 1 is a schematic diagram of an exemplary embodiment of a dynamic inductance energy storage element using series-connected Josephson junctions when the DAC is inductively coupled to a target device. [Figure 25B] FIG. 1 is a schematic diagram of an exemplary embodiment of a dynamic inductance energy storage element using series-connected Josephson junctions when the DAC is galvanically coupled to a target device. [Figure 26A] FIG. 2 is a schematic diagram of an exemplary embodiment of two exemplary cascaded DACs with series-connected Josephson junctions. [Figure 26B]FIG. 2 is a schematic diagram of an exemplary embodiment of three exemplary cascaded DACs with series-connected Josephson junctions. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS
[0063] In the following description, specific details are set forth to provide a thorough understanding of the various disclosed aspects. However, one of ordinary skill in the relevant art will understand that aspects may be practiced without one or more of these specific details, or with other methods, components, materials, etc. In other instances, well-known structures related to computer systems, server computers, communication networks, superconducting circuits, and / or resonators have not been shown or described in detail to avoid unnecessarily obscuring the description of the aspects. Throughout this specification and the appended claims, the terms "element" and "( The term "elements" is used to encompass all such structures, systems and devices related to, but not limited to, computer systems, server computers, communication networks, superconducting circuits and resonators.
[0064] Unless the context requires otherwise, throughout this specification and the following claims, the words "comprising" and "comprising" are synonymous with the words "including" and "including" and are intended to include or include any other term. It is open-ended (i.e., it does not exclude additional, unrecited elements or method acts).
[0065] Throughout this specification, references to "one embodiment," "an embodiment," "an other embodiment," "an example," "an example," "an other example," "one aspect," "an other aspect," and the like mean that the particular associated feature, structure, or characteristic described in connection with that embodiment, example, or aspect is included in at least one embodiment, example, or aspect. Thus, appearances of the phrases "in one embodiment," "in an embodiment," "in another embodiment," and the like in various places throughout this specification do not necessarily refer to the same embodiment, example, or aspect. Moreover, references to more than one embodiment, example, or aspect may be used interchangeably. The particular features, structures or characteristics may be combined in any combination in an embodiment, example or aspect. The singular forms "an" and "the" are used unless the context clearly indicates In addition, the term "or" is generally used in its sense including "and / or" unless the context clearly indicates otherwise.
[0066] The headings and abstract of the disclosure provided herein are for convenience only and do not interpret the scope or meaning of the embodiments.
[0067] Additionally, certain figures herein depict various voltage and current waveforms that are intended for purposes of understanding the operation of the embodiments and are not intended to be scaled and / or to accurately and precisely depict the waveform behavior in terms of shape, amplitude, duty cycle, frequency, distortion, or other characteristics.
[0068] In classical electrical circuits, complex and laborious operations can be performed by specific arrangements of simple switching devices such as transistors. Examples of such arrangements include shift registers, memory arrays, adders, flip-flops, etc. In the developing field of superconducting electronics, and especially superconducting quantum computing, it is essential to develop circuits that realize superconducting analogs of each of these arrangements. These arrangements can be constructed using simple switching devices as well, but classical switching devices such as transistors are not relevant in superconducting regimes. It is therefore important to establish basic superconducting logic devices that can realize many other circuits and operations in many superconducting systems.
[0069] An application of superconducting electronics of particular interest is in the field of quantum computing. Superconducting circuits are capable of exploiting quantum effects on a macroscopic scale, providing a much more manageable mechanism for implementing quantum computing than some of the alternative technologies. For purposes of explanation, the basic unit of quantum computing is the qubit. Superconducting qubits can take a variety of forms, including superconducting flux qubits. Superconducting flux qubits can be realized in the form of a superconducting loop that includes at least one switching device known as a Josephson junction. A quantum processor can then include any number of superconducting flux qubits. Thus, in implementing such a quantum processor, a solid understanding of the operation of such devices can be developed, along with techniques for fabricating large numbers of superconducting flux qubits. The present systems, methods and apparatus describe the use of a basic superconducting flux qubit structure as a superconducting switching device in a quantum processor. In an exemplary embodiment, an arrangement of superconducting flux qubits can be implemented as a superconducting flux-based shift register in a superconducting quantum processor.
[0070] The present disclosure relates to a superconducting quantum processor with a superconducting digital-to-analog converter (DAC). The DAC includes an energy storage element that stores energy using dynamic inductance. In various embodiments, single-loop and multi-loop (or "cascaded") DAC designs are disclosed. Thin-film and Josephson junction-based energy storage elements with dynamic inductance are disclosed, including embodiments with serpentine layouts and / or galvanic coupling.
[0071] Although the term DAC is used throughout, it will be understood that the described devices can be used for a variety of purposes, not necessarily limited to converting digital signals to analog signals (and in some embodiments not including such conversion at all). For example, as discussed above, a quantum processor may use a superconducting DAC to store a signal for a period of time (e.g., thereby acting as a form of memory).
[0072] To aid in understanding the reader, the operation of an example superconducting quantum processor is described below, providing circumstances in which a superconducting DAC may perform and exhibit at least some exemplary functions of such a DAC.
[0073] Exemplary Hybrid Computing System FIG. 1 illustrates at least one example embodiment of a hybrid computing system 100 including a classical or digital computer 102 and a quantum computer 104 that may incorporate various logic devices and / or implement various addressing techniques as described herein.
[0074] The digital computer 102 includes one or more digital processors 106, e.g., one or more single-core or multi-core microprocessors, central processing units (CPUs), graphical processor units (GPUs), digital signal processors (DSPs), or application specific integrated circuits (ASICs). The digital computer 102 may include one or more user interface components, e.g., one or more displays 108a, pointer devices 108b (e.g., computer mouse, trackball), and keypads or keyboards 108c, collectively 108. The digital computer 102 may include one or more non-transitory computer-readable or processor-readable media, e.g., one or more memories (e.g., volatile memory, static memory, read-only memory (ROM), random access memory (RAM)) 110 and / or one or more storage devices (e.g., magnetic hard disk drives (HDDs), optical disk drives, solid-state drives (SSDs), and / or flash drives) 112. Digital computer 102 may include one or more quantum computer control subsystems 114 that interface with quantum computer 104. Processor 106, memory 110, storage 112, and quantum computer control subsystem 114 may be communicatively coupled via one or more communication channels, such as, for example, one or more buses 116 (e.g., a power bus, a communication bus, an instruction bus, an address bus).
[0075] A non-transitory computer-readable medium or processor-readable medium, for example one or more memories 110, stores processor-readable instructions and data that, when executed by one or more processors 106, cause the processor 106 to perform one or more of the various algorithms described herein. The processor-executable instructions and data may include, for example, a basic input / output system instruction set or "module" 118a that configures the digital computer 102 to operate upon startup to provide various file management and user interface services. The processor-executable instructions and data may include, for example, a server instruction set or "module" 118c that configures the digital computer 102 to operate as a server that provides access to information and services by other computers. The processor-executable instructions and data may include, for example, a computational instruction set or "module" 118d that configures the digital computer 102 to perform various computations associated with transforming a problem into a problem graph and / or post-processing of a potential solution generated by the quantum computer 104. The processor executable instructions and data may include, for example, a quantum processor instruction set or "module" 118d that configures the digital computer 102 to map a problem from a problem graph to a hardware graph for incorporation into the quantum processor 120 of the quantum computer 104 for execution. The processor executable instructions and data may include, for example, a read instruction set or "module" 118f that configures the digital computer 102 to perform various read functions associated with retrieving a latent solution to a problem from the quantum computer 104.
[0076] Quantum computer 104 comprises a quantum processor 120, typically including a plurality of qubits and / or a plurality of couplers, each coupler selectively operable to couple a respective pair of qubits. Quantum computer 104 includes a qubit control system 122 operably coupled to control each of the qubits, e.g., via various interfaces, i.e., inductive interfaces, that selectively couple magnetic flux to the qubits. Quantum computer 104 includes a coupler control system 124 operably coupled to control each of the couplers, e.g., via various interfaces, i.e., inductive interfaces, that selectively couple magnetic flux to the couplers to set the coupling strength, i.e., "coupling strength," of the couplers. Quantum computer 104 includes a readout control system 126 operably coupled to control the various interfaces, which are operable to read out the state of each of the qubits.
[0077] Various systems, components, structures and algorithms are described herein that implement the above. Many of the described systems, components, structures and algorithms can be implemented separately and some can be implemented in combination with each other.
[0078] Quantum Flux Parametron (QFP)-Based Shift Registers Existing systems address the DACs using an XYZ addressing scheme, where Z is conveniently called the power supply (all DAC stages in a C2 size processor are connected in series), and X and Y are conveniently called the address (ADDR) and trigger (TRIG). The C2 block is powered up, ADDR is asserted, and TRIG is toggled a few times to write a corresponding number of pulses to one uniquely selected DAC stage. This addressing scheme can handle the thousands of qubit level (e.g., 8,000 or 16,000). To move to the next level of complexity, e.g., with 100,000 qubits or more, one can use a method of loading data serially into the PMM and write long bit streams to the DACs using only a few wires. SFQ-based shift registers dissipate excessive power on chip, so a QFP-based scheme is preferred.
[0079] In at least one existing approach, ADDR and TRIG are each Φ 0 (where Φ 0 is the superconducting flux quantum), the selected DAC stage (where ADDR and TRIG correspond to the direction of the flux and the DAC stage is powered) is 0Consider about 1 / 2 of the total signal, and pass the other single flux quantum (SFQ) through the corresponding DAC storage inductor. In principle, one of these signals (e.g., ADDR) could be fed by the QFP stage, either magnetically, or galvanically if it is connected to a QFP shift register that is magnetically coupled to half of the DAC stage, rather than from a dedicated room temperature trace. Figures 2 and 3 show two possible arrangements. In particular, Figure 2 shows a DAC magnetically coupled to the QFP-SR stage, while Figure 3 shows an additional QFP galvanically coupled to the DAC, which feeds more signals into the DAC at the expense of additional junction / body footprint, acting as a flux amplifier. Since all DACs are connected in series (via the POWER lines), it may be preferable for the QFP shift register (QFP-SR) to have a galvanic connection between stages, which could improve margins in the presence of possible flux offsets, and there should be a (less effective) magnetic connection somewhere between the DAC and the QFP-SR.
[0080] FIG. 2 illustrates a quantum flux parametron (QFP)-based shift register 200 communicatively coupled to a set of digital-to-analog converters (DACs) 202a, 202b (only two shown, collectively referred to as 202) according to at least one illustrated embodiment.
[0081] The QFP-based shift register includes a number of QFP-based shift register elements 204a, 204b, 204c (only three of which are called out and collectively referred to as 204) and an inductor ladder circuit 206 that inductively or galvanically couples to the DACs 202a, 202b of the set of DACs 202.
[0082] The QFP-based shift register elements 204 are individual QFPs each including an individual material loop 208, a pair of Josephson junctions 210a, 210b (two called out and collectively referred to as 210) interposed in the individual material loop 208, and an individual interface 212, e.g., an inductive interface. The material loops 208 may comprise a material loop that undergoes superconductivity at a critical temperature. The interface 212 may be arranged to communicatively couple a signal (e.g., magnetic flux) from an address line 214 (only one called out). The QFP-based shift register 200 preferably includes at least three QFP register elements 204a, 204b, 204c for each DAC 202a of the set of DACs 202.
[0083] The inductor ladder circuit 206 includes a material path 216 with a pair of rails 216a, 216b and multiple inductors 218 (only one called out) in parallel coupled between the rails 216a, 216b as "rungs." The inductor ladder circuit 206 also includes multiple interfaces (e.g., inductive interfaces) 220 (only one called out) arranged to communicatively couple signals (e.g., magnetic flux) to the respective DACs 202a, 202b via a complementary interface 222 (only one called out) of the DAC 202a. The material path 216 may be comprised of a material path that undergoes superconductivity at a critical temperature.
[0084] Each of the DACs 202a, 202b in the set of DACs 202 includes an individual material loop 224 (only one is called out) and an individual pair of Josephson junctions 226a, 226b (only one is called out) interposed in the individual material loop 224, with the Josephson junctions 226a, 226b electrically coupled in parallel between electrical connections 225a, 225b of the material loop 224 by a power line 232. Each of the DACs 202a, 202b in the set of DACs 202 includes a number of interfaces 222, 228, 230, e.g., inductive interfaces, for coupling signals thereto. Each of the 200 sets of DACs 202a, 202b is independently addressable successive times to store a variable number of flux quanta by a triplet of three signals including a first signal obtained via the QFP-based shift register element 204 of the QFP-based shift register 200, a second signal obtained via a power supply line 232, and a third signal obtained via a trigger line 234.
[0085] Control circuitry 236 is communicatively coupled to provide signals to power line 232, trigger line 234, and address line 214. Control circuitry 236 provides signals to power line 232, trigger line 234, and address line 214 to load information (e.g., flux quanta) into selected ones of DACs 202a, 202b.
[0086] Although shown as a first set of DACs 202 and a first shift register 200, additional sets of DACs and additional associated shift registers may be used.
[0087] FIG. 3 illustrates a QFP-based shift register 200 communicatively coupled to a set of DACs 202a, 202b (only two shown, collectively referred to as 202) via multiple intermediate QFPs 340a, 340b (only two shown, collectively referred to as 340), according to at least one illustrated embodiment.
[0088] Many of the structures shown in Figure 3 are similar or even identical to the structures shown in Figure 2 and therefore are numbered the same. For the sake of brevity and therefore clarity, only the significant differences between Figure 3 and Figure 2 are described below.
[0089] The intermediate QFPs 340a, 340b act as QFP flux amplifiers that galvanically couple to the individual DACs 202a, 202b of the first set of DACs 202. Each of the QFPs 340a, 340b includes an individual material loop 342 (only one callout) and a corresponding pair of Josephson junctions 344a, 344b (only one callout) interposed in the respective loop 342. Each of the QFPs 340a, 340b includes an individual interface (e.g., inductive interface) 346 (only one callout) that communicatively couples signals thereto. The individual material loops 342 may undergo superconductivity at a critical temperature. Each of the QFPs 340a, 340b is communicatively coupled (e.g., galvanically or inductively) between an individual DAC of the DACs 202a, 220b and a QFP-based shift register element 204a, 204b, 204c of the QFP-based shift register 200 via the inductor ladder circuit 206.
[0090] 4 illustrates a method 400 of operating a circuit utilizing a QFP-based shift register, according to at least one illustrated embodiment. Method 400 may utilize, for example, QFP-based shift register 200 of FIG. 2 or QFP-based shift register 300 of FIG. 3.
[0091] The method 400 begins at 402, for example, by invoking a routine or program in response to, for example, powering on, presenting a problem, or requesting activation.
[0092] At 404, the control circuit resets all of the DACs in the first DAC set. To reset all of the DACs, the control circuit may repeatedly provide a signal to or through the trigger line to release all of the magnetic flux in the storage inductors.
[0093] At 406, the control circuitry provides a sustaining current in a first direction (e.g., clockwise) to a first number of QFP-based shift registers coupled to respective DAC stages having flux quanta added therein, and at 408, the control circuitry provides a sustaining current in an opposite direction (e.g., counterclockwise) to a second number of QFP-based shift registers coupled to respective DAC stages having no flux quanta loaded therein.
[0094] At 410, the control circuit provides a signal to the power supply line. At 412, the control circuit provides a signal to the trigger line a first number of times, the first number being at least proportional to the total number of flux quanta to be added. Typically, the first number is equal to the total number of flux quanta to be loaded. Thus, it is possible to produce pulses on the trigger line TRIG with a polarity that adds to the clockwise current in the selected QFP stage and subtracts from the counterclockwise current in the unselected stages, a total number of times equal to the number of pulses to be written, increasing the number of SFQ quanta in the storage inductor. If, at 414, the control circuit determines that the signal has not yet been provided to the trigger line the first number of times, the method 400 returns to 412.
[0095] Once the control circuitry determines that the signal has been applied to the trigger line the first number of times at 414, the method 400 stops at 416 until it is called or initiated again. Alternatively, the method 400 can operate continuously.
[0096] 5 illustrates a method 500 of operating a circuit utilizing a QFP-based shift register, according to at least one illustrated embodiment. Method 500 may be a specific implementation of method 400. (FIG. 4). The method 500 may utilize, for example, the QFP-based shift register 200 of FIG. 2 or the QFP-based shift register 300 of FIG.
[0097] The method 500 begins at 502, for example, by invoking a routine or program in response to, for example, powering on, presenting a problem, or requesting activation.
[0098] At 504, the control circuit sorts all of the DACs based on the number of flux quanta loaded into the DACs. At 506, the control circuit resets all of the DACs. The control circuit may repeatedly provide a signal to or through the trigger line to reset all of the DACs.
[0099] At 508, the control circuitry provides a sustaining current in a first direction (e.g., clockwise) to a plurality of QFP-based shift registers coupled to respective DAC stages having at least one flux quantum loaded therein. At 510, the control circuitry provides a sustaining current in an opposite direction (e.g., counterclockwise) to a plurality of QFP-based shift register stages coupled to respective DAC stages having less than one flux quantum loaded therein. At 512, the control circuitry provides a signal to or through a power supply line. At 514, the control circuitry provides a signal to or through a trigger line.
[0100] At 516, the control circuitry then provides a sustaining current in a first direction to a number of QFP-based shift registers coupled to respective DACs having at least two flux quanta loaded therein. At 518, the control circuitry provides a sustaining current in the opposite direction to a second number of QFP-based shift registers coupled to respective DACs having less than two flux quanta loaded therein. At 520, the control circuitry provides a signal to the power supply line. At 522, the control circuitry then provides a signal to or through the trigger line.
[0101] The method 500 stops at 524, for example, until it is called or initiated again.
[0102] 6 illustrates a method 600 of operating a circuit utilizing a QFP-based shift register, according to at least one illustrated embodiment. Method 600 may be a generalization of method 400 (FIG. 4). Method 600 may utilize, for example, QFP-based shift register 200 of FIG. 2 or QFP-based shift register 300 of FIG. 3.
[0103] To save programming time (e.g., programming tens of thousands of qubits, 8 DACs per qubit, 16 DAC stages per qubit), the DAC stages can first be sorted by the number of pulses (up to the maximum number of single flux quanta (MAXSFQ)) to be added to each stage, e.g., about 20. Hence, many stages with the same target number of pulses can occur. For example, some stages need 1 pulse, some need 2 pulses, others need 3 pulses, and so on, eventually resulting in a stage needing 20 pulses. First, select the stages that need at least 1 pulse and pulse TRIG once. The new pattern excludes stages that only needed 1 SFQ and pulses TRIG twice. Note that all other stages need 3 or more SFQs. This continues until only stages that need 20 SFQs remain, which are then selected and filled. This process requires, in the worst case scenario, a MAXSFQ reprogramming cycle in which all QFP-SRs write a new "serial program" onto the processor. The process can be further sped up by breaking the QFP-SR into shorter sections, resulting in a small increase in the number of interconnects approaching room temperature.
[0104] Also, the QFP-SR can be used for quantum bit readout by NDRO. Without using a separated structure, data can be brought close to the quantum bit to program the DAC of the quantum bit. For reliability, instead of using one long shift register, a grid-shaped or array-shaped shift register stage may be used. Therefore, there are multiple possible paths to reach any point through the QFP-SR, and the QFP-SR can tolerate devices that fail at a few percent.
[0105] Method 600 starts, in 602, for example, by calling a routine or program in response to power-on, problem provision, or startup request.
[0106] In 604, the control circuit sorts all of the DAC stages based on the number of flux quanta loaded into the DAC. In 606, the control circuit performs a reset of all of the DACs. The control circuit may repeatedly supply a signal to a trigger line, or via a trigger line, to reset all of the DACs.
[0107] In 608, for the number of times i of the programmed integer that takes values from 1 to the maximum number n of flux quanta, the control circuit supplies a continuous current in a first direction (e.g., clockwise) to a plurality of QFP-based shift registers coupled to individual DAC stages into which at least i flux quanta are loaded internally. In 610, when the control circuit determines that i < n, the control of method 600 returns to 608. In 610, when the control circuit determines that i = n, the control of method 600 proceeds to 612.
[0108] In 612, the control circuit also supplies a continuous current in the opposite direction (e.g., counterclockwise) to a plurality of QFP-based shift registers coupled to individual DAC stages into which less than i flux quanta are additionally loaded internally. In 614, when the control circuit determines that i < n, the control of method 600 returns to 612. In 614, when the control circuit determines that i = n, the control of method 600 proceeds to 616.
[0109] At 616, the control circuit provides a signal to or through the power line. At 618, the control circuit provides a signal to or through the trigger line. The maximum number of flux quanta, n, can be, for example, 18-22.
[0110] The method 600 stops at 620, for example, until it is called or initiated again.
[0111] Quantum Flux Parametron (QFP) Digital-to-Analog Converter (DAC) It may be desirable to achieve 4-bit control over all parameters, some reduced thermal occupancy of excited states compared to previous quantum processor designs, more reliable permanent magnet memory (PMM), and a higher total energy scale. At large integration scales, performance may be limited by 4-bit control over the qubit Ip, thermal occupancy, and nonlinearities (which increase thermal occupancy at the expense of lowering the energy scale to compensate and reduce processor operation time).
[0112] Here, we describe various important modifications to the architecture that improve qubit performance (e.g., improved qubit Ip nonlinearity and shorter length leading to faster devices), speed up calibration (e.g., no on-chip power dissipation due to readout or PMMs), and improve control precision (e.g., higher maximum number of SFQ-DACs). The described platform also features parallel programming by using high bandwidth wiring and by freeing up input / output (I / O) lines and / or adding I / O lines. This provides the basis for speeding up which operations are performed by logging and reading.
[0113] Such an approach could lead to much faster programming, calibration, readout, higher control precision, and larger processors. Furthermore, free spins with internal energy splitting greater than, for example, 100 mK would not have the thermal energy available to switch states, i.e., the 1 / f corner would move to lower frequencies, reducing power dissipation on the chip and reducing low frequency flux noise. This, combined with shorter qubits on a quantum processor chip, could allow for higher control precision without the need to significantly reduce manufacturing noise.
[0114] While lower temperatures and shorter qubits provide the performance improvements discussed above, there are further reasons to utilize the approaches described herein. Both problem-solving and calibration times are major areas for improvement. Calibration and problem-solving times depend in different ways on 1) DC SQUID readout time, 2) the number of DC SQUIDs that can be read out in parallel, 3) the cooling time after DC SQUID readout, 4) the cooling time after PMM programming, and 5) the equilibrium temperature on the chip.
[0115] Quantum Flux Parametron DAC (QFP-DAC) calibration and offset flux measurements, which account for approximately one third of the total calibration, can utilize DAC lock-in measurements. For this third of the calibration, the fourth and fifth items above dominate time on previous quantum processor architectures. An improved PMM can address this issue.
[0116] The remaining part of the calibration is time limited by the read and the cooling after the read. At least one approach described herein overcomes the limitations of the cooling after the read and can also reduce the read time.
[0117] The designs described herein facilitate parallel calibration and parallel readout, allowing quantum processors with 1000 or more qubits to be calibrated and operational in a finite amount of time.
[0118] At all moderate integration scales the problem-solving time depends on the qubit length (which sets the energy scale), the temperature (the 5 items above, going through the required iterations), and at smaller scales the problem-solving time depends on the readout time and, for some parameter ranges, the PMM programming time.
[0119] Here, various embodiments of a QFP-Demux and DAC are described that work similarly to existing SFQ-DACs, but advantageously do not dissipate power, thus removing some significant portion of the current cycle time. This QFP-DAC / DEMUX does not actually improve loading time, in fact it is slightly slower for the same I / O lines. Improvements to I / O lines and parallel loading are one way to speed things up. The final parameter design of the QFP-DAC / DEMUX is highly dependent on manufacturing capabilities.
[0120] A novel readout scheme is also described that allows for the readout of qubits in parallel and at a greatly enhanced speed (e.g., about 100 times faster) than existing designs. Design parameters for the readout scheme are detailed.
[0121] To address anticipated requirements, it is possible to operate a lock-in measurement on the DAC when the DAC needs to quickly reverse state, and this section describes how this can reduce input / output (I / O) bandwidth or the need for tens of milliseconds of cooling between measurements, which is currently required. We propose a non-dissipative PMM method that limits the measurement by the readout time.
[0122] The circuit shown in FIG. 7A generates the flux quanta required for the DAC from a QFP used in conjunction with the QFP demultiplexer tree. FIG. 7A shows a portion of the final branch of the demultiplexer tree addressed by LATCH1 and (LATCH1). The OFFSET signal is selected to be the "off" direction of the QFP, and is greater than the flux coupled between the QFPs. The operation in this small circuit is to set OFFSET to 0, apply LATCH1, set the offset to a large value, and apply (LATCH1). When the LATCH signal is applied, it changes the flux quantum from Φ0 / 2 to Φ0, and when the LATCH signal is not applied, the flux quantum remains at Φ0 / 2. This can be easily achieved using only one wire per level of the demultiplexer tree with an additional shared LATCH "offset" line (except that the LATCH "offset" is time dependent).
[0123] In that case, assert LATCH2 without applying an input to TIP. In principle, the last two QFP stages can be coupled together, where only one of the left parts of the QFP-DAC has a positive (negative) flux quantum in the QFP-DAC, and the rest have a negative (positive) flux quantum. Provide the input bias current to the current bias shared by all DAC junctions. This can be done because only one DAC switches at a time. Finally, provide a chip pulse that is added to the signal in the selected DAC and subtracted from the others. The selected DAC will momentarily exceed the critical current of the DAC and therefore will introduce a flux quantum into the DAC from the QFP DAC loop. The other DACs will not introduce a flux quantum. The LATCH2 / TIP cycle can be repeated to load as many flux quanta into the DACs as desired without the need to rerun the demultiplexer tree.
[0124] Since the quantum flux parametron QFP-DAC does not require power on chip, the current Ic of all junctions can be increased almost without cost, but this may require a new three layers to limit the growth of the demux / DAC at some point as the DAC approaches the limiting junction size. This allows the inductance of the DAC to be reduced, shrinking the footprint of the DAC and shortening the qubit length. This can advantageously result in better qubits and faster processors. In particular, another metal layer or three layers may be required to implement a smaller DAC. An improved reset technique, described herein, may facilitate simultaneously reducing the DAC step size and improving accuracy.
[0125] Quantum Flux Parametron Digital-to-Analog Converters (QFP-DACs) advantageously facilitate the use of weak signals (QFP signals) to multiple flux quantum DACs. This can be advantageously used, for example, in large adiabatic quantum processors, or any quantum processor where quasi-DC magnetic fields are used. This may be used with high bandwidth wiring to operate at update rates of multiple GHz, for example, in finite size trees with DACs updated at rates of hundreds of MHz.
[0126] 7A illustrates a latch-controlled quantum flux parametron demultiplexer (QFP demux) circuit 700a coupled to QFP-digital-to-analog converters (QFP-DACs) 702a, 702b, according to at least one illustrated embodiment. In particular, FIG. 7A illustrates the final stage of the QFP-demux tree.
[0127] The latch control QFP demux circuit 700a includes a plurality of QFPs 704 (one shown) and a set or row of a plurality (e.g., two) quantum flux parametron latches (QFP latches) 706a, 706b for each QFP 704 (each row is shown in FIG. 7A with one QFP latch). Each set or column of QFP latches 706a, 706b is selectively operable to communicatively couple magnetic flux between the QFP 704 and a respective one of the first and second QFP-DACs 702a, 702b of a pair of respective QFP-DACs 702a, 702b. A current bias line 710 is arranged to provide a current bias to at least both of the QFP-DACs 702a, 702b of the pair of QFP-DACs 702, advantageously without the use of bias transistors.
[0128] The QFP-DAC 702 includes an asymmetric DC SQUID (illustrated on the right side of FIG. 7A) connected by a current bias. It may be possible to eliminate the current bias and use a magnetic bias in the CJJ. Otherwise, the current bias advantageously does not require a bias resistor and does not generate heat, but it is indeed necessary to bias all the QFP-DACs 702 in series. In particular, the bias line 710 can also be used to address the different trees.
[0129] The QFP 704 may act as an input or node in the tree structure between the branches of the tree structure. The QFP 704 includes a first loop of material 712a, a second loop of material 712b interposed with the first loop of material 712a, and a pair of Josephson junctions 714a, 714b interposed with the second loop of material 712b. The second loop 712b includes an interface (e.g., inductive interface) 716 for receiving a signal, for example, designated LATCH0. The first loop of material 712a includes a pair of interfaces (e.g., inductive interfaces) 718a, 718b arranged to communicatively couple a signal (e.g., magnetic flux) to interfaces 720a, 720b of respective first QFP latches 706a, 706b of a first and second set or row of QFP latches 706a, 706b (only one shown in each set or row). The first material loop 712a and the second material loop 712b of the QFP 704 may be made of, for example, a material that exhibits superconductivity below a critical temperature.
[0130] The first QFP latches 706a, 706b of the first and second sets or rows of QFP latches 706a, 706b each include a first material loop 722a (only one is called out in FIG. 7A), a second material loop 722b (only one is called out in FIG. 7A) interposed in the first material loop 722a, and a pair of Josephson junctions 724a, 724b (only one is called out in FIG. 7A) interposed in the second material loop 722b. The first loop includes an interface (e.g., an inductive interface) 725 for receiving a signal, e.g., designated OFFSET. The second loop 722b includes an interface (e.g., an inductive interface) 726 for receiving a signal, e.g., designated LATCH1 or (LATCH1). Opposite current flows (e.g., clockwise and counterclockwise, as shown by arrows in the first loop 722a of the first QFP latches 706a, 706b) can be established in the first loop 722a by providing opposite signals LATCH1 or (LATCH1) to the interface 726 of the second loop 722b of the first QFP latches 706a, 706b. The first material loop 722a includes further interfaces (e.g., inductive interfaces) 728 (only one is called out in FIG. 7A ) arranged to communicatively couple signals (e.g., magnetic flux) to interfaces 730 (only one is called out in FIG. 7A ) of either an individual second QFP latch of the set or column of QFP latches 706a, 706b or of the QFP-DAC 702a, 702b. The first material loop 722a and the second material loop 722b of the first QFP latch 706a, 706b may be made of, for example, a material that exhibits superconductivity below a critical temperature.
[0131] Each of the QFP-DACs 702a and 702b includes a first material loop 738a (only one is called out in FIG. 7A) and a second material loop 738b interposed in the first material loop 738a. The first loop 738a includes a plurality of interfaces (e.g., three inductive interfaces) 742 for reading signals stored in, for example, the QFP-DACs 702a, 702b. The second loop 738b includes an interface (e.g., an inductive interface) 744 for receiving a signal operable to reset the value of the QFP-DACs 702a, 702b, for example, a signal designated RESET. The set or string of QFP latches 706a, 706b transmit the signal received via the QFP 704 to the QFP-DACs 702a, 702b. 702a, 702b, and the signal can eventually be read out from the QFP-DACs 702a, 702b. The first and second material loops of the QFP-DACs 702a, 702b can be made of, for example, a material that exhibits superconductivity below a critical temperature.
[0132] Each QFP-DAC includes a third material loop 738c (only one is called out in FIG. 7A) interposed in a first material loop 738a, and a pair of Josephson junctions 746a, 746b (only one is called out in FIG. 7A) interposed in the third material loop 738c. The first loop 738a includes an interface (e.g., an inductive interface) 736a for receiving a signal, e.g., a signal designated TIP. The third loop 738c includes an interface (e.g., an inductive interface) 736b for receiving a signal, e.g., a signal designated LATCH2. Opposing current flows (e.g., clockwise and counterclockwise as shown by the arrows in the first loop 738a of the QFP-DAC 702a, 702b) can be established by establishing opposing current flows in the first loop 722a of the QFP latch 706a, 706b. The first material loop 738a, the second material loop 738b and the third material loop 738c of the QFP-DACs 702a, 702b may be made of, for example, a material that exhibits superconductivity below a critical temperature.
[0133] 7B illustrates a quantum flux parametron demultiplexer (QFP-Demux) circuit 700b controlled by body flux (address) coupled to a digital-to-analog converter (DAC) according to at least one illustrated embodiment. In particular, FIG. 7B illustrates the final stage of a QFP demux tree.
[0134] Many of the structures shown in Figure 7B are similar or even identical to the structures shown in Figure 7A and therefore are numbered the same, and for the sake of brevity and therefore clarity, only the significant differences between Figures 7A and 7B are described below.
[0135] In contrast to the embodiment of FIG. 7A, the embodiment of FIG. 7B provides the same LATCH1 signal to both first QFP latches 706a, 706b, and rather than providing the same OFFSET signal to both first QFP latches 706a, 706b, the embodiment of FIG. 7B controls the direction of current flow by providing opposing address signals, designated ADDR1 and (ADDR1), via address lines to the first QFP latches 706a, 706b.
[0136] Note that the demultiplexer can also operate in broadcast mode and can use the OFFSET line to generate a signal to start in the middle of the tree. The demultiplexer can load positive and negative pulses. Reset is achieved in the same way as the SFQ-DAC. Optionally, all of the QFP-DACs can be unloaded by over-biasing the QFP-DAC's Josephson junctions, where the bias current is varied to inhibit the QFP-DAC's Josephson junctions and release metastable flux quanta. Sufficient variation of the bias current can produce a known state.
[0137] The QFP demux can be collapsed by about 5 bits using direct addressing (see FIG. 7B). For example, 5 address lines can be logarithmically aggregated within the QFP body. This can be used to advantage to process physically larger parts of the tree (e.g., the first fewer addresses).
[0138] In some embodiments, the QFP-DAC is operated in an uncontrolled MFQ mode to increase the loading speed by over-current biasing during tipping without shunting the DAC junction, which can be useful, for example, for a QFP-DAC used as nothing more than a nonlinear amplifier in a readout chain.
[0139] The LATH and (LATCH) pair is optimally implemented with a global LATCHOFFSET line (not shown in FIG. 7B) and a single LATCH per tree level.
[0140] It is possible to operate the demultiplexer addressed by the LATCH as a multiplexer that can be used for reading.
[0141] Also, if we were truly wire-constrained, the compound Josephson junctions (CJJs) in QFP DACs could be intentionally made asymmetric to allow the omission of a global offset line in a latch-addressed implementation (Figure 7A), but this would reduce margins since the "offset" cannot be toggled on and off any more.
[0142] It is possible to design a QFP demux circuit with a loading bandwidth that is one-third to one-half that of an equivalent SFQ demux circuit.
[0143] The new reset technique can be implemented as part of the QFP demux circuit of FIG. 7A and / or FIG. 7B. In particular, it can utilize four branched junctions in parallel for each loop, which are biased equally to provide improved reset that would not be possible otherwise. This can be effectively thought of as a DC SQUID with both the junctions and the main loop suppressed. Junction suppression eliminates the asymmetry problem and the β-limited modulation. In the presence of asymmetry, the reset appears to be slightly better than the normal shunt reset. With the QFP-DAC, the MAX-SFQ can be significantly over-designed without penalty, thereby reducing the reset junction size and achieving a more robust reset. The benefit of this reset is that it resets to zero flux quantum even for significant asymmetry in the loops, while the branched two-junction reset assumes a final state that depends on the asymmetry of the two junctions.
[0144] The use of sufficiently short power pulses for fast lock-in measurements can facilitate a reduction in the typical cooling times of, for example, tens of milliseconds to tens of microseconds used in permanent magnet memories.
[0145] FIG. 8 shows a plot 800 of signals associated with the latch controlled QFP demux circuit of FIG. 7A.
[0146] In particular, plot 800 is a WRSPICE simulation of a QFP DAC without a demux, where the demux is simulated as a flux input.
[0147] Plot 800 shows values of phase 802 and current 804 near the bottom of plot 800 as three pulses are loaded into the QFP-DAC. Plot 800 also shows values of a pseudo demux input signal 806, a latch signal 808, a chip pulse 810, and a bias current 812. 8. Plot 800 also shows the value of QFP junction phase 814. Note that the reset is noisy in the QFP-DAC because the proper input signal was not provided over the chip lines to make the QFP-DAC invertible, and therefore the QFP-DAC loading is not adiabatic. When inverting the input signal (false final demux stage), there is no flux pumped into the loop.
[0148] Multijunction Superconducting Quantum Interface Device (SQUID) Flux Pumps Useful for Multiphase Flux DACs with Galvanic Selection-Free Addressing Existing architectures use an XYZ addressing scheme to select which DAC to program. One of these wires, sometimes called the POWER line, is galvanically connected to the DAC SQUID loop. However, the galvanic connection from the POWER line to the DAC makes it difficult to use some alternative inter-DAC coupling schemes, such as galvanically shared inductance, that might otherwise be useful when using Josephson DACs or DACs constructed from high dynamic inductance materials.
[0149] The other two lines, called ADDRESS and TRIGGER, are degenerate in their effect on selecting the DAC, meaning that the DAC is selectable (subject to POWER) when the sum of the biases on the two lines exceeds a threshold. At the operating point, the POWER signal is not orthogonal with respect to the ADDRESS and TRIGGER lines. The operating margin using these three relatively non-orthogonal biases is, in a sense, reduced compared to the operating margin using XY addressing, and therefore the present approach does not scale well to higher order addressing schemes, with the available operating margin reduced by approximately 1 / (addressing order).
[0150] Both of these problems are ameliorated to some extent by the addressing scheme described below: (1) no galvanic connection to the DAC is required by the addressing circuit, and (2) the control signals for the flux pumps are relatively more orthogonal.
[0151] FIG. 9 illustrates a three-junction two-loop flux pump circuit 900 addressing a digital-to-analog converter (DAC) 902 according to at least one illustrated embodiment.
[0152] The illustrated three-junction two-loop flux pump circuit 900 includes two loops 904a, 904b (collectively 904). Each of the loops 904 has a shared portion 906a, 906b (collectively 906) that is shared by successively adjacent ones of the loops 904. The three-junction two-loop flux pump circuit 900 includes three Josephson junctions 908a, 908b, 908c. Each of the shared portions 906 of the loops 904 that are shared by successively adjacent ones of the loops 904 is interposed with a respective one of the Josephson junctions 908a, 908b, 908c. Each of the loops 904 of the three-junction two-loop flux pump 900 has a respective storage inductance 910a, 910b.
[0153] The two interfaces 912a and 912b are connected to the respective storage inductances 910a and 910b of the loop 904, and the magnetic flux Φ xa , Φ xb Each loop 904 may be constructed from a material that exhibits superconductivity at least at a critical temperature.
[0154] The DAC 902 is coupled to the most extreme one of the loops 904b of the first multi-junction SQUID flux pump circuit 900. The loops of the DAC include a loop 914 that includes a shared portion 906b that is shared by the most extreme loop 904b of the first multi-junction SQUID flux pump circuit 900. The DAC 902 includes a storage inductance 916. The storage inductance 916 is at least one of a magnetic inductance, a kinetic inductance, a Josephson inductance, or a combination of two or more of a magnetic inductance, a kinetic inductance, and a Josephson inductance. The DAC 902 may be comprised of a loop of material that exhibits superconductivity at least at a critical temperature.
[0155] Figure 9 shows the most basic embodiment, where a two-loop three-junction flux pump is connected to a single superconducting storage inductor. The size of the inductor and the flux pump junctions determine the maximum useful capacity of the DAC. Figure 10 shows the applied flux waveforms 1002 and 1004, which are programming / deprogramming patterns, for the three-junction two-loop flux pump of Figure 9. Also shown is the Josephson phase 1006 of junction 908c of Figure 9, which is 2π times the total number of flux quanta in the loop.
[0156] One mode of programming operation shown in FIG. 10 is to supply a magnetic flux Φ xa How can the magnetic flux be pushed into and out of the loop 908a / 908b by using the supplied magnetic flux Φ xb We show how the same can be done for the 908b / 908c loop using the clock Φ xa The phase of is Φ xb When operating in sync with a π / 2 advance on the programming signal, one Φ 0 The magnetic flux is forced into the DAC by the pump using Φ xa Φ xb If the loop is delayed by π / 2 with respect to NΦ, then flux is pumped out of the loop, or negative flux is pumped into the loop. 0 is fed into the loop, and 2NΦ 0 is pumped out of the loop, and NΦ 0 is pumped in again.
[0157] FIG. 11 illustrates a four-junction, three-loop flux pump circuit 1100 addressing a digital-to-analog converter (DAC) 902 according to at least one illustrated embodiment.
[0158] The four junction, three loop flux pump circuit 1100 is similar to the three junction, three loop flux pump 900 (FIG. 9) but with an additional loop. Similar or even identical structures are identified with the same reference numbers as used in FIG. 9 and the accompanying description of FIG. 12A. For brevity and therefore clarity, only the significant differences between FIG. 11 and FIG. 9 are described below.
[0159] The illustrated four-junction, two-loop flux pump adds a third loop 904c, a fourth Josephson junction 908d, and a third storage inductance 901c in the third loop 904c. A third interface 912c is connected to a third inductance 910c to provide a third inductance with a magnetic flux Φ xc As noted above, each loop 904 may be comprised of a material that exhibits superconductivity at least at a critical temperature.
[0160] FIG. 12A shows a plot 1200a of applied flux waveforms 1202-1206 and Josephson phase 1208 of an endmost junction 908d for a programming / deprogramming pattern for the four-junction, two-loop flux pump of FIG. 11 according to at least one illustrated embodiment.
[0161] Although three loops and four Josephson junctions are shown, along with two loops and three Josephson junctions, the teachings herein are extendable to many more loops and Josephson junctions. Increasing the number of loops may allow flux to build up in the pump without supplying a bias, and a small number of clock activations may be sufficient to push the flux into the loops.
[0162] One advantage of avoiding a galvanic select line (e.g., the POWER line in at least some existing implementations) is that by providing a galvanic select line, continuous DAC accumulation can be avoided. The problem is that the loops are galvanically connected independently of other galvanic connections on the supply side, which in turn can create other undesirable current return paths. This can be particularly beneficial for Josephson DACs or multi-stage DACs with dynamic inductance-based storage elements, such as those with large λ thin-film TiN, NbN, NbTiN, or granular aluminum. In this case, the design of the division ratio can be simplified by using a shared inductance in the linked storage loops and avoiding magnetically coupling each to a linked washer, as is done in at least some existing architectures.
[0163] 12B is a schematic diagram of a circuit 1200 including a direct galvanic connection between two flux DACs according to at least one illustrated embodiment. Junctions 1210 and 1212 are input junctions to the individual DACs, e.g., corresponding to 908d in FIG. 11. The division ratio can be determined from the values of the shared inductance 1214 of the two DAC loops and their respective stored inductances 1216 and 1218. Circuit 1200b includes an inductance 1220.
[0164] For multiple phases, an XYZ addressing scheme is possible, for example as shown in Figure 13. Figure 13 illustrates eight sets of four-junction, three-loop flux pump circuits 1100a, 1100b, 1100c, 1100d, 1100e, 1100f, 1100g, 1100h (collectively 1100) addressing individual digital-to-analog converters (DACs) 902a, 902b, 902c, 902d, 902e, 902f, 902g, 902h (collectively 902) via multiple address lines 1352 according to at least one illustrated embodiment.
[0165] The four junction, three loop flux pump circuit 1100 may be, for example, identical to the four junction, three loop flux pump circuit 1100 (FIG. 11). The DAC 902 may be, for example, identical to the DAC 902 (FIG. 11). Similar or even identical structures are identified using the same reference numbers as used in FIG. 11 and the accompanying description of FIG. 14. For brevity and therefore clarity, the description of the structures will not be repeated and the present description will focus on the XYZ addressing scheme used to address a set or array of four junction, three loop flux pump circuits.
[0166] In particular, the eight DACs 902 are addressed using eight three-phase flux pumps 1100 operated by (2+2+2) clock lines. Within a device set, a particular phase is operated by one of N lines (two in FIG. 13, shown by solid vs. dashed lines). Selection is made to completely specify which DAC the wiring for each phase accesses. Selection is made at each layer, choosing between solid or dashed lines of each color.
[0167] The first control line set may include a first control line subset addressing a first multi-junction SQUID flux pump circuit subset, e.g., a first column of multi-junction SQUID flux pump circuits 1100a-1100d, and a second control line subset addressing a second multi-junction SQUID flux pump circuit subset, e.g., a second column of multi-junction SQUID flux pump circuits 1100e-1100h.
[0168] For example, the first control line subset may include a first control line 1352a coupled to an individual interface of the first loop of each multi-junction SQUID flux pump circuit 1100a-1100d of the first subset, a second control line 1352b coupled to an individual interface of the second loop of each multi-junction SQUID flux pump circuit 1100a-1100d of the first subset, and a third control line 1352c coupled to an individual interface of the third loop of each multi-junction SQUID flux pump circuit 1100a-1100d of the first subset.
[0169] For example, the second control line subset may include a first control line 1352d coupled to an individual interface of the first loop of each multi-junction SQUID flux pump circuit 1100e-1100h of the second subset, a second control line 1352e coupled to an individual interface of the second loop of each multi-junction SQUID flux pump circuit 1100e-1100g of the second subset, and a third control line 1352f coupled to an individual interface of the third loop of each multi-junction SQUID flux pump circuit 1100e-1100h of the third subset.
[0170] A first set of control lines 1352a-1352f (collectively 1352) includes N ctrl number of control lines, where N ctrl = 2 × N, the total number is N fp There are N multi-junction SQUID flux pumps. fp is equal to 2(N+1).
[0171] The control circuit 1360 is communicatively coupled and operable to perform multi-phase radio frequency (RF) clocking of the flux signals without direct current (DC) bias for the loops of the multi-junction SQUID flux pump circuit 1100 .
[0172] Multi-phase radio frequency (RF) clocking of the flux signal sequentially pushes flux through the loops of the first multi-junction SQUID flux pump into the DAC 902. Multi-phase radio frequency (RF) clocking of the flux signal sequentially pushes flux out of the DAC 902 through the loops of the multi-junction SQUID flux pump circuit 1100. The total number of phases of the multi-phase radio frequency (RF) clocking of the flux signal is equal to the total number N of loops of the multi-junction SQUID flux pump circuit 1100 into the DAC 902.
[0173] For example, the control circuit 1360 may sequentially supply flux signals to successive loops of the multi-junction SQUID flux pump circuit 1100 via the first N interfaces, where the individual flux signals supplied to each interface lead by π / 2 relative to the individual flux signals supplied to the immediately succeeding loop along the linearly successive loops of the multi-junction SQUID flux pump circuit 1100.
[0174] Also for example, the control circuit 1360 can sequentially supply flux signals to successive loops of the multi-junction SQUID flux pump circuit 1100 via the first N interfaces, where the individual flux signals supplied to each interface are delayed by π / 2 with respect to the individual flux signals supplied to the immediately succeeding loop along the linearly successive loops of the multi-junction SQUID flux pump circuit 1100.
[0175] In general, with an n-phase clocking scheme, imperfect patterns (e.g., patterns in which clock signals are missing in one or more of the phases) will interfere with the transmission of flux quanta through the pump. At least for some parameter values and for some combinations of clocks, this interference is not perfect and pulses can be sent with partial clocking. However, blocking pulses can be effectively used on otherwise dormant wires to enable, for example, normal XYZ selection operation. This is illustrated in plot 1400 of FIG. 14.
[0176] FIG. 14 illustrates a plot 1400 of the applied flux waveform versus Josephson phase of the endmost junction 908d for a programming / deprogramming pattern for the four-junction, two-loop flux pump of FIG. 11 according to at least one illustrated embodiment.
[0177] In particular, plot 1400 shows three clock signals 1402, 1404, and 1406, and one phase 1408 of the last Josephson junction 908d in the flux pump, illustrating the XYZ selection performance. In particular, the last sequence shows a phase shift with an inverse magnitude to the third clock phase. Use a blocking pulse of
[0178] The sequence shows one normal addressing of the DAC, with an associated lead at the output junction of the flux pump, and three partial (2 / 3) selections that produce no lead. The third partial selection is a fault case and requires a blocking pulse (a clock pulse of opposite magnitude) to prevent addressing of the DAC. The need for a blocking pulse does not limit the XYZ addressing scheme, but it does impose an additional constraint when calculating a parallel DAC loading scheme.
[0179] 9-14 and the accompanying description, it is clear that a multi-junction superconducting quantum interface device (SQUID) flux pump may comprise at least N loops, N being an integer equal to or greater than 2, each loop including a material that exhibits superconductivity at least at a critical temperature, each loop having a portion shared by successive adjacent loops, M Josephson junctions, M being greater than N, each loop having a portion shared by successive adjacent loops interposed with at least one of the Josephson junctions, each loop having a respective storage inductance, and a first N interfaces positioned to selectively communicatively couple to the respective loops magnetic flux of the respective loop. The loops of the multi-junction SQUID flux pump form a linear array of loops. The number N may be an integer, for example, 2 to 4 inclusive. The number M is equal to the sum N+1.
[0180] The DAC is coupled to the end loops of the multi-junction SQUID flux pump, the DAC comprising a material loop and a storage inductance, the DAC loop including a portion shared by the end loops of the multi-junction SQUID flux pump, and the interfaces can be individual inductive interfaces each located proximate to the storage inductance of an individual loop of the multi-junction SQUID flux pump.
[0181] FIG. 15 illustrates a method 1500 of operating a multiple junction superconducting quantum interface device (SQUID) flux pump according to at least one illustrated embodiment.
[0182] The method begins at 1502 upon initiation, such as by powering on, presenting a question, or invoking a routine or program.
[0183] At 1504, the controller circuit provides a flux signal to a first loop of the multi-junction SQUID flux pump via a respective first interface. At 1506, the controller circuit then, in turn, provides a flux signal to a second loop of the multi-junction SQUID flux pump via a respective second interface, the flux signal provided to the second loop being out of phase with respect to the first flux signal provided to the first loop.
[0184] Optionally, the flux pump includes a third loop, and at 1508, the controller circuit sequentially provides a flux signal to at least a third loop of the first multi-junction SQUID flux pump via a separate third interface, the flux signal provided to the third loop being out of phase with respect to the flux signal provided to the second loop.
[0185] Optionally, the flux pump includes a fourth loop, and at 1510, the controller circuit sequentially provides a flux signal to at least the fourth loop of the first multi-junction SQUID flux pump via a respective fourth interface, the flux signal provided to the fourth loop being out of phase with respect to the flux signal provided to the third loop.
[0186] For example, the controller circuit may include a first interface, a second interface, A separate flux signal can be provided to each of the first and second interfaces, the flux signals being out of phase by π / 2 with respect to the separate flux signals provided to the immediately succeeding loops along the linear succession of loops of the multi-junction SQUID flux pump.
[0187] Providing flux signals to the first, second and optional further (e.g., third) loops of the multi-junction SQUID flux pump may include providing flux signals that sequentially push flux through the loops of the multi-junction SQUID flux pump to a digital-to-analog converter (DAC). For example, the controller circuit may provide flux signals to the first, second and third loops of the multi-junction SQUID flux pump, where the individual flux signals provided to each of the individual first, second and third interfaces lead by π / 2 with respect to the individual flux signals provided to the immediately succeeding loop along the linear train of the multi-junction SQUID flux pump.
[0188] Alternatively, providing flux signals to the first loop, the second loop, and optional further (e.g., third) loops of the multi-junction SQUID flux pump may include providing flux signals that sequentially push flux through the loops of the multi-junction SQUID flux pump to a digital-to-analog converter (DAC). For example, the controller circuit may provide flux signals to the first loop, the second loop, and the third loop of the multi-junction SQUID flux pump, where the individual flux signals provided to each of the individual first interface, second interface, and third interface are delayed by π / 2 with respect to the individual flux signal provided to the immediately succeeding loop along the linear series of loops in the multi-junction SQUID flux pump.
[0189] Providing a flux signal to the first loop, the second loop and the third loop of the first multi-junction SQUID flux pump may include providing a flux signal having a first sign and a first magnitude via a first control line, a second control line and a third control line, and subsequent to providing the flux signal to the first loop, the second loop and the third loop of the first multi-junction SQUID flux pump, providing a flux signal having a second sign and a first magnitude to at least one of the loops of the second multi-junction SQUID flux pump via a fourth control line, where the second sign is opposite to the first sign.
[0190] The method 1500 stops at 1512, for example, until invoked again.
[0191] FIG. 16 illustrates a method 1600 of operating a multiple junction superconducting quantum interface device (SQUID) flux pump according to at least one illustrated embodiment.
[0192] The method begins at 1602 upon initiation, such as by powering on, presenting a question, or invoking a routine or program.
[0193] The controller circuit simultaneously supplies a flux signal having a first sign and a first magnitude to a first loop of the multi-junction SQUID flux pump via a respective first interface at 1604 and supplies flux signals having a first sign and a first magnitude to all of the other loops of the multi-junction SQUID flux pump via respective second interfaces at 1606, the flux signals supplied to all of the loops being in phase with each other.
[0194] Providing a flux signal having a first sign and a first magnitude to all of the other loops may include providing a flux signal having a first sign and a first magnitude to the first loop, the second loop and at least a third loop of the first multi-junction SQUID flux pump via respective first control lines, second control lines and third control lines.
[0195] The method 1600 stops, for example, at 1608 until invoked again.
[0196] Resetting the flux DAC is a difficult operation to implement as a simple reset scheme, as shown in plot 1600b of FIG. 16B, and involves activating all clock phases 1610, 1612, and 1614 at the same time, and in phase to deprogram the DAC (see 1616). As shown in FIG. 16B, an initial clock sequence 1618 programs the DAC loop, and a subsequent second sequence 1620 unloads it. The essential difference in the second sequence is that the pulses occur simultaneously, as opposed to the staggered pulses of the programming sequence. When they occur simultaneously, there is no preferred direction for the flux, so a pulse in either direction will unload it. Therefore, the DAC is first programmed with the clock sequence as described. The DAC is then emptied by running the clocks simultaneously.
[0197] An alternative approach would be to work with DACs that have equivalent design capacities, but overfill the DACs, bring them to capacity, and then attempt to deprogram them by a known amount. This approach can be applied even if the design capacities of the various DACs are different from each other, and it may simply be that the starting point for programming some DACs can be, for example, +10Φ0, rather than an intermediate scale. In fact, it is possible to fill all DACs to capacity and then not deprogram them, but to treat the full-to-capacity state as a starting point. However, this approach is disadvantageously prone to increasing the total number of Φ0s that must be sent to achieve a given processor state.
[0198] Using capacitance as a reference point would probabilistically result in one of two possible states at the next reset, since the DAC's capacitance is exactly on the boundary between the two levels and would not avoid the DAC's problems. This behavior can be addressed by performing a customized per-DAC reset recovery for the DAC that just went into an abnormal state when nominally reset. If flux pumps were used to program the DACs to capacity (or any other reset scheme), this could similarly be addressed by running the individual flux pumps at customized levels in the reset recovery sequence.
[0199] Extending XYZ Addressing to Multicore Processors The XYZ addressing scheme described above can be extended to the case of processors with multiple cores, e.g. quantum processors. In particular, adding two or more serial processor cores allows them to use some of the same signal lines (e.g. IO lines driving a particular device).
[0200] The signal lines may be connected in series on a sample holder, carrier PCB or multi-chip module, while the individual cores may be tested and calibrated separately prior to assembly.
[0201] An example embodiment of such a multi-core processor may have N cores, each core being an array of qubit cells, where N may be greater than or equal to 2. A qubit cell may have Q qubits, e.g., 8 qubits, and a core may have C cells, e.g., 24 cells.
[0202] In an embodiment with 8 qubits per cell, there are 9 DACs per qubit, and therefore 5 trigger lines, 15 address lines, and one common power line within the cell. The power line selects an array of tiles, and the address and trigger lines select different cells. Considering an embodiment with 24 cells per core with 8 qubits and N cores, the total number of qubits is 9×8×(24×24) * There are N=41,472N addressed DACs. With 16 cores, for example, there are 661,248 DACs. For illustrative purposes only, an embodiment with 16 cores is described herein, and one of ordinary skill in the art will recognize that fewer and more cores are possible.
[0203] XYZ addressing requires 3 times the cube root of the number of wires rounded up, which is 264 in the above example. To make the layout more regular, the user power lines may be allowed to select a quarter of the cells on each processor core, so in the above example there are 64 power lines, each line selecting 6×24=144 cells. To address each DAC in a group of 144 cells requires 24×5=120 trigger lines and 6×15=90 address lines. Thus, in the above example, the total number of wires for XYZ programming of all DACs is 64+120+90=274.
[0204] Braid control line 17A illustrates a set of four digital-to-analog converters (DACs) 1700a, 1700b, 1700c, 1700d (collectively referred to as 1700) and a number of signal lines 1702a-1702c (collectively referred to as 1702, with only three called out in FIG. 17A for clarity) arranged to individually address each of the DACs 1704(i,j) (where i is an integer from 1 to n and j is an integer from 1 to n, with only six called out in FIG. 17A for clarity of illustration, collectively referred to as 1704). A control circuit 1708 is communicatively coupled to provide signals to selected ones of the DACs 1704 via individual pairs of the signal lines 1702 and one of four power lines (not shown) according to at least one illustrated embodiment.
[0205] In the illustrated embodiment, the DACs 1704 in each set of DACs 1700a-1700d are arranged in a separate two-dimensional array with multiple rows (e.g., extending horizontally across the page) and multiple columns (e.g., extending vertically across the page). Each DAC set or array 1700a-1700d is controlled or selected via a separate one of four signal lines (not shown), which may be referred to as a "power line" for ease of reference. Similarly, each DAC set or array 1700a-1700d may be referred to as a power domain, with each DAC in a power domain being controlled or selected by the same power line. The DACs 1704 in each DAC set or array 1700a-1700d are arranged in a 6×6 grid or array (i.e., four DACs) of quartets of DACs 1710a, 1710b (only two are called out, one from set or array 1700a and one from set or array 1700b, collectively referred to as 1710), for a total of 36 quartets of DACs 1710 and a total of 144 DACs per power domain. 1704 (e.g., DAC sets or arrays 1700a-1700d). Although generally shown as an ordered arrangement, many of the techniques and structures described herein can be utilized with non-ordered arrangements or sets of DACs and / or other devices.
[0206] In addition to the "power lines," there are other signal lines 1702 that, in combination with the power lines, uniquely address each of the 144 DACs 1704 in the four power domains 1700a-1700d. These additional signal lines 1702 may be referred to as "control lines" for ease of reference. As shown and described below, these control lines 1702 are arranged in a braided pattern or configuration across individual sets or arrays or power domains 1700a-1700d, allowing for a relatively limited total number of signal lines 1702 addressing a given number of DACs 1704 to achieve a very efficient addressing scheme. For example, signal The signal lines 1702 are 4 (N signal -1) 2 Add to DAC 1704 Each DAC 1704 can address one or more of the DACs, which are communicatively coupled to a respective pair of signal lines 1702 and one power line, and are controlled by signals provided by control circuitry 1708 and carried by the respective pair of signal lines 1702 and one power line. Thus, control circuitry 1708 can uniquely address every DAC 1704 via a pair of signal lines 1702 and one of the four power lines.
[0207] N signal The signal lines 1702 may include a set of control lines 1702 and a set of power lines (not shown), with each triplet consisting of a unique combination of two of the control lines 1702 and one of the power lines. The signal lines 1702 and the power lines may each be composed of a material that exhibits superconductivity at least at a critical temperature.
[0208] As described above, the DACs 1704 are arranged in a plurality of two-dimensional arrays, each of the two-dimensional arrays including a respective plurality of DACs 1704, a plurality of signal lines 1702, and a first subset of signal lines, each signal line of the first subset of signal lines being communicatively coupled to a DAC of the respective two-dimensional array. Each signal line of the first subset of signal lines may be referred to as a power line and is communicatively coupled to all of the DACs of the respective two-dimensional array or power domain 1700a-1700d. Each signal line 1702 may be referred to as a control line 1702 and is communicatively coupled to a DAC subset 1704 in each of two or more of the two-dimensional arrays 1700a-1700d. For example, each signal line 1702 of the second subset of signal lines is communicatively coupled to all of the DAC subsets of the two-dimensional arrays 1700a-1700d.
[0209] As shown, the first control line is operatively proximate (e.g., close enough to transmit magnetic flux) to at least one DAC in each of at least three rows and at least three columns of the first DAC two-dimensional array so as to communicatively couple to an individual DAC in each of at least three rows and at least three columns of the first DAC two-dimensional array. As shown, the second control line is operatively proximate to at least one DAC in each of at least three rows and at least three columns of the first DAC two-dimensional array so as to communicatively couple to an individual DAC in each of at least three rows and at least three columns of the first DAC two-dimensional array. The first control line and the second control line may be operatively proximate to different DACs without a common DAC. Alternatively, the first control line and the second control line may be operatively proximate to the same at least one of the DACs with at least one common DAC.
[0210] FIG. 17B shows four digital-to-analog converter (DAC) sets 1700a, 1700b, 1700c, 1700d (collectively referred to as 1700) and control lines 1702, e.g., 1702a, 1702b, and 1702c of FIG. 17A (not specifically called out in FIG. 17B to avoid clutter), and also shows multiple diagonal lines representing the direction of the braid control lines 1702.
[0211] The method advantageously includes signal 4(N signal -1) 2 It is possible to control DACs.
[0212] With respect to the braiding of the control lines 1702, the first control lines are disposed proximate to each DAC along the first diagonal line 1770a of the first DAC two-dimensional array 1700a such that they communicatively couple to individual DACs along the first diagonal line 1770a of the first DAC two-dimensional array 1700a. The first control lines are also disposed proximate to each DAC along the first diagonal line 1770b of the second DAC two-dimensional array 1702b such that they communicatively couple to individual DACs along the first diagonal line 1770b of the second DAC two-dimensional array 1700b. Similarly, the first control lines are also disposed proximate to each DAC along the first diagonal line 1770b of the third DAC two-dimensional array 1700c. 770c and the first diagonal line 1770d of the fourth DAC two-dimensional array 1700d may be positioned proximate to each DAC along the first diagonal line 1770c of the third two-dimensional array 1700c and the first diagonal line 1770d of the fourth DAC two-dimensional array 1700d so as to be communicatively coupled to individual DACs along the first diagonal line 1770c of the third two-dimensional array 1700c and the first diagonal line 1770d of the fourth DAC two-dimensional array 1700d.
[0213] When the DACs of the first DAC two-dimensional array 1700a are arranged in multiple rows and columns, the first diagonal line 1770a of the first DAC two-dimensional array 1700a may extend across all rows and columns of the first DAC two-dimensional array 1700a, for example. When the DACs of the second DAC two-dimensional array 1700b are arranged in multiple rows and columns, the first diagonal line 1770b of the second DAC two-dimensional array 1700b may extend across all rows and columns of the second DAC two-dimensional array, for example. Similarly, the first diagonal line 1770c of the third DAC two-dimensional array 1700c and the first diagonal line 1770d of the fourth DAC two-dimensional array 1700d may extend across all rows and columns of the third DAC two-dimensional array 1700c and the fourth DAC two-dimensional array 1700d, for example.
[0214] In some embodiments, the first diagonal lines 1770b of the second two-dimensional array 1702b are perpendicular to the first diagonal lines 1770a of the first two-dimensional array 1700a, and the first diagonal lines 1770c of the third two-dimensional array 1700c are perpendicular to the first diagonal lines 1770b of the second two-dimensional array 1700b. In some embodiments, the first diagonal lines 1770d of the fourth two-dimensional array 1700d are perpendicular to the first diagonal lines 1770a of the first two-dimensional array 1700a, and the first diagonal lines 1770d of the fourth two-dimensional array 1700d are perpendicular to the first diagonal lines 1770c of the third two-dimensional array 1700c.
[0215] In particular, a first control line may follow a serpentine path along a first diagonal line 1770a of the first DAC two-dimensional array 1700a. A first control line may follow a serpentine path along a first diagonal line 1770b of the second DAC two-dimensional array 1700b. A first control line may follow a serpentine path along a first diagonal line 1770c of the third DAC two-dimensional array 1700c. A first control line may follow a serpentine path along a first diagonal line 1770d of the fourth DAC two-dimensional array 1700d. Similarly, additional control lines may follow serpentine paths along individual diagonals of the first, second, or other DAC two-dimensional arrays 1700a-1700d.
[0216] With respect to the braiding of the control lines, the second control lines are disposed proximate to each DAC along the second diagonal line 1772a of the first DAC two-dimensional array 1700a such that they are communicatively coupled to individual DACs along the second diagonal line 1772a of the first DAC two-dimensional array 1700a. The second control lines are also disposed proximate to each DAC along the second diagonal line 1772b of the second DAC two-dimensional array 1700b such that they are communicatively coupled to individual DACs along the second diagonal line 1772b of the second DAC two-dimensional array 1700b. Similarly, the second control lines may also be positioned proximate to each DAC along the second diagonal line 1772c of the third DAC two-dimensional array 1700c and the second diagonal line 1772d of the fourth DAC two-dimensional array 1700d so as to communicatively couple to individual DACs along the second diagonal line 1772c of the third DAC two-dimensional array 1700c and the second diagonal line 1772d of the fourth DAC two-dimensional array 1700d.
[0217] Where the DACs of the first DAC two-dimensional array are arranged in multiple rows and columns, the second diagonal line 1772a of the first DAC two-dimensional array 1700a may extend across, for example, all of the rows and all of the columns of the first DAC two-dimensional array 1700a. Alternatively, the second diagonal line 1772a of the first DAC two-dimensional array 1700a may extend across, for example, a subset (i.e., less than all) of the rows and / or a subset (i.e., less than all) of the columns of the first DAC two-dimensional array 1700a. Thus, any given control line may extend along a first diagonal line across a portion of DAC two-dimensional arrays 1700a-1700d, then switch direction and extend along a second diagonal line across the remainder of DAC two-dimensional arrays 1700a-1700d. The second diagonal line may extend at a non-zero angle, e.g., a right angle or 90°, from the first diagonal line. In some embodiments, e.g., with a sufficiently large DAC array, a given control line may extend along one, two, three, or more diagonals, with each successive diagonal line extending at a non-zero angle relative to the front and rear diagonals. The successive angles may all be the same, or one or more angles may differ from one another in magnitude and / or direction. For example, the non-zero angles may include, e.g., 90°, 60°, 45°, or 30°. Also, for example, the direction of the angles may be clockwise or counterclockwise relative to the front and rear diagonals. Therefore, some of the control lines may take illegal, zig-zag paths across one or more DAC sets or arrays 1700a-1700d.
[0218] With respect to the braiding of the control lines, the third control lines are disposed adjacent to each DAC along the third diagonal line 1774a and the fourth diagonal line 1776a of the first DAC two-dimensional array 1700a so as to communicatively couple to individual DACs along the third diagonal line 1774a and the fourth diagonal line 1776a of the first DAC two-dimensional array 1700a. The fourth diagonal line extends at a non-zero angle (e.g., perpendicular, right angle, 90°) to the third diagonal line. For example, if the DACs of the first DAC two-dimensional array 1700a are arranged in rows and columns, the third diagonal line 1774a of the first DAC two-dimensional array 1700a may extend across a first plurality of rows and a first plurality of columns of the first DAC two-dimensional array 1700a, and the fourth diagonal line 1776a of the first DAC two-dimensional array 1700a may extend across a second plurality of rows and a second plurality of columns of the first DAC two-dimensional array 1700a. In some aspects, the combination of the first plurality of rows and the second plurality of rows includes all of the rows of the first DAC two-dimensional array 1700a, and the combination of the first plurality of columns and the second plurality of columns includes all of the columns of the first DAC two-dimensional array 1700a.
[0219] The third control lines may also be disposed proximate to each DAC along a third diagonal line (not shown) and a fourth diagonal line (not shown) of the second DAC two-dimensional array 1700b to communicatively couple to individual DACs along the third diagonal line and the fourth diagonal line of the second DAC two-dimensional array, the fourth diagonal line of the second DAC two-dimensional array extending at a non-zero angle (e.g., perpendicular, or orthogonal) to the third diagonal line of the second DAC two-dimensional array. The third control lines may also be disposed proximate to each DAC along the third diagonal line and the fourth diagonal line of the third dimensional array of DACs 1700c to communicatively couple to individual DACs along the third diagonal line and the fourth diagonal line of the third DAC two-dimensional array 1700c. The third control lines may also be positioned closer to each DAC along the third and fourth diagonal lines of a fourth dimensional array of DACs 1700d so as to communicatively couple to individual DACs along the third and fourth diagonal lines of the fourth DAC two-dimensional array 1700d, and similarly, the fourth diagonal lines of the fourth DAC two-dimensional array 1700d extend at a non-zero angle relative to the third diagonal lines of the fourth DAC two-dimensional array 1700d.
[0220] In this embodiment of the concept, there is little difference between having an even number of DACs per column and an odd number of DACs. Figure 18 shows a repeatable tiling for blade crossing pairs with scalability to an arbitrarily large number of lines.
[0221] FIG. 18 illustrates a quartet 1800 of DACs in accordance with at least one illustrated embodiment; 18 shows three control lines 1802a, 1802b, 1802c. In the quartet of DACs 1800, the three control lines 1802a, 1802b, 1802c (collectively 1802) may be similar or even identical to the corresponding structures in Figures 17A and 17B. Figure 18 better illustrates the physical placement of the control lines 1802 that individually address each of the DACs in the quartet 1800.
[0222] The quartet 1800 includes a first DAC 1804 a,a, a second DAC 1804 a,b, a third DAC 1804 b,b, and a fourth DAC 1804 b,a. Each of the DACs includes a material loop, a Josephson junction pair interposed in the material loop, and an interface pair (e.g., an inductive interface).
[0223] A first control line 1802a passes in operative proximity through the first DAC 1804a,a, the fourth DAC 1804b,a, the third DAC 1804b,b and the second DAC 1804a,b of the quartet 1800. The first control line 1802a includes an interface (e.g., an inductive interface) in operative proximity to complementary interfaces of the first DAC 1804a,a, the fourth DAC 1804b,a, the third DAC 1804b,b and the second DAC 1804a,b to provide a signal (e.g., magnetic flux) between the complementary interfaces.
[0224] The second control line 1802b passes in operative proximity through the fourth DAC 1804b,a and the first DAC 1804a,a of the quartet 1800. The second control line 1802b includes an interface (e.g., an inductive interface) in operative proximity to the complementary interfaces of the fourth DAC 1804b,a and the first DAC 1804a,a to provide a signal (e.g., magnetic flux) between the complementary interfaces.
[0225] The third control line 1802c passes in operative proximity through the third DAC 1804b,b and the second DAC 1804a,b of the quartet 1800. The third control line 1802c includes an interface (e.g., an inductive interface) in operative proximity to the complementary interfaces of the third DAC 1804b,b and the second DAC 1804a,b to provide a signal (e.g., magnetic flux) between the complementary interfaces.
[0226] FIG. 19 illustrates a method 1900 of operation for utilizing braided control lines to individually address DACs in a two-dimensional set (eg, an ordered array) of DACs, according to at least one illustrated embodiment.
[0227] The method 1900 begins at 1902, for example, in response to powering up, providing or installing a problem, or in response to initiation by invoking a routine or program.
[0228] During a first time period, at 1904, the control circuit provides signals to a first set of DACs (eg, a power domain) via a first power line.
[0229] During a first time period, at 1906, the control circuit provides a signal over a first control line to a first subset of the first DAC set, simultaneously with providing a signal over the first power supply line.
[0230] During a first period, at 1908, the control circuit provides signals to a second subset of the first set of DACs via second control lines, simultaneously with providing signals via the first power line and providing signals via the first control line. The second subset of DACs has a single DAC in common with the first subset of DACs, such that the single DAC is supplied by a triplet of one power line and two control lines (i.e., three signal lines). Each of the memory cells is individually addressed by a signal.
[0231] The method 1900 stops at 1910, for example, until it is called or invoked again.
[0232] Low-field magnetic imaging device It is desirable to measure magnetic fields at low temperatures in close proximity to quantum anneal processors. Existing implementations use DC SQUIDs to measure magnetic fields around the processor. Typically, 13 bias lines are used for 35 DC SQUIDs, with 5 of the bias lines requiring dedicated filtering and electronic boards. If the DC SQUIDs are removed on the quantum processor, the magnetometer remains simply for the reason of maintaining the old filters and electronic boards. N SQUIDs typically require approximately sqrt(N) bias lines. DC SQUIDs can occasionally latch up to unintended voltage states, which undesirably generates heat on the chip. Additionally, DC SQUID designs are limited from making the body inductance too large, as this reduces the modulation depth and limits the magnetic field sensitivity of the SQUIDs.
[0233] Using QFP magnetometers hung in a shift register and readout via an NDRO, reliance on DC SQUIDs can be advantageously eliminated. The number of additional wires required is approximately 6 for any number of linearly connected QFP magnetometers. Two-dimensional array QFP magnetometers can be used with up to 18 wires.
[0234] A QFP magnetometer is a QFP whose body is not active under a ground plane, so its degeneracy point is a function of the external magnetic field. A single read of the QFP gives only a 0 or a 1, and while reading the QFP many times, the flux bias on the QFP body is swept and the output is monitored until it switches from a 0 to a 1. By fitting the results to a tanh shape, the degeneracy point can be found very accurately (where the density of 0s and 1s is equal).
[0235] +Φ 0 / 2 and +Φ 0 Between and -Φ 0 / 2 and -Φ 0 The asymmetry can be compensated for by annealing the QFP between and and averaging the resulting degeneracy points. It is possible to measure and compensate for the interference of the QFP anneal lines to the body.
[0236] QFP magnetometers can also be used as elements of a shift register if the QFP magnetometer is instrumented with flux bias lines or if the supplied flux is less than the coupling flux between the shift register elements.
[0237] This allows, for example, a grid of QFP magnetometers to be set up as an XY shift register used for magnetic field imaging to read out a QFP at a time. In the simplest case, one must find the degeneracy point of each QFP individually. The QFPs can also be read out one column at a time if the magnetic field is small enough and the flux bias for each column is scanned to find the degeneracy points of the QFPs in the column in parallel.
[0238] The simplest approach is to replace the existing X, Y, small, medium and large DC SQUIDs with the equivalent QFP versions. The body will be the same except that the SQUIDs will be communicatively coupled to nearby SQUIDs or to small shift registers that connect the SQUIDs to a shift register external to the processor and plug into the NDRO at the corner of the external shift register.
[0239] A QFP with a flux bias acting as a magnetometer can be used, for example, to compensate for residual magnetic fields (e.g., cycle to 9.2K, cool, read at 4K, compensate, cycle to 9.2K, It can be used for cooling, etc.
[0240] In at least one embodiment, a system includes a first quantum flux parametron (QFP) magnetometer set, each QFP magnetometer of the first QFP magnetometer set including an individual first material loop, an individual second material loop, and an individual compound Josephson junction with a Josephson junction pair, the individual first material loop and the individual second material loop undergo superconductivity at a critical temperature, the second material loop is interposed in the individual first material loop, the individual Josephson junction pair is interposed in the second material loop, the individual second material loop is parallel to each other in the second material loop with respect to a node interposed in the individual first material loop, and the individual second material loop has an individual second loop inductively receiving a control signal to control the individual compound Josephson junction. the first material loop includes a first quantum flux parametron (QFP) magnetometer set, the individual first material loop including a plurality of first loop inductive interfaces, at least one of the first loop inductive interfaces being exposed to magnetic flux from an external magnetic field without an intervening ground plane to measure the external magnetic field with an individual QFP magnetometer, a control line set arranged to selectively communicatively couple a control signal to the second loop inductive interface, and a flux bias line set arranged to selectively communicatively couple a flux bias to at least one of the first loop inductive interfaces of each QFP magnetometer of the first QFP magnetometer set that is subject to flux feedback for measuring the external magnetic field. In some embodiments, for each QFP magnetometer of the first QFP magnetometer set, at least some of the first loop inductive interfaces couple the individual QFP magnetometer to individual first loop inductive interfaces of nearby QFP magnetometers of the first QFP magnetometer set.
[0241] The system may further include a processor chip with a plurality of qubits and a plurality of couplers, each coupler operable to selectively communicatively couple a respective pair of qubits, and the first QFP magnetometer set is arranged to detect an external magnetic field. The first QFP magnetometer set may be an integrated component of the processor chip. Control circuitry communicatively coupled to the QFP magnetometers of the first magnetometer set via flux bias lines may be operable to sweep a flux bias supplied to a body of each of the QFP magnetometers of the first magnetometer set while taking a plurality of readouts of the QFP magnetometers of the first magnetometer set. The control circuitry may determine a respective degeneracy point of each of the QFP magnetometers of the first magnetometer set based at least in part on the plurality of readouts. For example, for each QFP magnetometer in the first magnetometer set, the control circuitry fits the respective set of readout results to a tanh shape to determine the respective degeneracy point for each QFP magnetometer in the first magnetometer set.
[0242] The control circuit is further communicatively coupled to the QFP magnetometers of the first magnetometer set via the first set of anneal lines to control a +Φ 0 / 2 and +Φ 0 Between and -Φ 0 / 2 and -Φ 0 and operative to anneal each of the QFP magnetometers of the first magnetometer set between.
[0243] The control circuitry may further determine a measurement of the anneal line / QFP magnetometer interference and compensate for the determined anneal line / QFP magnetometer interference.
[0244] The system may further include a first shift register communicatively coupled to the first set of QFP magnetometers, The first shift register may take the form of a QFP-based shift register, which is a shift register magnetically or galvanically coupleable to individual ones of the first set of QFP magnetometers.
[0245] The system may further include a number of non-destructive readouts (NDROs).
[0246] In operation, the control circuit is capable of sweeping the flux bias applied to the body of the individual QFP magnetometer for each of the QFP magnetometers in the first QFP magnetometer set, and taking multiple readings of the individual QFP magnetometer while sweeping the flux bias applied to the body of the individual QFP magnetometer.
[0247] In operation, the control circuitry can further provide a flux bias signal to sequentially shift quantum flux values between successive shift register elements.
[0248] During operation, the control circuitry may further determine an individual degenerate point for each of the QFP magnetometers of the first set of QFP magnetometers based at least in part on the multiple readouts. Determining the individual degenerate points for each of the QFP magnetometers of the first set of magnetometers based at least in part on the multiple readouts may include, for example, fitting the set of individual readout results to a tanh shape to determine the individual degenerate points for each of the QFP magnetometers of the first set of magnetometers.
[0249] In operation, the control circuit further provides signals via the first set of anneal lines to the QFP magnetometers of the first set of magnetometers to provide +Φ 0 / 2 and +Φ 0 Between and -Φ 0 / 2 and -Φ 0 Each of the QFP magnetometers of the first magnetometer set can be annealed between and.
[0250] In operation, the control circuitry can further determine a measurement of the anneal line / QFP magnetometer interference based at least in part on the determined measurement, and compensate for the determined anneal line / QFP magnetometer interference.
[0251] Exemplary Superconducting Quantum Processor FIG. 20 is a schematic diagram of a portion of an exemplary superconducting quantum processor 2000 designed for quantum annealing (and / or adiabatic quantum computing) components that may be used to implement the present systems and apparatus. The portion of superconducting quantum processor 2000 shown in FIG. 20 includes two superconducting qubits 2001 and 2002. Also shown is a tunable coupling (diagonal coupling) between qubit 2001 and qubit 2002 (i.e., providing two local interactions) via coupler 2010. Although the portion of quantum processor 2000 shown in FIG. 20 includes only two qubits 2001, 2002 and one coupler 2010, one skilled in the art will appreciate that quantum processor 2000 may include any number of qubits and any number of couplers that couple information between the qubits.
[0252] As shown in FIG. 20, portions of quantum processor 2000 may be implemented to physically realize quantum annealing and / or adiabatic quantum computation. Quantum processor 2000 includes a number of interfaces 2021-2025 that are used to configure and control the state of quantum processor 2000. Each of interfaces 2021-2025 may be part of a programming subsystem and / or a time evolution subsystem, as shown. 2000。 Alternatively, or additionally, interfaces 2021-2025 may be implemented by galvanic coupling structures, e.g., as described in more detail elsewhere herein. In some embodiments, one or more of interfaces 2021-2025 may be driven by one or more DACs. Such programming and / or time evolution subsystems may be separate from quantum processor 2000 or may be included locally (i.e., on-chip with quantum processor 2000).
[0253] In operation of quantum processor 2000, interfaces 2021 and 2024 are used to couple flux signals into the respective compound Josephson junctions 2031 and 2032 of qubits 2001 and 2002, respectively, thereby introducing tunable inputs into the Hamiltonian of the system. The possible tunneling term (Δ i This coupling can be realized by the off-diagonal coupling term σ x and these flux signals are examples of “delocalized signals.” Examples of Hamiltonians (and Hamiltonian terms) used in quantum computing are described in more detail, for example, in U.S. Patent Application Publication No. 20140344322.
[0254] Similarly, interfaces 2022 and 2023 are used to provide flux signals to individual ones of the qubits 2001 and 2002, respectively, thereby introducing h i This coupling can be realized by the diagonal terms σ z We also use interface 2025 to couple the flux signal into coupler 2010, which gives us J ij This coupling can be realized by the diagonal term σ i z σ j z Gives.
[0255] In Figure 20, the contribution of each of the interfaces 2021-2025 to the Hamiltonian of the system is shown in boxes 2021a-2025a, respectively. As shown, in the example of Figure 20, boxes 2021a-2025a are elements of the time-varying Hamiltonian for quantum annealing and / or adiabatic quantum computation.
[0256] Throughout this specification and the appended claims, the term "quantum processor" is used generally to describe a collection of physical qubits (e.g., qubits 2001 and 2002) and couplers (e.g., coupler 2010). The physical qubits 2001 and 2002 and coupler 2010 are referred to as the "programmable devices" of quantum processor 2000, and corresponding parameters of the qubits and couplers (e.g., qubit h i Value and coupler J ij These programmable parameters (values) are referred to as the "programmable parameters" of the quantum processor. In the context of a quantum processor, the term "programming subsystem" is used generally to describe the application of programmable parameters to programmable devices of quantum processor 2000 and other associated control circuitry and / or instructions using interfaces (e.g., "programming interfaces" 2022, 2023, and 2025).
[0257] As previously mentioned, the programming interface of the programming subsystem may communicate with other subsystems that may be separate from the quantum processor or may be included locally on the processor. As described in more detail below, the programming subsystem may be configured to receive machine language programming instructions for the quantum processor and execute the programming instructions to program the programmable device in accordance with the programming instructions. Similarly, in the context of a quantum processor, the term "time evolution subsystem" generally includes time evolving the programmable devices and other associated control circuitry and / or instructions of the quantum processor 2000 using an interface (e.g., "time evolution interfaces 2021 and 2024). For example, the time evolution subsystem may include anneal signal lines and corresponding interfaces (2021, 2024) to the qubits (2001, 2002).
[0258] Quantum processor 2000 also includes readout devices 2051 and 2052, where readout device 2051 is associated with qubit 2001 and readout device 2052 is associated with qubit 2002. In some embodiments, as shown in FIG. 20, readout devices 2051 and 2052 each include a DC SQUID inductively coupled to a corresponding qubit. In the context of quantum processor 2000, the term "readout subsystem" is used generally to describe the use of readout devices 2051, 2052 to read out the final states of qubits (e.g., 2001 and 2002) in the quantum processor to generate a bit string. The readout subsystem also includes routing circuitry (e.g., The qubit readout may also be performed using alternative circuits such as those described in PCT Patent Application Publication WO2012064974.
[0259] 20 shows only two physical qubits 2001, 2002, one coupler 2010, and two readout devices 2051, 2052, a quantum processor (e.g., processor 2000) may utilize any number of qubits, couplers, and / or readout devices, including many (e.g., hundreds, thousands, or more) of qubits, couplers, and / or readout devices. Applying the teachings herein to a processor to a different (e.g., greater) number of computational components will be readily apparent to one of ordinary skill in the art.
[0260] Examples of superconducting qubits include superconducting flux qubits and superconducting charge qubits. In superconducting flux qubits, the Josephson energy dominates or is equal to the charge energy. In charge qubits, the opposite is true. Examples of flux qubits that can be used include radio frequency SQUIDs, which contain a superconducting loop interleaved with one Josephson junction, and persistent current qubits, which contain a superconducting loop interleaved with three Josephson junctions.
[0261] Exemplary Superconducting Dynamic Inductance DAC FIG. 21 illustrates an example DAC 2100 with a superconducting loop 2108 that carries a superconducting current during operation. The superconducting loop 2108 may include any suitable superconducting material, such as, for example, niobium, titanium, and / or the like. The superconducting loop 2108 may be coupled to an inflow line 2120 and an outflow line 2122 that receive a current, and a voltage is induced in the superconducting loop 2108 by the inflow line 2120 and the outflow line 2122. The outflow line 2122 may optionally supply current to other devices (e.g., by acting as an inflow line 2120 to other devices). For example, in some embodiments, the DACs are coupled together in series by the inflow line 2120 and the outflow line 2122, thereby supplying current to multiple DACs.
[0262] The superconducting loop 2108 is interleaved with one or more Josephson junctions. For example, as shown in FIG. 21, the superconducting loop 2108 may be interleaved with a compound Josephson junction ("CJJ") 2102. The CJJ 2102 comprises a plurality of Josephson junctions 2104a, 2104b (individually and collectively referred to as 2104, with the embodiment shown showing two Josephson junctions) at least partially arranged in parallel. The parallel branches of the CJJ 2102 may have separate bias elements 2106a, 2106b (individually and collectively referred to as 2106), e.g., Josephson junction 2104a is serially coupled to bias element 2106a and Josephson junction 2104b is serially coupled to bias element 2106b. The biasing elements 2106 may be inductively, galvanically, or otherwise coupled to an electrical signal to bias the current passing through the individual Josephson junctions 2104, thereby programmably altering the behavior of the CJJs 2102 (and thus the behavior of the DAC 2100). The operation of Josephson junctions and / or CJJs in a DAC is described in more detail in, for example, U.S. Patents 7,876,248 and 8,098,179.
[0263] The superconducting loop 2108 is also interposed with a coupling element 2110 and an energy storage element 2112 that receive the superconducting current mediated by the state of the CJJ 2102. The coupling element 2110 couples the DAC 2100 to a target device 2130 (e.g., a qubit of a quantum processor, a coupler, and / or other programmable device). The coupling element 2110 may be an inductor or a gyroscope between the DAC 2100 and the target device 2130. 21 illustrates an example embodiment in which the coupling element 2110 comprises an inductor. An example embodiment in which the coupling element 2110 comprises a galvanic connection is described in more detail below.
[0264] The energy storage element 2112 provides kinetic inductance by having a lower charge carrier density than the superconducting loop 2108. As charge carriers (e.g., Cooper pairs of electrons) pass through the energy storage element 2112, the density of the charge carriers decreases, resulting in a corresponding increase in the velocity of the charge carriers, such that the total current is maintained. The increased velocity of the charge carriers stores kinetic energy, thereby providing a kinetic inductance that may be used instead of or in addition to a magnetic inductance (such as conventional magnetic inductance energy storage described elsewhere herein).
[0265] In ordinary metals, scattering effects dominate, so this effect is generally negligible at sub-THz frequencies, and the energy is quickly released by Joule heating. In superconductors, however, scattering is significantly reduced. Kinetic inductance effects in superconductors can contribute significantly to the total inductance in frequency ranges that include ordinary AC frequencies.
[0266] In some embodiments, the energy storage element 2112 and the coupling element 2110 are combined into a single element. For example, if the energy storage element 2112 also provides sufficient magnetic flux, all or a portion of that flux can be used as all or a portion of the inductive coupling element. Alternatively, or in addition, all or a portion of the energy storage element 2112 may include all or a portion of a galvanic coupling element.
[0267] The energy storage element 2112 may include, for example, superconducting thin film embodiments, Josephson junction based embodiments, and / or other embodiments. Several embodiments of the energy storage element 2112 are described in more detail below. Certain embodiments are prone to exhibit particular behavior (and therefore may offer particular design opportunities) in conjunction with cascaded DACs, and therefore this disclosure is directed initially to cascaded DAC embodiments.
[0268] Exemplary Cascaded DAC In some embodiments, multiple DACs (including, for example, DAC 2100) are coupled in series such that at least one of the DACs is directly coupled to the target device and the remaining DACs are indirectly coupled to the target device via the directly coupled DAC. In some embodiments, a first DAC may be directly coupled to the target device to encode the most significant digit of the signal represented by the DAC, and the other DACs may be coupled in series to the first DAC to encode the least significant digits of the more distant DACs (relative to the first DAC). Aspects of cascaded DACs are described in more detail, for example, in U.S. Patents 7,876,248 and 8,098,179.
[0269] 22 shows a schematic example of a cascaded DAC 2200 having a first DAC 2232a (representing the most significant digit) and a second DAC 2232B (representing the least significant digit). DACs 2232a and 2232b (individually or collectively DACs 2232) are programmable via CJJs 2202a and 2202b (individually or collectively CJJs 2202) and each have an energy storage element 2212a and 2212b (individually or collectively energy storage element 2212). DAC 2232 receives current via inflow line 2220 and sends current to ground via outflow line 2222. DAC 2232 is connected to a target device 2230 via a coupling 2210 (which may be implemented inductively, galvanically or in other ways, as described above).
[0270] The DAC 2232 operates according to the dynamic inductance of the energy storage element 2212. The DACs 2232 share a galvanic connection 2216 through which the superconducting loops of the DACs 2232 are affected. The galvanic connection 2216 couples the superconducting loops of the DACs to create a shared portion 2214 that is part of both superconducting loops of the DACs. The shared portion 2214 also provides a dynamic inductance, and the dynamic inductance of the shared portion 2214 defines, at least in part, the bit weight ratio between the DACs 2232 (i.e., the relative significance of the bits of the DACs 2232 to the output signal with which they are combined). In some embodiments, the dynamic inductance of the shared portion 2214 is proportional to the dynamic inductance of the energy storage element 2212 of the more significant DAC 2232 (e.g., energy storage element 2212a of DAC 2232a).
[0271] In some embodiments, the dynamic inductance of the shared portion 2214 is proportional to the sum of the dynamic inductance of the shared portion 2214 and the dynamic inductance of the DAC 2232 that represents the more significant digit (relative to the other DAC 2232 that shares the shared portion 2214). That is, the total dynamic inductance may be divided according to the ratio of the shared portion 2214 to the more significant DAC 2232 based on the desired bit weight of the less significant DAC 2232. When current flows from the energy storage element 2212b through the galvanic connection 2216, the current is divided between the shared portion 2214 and the energy storage element 2212a in proportion to the ratio of the dynamic inductance between them. Effectively, the shared portion 2214 attenuates the signal to the more significant DAC 2232 (e.g., DAC 2232a) by siphoning current from the output of the less significant DAC 2232 (e.g., DAC 2232b). The lower the dynamic inductance of the shared portion 2214 for the more significant DACs 2232, the more current is siphoned off (because current has an inverse relationship to inductance), weakening the signal of the less significant DACs 2232.
[0272] For example, if a 12-bit cascaded DAC 2200 is desired, then two 6-bit DACs 2232 can be provided, with the less significant DAC 2232b having a bit weight 1 / 64th the bit weight of the more significant DAC 2232a. This allows the energy storage element of each DAC to have twice the total inductance (dynamic and / or magnetic) in the form of 6 =64 single flux quanta Φ 0 DAC 2232b, each of which is given by the less significant DAC 2232b. 0 The overall output signal is equivalent to one Φ 0 This may be accomplished, for example, by splitting the dynamic inductance between the shared portion 2214 and the energy storage element 2212a by 64 units, such that the shared portion 2214 has 1 unit of dynamic inductance and the energy storage element 2212a has 63 units of dynamic inductance, thereby shifting 1 / 64 of the current output by the less significant DAC 2232b to the energy storage element 2212a.
[0273] In at least some embodiments of cascaded DACs 2200 with a galvanic connection 2216 and a shared portion 2214 having a dynamic inductance proportional to the dynamic inductance of the energy storage element 2212, the systems and methods of the present disclosure may define bit weights between DACs 2232 with relatively greater accuracy and / or reliability than at least some magnetic inductance-based DACs. As described in more detail below, the energy storage element 2212 and the shared portion 2214 (generally defined as area 2240 in FIG. 22) may additionally or alternatively require relatively less space than at least some magnetic inductance-based DACs.
[0274] Exemplary Thin Film Energy Storage Elements FIG. 23 shows an example energy storage region 2300 having a superconducting film 2302 coupled to a wire 2304. The wire 2304 may be part of a superconducting loop 2108. 2106 and / or coupled to a superconducting loop 2108. Superconducting film 2302 is an example embodiment of energy storage element 2112 and / or 2212.
[0275] The superconducting film 2302 has a dynamic inductance L as a portion of the total inductance L of the superconducting film. K In general, the total inductance L is L = L K +L G is given by L G is the geometric inductance provided by the magnetic field of the superconducting film 2302. In general, L K , L K is the complex surface impedance Z s =R s +iX s This can be difficult to calculate because it is proportional to R s is the resistance of the material in the superconducting state, and X s is the impedance Z s (including geometric and dynamic components). According to the Drude model of charge transport, R s Solving this gives the complex conductivity
[0276]
number
[0277] Therefore, the dynamic inductance of a superconducting film near zero temperature is given by the effective penetration depth λ eff In particular, for a film with a given thickness t, the dynamic inductance of the film is proportional to the ratio of the film's width W to the film's length L, where the length is in the direction of the current flow and the width is orthogonal to the length (note that both width and length are orthogonal to the dimension measuring the thickness). That is, for a superconducting film with a given thickness,
[0278]
number
[0279] In some embodiments, the membrane 2302 comprises a high penetration depth material, such as NbN, NbTiN, or TiN. Alternatively or additionally, a high λ effOther materials having a .DELTA..times ...
[0280] In some aspects, membrane 2302 is disposed entirely within a single fabrication layer of the quantum processor and is substantially flat. The flat placement of membrane 2302 can, for example, reduce the likelihood of phase drops due to magnetic flux and / or reduce the sensitivity of the dynamic inductance of membrane 2302 to thickness variations between fabrication layers (so that membrane 2302 has a constant λ ef f and more predictable dynamic inductance).
[0281] At least some such embodiments are directed to compactly arranging the energy storage elements 212 and / or 312 and / or reducing the strength of any flux provided by the membrane 2302. For example, as shown in the example embodiment of FIG. 23, the membrane 2302 is serpentine in the flat region. In particular, the membrane 2302 is serpentine in the region 2310 and extends in a length dimension 2312 and a width dimension 2314. The length portions 2320 of the membrane 2302 extend across the length dimension and are spaced apart in the width dimension 2314. The width portions 2322 connect the length portions 2320. Such an arrangement may reduce the total distance in the width dimension 2314 required to obtain a particular dynamic inductance (relative to a linearly arranged membrane 2302). Alternatively, or in addition, such an arrangement may allow adjacent longitudinal portions 2320 to carry current in opposite directions, thereby causing the membrane 2302 to destructively interfere with its own magnetic field (potentially reducing or eliminating the need for shielding and / or increasing the predictability of the behavior of the membrane 2302).
[0282] However, when the current through the membrane 2302 changes direction at a corner 2324, the dynamic inductance contributed by the length of the membrane 2302 that includes the corner is generally less than the dynamic inductance contributed by a straight section (such as the central region of the lengthwise section 2314). For example, in some embodiments, as shown in FIG. 23, the contribution of the corner 2324, where the current turns 90 degrees, is estimated to be about 0.55 of the contribution of the normal region where the current does not turn. In some embodiments, the length of the meandering wire is increased to compensate. For example, if the corner 2324 comprises about one-eighth of the area of the membrane 2302, the length of the membrane 2302 (measured as the current flows, not necessarily in either direction of the dimension 2312 or 2314) may be increased by about 7% (e.g., by an amount in the range of 5% to 10%) to compensate for the effect of the corner 2324.
[0283] Fig. 24 illustrates an example embodiment of a cascaded energy storage region 2400 having two superconducting membranes 2412a and 2412b (collectively or individually, membranes 2412) galvanically coupled to a shared portion 2414. Features of region 2240 of Fig. 22 may be implemented, for example, by features of Fig. 24, e.g., the energy storage element may be implemented, at least in part, by membrane 2412 and shared portion 2214 may be implemented, at least in part, by shared portion 2414. Membrane 2412 may be galvanically coupled to wires 2404a, 2404b and / or 2404c (shown in dashed lines in Fig. 24, collectively or individually, wires 2404).
[0284] In some aspects, it may be desirable to provide membrane 2412 with a low dynamic inductance across shared portion 2414. One way to achieve this is to make shared portion 2414 much shorter (i.e., shorter in length in the direction of current flow) than membrane 2412 (but otherwise of the same material, width and thickness). However, manufacturing constraints may not permit the formation of such a short shared portion 2412 (as typical manufacturing techniques include minimum manufacturing dimensions). Additionally, more area may be necessary or desirable to reliably couple the shared portion to other manufactured features, such as wire 2404 and / or superconductor 2420.
[0285] An alternative (or additional) approach is to widen the shared portion 2414 (i.e., increase the area of the shared portion 2414 in a direction perpendicular to the direction in which it currently extends). As discussed above, dynamic inductance is proportional to the ratio of length to width, so (for example) the width of the shared portion 2414 can be doubled to halve its dynamic inductance. However, this may require adding a large additional area if the length of the shared portion 2414 is large.
[0286] In some embodiments, portions of the membrane 2412 and / or the shared portion 314 are bonded to one or more low dynamic inductance superconducting wires 2404 and / or other superconductors 2420. For example, in FIG. 24, a region of the shared portion 2416 is bonded to wire 2404c, and an additional region is bonded to the superconductor 2420. The superconductor 2420 may include a plate, a via, or any other suitable structure. It may be positioned arbitrarily, but is not necessarily bonded to other structures.
[0287] In the area of the bonded shared portion 2414, current does not flow primarily through the high inductance material of the shared portion 2416 since it has an alternative low inductance path. Therefore, such bonded areas do not contribute significantly to the dynamic inductance of the shared portion 2416. The remaining unbonded area 2416 has a dynamic inductance determined by its length, width, thickness and material, as discussed above. This provides a convenient way of providing shared portions 2216 of various sizes, while still achieving the desired dynamic inductance by shaping the unbonded area 2416.
[0288] Returning to the example cascaded DAC 2200 from above, where there is a desired bit weight of 1 / 64 between DACs 2232, let us assume that area 2400 is fabricated with the constraint that all features be formed from square material with a minimum side length of 0.25 μm. If each membrane 2412 snakes through an area of 1.5 μm by 2.5 μm (i.e., 6 squares by 10 squares), with five 6-square lengthwise sections (all sections one square wide) connected by four 1-square widthwise sections, plus two more squares (each end of each membrane 2412 is adjacent to wire 2412 and superconductor 2420), then each membrane 2412 contains 36 squares. With the help of the eight corner squares, each membrane 2412 has an effective dynamic inductance equivalent to the dynamic inductance created by approximately 32 linearly arranged squares.
[0289] To obtain 1 / 64th of the bit weight between the DACs 2232, it may be desirable for the shared portion 2414 (and / or the non-bonded region 2414) to have 1 / 64th of the dynamic inductance of each of the membranes 2412. A shared portion 2414 with a width of one square would require a length of half a square (i.e., 0.125 μm, which violates the example manufacturing constraints mentioned above). The desired bit weight could be obtained by doubling the length of the membrane 2412, so that a shared portion of one square would suffice, but this would require a significant expansion of the area of the extending membrane 2412. Alternatively, the width of the shared portion 2412 (and / or the non-bonded region 2416) could be doubled, thereby achieving substantially the same bit weight without expanding the membrane 2412, resulting in a shared portion 2412 (and / or the non-bonded region 2416) that is two squares wide and one square long.
[0290] For comparison, at least some magnetic inductance energy storage devices occupy areas having side lengths on the order of about 10 μm using conventional manufacturing techniques (and some embodiments are significantly larger). Thus, compared to at least some magnetic inductance energy storage devices, an equivalent dynamic inductance energy storage element may provide a savings in planar area of over 95%. This is even more significant when considering that the disclosed dynamic inductance energy storage elements may be located entirely within one manufacturing layer and emit a relatively weak magnetic field, whereas at least some magnetic inductance energy storage devices may use multiple manufacturing layers (and in some embodiments, all that is available) and emit a relatively strong magnetic field, which may require additional shielding and / or spacing of other devices away from the magnetic inductance energy storage device.
[0291] The dynamic inductance response of superconducting thin-film energy storage elements to changes in current is substantially It is believed, and in some cases may be preferred, that the response is generally linear, although this feature (as well as other potentially desirable features of the dynamic inductance energy storage elements described elsewhere herein) is not required, e.g., in at least some embodiments, at least some embodiments of the Josephson junction energy storage elements described herein do not generally provide a linear response.
[0292] Exemplary Josephson Junction Energy Storage Elements FIG. 25A illustrates an example Josephson junction energy storage element 2500a that includes a pulse generator 2502 (which may include, for example, a flow-in wire, CJJ, etc.), a storage loop 2512 interposed with one or more Josephson junctions 2504 (14 are shown, only one is called out in FIG. 25A), and a coupling 2532a to a target device 2530a. The example Josephson junction energy storage element 2500a is inductively coupled to a target device 2530. FIG. 25B illustrates a Josephson junction energy storage element 2500b that is galvanically coupled to a target device 2530b by a coupling 2532b. In other respects, the Josephson junction energy storage elements 2500a and 2500b are substantially equivalent, and for convenience are referred to herein collectively and individually as a Josephson junction energy storage element 2500 having a coupling 2532a to a target device 2530a.
[0293] The storage loop 2512 comprises N series-connected Josephson junctions 2504. Each Josephson junction 2504 has a critical current I c This induces a Josephson inductance (a form of dynamic inductance) that depends on the junction resistance R N and the product R of the joint area A N Each Josephson junction 2504 has the same critical current I c , the phase drop across each Josephson junction 2504 is 2π / N, and p Φ 0The current circulating in the storage loop 2512 corresponds to I = I c sin(2pπ / N).
[0294] Single flux quantum Φ 0 It can be observed that for small N (roughly N<4), one Φ 0 The storage of PΦ is generally unstable due to the negative curvature of the free energy of the Josephson junction 2504. The capacitance of the storage loop 2512 is approximately linear in N, so that PΦ 0 The minimum magnitude of N for a given storage loop 2512 designed to store an energy equivalent to c Since N may vary due to manufacturing variations and / or other factors, it may be desirable to provide Josephson junctions 2504 greater than 4P (to avoid the possibility that the circuitry used to inject magnetic flux into the N=4P ring will need to operate at a lower phase threshold due to such variations). For example, in some embodiments, the storage loop 2512 is provided with N=8P Josephson junctions 2504.
[0295] Although the energy storage of loop 2512 increases linearly with N, the individual Josephson junctions do not respond linearly with p (i.e., with respect to the amount of stored energy). Whether this affects the linearity of the signal provided to target device 2530 (referred to as the "DAC response") depends on coupling 2532. If coupling 2532 comprises a linear inductance (e.g., as shown in FIG. 25A), then the DAC response is c It changes nonlinearly as sin(2pπ / N). Therefore, I c R N Since this is proportional to A, this situation is R N In some embodiments, the R N A is the target R by the threshold amount or less. N Deviate from A.
[0296] However, the coupling 2532 is a shared Josephson junction 2504 (e.g., as shown in FIG. 25B). ) the DAC response varies approximately linearly (i.e., linearly with respect to the stored flux, or the kernel of sin(2pπ / N)). Therefore, such an embodiment is generally preferred since the entire storage loop 2512 is connected to the Josephson junction at the minimum critical current I c R, except that it is constrained by a current above which unpredictable behavior may occur. N It is not sensitive to changes in A.
[0297] Figures 26A and 26B show example cascaded Josephson junction DACs 2600a and 2600b. Figure 26A shows an example 2-DAC embodiment that may be used, for example, to implement cascaded DAC 2200 of Figure 22. Figure 26B shows an example 3-DAC embodiment that illustrates the possibility of extending to more than two cascaded DACs. 26A and 26B show DACs 2620a, 2620b, and 2620c (FIG. 26B only) (collectively and individually referred to herein as DAC 2620) having storage loops 2612a, 2612b, and 2612c (FIG. 26B only) (collectively and individually referred to herein as storage loops 2612) driven by respective pulse generators 2602a, 2602b, and 2602c (FIG. 26B only) and coupled by a shared portion 2614a and 2614b (FIG. 26B only) (collectively and individually referred to herein as shared portion 2614). Loop 2612 includes a non-shared Josephson junction 2604a, and the shared portion includes a shared Josephson junction 2604b (collectively and individually referred to herein as Josephson junction 2604).
[0298] DAC 2620a is directly coupled to target device 2630 via coupling 2632, thereby representing the most significant digit. DACs 2620b and 2620c are connected in series with DAC 2620a such that DAC 2620c represents the least significant digit and DAC 2620b represents a less significant digit than DAC 2620a and a more significant digit than DAC 2620c. Shared portion 2614 comprises a Josephson junction 2604 that is common to adjacent 2620. For example, storage loops 2612a and 2612b each comprise a shared portion 2614a and storage loops 2612b and 2612c each comprise a shared portion 2614b.
[0299] The relative bit weights of the DACs 2612 are determined in part by the shared portion 2614. Each of the two adjacent loops 2612 includes N Josephson junctions 2604, and M of the N Josephson junctions 2604 are common (i.e., the shared portion 2614 has M Josephson junctions 2604). As described above, pΦ 0 For an energy equivalent to a flux of qΦ, there is a phase drop of 2pπ / N across each Josephson junction 2604. Thus, there is a total phase drop of 2Mpπ / N across all of the Josephson junctions 2604 in the shared portion 2614. This induces a corresponding phase drop of 2Mpπ / N(NM) across the other loop 2612b. 0 is stored across the other loop 2612, the phase drop across the non-shared Josephson junction 2604a in loop 2612a is (2π / N)(pq·M / (NM)), resulting in an effective bit weight ratio of M / (NM). For example, for N=80 and M=8, the bit weight ratio is 1 / 9.
[0300] Because energy stored in one loop 2612 effectively "spills over" to adjacent loops 2612 (potentially exceeding if additional loops 2612 are cascaded together), each Josephson junction 2604 can experience a phase drop greater than the approximately 2πp / N expected for the single-loop case. Instead, for the two-loop embodiment, the maximum phase drop across the non-shared Josephson junction 2604a (found when p=-q for maximum p) is approximately 2πp / (NM), and the maximum phase drop across the shared Josephson junction 2604b is 4πp / N, roughly twice the maximum phase drop for the single-loop case.
[0301] Thus, in some embodiments, the area of the shared Josephson junction 2604b is greater than the area of the non-shared Josephson junction 2604a. For example, the shared Josephson junction 2604b may have an area twice that of the non-shared Josephson junction 2604a. As another example, the shared Josephson junction 2604b may have an area 2N / (NM) times that of the non-shared Josephson junction 2604a (although for N>>M, this amount will be roughly the same).
[0302] Because the phase drop across each Josephson junction 2604 is proportional to the area of the junction, increasing the area of the shared Josephson junction 2604b results in a smaller total phase drop across the shares 2614 for a given M. This effect can be mitigated by increasing the magnitude of M in proportion to the increased area of the shares 2614. For example, if the area of the shared Josephson junction 2604b is twice the area of the non-shared Josephson junction 2604a, then the magnitude of M can be doubled to accommodate the same maximum energy storage.
[0303] Alternatively, Josephson junctions 2604a and 2604b can be sized identically and the current supplied to loops 2612a and 2612b can be limited to avoid overloading shared Josephson junction 2604b.
[0304] (For example), as shown in FIG. 26B, three or more Josephson junction DACs 2620 may be connected in cascade. Generally, when N Josephson junctions 2604 are arranged in series in a single loop 2612, the loop has a dynamic range of R. However, when the loop 2612 with a dynamic range of D including N Josephson junctions 2604 is divided among the Josephson junctions, it is about (R / D) D is. For example, for each Φ of the four loops 2612 with a capacitance 0 assuming eight Josephson junctions 2604, about 180 Josephson junctions 2604 can achieve a dynamic range of 1000Φ 0 whereas, to achieve the same dynamic range, about 240 Josephson junctions 2604 are required for three loops 2604, and about 500 Josephson junctions 2604 are required for two loops 2604.
[0305] As is clear from the above description, the design of a cascade-connected Josephson junction-based DAC having a desired bit weight may be substantially simpler than the design of an equivalent magnetic inductance DAC, which is generally performed by simulating the magnetic field around a complex three-dimensional magnetic structure. Furthermore, the bit weight ratio is generally relatively insensitive to changes in layer thickness and R N A, which can give relatively predictable behavior compared to a specific magnetic inductance DAC. In addition, although some aspects of the Josephson junction DAC may require a relatively larger area than some aspects of the thin-film energy storage DAC, both use a substantially reduced planar occupancy area, use fewer layers, and have a substantially lower degree of magnetic interference than at least some magnetic inductance DACs despite storing an equal amount of energy.
[0306] Enumeration of Exemplary Embodiments The following enumerated exemplary embodiments provide examples of features and combinations of features and are non-limiting exemplary embodiments of the present invention.
[0307] Exemplary embodiment 1 At least a first multi-junction superconducting quantum interface device (SQUID) flux pump, the first multi-junction SQUID flux pump having N loops, N being an integer greater than or equal to 2, each loop including a material that exhibits superconductivity at least at a critical temperature, the loops N loops, each of which has a portion shared by consecutive adjacent ones of the loops; and M Josephson junctions, where M is greater than N, each of which has at least one of the Josephson junctions interposed in the portion of the loop shared by consecutive adjacent ones of the loops, each of which has a respective storage inductance. a first N interfaces positioned to selectively communicatively couple magnetic flux of the loops to respective ones of the loops.
[0308] Exemplary embodiment 2 The system of exemplary embodiment 1 further comprises a digital-to-analog converter (DAC) coupled to the most extreme of the loops of the first multi-junction SQUID flux pump, the DAC comprising a material loop that produces superconductivity at least at a critical temperature and a storage inductance.
[0309] Exemplary embodiment 3 The system of exemplary embodiment 2, wherein the storage inductance of the DAC is at least one of a magnetic inductance, a dynamic inductance, a Josephson inductance, or a combination of two or more of a magnetic inductance, a dynamic inductance, and a Josephson inductance.
[0310] Exemplary embodiment 4 The system of Exemplary Embodiment 1, wherein the loop of the DAC includes a portion shared by an endmost loop of a first multi-junction SQUID flux pump.
[0311] Exemplary embodiment 5 The system of exemplary embodiment 1, wherein each of the first N interfaces is a respective inductive interface positioned in proximity to a storage inductance of a respective one of the loops of the first multi-junction SQUID flux pump.
[0312] Exemplary embodiment 6 13. The system of claim 1, wherein the loops of the first multi-junction SQUID flux pump form a linear array of loops.
[0313] Exemplary Embodiment 7 The system of exemplary embodiment 1, wherein the number N is greater than or equal to 2 and less than or equal to 4.
[0314] Exemplary Embodiment 8 The system of exemplary embodiment 1, wherein the number M is equal to N+1.
[0315] Exemplary Embodiment 9 The system of any one of exemplary embodiments 1 to 8, further comprising a control circuit operable to perform multi-phase radio frequency (RF) clocking of the flux signal without a direct current (DC) bias for the loop of the first multi-junction SQUID flux pump.
[0316] Exemplary Embodiment 10 10. The system of claim 9, wherein multi-phase radio frequency (RF) clocking of the flux signal sequentially pushes the flux through a loop of a first multi-junction SQUID flux pump to a DAC.
[0317] Exemplary embodiment 11 Multiphase radio frequency (RF) clocking of the flux signal is performed by the first multijunction SQUID flux pump. 10. The system of exemplary embodiment 9, which sequentially pushes magnetic flux out of the DAC through a loop.
[0318] Exemplary embodiment 12 10. The system of example embodiment 9, wherein a total number of phases of the multi-phase radio frequency (RF) clocking of the flux signal is equal to a total number N of loops of the first multi-junction SQUID flux pump into the DAC.
[0319] Exemplary Embodiment 13 The system of any one of exemplary embodiments 1 to 8, further comprising a control circuit communicatively coupled to sequentially supply flux signals to the loops of the first multi-junction SQUID flux pump via the first N interfaces, wherein the individual flux signals supplied to each interface are advanced by π / 2 relative to the individual flux signals supplied to the immediately preceding loop along the linear series of loops of the first multi-junction SQUID flux pump.
[0320] Exemplary Embodiment 14 The system of any one of exemplary embodiments 1 to 8, further comprising a control circuit communicatively coupled to sequentially supply flux signals to the loops of the first multi-junction SQUID flux pump via the first N interfaces, wherein the individual flux signals supplied to each interface are delayed by π / 2 with respect to the individual flux signals supplied to the immediately succeeding loops along the linear series of loops of the first multi-junction SQUID flux pump.
[0321] Exemplary Embodiment 15 The system of exemplary embodiment 1 further comprising a first set of control lines, the first set of control lines being comprised of a first control line coupled to an interface of a first one of the loops of the first multi-junction SQUID flux pump, a second control line coupled to an interface of a second one of the loops of the first multi-junction SQUID flux pump, and a third control line coupled to an interface of a third one of the loops of the first multi-junction SQUID flux pump.
[0322] Exemplary Embodiment 16 a plurality of further multi-junction SQUID flux pumps each including a respective N loops, each of the loops having a portion shared by successive adjacent ones of the loops, and a respective M Josephson junctions, at least one of the Josephson junctions being interposed in each of the portions of the loop shared by successive adjacent ones of the loops, each of the loops having a respective storage inductance; The system of exemplary embodiment 1 further comprising: a further N interfaces for each of the further multi-junction SQUID flux pumps positioned to selectively communicatively couple magnetic flux carried by the loops to respective ones of the loops of respective ones of the further multi-junction SQUID flux pumps.
[0323] Exemplary Embodiment 17 Further comprising a first control line set, the first control line set including a first control line subset, the first control line subset including a first control line coupled to an interface of a respective first one of the loops of each multi-junction SQUID flux pump of the first multi-junction SQUID flux pump subset, a second control line coupled to an interface of a respective second one of the loops of each multi-junction SQUID flux pump of the first multi-junction SQUID flux pump subset, and a third control line coupled to an interface of a respective third one of the loops of each multi-junction SQUID flux pump of the first multi-junction SQUID flux pump subset. 17. The system of exemplary embodiment 16, comprising:
[0324] Exemplary Embodiment 18 The system of exemplary embodiment 17, wherein the first control line set includes a second control line subset, the second control line subset including a first control line coupled to an interface of a respective first loop of the loops of each multi-junction SQUID flux pump of the second multi-junction SQUID flux pump subset, a second control line coupled to an interface of a respective second loop of the loops of each multi-junction SQUID flux pump of the second multi-junction SQUID flux pump subset, and a third control line coupled to an interface of a respective third loop of each multi-junction SQUID flux pump of the second multi-junction SQUID flux pump subset, the second multi-junction SQUID flux pump subset including at least three multi-junction SQUID flux pumps.
[0325] Exemplary Embodiment 19 The system of exemplary embodiment 16 further comprising a first set of J control lines, where J=2×N, and there are a total of I multi-junction SQUID flux pumps, where I is equal to 2(N+1).
[0326] Exemplary Embodiment 20 1. A method of operating a system comprising a plurality of multi-junction superconducting quantum interface device (SQUID) flux pumps, each of the multi-junction SQUID flux pumps comprising: N loops, N being an integer greater than or equal to two, each of the loops having a portion shared by consecutive adjacent ones of the loops; and M Josephson junctions, M greater than N, each of the portions of the loop shared by consecutive adjacent ones of the loops having at least one of the Josephson junctions interposed therein, each of the loops having a respective storage inductance; and wherein for each of the multi-junction SQUID flux pumps, N interfaces are positioned to selectively communicatively couple magnetic flux of the respective multi-junction SQUID flux pump to a respective one of the loops of the multi-junction SQUID flux pump; and the method comprises: Sequentially, providing a flux signal to a first one of the loops of a first one of the multi-junction SQUID flux pumps via a respective first one of the interfaces; providing a flux signal to a second one of the loops of a first one of the multi-junction SQUID flux pumps via a respective second one of the interfaces, wherein the flux signal provided to the second one of the loops is out of phase with respect to the flux signal provided to the first one of the loops.
[0327] Exemplary Embodiment 21 21. The method of exemplary embodiment 20, further comprising: supplying a flux signal to at least a third one of the loops of a first one of the multi-junction SQUID flux pumps via a separate third one of the interfaces, wherein the flux signal supplied to the third one of the loops is out of phase with respect to the flux signal supplied to the second one of the loops.
[0328] Exemplary Embodiment 22 via a respective fourth one of the interfaces, at least a fourth one of the loops of the first one of the multi-junction SQUID flux pumps; 22. The method of exemplary embodiment 21, further comprising: providing flux signals to a fourth one of the loops, the flux signal provided to a fourth one of the loops being out of phase with respect to the flux signal provided to the third one of the loops.
[0329] Exemplary Embodiment 23 The method of exemplary embodiment 21, wherein supplying a flux signal to a first loop, a second loop, and a third loop of the loops of a first multi-junction SQUID flux pump of the multi-junction SQUID flux pump includes supplying to each of the individual first interface, second interface, and at least a third interface an individual flux signal that is out of phase by π / 2 with respect to an individual flux signal supplied to an immediately succeeding one of the loops along the linear series of loops of the first multi-junction SQUID flux pump of the multi-junction SQUID flux pump.
[0330] Exemplary Embodiment 24 Providing a flux signal to a first loop, a second loop, and a third loop of the loops of a first multi-junction SQUID flux pump of the multi-junction SQUID flux pump includes: 24. The method of exemplary embodiment 23, comprising providing flux signals to a first loop, a second loop and at least a third loop of the loops of a first multi-junction SQUID flux pump of the multi-junction SQUID flux pump, wherein the individual flux signals provided to each of the individual first interface, second interface and third interface lead by π / 2 with respect to the individual flux signals provided to the immediately succeeding loops along the linear series of loops of the first multi-junction SQUID flux pump.
[0331] Exemplary Embodiment 25 Providing a flux signal to a first loop, a second loop, and a third loop of the loops of a first multi-junction SQUID flux pump of the multi-junction SQUID flux pump includes: 24. The method of example embodiment 23, comprising providing flux signals to a first loop, a second loop, and a third loop of a first multi-junction SQUID flux pump of the multi-junction SQUID flux pump, wherein the individual flux signals provided to each of the individual first interface, second interface, and at least third interface are delayed by π / 2 with respect to the individual flux signal provided to an immediately succeeding one of the loops along the linear series of the first multi-junction SQUID flux pump loops.
[0332] Exemplary Embodiment 26 22. The method of exemplary embodiment 21, wherein providing a flux signal to a first loop, a second loop, and a third loop of the loops of a first multi-junction SQUID flux pump of the multi-junction SQUID flux pump includes providing a flux signal that sequentially pushes flux through the loops of the first multi-junction SQUID flux pump to a digital-to-analog converter (DAC).
[0333] Exemplary Embodiment 27 22. The method of exemplary embodiment 21, wherein providing a flux signal to a first loop, a second loop, and a third loop of the loops of a first multi-junction SQUID flux pump of the multi-junction SQUID flux pump includes providing a flux signal that sequentially pushes flux through the loops of the first multi-junction SQUID flux pump out of a digital-to-analog converter (DAC).
[0334] Exemplary Embodiment 28 providing a flux signal to the first loop, the second loop, and the third loop of the loops of the first multi-junction SQUID flux pump of the multi-junction SQUID flux pump includes providing a flux signal having a first sign and a first magnitude via a first control line, a second control line, and a third control line; 22. The method of exemplary embodiment 21, further comprising: supplying a flux signal having a second sign and a first magnitude to at least one of the loops of a second multi-junction SQUID flux pump of the multi-junction SQUID flux pumps via a fourth control line following supplying the flux signal to the first loop, the second loop, and the third loop of the loop of a first multi-junction SQUID flux pump of the multi-junction SQUID flux pumps, wherein the second sign is opposite to the first sign.
[0335] Exemplary Embodiment 29 1. A method of operating a system comprising a plurality of multi-junction superconducting quantum interface device (SQUID) flux pumps, each of the multi-junction SQUID flux pumps comprising: N loops, N being an integer greater than or equal to two, each of the loops having a portion shared by consecutive adjacent ones of the loops; and M Josephson junctions, M greater than N, each of the portions of the loop shared by consecutive adjacent ones of the loops having at least one of the Josephson junctions interposed therein, each of the loops having a respective storage inductance; and wherein for each of the multi-junction SQUID flux pumps, N interfaces are positioned to selectively communicatively couple magnetic flux of the respective multi-junction SQUID flux pump to a respective one of the loops of the multi-junction SQUID flux pump; and the method comprises: at the same time, providing a flux signal having a first sign and a first magnitude to a first loop of the loops of a first one of the multi-junction SQUID flux pumps via a respective first one of the interfaces; and supplying a flux signal having a first sign and a first magnitude to all of the other loops of a first multi-junction SQUID flux pump of the multi-junction SQUID flux pumps via a respective second interface of the interfaces, wherein the flux signals supplied to all of the loops are in phase with each other.
[0336] Exemplary embodiment 30 30. The method of exemplary embodiment 29, wherein applying a flux signal having a first sign and a first magnitude to all of the other loops of a first multi-junction SQUID flux pump of the multi-junction SQUID flux pumps includes first applying a flux signal having a first sign and a first magnitude to the first loop, the second loop, and at least a third loop of the first multi-junction SQUID flux pump of the multi-junction SQUID flux pumps via respective ones of a first control line, a second control line, and a third control line.
[0337] Exemplary embodiment 31 1. A method of operating a computing system comprising: a plurality of digital-to-analog converters (DACs); a plurality of quantum flux parametron (QFP) based shift registers, each having a respective set of QFP based shift register elements, each of which is magnetically or galvanically couplable to a respective one of the DACs; a plurality of power supply lines; and a plurality of trigger lines, each of the DACs being independently addressable successive times to store a variable number of flux quanta by a triplet of three signals, the triplet of three signals including a first signal obtained via the shift register elements, a second signal obtained via the power supply lines, and a third signal obtained via the trigger lines, the method comprising: Resetting all of the DACs and providing a signal to one or more of the power lines; providing a clockwise sustaining current to a first plurality of QFP-based shift register elements coupled to a DAC to which the magnetic flux quanta are loaded; A second multiple QFP-based shift register coupled to a DAC that is not loaded with flux quanta providing a counterclockwise sustaining current to the starter element; applying a signal to the trigger line a first plurality of times, the first plurality of times being equal to a total number of flux quanta to be loaded.
[0338] Exemplary embodiment 32 Sorting all of the DACs based on the number of flux quanta loaded into the DAC; For an integer number of programmed iterations, i, ranging from 2 to the maximum number of flux quanta, n, providing a clockwise sustaining current to a plurality of QFP-based shift register elements coupled to a DAC loaded with at least i magnetic flux quanta; providing a counterclockwise sustaining current to a plurality of QFP-based shift register elements coupled to a DAC loaded with less than i magnetic flux quanta; 32. The method of exemplary embodiment 31, further comprising: providing a signal to the trigger line.
[0339] Exemplary Embodiment 33 33. The method of exemplary embodiment 32, wherein the maximum number of flux quanta, n, is between 18 and 22.
[0340] Exemplary Embodiment 34 1. A method of operating a system comprising: a plurality of quantum flux parametrons (QFPs); at least one quantum flux parametron digital-to-analog converter (QFP-DAC) pair; a current bias line that provides a current bias to both of the QFP-DACs of the at least one QFP-DAC pair without a bias resistor; and, for each QFP of the plurality of QFPs, a respective first plurality of quantum flux parametron latches (QFP latches) selectively operable to communicatively couple flux between the QFP and a first QFP-DAC of the at least one QFP-DAC pair and a respective second plurality of QFP latches selectively operable to communicatively couple flux between the QFP and a second QFP-DAC of the at least one QFP-DAC pair, the method comprising: During the first period, simultaneously: providing a signal to a first QFP latch of the first plurality of QFP latches; providing a signal to a first QFP latch of the second plurality of QFP latches; A method comprising transferring a flux quantum state between at least one of the QFP-DACs of the at least one QFP-DAC pair and an individual one of the flux quantum parametrons.
[0341] Exemplary Embodiment 35 providing a signal to a first one of the QFP-DACs; 35. The method of exemplary embodiment 34, further comprising: providing a signal to a second one of the QFP-DACs.
[0342] Exemplary Embodiment 36 36. The method of exemplary embodiment 35, wherein providing a signal to a first QFP latch of the first plurality of QFP latches and providing a signal to a first QFP-DAC of the QFP-DACs includes providing a signal sequence to load information into a first DAC of the DACs, and providing a signal to a first QFP latch of the second plurality of QFP latches and providing a signal to a second QFP-DAC of the QFP-DACs includes providing a signal sequence to load information into a second DAC of the DACs.
[0343] Exemplary Embodiment 37 providing a signal to a second one of the first pair of QFP latches; providing a signal to a second QFP latch pair of the second QFP latch pair; and 35. The method of exemplary embodiment 34, comprising:
[0344] Exemplary Embodiment 38 providing a signal to a first QFP-DAC of the QFP-DAC subsequent to providing a signal to a first QFP latch and a second QFP latch of the first plurality of QFP latches; 38. The method of exemplary embodiment 37, further comprising: subsequent to providing the signal to the first QFP latch and the second QFP latch of the second plurality of QFP latches, providing the signal to a second QFP-DAC of the QFP-DAC.
[0345] Exemplary Embodiment 39 The method of exemplary embodiment 38, wherein providing a signal to a first QFP-DAC of the QFP-DACs includes providing a separate second latch signal to the first QFP-DAC of the QFP-DACs, and providing a signal to a second QFP-DAC of the QFP-DACs includes providing a separate second latch signal to the second QFP-DAC of the QFP-DACs.
[0346] Exemplary Embodiment 40 The method of exemplary embodiment 38, wherein providing a signal to a first QFP-DAC of the QFP-DACs includes providing a separate first chip signal to the first QFP-DAC of the QFP-DACs simultaneously with providing a second latch signal to the first QFP-DAC of the QFP-DACs, and providing a signal to a second QFP-DAC of the QFP-DACs includes providing a separate first chip signal to the second QFP-DAC of the QFP-DACs simultaneously with providing a second latch signal to the second QFP-DAC of the QFP-DACs.
[0347] Exemplary embodiment 41 providing a reset signal to at least one of the QFP-DACs; 41. The method of exemplary embodiment 40, further comprising: providing a separate offset signal to at least one of the QFP-DACs.
[0348] Exemplary embodiment 42 38. The method of exemplary embodiment 37, wherein providing signals to a first QFP latch and a second QFP latch of the first plurality of QFP latches and providing signals to a first QFP latch and a second QFP latch of the second plurality of QFP latches includes providing signals to the QFP latches to load information into a QFP-DAC.
[0349] Exemplary Embodiment 43 43. The method of exemplary embodiment 42, further comprising providing a reset signal to the QFP-DAC before providing a signal to the QFP latch in a given cycle to load information into the QFP-DAC.
[0350] Exemplary Embodiment 44 38. The method of exemplary embodiment 37, wherein providing signals to a first QFP latch and a second QFP latch of the first plurality of QFP latches and providing signals to a first QFP latch and a second QFP latch of the second plurality of QFP latches includes providing signals to the QFP latches to operate as a quantum flux parametron demultiplexer (QFP-Demux).
[0351] Exemplary Embodiment 45 45. The method of exemplary embodiment 44, further comprising: providing a reset signal to the QFP-DAC before providing a signal to the QFP latch in a given cycle to operate as a QFP-Demux.
[0352] Exemplary Embodiment 46 38. The method of exemplary embodiment 37, wherein providing signals to a first QFP latch and a second QFP latch of the first plurality of QFP latches and providing signals to a first QFP latch and a second QFP latch of the second plurality of QFP latches includes providing signals to the QFP latches to load information from a QFP-DAC.
[0353] Exemplary Embodiment 47 38. The method of exemplary embodiment 37, wherein providing signals to a first QFP latch and a second QFP latch of the first plurality of QFP latches and providing signals to a first QFP latch and a second QFP latch of the second plurality of QFP latches includes providing signals to operate as a quantum flux parametron multiplexer (QFP-Mux).
[0354] Exemplary Embodiment 48 35. The method of exemplary embodiment 34, wherein providing a signal to a first QFP latch of the first plurality of QFP latches includes providing a respective first latch signal to a first QFP latch of the first plurality of QFP latches, and providing a signal to a first QFP latch of the second plurality of QFP latches includes providing a logical inverse of the respective first latch signal to a first QFP latch of the second plurality of QFP latches.
[0355] Exemplary Embodiment 49 49. The method of exemplary embodiment 48, wherein providing a signal to a first QFP latch of the first plurality of QFP latches includes providing a respective first offset signal to a first QFP latch of the first plurality of QFP latches, and providing a signal to a first QFP latch of the second plurality of QFP latches includes providing a respective first offset signal to a first QFP latch of the second plurality of QFP latches.
[0356] Exemplary embodiment 50 35. The method of exemplary embodiment 34, wherein providing a signal to a first QFP latch of the first plurality of QFP latches includes providing a respective first address signal to a first QFP latch of the first plurality of QFP latches, and providing a signal to a first QFP latch of the second plurality of QFP latches includes providing a logical inverse of the respective first address signal to a first QFP latch of the second plurality of QFP latches.
[0357] Exemplary embodiment 51 51. The method of exemplary embodiment 50, further comprising logarithmically aggregating a plurality of address lines into one of the QFPs.
[0358] Exemplary Embodiment 52 The method of exemplary embodiment 50, wherein providing a signal to a first QFP latch of the first plurality of QFP latches includes providing a respective first latch signal to a first QFP latch of the first plurality of QFP latches, and providing a signal to a first QFP latch of the second plurality of QFP latches includes providing a respective first latch signal to a first QFP latch of the second plurality of QFP latches.
[0359] Exemplary Embodiment 53 A plurality of N signal lines; a plurality of 4(N-1)2 digital-to-analog converters (DACs), the DACs communicatively coupled to respective triplets of signal lines and controlled via signals carried by the respective triplets of signal lines, each DAC of the plurality of DACs comprising a respective material loop and a respective Josephson junction pair interposed in the respective loop and electrically connected in parallel with each other within the material loop.
[0360] Exemplary Embodiment 54 The system of exemplary embodiment 53, wherein the plurality of N signal lines includes a set of control lines and a set of power lines, each triplet consisting of a unique combination of two of the control lines and one of the power lines.
[0361] Exemplary Embodiment 55 The system of exemplary embodiment 53, wherein each of the signal lines is made of a material that exhibits superconductivity at least at a critical temperature, and each individual material loop of the DAC is made of a material that exhibits superconductivity at least at a critical temperature.
[0362] Exemplary Embodiment 56 The system of exemplary embodiment 53, wherein the DACs are arranged in a plurality of two-dimensional arrays, each of the two-dimensional arrays including a respective plurality of DACs, and the plurality of signal lines includes a first subset of signal lines, each signal line of the first subset of signal lines being communicatively coupled to a DAC of a respective one of the two-dimensional arrays.
[0363] Exemplary Embodiment 57 57. The system of exemplary embodiment 56, wherein each signal line of the first subset of signal lines is communicatively coupled to all of the DACs of a respective one of the two-dimensional arrays.
[0364] Exemplary Embodiment 58 The system of exemplary embodiment 56 or 57, wherein the plurality of signal lines includes a second subset of signal lines, each signal line of the second subset of signal lines being communicatively coupled to a subset of DACs for each two or more of the two-dimensional array.
[0365] Exemplary Embodiment 59 59. The system of exemplary embodiment 58, further comprising a control circuit communicatively coupled to provide signals to selected ones of the DACs via respective triplets of signal lines.
[0366] Exemplary embodiment 60 The system of exemplary embodiment 56 or 57, wherein the plurality of signal lines includes a second subset of signal lines, each signal line of the second subset of signal lines being communicatively coupled to a subset of all DACs in the two-dimensional array.
[0367] Exemplary embodiment 61 a first two-dimensional array of digital-to-analog converters (DACs), the first two-dimensional array including a plurality of DACs; a second two-dimensional array including at least a second two-dimensional array of DACs that are not DACs in the first two-dimensional array; a first power supply line coupled to selectively supply current to the DACs of the first two-dimensional array of DACs; a second power supply line coupled to selectively supply current to the DACs of the second two-dimensional array of DACs; and a control line set, the control line set including at least a first control line disposed proximate to each DAC along a first diagonal line of the first DAC two-dimensional array and communicatively coupled to individual DACs along the first diagonal line of the first DAC two-dimensional array, the first control line also disposed proximate to each DAC along the first diagonal line of the second DAC two-dimensional array and communicatively coupled to individual DACs along the first diagonal line of the second DAC two-dimensional array.
[0368] Exemplary embodiment 62 The system described in exemplary embodiment 61, wherein the DACs of the first DAC two-dimensional array are arranged in multiple rows and multiple columns, and a first diagonal line of the first DAC two-dimensional array extends across all rows and all columns of the first DAC two-dimensional array.
[0369] Exemplary embodiment 63 The system described in exemplary embodiment 62, wherein the DACs of the second DAC two-dimensional array are arranged in multiple rows and multiple columns, and a first diagonal line of the second DAC two-dimensional array extends across all rows and all columns of the second DAC two-dimensional array.
[0370] Exemplary Embodiment 64 a third two-dimensional array of DACs, the third two-dimensional array including a plurality of DACs that are not DACs in the first two-dimensional array or DACs in the second two-dimensional array; and a fourth two-dimensional array of DACs, the fourth two-dimensional array including a plurality of DACs that are not DACs in the first, second, or third two-dimensional array; and a third power supply line coupled to selectively supply current to the DACs of the third two-dimensional array of DACs; a fourth power supply line coupled to selectively supply current to the DACs of the fourth two-dimensional array of DACs; The system described in exemplary embodiment 61, wherein the first control line is positioned proximate to each DAC along a first diagonal line of the third DAC dimensional array and is communicatively coupled to individual DACs along the first diagonal line of the third DAC two-dimensional array, and the first control line is positioned further proximate to each DAC along the first diagonal line of the fourth DAC dimensional array and is communicatively coupled to individual DACs along the first diagonal line of the fourth DAC two-dimensional array.
[0371] Exemplary Embodiment 65 The system described in exemplary embodiment 64, wherein the DACs of the first DAC two-dimensional array are arranged in multiple rows and multiple columns, and a first diagonal line of the first DAC two-dimensional array extends across all rows and all columns of the first DAC two-dimensional array.
[0372] Exemplary Embodiment 66 The system described in exemplary embodiment 65, wherein the DACs of the second DAC two-dimensional array are arranged in multiple rows and multiple columns, and a first diagonal line of the second DAC two-dimensional array extends across all rows and all columns of the second DAC two-dimensional array.
[0373] Exemplary Embodiment 67 The system described in exemplary embodiment 66, wherein the DACs of the third DAC two-dimensional array are arranged in multiple rows and multiple columns, and a first diagonal line of the third DAC two-dimensional array extends across all rows and all columns of the third DAC two-dimensional array.
[0374] Exemplary Embodiment 68 the DACs of the fourth two-dimensional array of DACs are arranged into a plurality of rows and a plurality of columns; The system of exemplary embodiment 67, wherein the first diagonal line of the fourth DAC two-dimensional array extends across all rows and all columns of the fourth DAC two-dimensional array.
[0375] Exemplary Embodiment 69 The system described in exemplary embodiment 65, wherein the first diagonal line of the second two-dimensional array is perpendicular to the first diagonal line of the first two-dimensional array, and the first diagonal line of the third two-dimensional array is perpendicular to the first diagonal line of the second two-dimensional array.
[0376] Exemplary embodiment 70 The system described in exemplary embodiment 65, wherein the first diagonal line of the fourth two-dimensional array is perpendicular to the first diagonal line of the first two-dimensional array, and the first diagonal line of the fourth two-dimensional array is perpendicular to the first diagonal line of the third two-dimensional array.
[0377] Exemplary embodiment 71 The system of exemplary embodiment 64, wherein the first control line follows a serpentine path along a first diagonal line of the first DAC two-dimensional array, follows a serpentine path along the first diagonal line of the second DAC two-dimensional array, follows a serpentine path along the first diagonal line of the third DAC two-dimensional array, and follows a serpentine path along the first diagonal line of the fourth DAC two-dimensional array.
[0378] Exemplary Embodiment 72 The control line set is The system of exemplary embodiment 64, further comprising a second control line, the second control line being positioned proximate to each DAC along a second diagonal line of the first DAC two-dimensional array and communicatively coupled to individual DACs along the second diagonal line of the first DAC two-dimensional array, the second control line also being positioned proximate to each DAC along the second diagonal line of the second DAC two-dimensional array and communicatively coupled to individual DACs along the second diagonal line of the second DAC two-dimensional array, the second control line being positioned further proximate to each DAC along the second diagonal line of the third DAC two-dimensional array and communicatively coupled to individual DACs along the second diagonal line of the third DAC two-dimensional array, and the second control line being positioned even further proximate to each DAC along the second diagonal line of the fourth DAC dimensional array and communicatively coupled to individual DACs along the second diagonal line of the fourth DAC two-dimensional array.
[0379] Exemplary Embodiment 73 The system of exemplary embodiment 72, wherein the DACs of the first DAC two-dimensional array are arranged in multiple rows and multiple columns, and the second diagonal line of the first DAC two-dimensional array extends across all rows and all columns of the first DAC two-dimensional array.
[0380] Exemplary Embodiment 74 The control line set is The system of exemplary embodiment 72 further comprises a third control line, the third control line being positioned proximate to each DAC along a third diagonal line and a fourth diagonal line of the first DAC two-dimensional array and communicatively coupled to individual DACs along the third diagonal line and the fourth diagonal line of the first DAC two-dimensional array, the fourth diagonal line being perpendicular to the third diagonal line.
[0381] Exemplary Embodiment 75 The DACs of the first DAC two-dimensional array are arranged in a plurality of rows and a plurality of columns, a third diagonal line of the first DAC two-dimensional array extends across a first plurality of rows and a first plurality of columns of the first DAC two-dimensional array, a fourth diagonal line of the first DAC two-dimensional array extends across a second plurality of rows and a second plurality of columns of the first DAC two-dimensional array, and a combination of the first plurality of rows and the second plurality of rows forms a first DAC. The system of exemplary embodiment 74, wherein the first plurality of columns and the second plurality of columns in combination include all the columns of the first DAC two-dimensional array.
[0382] Exemplary Embodiment 76 The third control lines are also disposed proximate to each DAC along a third diagonal line and a fourth diagonal line of the second DAC two-dimensional array and are communicatively coupled to individual DACs along the third diagonal line and the fourth diagonal line of the second DAC two-dimensional array, the fourth diagonal line of the second DAC two-dimensional array being perpendicular to the third diagonal line of the second DAC two-dimensional array; third control lines are disposed closer to each DAC along a third diagonal line and a fourth diagonal line of the third DAC dimensional array and are communicatively coupled to the individual DACs along the third diagonal line and the fourth diagonal line of the third DAC two dimensional array, the fourth diagonal line of the third DAC two dimensional array being perpendicular to the third diagonal line of the third DAC two dimensional array; The system of exemplary embodiment 74, wherein the third control line is positioned even closer to each DAC along a third diagonal line and a fourth diagonal line of the fourth DAC two-dimensional array and is communicatively coupled to individual DACs along a third diagonal line and a fourth diagonal line of the fourth two-dimensional array of DACs, the fourth diagonal line of the fourth DAC two-dimensional array being perpendicular to the third diagonal line of the fourth DAC two-dimensional array.
[0383] Exemplary Embodiment 77 The system of any one of exemplary embodiments 64-74, wherein there are a total of N signal lines, and each of the first array of DACs, the second array of DACs, the third array of DACs, and the fourth array of DACs includes 4(N-1)2 DACs.
[0384] Exemplary Embodiment 78 The system of any one of exemplary embodiments 64-74, further comprising control circuitry communicatively coupled to provide signals to selected ones of the DACs via individual triplets of one power line and two of the signal lines, the triplets being uniquely communicatively coupled to operate a single individual DAC.
[0385] Exemplary Embodiment 79 1. A method of operating a system, the system comprising: a first two-dimensional array of digital-to-analog converters (DACs), the first two-dimensional array including a plurality of DACs; at least a second DAC two-dimensional array, the second two-dimensional array having a plurality of DACs that are not DACs in the first two-dimensional array; a first power supply line coupled to selectively supply current to DACs of the first DAC two-dimensional array; a second power supply line coupled to selectively supply current to DACs of the second DAC two-dimensional array; and a set of control lines, at least a first control line disposed adjacent each DAC along a first diagonal line of the first DAC two-dimensional array and adapted to selectively supply current to each DAC of the second DAC two-dimensional array. a control line set having a first control line communicatively coupled to individual DACs along a first diagonal line and further disposed proximate to each DAC along the first diagonal line of the second DAC two-dimensional array and communicatively coupled to individual DACs along the first diagonal line of the second DAC two-dimensional array; and a second control line disposed proximate to each DAC along a second diagonal line of the first DAC two-dimensional array and communicatively coupled to individual DACs along the second diagonal line of the first DAC two-dimensional array, and further disposed proximate to each DAC along the second diagonal line of the second DAC two-dimensional array and communicatively coupled to individual DACs along the second diagonal line of the second DAC two-dimensional array, the method comprising: During the first period, simultaneously: Providing a signal to a first DAC via a first power line; Providing a signal to a first DAC via a first control line; The method includes providing a signal to a first DAC via a second control line.
[0386] Exemplary embodiment 80 A plurality of N signal lines; a first plurality of analog-to-analog converters (DACs) arranged in a first two-dimensional array, each DAC of the first plurality of DACs including a respective loop of material and a respective pair of Josephson junctions interposed in the respective loop and electrically coupled in parallel to each other within the material loop; a plurality of signal lines arranged in a blade configuration through the first two-dimensional array of the first plurality of DACs and communicatively coupled to the DACs of the first plurality of DACs.
[0387] Exemplary embodiment 81 a second plurality of analog-to-analog converters (DACs) arranged in a second two-dimensional array, each DAC of the second plurality of DACs including a respective loop of material and a respective pair of Josephson junctions interposed in the respective loop and electrically coupled in parallel to each other within the material loop; a third plurality of analog-to-analog converters (DACs) arranged in a third two-dimensional array, each DAC of the third plurality of DACs including a respective loop of material and a respective pair of Josephson junctions interposed in the respective loop and electrically coupled in parallel to each other within the material loop; The system of exemplary embodiment 80 further comprises a fourth plurality of analog-to-analog converters (DACs) arranged in a fourth two-dimensional array, each of the DACs of the fourth plurality of DACs including a respective loop of material and a respective Josephson junction pair interposed in the respective loop and electrically coupled in parallel to each other within the material loop, and a plurality of signals and a plurality of signal lines are arranged in a braid configuration through the second two-dimensional array, the third two-dimensional array, and the fourth two-dimensional array of the second plurality of DACs, the third plurality of DACs, and the fourth plurality of DACs, respectively, and communicatively coupled to the DACs of the second plurality of DACs, the DACs of the third plurality of DACs, and the DACs of the fourth plurality of DACs.
[0388] Exemplary embodiment 82 The system of exemplary embodiment 80 or 81, wherein the DACs of the first plurality of DACs, the DACs of the second plurality of DACs, the DACs of the third plurality of DACs, and the fourth plurality of DACs total 4(N-1)2 DACs, the total DACs being communicatively coupled to respective triplets of signal lines and controlled via signals carried by the respective triplets of signal lines.
[0389] Exemplary Embodiment 83 The system of exemplary embodiment 72, wherein the plurality of signal lines includes a plurality of N signal lines, the plurality of N signal lines including a set of control lines and a set of power lines, each DAC being controlled via a respective one of the control lines, and each triplet consisting of a unique combination of two of the control lines and one of the power lines.
[0390] Exemplary Embodiment 84 a first two-dimensional array of digital-to-analog converters (DACs), the first two-dimensional array having a plurality of DACs arranged in a plurality of rows and a plurality of columns in the first two-dimensional array; a first power supply line coupled to selectively supply current to the DACs of the first two-dimensional array of DACs; and a control line set, the control line set including at least A first control line, the first control line being connected to at least three of the first DAC two-dimensional array a first DAC in each of the rows and at least three columns in a two-dimensional array of DACs, the first DAC being operatively proximate to at least one DAC in each of the rows and at least three columns and communicatively coupled to individual DACs in each of the at least three rows and at least three columns of the two-dimensional array of DACs;
[0391] Exemplary Embodiment 85 a second DAC two-dimensional array having at least a plurality of DACs that are not DACs in the first two-dimensional array, the plurality of DACs including a plurality of DACs arranged in a plurality of rows and a plurality of columns of the second two-dimensional array; The system of exemplary embodiment 84 further comprises: a second power supply line coupled to selectively supply current to the DACs of the second DAC two-dimensional array; and the first control line also disposed in operative proximity to at least one DAC in each of the at least three rows and at least three columns of the second DAC two-dimensional array and communicatively coupled to individual DACs in each of the at least three rows and at least three columns of the second DAC two-dimensional array.
[0392] Exemplary Embodiment 86 The system of exemplary embodiment 84 or 85, further comprising a second control line, the second control line being operatively proximate to at least one DAC in each of the at least three rows and at least three columns of the first DAC two-dimensional array and communicatively coupled to individual DACs in each of the at least three rows and at least three columns of the first DAC two-dimensional array.
[0393] Exemplary Embodiment 87 87. The system of exemplary embodiment 86, wherein the first control line and the second control line are operatively proximate to different DACs without a common DAC.
[0394] Exemplary Embodiment 88 88. The system of exemplary embodiment 87, wherein the first control line and the second control line are operatively adjacent to at least one same DAC with at least one common DAC.
[0395] Exemplary Embodiment 89 A first quantum flux parametron (QFP) magnetometer set, each QFP magnetometer of the first QFP magnetometer set includes a respective first material loop, a respective second material loop, and a respective compound Josephson junction with a Josephson junction pair, the respective first material loop and the respective second material loop exhibiting superconductivity at a critical temperature, the second material loop being interposed between the respective first material loop, the respective Josephson junction pair being interposed between the second material loop, and the respective second material loop being interposed between the respective first material loop. a first set of quantum flux parametron (QFP) magnetometers in parallel with each other in a second material loop with respect to the selected node, each second material loop including a respective second loop inductive interface for inductively receiving a control signal for controlling a respective compound Josephson junction, each first material loop including a plurality of first loop inductive interfaces, at least one of the first loop inductive interfaces being exposed to magnetic flux from an external magnetic field without an intervening ground plane to measure the external magnetic field with a respective QFP magnetometer; a set of control lines arranged to selectively communicatively couple control signals to the second loop induction interface; a set of flux bias lines arranged to selectively communicatively couple a flux bias to at least one of the first loop inductive interfaces of each QFP magnetometer of the first set of QFP magnetometers that provides flux feedback to measure an external magnetic field.
[0396] Exemplary embodiment 90 The system of exemplary embodiment 89 further comprises a processor chip including a plurality of quantum bits and a plurality of couplers, each coupler operative to selectively communicatively couple an individual pair of quantum bits, and the first QFP magnetometer set configured to detect an external magnetic field.
[0397] Exemplary embodiment 91 The system of exemplary embodiment 90, wherein the first QFP magnetometer set is an integrated component of the processor chip.
[0398] Exemplary embodiment 92 The system of any one of exemplary embodiments 89-91, further comprising a first shift register communicatively coupled to the first QFP magnetometer set.
[0399] Exemplary embodiment 93 The system of exemplary embodiment 89, wherein the first shift register is a QFP-based shift register that is a shift register that can be magnetically or galvanically coupled to individual ones of the first QFP magnetometer set.
[0400] Exemplary Embodiment 94 94. The system of exemplary embodiment 93, further comprising multiple non-destructive readouts (NDROs).
[0401] Exemplary Embodiment 95 The system of exemplary embodiment 93, further comprising a control circuit communicatively coupled to the QFP magnetometers of the first magnetometer set via a flux bias line and operable to sweep the flux bias supplied to the body of each of the QFP magnetometers of the first magnetometer set while taking multiple readings of the QFP magnetometers of the first magnetometer set.
[0402] Exemplary Embodiment 96 96. The system of exemplary embodiment 95, wherein the control circuitry further determines individual degeneracy points for each of the QFP magnetometers of the first magnetometer set based at least in part on the multiple readouts.
[0403] Exemplary Embodiment 97 The system of exemplary embodiment 96, wherein for each QFP magnetometer of the first magnetometer set, the control circuit fits an individual set of readout results to a tanh shape to determine an individual degenerate point for each QFP magnetometer of the first magnetometer set.
[0404] Exemplary Embodiment 98 The control circuitry is further communicatively coupled to the QFP magnetometers of the first magnetometer set via the first set of anneal lines to control the QFP magnetometers of the first set of magnetometers to receive a signal from the control circuitry. 0 / 2 and +Φ 0 Between and -Φ 0 / 2 and -Φ 0 103. The system of exemplary embodiment 97, operable to anneal each of the QFP magnetometers of the first magnetometer set between
[0405] Exemplary embodiment 99 The control circuit includes: Further measurements of annealed wire / QFP magnetometer interference were obtained, The system of exemplary embodiment 98 further compensates for the determined anneal line / QFP magnetometer interference.
[0406] Exemplary embodiment 100 The system of exemplary embodiment 89, wherein for each QFP magnetometer of the first QFP magnetometer set, at least some of the first loop inductive interfaces couple the individual QFP magnetometer to individual first loop inductive interfaces of nearby QFP magnetometers of the first QFP magnetometer set.
[0407] Exemplary embodiment 101 A method of operating a system, the system comprising: a first set of quantum flux parametron (QFP) magnetometers, each QFP magnetometer of the first set of QFP magnetometers including an individual first material loop, an individual second material loop, and an individual compound Josephson junction having a Josephson junction pair, the individual first material loop and the individual second material loop undergo superconductivity at a critical temperature, the second material loop being interposed in the individual first material loop, the individual Josephson junction pair being interposed in the second material loop, the individual second material loops being parallel to each other in the second material loop with respect to a node interposed in the individual first material loop, the individual second material loop having an individual second loop inductive input for inductively receiving a control signal to control the individual compound Josephson junction. the first material loop includes a first quantum flux parametron (QFP) magnetometer set, the first material loop including a first interface, the first material loop including a plurality of first loop inductive interfaces, at least one of the first loop inductive interfaces being exposed to magnetic flux from an external magnetic field without an intervening ground plane to measure the external magnetic field with a respective QFP magnetometer; a control line set arranged to selectively communicatively couple a control signal to the second loop inductive interface; a flux bias line set arranged to selectively communicatively couple a flux bias to at least one of the first loop inductive interfaces of each QFP magnetometer of the first QFP magnetometer set, the control line set providing flux feedback for measuring the external magnetic field; and a control circuit; For each QFP magnetometer in the first magnetometer set: Sweeping the flux bias applied to the body of the individual QFP magnetometers; and sweeping a flux bias applied to the body of the individual QFP magnetometer while taking multiple readings of the individual QFP magnetometer.
[0408] Exemplary embodiment 102 The system further includes a first shift register, and the method includes: communicatively coupling a first set of QFP magnetometers to a first shift register; 13. The method of exemplary embodiment 101, further comprising: providing a flux bias signal to sequentially shift quantum flux values between successive shift register elements.
[0409] Exemplary embodiment 103 13. The method of exemplary embodiment 102, wherein performing a plurality of readouts of the individual QFP magnetometers includes reading the individual QFP magnetometers by non-destructive readout (NDRO).
[0410] Exemplary embodiment 104 The method of any one of exemplary embodiments 101-103, further comprising determining an individual degeneracy point for each of the QFP magnetometers of the first magnetometer set based at least in part on the multiple readouts.
[0411] Exemplary Embodiment 105 The method of exemplary embodiment 104, wherein determining an individual degenerate point for each of the QFP magnetometers of the first magnetometer set based at least in part on the multiple readouts includes fitting the set of individual readout results to a tanh shape to determine an individual degenerate point for each of the QFP magnetometers of the first magnetometer set.
[0412] Exemplary embodiment 106 Supply a signal to the QFP magnetometers of the first magnetometer set via the first anneal line set to +Φ 0 / 2 and +Φ 0 Between and -Φ 0 / 2 and -Φ 0106. The method of exemplary embodiment 105, further comprising annealing each of the QFP magnetometers of the first magnetometer set between.
[0413] Exemplary Embodiment 107 Obtaining measurements of annealed wire / QFP magnetometer interference; and 107. The method of exemplary embodiment 106, further comprising compensating for the determined anneal line / QFP magnetometer interference.
[0414] Exemplary Embodiment 108 a first set of quantum flux parametron (QFP) magnetometers arranged as a first QFP shift register and communicatively coupled to each other, each QFP magnetometer of the first set of QFP magnetometers including an individual first material loop, an individual second material loop, and an individual compound Josephson junction having a Josephson junction pair, the individual first material loop and the individual second material loop undergo superconductivity at a critical temperature, the second material loop being interposed in the individual first material loop, the individual Josephson junction pair being interposed in the second material loop, the individual second material loops being parallel to each other in the second material loop with respect to a node interposed in the individual first material loop, the individual second material loop including an individual second loop inductive interface for inductively receiving a control signal to control the individual compound Josephson junction, and the individual first material loop including a plurality of first loop inductive interfaces; a set of control lines arranged to selectively communicatively couple control signals to the second loop induction interface; a set of flux bias lines arranged to selectively communicatively couple a flux bias to at least one of the first loop inductive interfaces of each QFP magnetometer of the first set of QFP magnetometers; and control circuitry communicatively coupled to sequentially provide magnetic flux signals to loops of a first QFP magnetometer set via a first loop inductive interface to shift values along at least one dimension of a first QFP shift register.
[0415] Exemplary embodiment 109 The system of exemplary embodiment 108, wherein the control circuit sequentially supplies magnetic flux signals to the loops of the first QFP magnetometer set via the first loop induction interface, and the individual magnetic flux signals supplied to each of the first loop induction interfaces are offset by π / 2 relative to the individual magnetic flux signals supplied to the immediately succeeding loop along the linear series of loops of the first QFP magnetometer set.
[0416] Exemplary embodiment 110 The system of exemplary embodiment 109, wherein the individual magnetic flux signals supplied to the first loop inductive interface of each loop are advanced by π / 2 relative to the individual magnetic flux signals supplied to the immediately preceding loop along the linear series of loops of the first QFP magnetometer set.
[0417] Exemplary embodiment 111 The system of exemplary embodiment 109, wherein the individual magnetic flux signals supplied to the first loop inductive interface of each loop are delayed by π / 2 with respect to the individual magnetic flux signals supplied to the immediately preceding loop along the linear series of loops of the first QFP magnetometer set.
[0418] Exemplary embodiment 112 The system of any one of exemplary embodiments 108-111, wherein the first QFP magnetometer set is arranged as a one-dimensional array shift register.
[0419] Exemplary embodiment 113 The system of any one of exemplary embodiments 108-111, wherein the first QFP magnetometer set is arranged as a two-dimensional array shift register.
[0420] Exemplary embodiment 114 The system of exemplary embodiment 108, wherein the control circuit reads out the shift register one QFP magnetometer at a time.
[0421] Exemplary Embodiment 115 The system of exemplary embodiment 108, wherein the control circuitry reads out the shift registers for all columns of the QFP magnetometer at once.
[0422] Exemplary Embodiment 116 The system of exemplary embodiment 108, 114 or 115, wherein the control circuitry is further operable to sweep the flux bias supplied to the body of each of the QFP magnetometers of the first magnetometer set while taking multiple readings of the QFP magnetometers of the first magnetometer set.
[0423] Exemplary Embodiment 117 17. The system of exemplary embodiment 116, wherein the control circuitry further determines an individual degeneracy point for each of the QFP magnetometers of the first magnetometer set based at least in part on the multiple readouts.
[0424] Exemplary Embodiment 118 The system of exemplary embodiment 117, wherein for each QFP magnetometer of the first magnetometer set, the control circuit fits an individual set of readout results to a tanh shape to determine an individual degenerate point for each QFP magnetometer of the first magnetometer set.
[0425] Exemplary Embodiment 119 The control circuitry is further communicatively coupled to the QFP magnetometers of the first magnetometer set via the first set of anneal lines to communicate with the QFP magnetometers of the first set of magnetometers ... 0 / 2 and +Φ 0 Between and -Φ 0 / 2 and -Φ 0 19. The system of exemplary embodiment 118, further operable to anneal each of the QFP magnetometers of the first magnetometer set between
[0426] Exemplary embodiment 120 The control circuit includes: Further measurements of annealed wire / QFP magnetometer interference were obtained, The system of exemplary embodiment 119 further compensates for the determined anneal line / QFP magnetometer interference.
[0427] 121 A hybrid computing system including a digital computer and a quantum computer, the quantum computer comprising a plurality of logic devices and / or control circuits implementing any one or more of the addressing methods described in any of the exemplary embodiments 1-120.
[0428] The above-described methods, processes, or techniques may be implemented by a series of processor-readable instructions stored on one or more non-transitory processor-readable media. Some examples of the methods or techniques are performed in part by a dedicated device, such as an adiabatic quantum computer or quantum annealing device, or a system that programs or controls the operation of the adiabatic quantum computer or quantum annealing device, such as a computer including at least one digital processor. The methods, processes, or techniques described above may include various acts, but those skilled in the art will appreciate that certain acts may be omitted and / or additional acts may be added in alternative examples. Those skilled in the art will appreciate that the order of acts shown is for illustrative purposes only and may be changed in alternative examples. Some of the exemplary acts or operations of the methods, processes, or techniques described above are performed iteratively. Some acts of the methods, processes, or techniques described above may be performed during each iteration, after multiple iterations, or at the end of all iterations.
[0429] The above description of the illustrated embodiments, including those described in the Abstract, are not intended to be exhaustive or to limit the embodiments to the precise forms disclosed. While specific embodiments and examples are described herein for illustrative purposes, those skilled in the art will recognize that various equivalent modifications may be made without departing from the spirit and scope of the disclosure. The teachings provided herein of the various embodiments are not limited to the exemplary quantum computing methods outlined above, but may also be applied to other quantum computing.
[0430] The various aspects, examples, and embodiments described above can be combined to provide further aspects, examples, and embodiments, respectively. To the extent not inconsistent with the specific teachings and definitions herein, all U.S. patents, U.S. patent application publications, U.S. patent applications, foreign patents, foreign patent applications, and non-patent publications referenced herein and / or listed in the Application Data Sheets are incorporated herein by reference, including, but not limited to, U.S. Provisional Patent Application No. 60 / 986,554, entitled "Systems, Devices and Methods for Analog Processing," filed November 8, 2007; U.S. Provisional Patent Application No. 62 / 161,780, filed May 14, 2015; U.S. Provisional Patent Application No. 62 / 288,251, filed January 28, 2016; U.S. Patent No. 8,854,074; U.S. Patent Application No. 12 / 017,995; U.S. Patent No. 8,169,231; No. 2006-0225165, U.S. Patent Application Publication No. 2006-0147154, U.S. Patent Application No. 12 / 013,192, U.S. Provisional Patent Application No. 62 / 331,287 filed May 3, 2016, U.S. Provisional Patent Application No. 62 / 405,027 filed October 6, 2016, International Patent Application No. PCT / US2016 / 031885, entitled "Frequency Multiplexed Resonator Input And / Or Output For A Superconducting Device", filed June 1, 2016, U.S. Patent No. 7,876,248, U.S. Patent No. 8,098,179, and U.S. Patent Application Publication No. 2014-0344322.
[0431] Aspects of the embodiments can be modified, if necessary, to employ systems, circuits, and concepts from various patents, applications, and publications to provide still other embodiments. These and other modifications can be made to the embodiments in light of the above detailed description. In general, in the following claims, the terms used should not be construed to limit the claims to the specific aspects disclosed in the specification and claims, but rather to include all possible aspects along with the full scope of equivalents to which such claims are entitled. Thus, the claims are not limited by this disclosure.
Claims
1. A plurality of quantum flux parametrons (QFPs) (704); at least one quantum flux parametron digital-to-analog converter (QFP-DAC) pair, each of the QFP-DACs including a respective primary material loop, a latch material loop interposed with a first pair of Josephson junctions and a latch interface, and a reset material loop interposed with a second pair of Josephson junctions and a reset interface, the primary material loops having the latch material loop interposed and the reset material loop interposed; a current bias line that provides a current bias to both of the QFP-DACs of the at least one QFP-DAC pair without a bias resistor; and for each QFP of the plurality of QFPs, a respective first plurality of quantum flux parametron latches (QFP latches) (706a) selectively operable to communicatively couple magnetic flux between the QFP and a first QFP-DAC (702a) of the QFP-DACs of the at least one QFP-DAC pair, and a respective second plurality of QFP latches (706b) selectively operable to communicatively couple magnetic flux between the QFP and a second QFP-DAC (702b) of the QFP-DACs of the at least one QFP-DAC pair.
2. 2. The system of claim 1, wherein each of the QFPs includes a respective first material loop (712a) interleaved with a respective second material loop (712b), the first material loop and the second material loop undergo superconductivity at a critical temperature, and each of the QFPs has a respective first inductive interface (718a) and a respective second inductive interface (718b), the second material loop having two Josephson junctions interleaved therewith, the first inductive interface (718a) arranged to communicatively couple magnetic flux to a first QFP latch of the respective first plurality of QFP latches (706a), and the second inductive interface (718b) arranged to communicatively couple magnetic flux to a first QFP latch of the respective second plurality of QFP latches (706b).
3. 2. The system of claim 1, wherein a first QFP latch of the first plurality of QFP latches (706a) and a first QFP latch of the second plurality of QFP latches (706b) each include a respective Josephson junction material loop having two Josephson junctions interposed therebetween.
4. 2. The system of claim 1, wherein each of the QFPs (704) includes an individual loop comprising a first inductive interface (718a) positioned to communicatively couple magnetic flux to a first QFP latch of the individual first plurality of QFP latches (706a) and a second inductive interface (718b) positioned to communicatively couple magnetic flux to a first QFP latch of the individual second plurality of QFP latches (706b).
5. 5. The system of claim 4, wherein the first QFP latches of the first plurality of individual QFP latches (706 a) include individual loops with individual interfaces arranged to communicatively couple magnetic flux to a second QFP latch of the first plurality of individual QFP latches (706 a), and the first QFP latches of the second plurality of individual QFP latches (706 b) include individual loops with individual interfaces arranged to communicatively couple magnetic flux to a second QFP latch of the second plurality of individual QFP latches (706 b).
6. 6. The system of claim 5, wherein the first QFP latches of the first plurality of individual QFP latches (706a) are responsive to a respective first latch signal and the first QFP latches of the second plurality of individual QFP latches (706b) are responsive to an inverted logic of the respective first latch signal.
7. 7. The system of claim 6, wherein the first QFP latches of the first plurality of discrete QFP latches (706a) are responsive to a respective first offset signal, and the first QFP latches of the second plurality of discrete QFP latches (706b) are responsive to the respective first offset signal.
8. 6. The system of claim 5, wherein the first QFP latches of the first plurality of discrete QFP latches (706a) are responsive to a respective first address signal and the first QFP latches of the second plurality of discrete QFP latches (706b) are responsive to an inverted logic of the respective first address signal.
9. 9. The system of claim 8, wherein the first QFP latches of the first plurality of discrete QFP latches (706a) are responsive to a respective first latch signal, and the first QFP latches of the second plurality of discrete QFP latches (706b) are responsive to the respective first latch signal.
10. The system of claim 7 or 9, wherein each of the at least one QFP-DAC pair is responsive to a respective second latch signal.
11. The system of claim 10, wherein each of the primary material loops of the QFP-DAC includes a first interface (730) arranged to be communicatively coupled to one of the first plurality of QFP latches or a respective one of the second plurality of QFP latches, and at least a second interface (742) to one of an input or output.
12. The system of claim 11 , wherein each of the QFP-DACs is responsive to a respective second latch signal.
13. The system of claim 12, wherein each of the QFP-DACs is responsive to a respective first chip signal.
14. The system of claim 13, wherein each of the QFP-DACs is responsive to a separate reset signal.
15. 15. The system of claim 14, wherein a first one of each of the QFPs is responsive to a respective offset signal.
16. The system of any one of claims 1 to 8, further comprising a control circuit communicatively coupled to control-supply signals to the QFP-DAC and the QFP latch.
17. 17. The system of claim 16, wherein the control circuit provides a signal to the QFP latch to load information into the DAC.
18. 17. The system of claim 16, wherein the control circuit provides a signal to the QFP latch to operate as a quantum flux parametron demultiplexer (QFP-Demux).
19. 17. The system of claim 16, wherein the control circuit provides a signal to the QFP latch to load information from the DAC.
20. 17. The system of claim 16, wherein the control circuit provides a signal to the QFP latch to operate as a quantum flux parametron multiplexer (QFP-Mux).
21. 17. The computing system of claim 16, wherein the QFP, the QFP-DAC and the QFP latch all form a non-dissipative programmable magnetic memory (PMM) residing on a single chip.
Citation Information
Patent Citations
Series structure of josephson gates and digital-analog converter using same
JP1993198856A
Qubit state readout system, method, and apparatus
JP2010541309A
Systems, methods, and apparatus for superconducting demultiplexer circuits
JP2011524131A
Systems, methods and apparatus for superconducting demultiplexer circuits
US20110065586A1
