Input / output system and method for superconducting devices
The hybrid computing system with FMR technology and interspersed grids addresses inefficiencies in quantum processor programming and readout, achieving constant time complexity for scalable and efficient operation.
Patent Information
- Application Number
- JP2021565846
- Authority / Receiving Office
- JP · JP
- Patent Type
- Patents
- Current Assignee / Owner
- Priority Date
- 2019-06-11
- Filing Date
- 2020-06-11
- Publication Date
- 2024-11-07
- Estimated Expiration
- 2040-06-11
AI Technical Summary
Existing quantum processors face inefficiencies in programming and readout times due to the high number of programming lines and qubit readout operations, which are proportional to the product of the number of tiles and qubits, limiting scalability and calibration efficiency.
Implementing a hybrid computing system with frequency multiplexed resonant readout (FMR) technology, incorporating a quantum processor with interspersed grids of qubits and shift registers, and using fewer analog lines for programming and control, allowing parallel programming and readout of qubits across multiple tiles.
Reduces programming and readout times to a constant time complexity independent of the number of tiles and qubits, enabling scalable and efficient operation of quantum processors with improved calibration and reduced latency.
Smart Images

Figure 0007579273000004 
Figure 0007579273000005 
Figure 0007579273000006
Abstract
Description
[Technical field]
[0001] Technical Field The present disclosure relates generally to input and / or output systems and methods for superconducting devices, such as superconducting quantum computers and superconducting classical computers, and more particularly to systems and methods for inputting data to a superconducting quantum processor and / or measuring the state of qubits in a superconducting quantum processor. [Background technology]
[0002] background Frequency Multiplexed Resonance (FMR) Readout Superconducting microwave resonators have been used in a variety of fields, including but not limited to quantum computing and astronomy. For example, in quantum computing, superconducting resonators have been used to detect the state of qubits. In astronomy, superconducting microwave resonators have been used in Microwave Kinetic Inductance Detectors (MKIDs). In both cases, many resonators (used as or in detectors) can be coupled to a common transmission line and integrated by frequency domain multiplexing (see glossary).
[0003] Using FMR techniques, superconducting resonators with different resonant frequencies can be used to read out multiple qubits, each of which can share a common microwave transmission line by using frequency domain multiplexing. Summary of the Invention [Means for solving the problem]
[0004] Quick Overview Briefly, a quantum processor may include a plurality of tiles, the plurality of tiles being arranged in a first grid, a first tile of the plurality of tiles including a first qubit, a shift register including at least one shift register stage communicatively coupled to a frequency multiplexed resonant (FMR) readout device, the qubit readout device, a plurality of digital-to-analog (DAC) buffer stages, and a plurality of shift-register-loadable digital-to-analog converters (DACs) arranged in a second grid. In some implementations, the first qubit is communicatively coupled to the qubit readout device. In some implementations, at least one of the plurality of DAC buffer stages is communicatively coupled to at least one of the plurality of shift-register-loadable DACs.
[0005] In some implementations, the first qubit is a superconducting qubit. In some implementations, the superconducting qubit is a superconducting flux qubit.
[0006] In some implementations, the quantum processor further comprises a microwave transmission line that is communicatively coupled to the FMR readout device.
[0007] In some implementations, the FMR readout device comprises a superconducting resonator.
[0008] In some implementations, the first grid and the second grid are interspersed with one another on the superconducting integrated circuit, with the term "interspersed" in this context referring to the first grid and the second grid (and the devices within each grid) being physically intermingled, woven, combined, twisted, or fused with one another on the superconducting integrated circuit.
[0009] In some implementations, the FMR readout device and multiple shift register loadable digital-to-analog converters (DACs) reside on the same superconducting integrated circuit.
[0010] In some implementations, the quantum processor further comprises a transmission line including at least one transmission line inductance, a superconducting resonator, and a coupling capacitance communicatively coupling the superconducting resonator to the transmission line.
[0011] In some implementations, the FMR readout device is communicatively coupled to at least one other element of the quantum processor by a superconducting via, hi some implementations, the FMR readout device is communicatively coupled to at least one other element of the quantum processor by at least one of a bump bond and / or a solder bond.
[0012] Briefly described, a hybrid computing system may include a quantum processor of the various implementations described above, and further includes at least one digital processor and at least one persistent, processor-readable medium communicatively coupled to the at least one digital processor that stores at least one of processor-executable instructions or data that, when executed by the at least one digital processor, enable the at least one digital processor to program the quantum processor.
[0013] In some implementations, at least one of the processor-executable instructions or data, when executed by the at least one digital processor, enables the at least one digital processor to program a first tile of the plurality of tiles, such that the at least one digital processor can program the quantum processor. In some implementations, at least one of the processor-executable instructions or data, when executed by the at least one digital processor, enables the at least one digital processor to program at least one of the plurality of shift register loadable DACs, such that the at least one digital processor can program a first tile of the plurality of tiles.
[0014] In some implementations, the at least one of the processor-executable instructions or data, when executed by the at least one digital processor, for enabling the at least one digital processor to program the quantum processor, further enables the at least one digital processor to program a second tile of the plurality of tiles. In some implementations, the at least one of the processor-executable instructions or data, when executed by the at least one digital processor, for enabling the at least one digital processor to program the second tile of the plurality of tiles, further enables the at least one digital processor to program a first tile and a second tile of the plurality of tiles in parallel. In various such implementations, enabling the at least one digital processor to program a first tile and a second tile of the plurality of tiles in parallel may enable a first time required to program a first tile to be at least approximately the same as a second time required to program a second tile of the plurality of tiles.
[0015] In some implementations, at least one of the processor-executable instructions or data, when executed by the at least one digital processor, further enables the hybrid computing system to read out a first qubit of a first tile of the plurality of tiles in parallel with a second qubit of the first tile of the plurality of tiles. In various implementations such as those described above, at least one of the processor-executable instructions or data, when executed by the at least one digital processor, further enables the hybrid computing system to read out a first qubit of a first tile of the plurality of tiles in parallel with a third qubit of a second tile of the plurality of tiles.
[0016] In some implementations, at least one of the processor-executable instructions or data, when executed by the at least one digital processor, further enables the hybrid computing system to read out a first qubit in a first tile of the plurality of tiles and a second qubit in a second tile of the plurality of tiles, wherein a first time required to read out the first qubit in the first tile of the plurality of tiles is at least approximately the same as a second time required to read out the second qubit in the second tile of the plurality of tiles.
[0017] In various implementations of the above, at least one of the processor-executable instructions or data, when executed by the at least one digital processor, further enables the at least one digital processor to calibrate at least a first portion of the quantum processor and a second portion of the quantum processor in parallel operation, wherein the first portion of the quantum processor is non-local with respect to the second portion of the quantum processor.
[0018] Briefly described, a quantum computer comprises a first superconducting integrated circuit including a quantum processor, the quantum processor including a plurality of superconducting flux qubits; and a second superconducting integrated circuit including an input / output system including a shift register including at least one shift register stage communicatively coupled to a Frequency Multiplexed Resonant Readout (FMRR) module, a qubit readout, a plurality of DAC buffer stages, and a plurality of shift register loadable digital-to-analog converters (DACs) arranged in a grid.
[0019] Briefly, a method of programming a hybrid computing system, the hybrid computing system including a quantum processor and a digital processor, the quantum processor including a plurality of tiles, the plurality of tiles arranged in a first grid, a first tile of the plurality of tiles including a first qubit, a shift register including at least one shift register stage communicatively coupled to a frequency multiplexed resonant (FMR) readout device, a qubit readout device, a plurality of digital-to-analog converter (DAC) buffer stages, and a plurality of shift register loadable DACs arranged in a second grid, may include programming the quantum processor with the digital processor and reading out the quantum processor with the digital processor.
[0020] In some implementations, programming the quantum processor with the digital processor includes programming a first tile of the plurality of tiles. In some implementations, programming the first tile of the plurality of tiles includes programming at least one of the plurality of shift register loadable DACs.
[0021] In some implementations, programming the quantum processor with the digital processor further includes programming a second tile of the plurality of tiles. In some implementations, programming the first and second tiles of the plurality of tiles includes programming the first and second tiles of the plurality of tiles in parallel. In various implementations such as those described above, programming the first tile of the plurality of tiles has a first duration and programming the second tile of the plurality of tiles has a second duration, the first duration being at least approximately the same as the second duration.
[0022] In some implementations, reading out the quantum processor by the digital processor includes reading out a first qubit of a first tile of the plurality of tiles in parallel with a second qubit of the first tile of the plurality of tiles. In various implementations described above, reading out the quantum processor by the digital processor includes reading out a first qubit of a first tile of the plurality of tiles in parallel with a third qubit of a second tile of the plurality of tiles. In some implementations, reading out the quantum processor by the digital processor includes reading out a first qubit of a first tile of the plurality of tiles and reading out a second qubit of a second tile of the plurality of tiles, wherein a first time taken to read out the first qubit of the first tile of the plurality of tiles is at least about the same as a second time taken to read out the second qubit of the second tile of the plurality of tiles.
[0023] In various implementations of the above, the method further includes calibrating at least a first portion of the quantum processor and a second portion of the quantum processor in parallel operation, the first portion of the quantum processor being non-local with respect to the second portion of the quantum processor.
[0024] A brief description of some of the figures in each drawing In the various drawings, the same reference numbers identify similar elements or acts. The size and relative position of each element in the drawings are not necessarily drawn to scale. For example, the shapes and angles of various elements are not necessarily drawn to scale, and some of these elements have been arbitrarily enlarged and positioned to improve drawing clarity. Furthermore, the specific shape of each depicted element does not necessarily convey any information about the actual shape of the specific element, but has been selected solely for ease of recognition in each drawing. [Brief description of the drawings]
[0025] [Figure 1] FIG. 1 is a schematic diagram illustrating an example arrangement of an FMRR array in a superconducting processor. [Diagram 2] FIG. 1 is a schematic diagram illustrating an exemplary implementation of a superconducting processor comprising a shift register. [Diagram 3] FIG. 1 is a schematic diagram illustrating an exemplary embodiment of a readout system for a superconducting circuit. [Figure 4] FIG. 1 is a schematic diagram of an exemplary hybrid computing system comprising a digital computer and a quantum computer that may incorporate the FMRR techniques described herein. [Diagram 5] 5 is a flow diagram of an exemplary method for programming a hybrid computing system, such as the hybrid computing system of FIG. 4. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS
[0026] Detailed Description Glossary Qubit: A qubit (also referred to in this application as a quantum bit) is the basic unit of quantum information and is a quantum version of a classical binary bit that can be physically realized in a two-state device. A qubit is a two-state quantum mechanical system. Qubit also refers to the actual physical device in which information is stored. For example, a superconducting qubit is a type of superconducting device that can be included in a superconducting integrated circuit. Superconducting qubits can take the form of, for example, charge-based or flux-based qubits.
[0027] Superconducting Devices: Superconducting devices are electronic devices that exploit the properties of superconducting materials, such as zero electrical resistance and emission of magnetic flux when cooled below the critical temperature characteristic of a superconducting material.
[0028] Superconducting Circuit: A superconducting circuit is a circuit that contains one or more superconducting devices.
[0029] Superconducting microwave resonators (also referred to in this application as superconducting microresonators): Superconducting microwave resonators are superconducting circuits that exhibit resonance at microwave frequencies. Superconducting microresonators may be formed by depositing a superconducting thin film on an insulating substrate and applying standard lithographic patterning techniques to form the resonator structure. Superconducting microresonators may be lumped element circuits or transmission line resonators. Microresonators can be attractive for detector applications due to their simplicity and because large arrays can be read out using frequency domain multiplexing (see FMRR below).
[0030] Microwave Transmission Line: A microwave transmission line is a cable or other structure comprising one or more conductors operable to carry alternating electrical current at microwave frequencies.
[0031] Frequency Domain Multiplexing (FDM): FDM is a technique used in which the communications bandwidth is divided into multiple non-overlapping sub-bands, with each sub-band carrying a separate signal.
[0032] Frequency Multiplexed Resonant Readout (FMRR): FMRR is a readout technique that includes a superconducting resonator that can operate in a frequency domain multiplexed mode. FMRR techniques include one or more Frequency Multiplexed Resonant (FMR) readouts.
[0033] SQUID (Superconducting Quantum Interference Device): A SQUID is a superconducting device that contains a superconducting loop containing one or more Josephson junctions. SQUIDs can be used as magnetometers that can measure very weak magnetic fields. DC SQUIDs have two Josephson junctions connected in parallel. RF-SQUIDs have a superconducting loop containing a single Josephson junction.
[0034] Lumped Element Design: In lumped element design, a spatially distributed physical system is described as a topology of discrete entities that approximates the behavior of the distributed system under certain assumptions. For example, this is useful in electrical and electronic systems.
[0035] Shift Register: A shift register is a sequential logic circuit operable to store and / or transfer data.
[0036] Quantum Flux Parametron (QFP): A QFP is a logic circuit that contains at least one superconducting Josephson junction and a resonant circuit whose oscillations can represent binary digits. Although the design utilizes quantum principles, QFPs are an element of classical computing technology, not quantum computing technology.
[0037] Hybrid Computer: A hybrid computer is a system that includes at least one digital processor and at least one analog processor (eg, a quantum processor).
[0038] Josephson Junction: A Josephson junction is a device that contains two electrodes of a material capable of superconducting below the critical temperature characteristic of the material, and a thin insulating barrier separating the electrodes.
[0039] Flux Digital-to-Analog Converter (DAC): A flux DAC is a superconducting device that establishes a flux quantum representation of a digital signal and converts it into an analog superconducting current, which can then be used as a complement to another device, e.g. a programmable device.
[0040] preamble In the following description, certain specific details are included to provide a thorough understanding of the various implementations and embodiments disclosed. However, those skilled in the art will understand that each implementation or embodiment may be practiced without one or more of these specific details, or with other methods, components, materials, etc. In other instances, well-known structures associated with superconducting circuits or superconducting resonators have not been shown or described in detail to avoid unnecessarily obscuring the description of the implementations and embodiments of the method. Throughout this specification and the appended claims, the terms "element" and "elements" are used to encompass, but not be limited to, any such structures, systems, and devices associated with superconducting circuits and superconducting resonators.
[0041] Unless the context requires otherwise, throughout the following specification and claims, the term "comprise" is synonymous with "including" and is inclusive or open-ended (i.e., does not exclude additional, unrecited elements or operations).
[0042] Throughout this specification, references to "one embodiment," "another embodiment," "one example," "one example," "another example," "one implementation," "another embodiment," or the like mean that the particular referenced feature, structure, or characteristic described in connection with that embodiment, example, or implementation is included in at least one embodiment, example, or implementation. Thus, the appearances of the phrases "in one embodiment," "in an embodiment," "another embodiment," and the like in various places throughout this specification do not necessarily all refer to the same embodiment, example, or implementation. Furthermore, particular features, structures, or characteristics may be combined in any suitable manner in one or more embodiments, examples, or implementations.
[0043] It should be noted that, as used in this specification and the appended claims, the singular forms "a," "an," and "the" include plural referents unless the context clearly dictates otherwise. Thus, for example, reference to a readout system including a "superconducting resonator" includes a single superconducting resonator or two or more superconducting resonators. It should also be noted that the term "or" is generally utilized in its "and / or" sense unless the context clearly dictates otherwise.
[0044] The headings provided herein are for convenience only and do not interpret the scope or meaning of each embodiment.
[0045] Detailed description of the technology In one conventional implementation of a quantum processor, the qubits are arranged in one or more tiles, and the tiles in the quantum processor are arranged in a grid, for example a square grid of N×N tiles.
[0046] In this application, a tile refers to a set of qubits. In one implementation, a tile includes M qubits aligned horizontally and an equal number of M qubits aligned vertically. In another embodiment, a tile includes at least one partial qubit (i.e., the tile includes at least one qubit where a portion of the qubit resides outside the boundary of the tile, or at least one qubit where the at least one qubit is exclusively a partial instantiation of the qubit). A portion of a qubit can be a segment of a qubit. A partial instantiation of a qubit can include only some (but not all) segments of a qubit. For some examples of quantum processor topologies, see the following reference International PCT Patent Publication No. WO 2017214331 A1, entitled "SYSTEMS AND METHODS FOR QUANTUM PROCESSOR TOPOLOGY."
[0047] In this application, a horizontally aligned qubit is a qubit that has a larger dimension along a horizontal axis than along a vertical axis, which is perpendicular to the horizontal axis. The horizontal axis may be defined, for example, as an axis parallel to one side of an integrated circuit that includes a quantum processor. In this application, a vertically aligned qubit is a qubit that has a larger dimension along a vertical axis than along the horizontal axis.
[0048] In an implementation of a quantum processor such as the quantum processor mentioned above (with N×N tiles and M×M qubits per tile), reading out a qubit in the quantum processor may take a readout time proportional to the product NM.
[0049] A quantum processor can be programmed using a number of programming lines to the quantum processor. In this application, the programming lines are also referred to as addressing lines and power lines. For a description of addressing lines and power lines, see, for example, "QUANTUM FLUX PARAMETRON BASED STRUCTURES (EG, MUXES, DEMUXES, SHIFT REGISTERS), ADDRESSING LINES AND RELATED METHODS" in International PCT Patent Application No. US 2018 / 054306 (International Publication No. WO 2019 / 070935A2).
[0050] A programming system for a quantum processor may include several digital-to-analog converters (DACs). In conventional implementations of quantum processors, such as the quantum processors mentioned above, the programming system may include approximately NM 2 It can be equipped with a DAC.
[0051] It is generally desirable to keep the number of programming lines (also referred to in this application as the total line count) of a quantum processor as low as practically possible. Typically, it is at least approximately
number
number
[0052] While the exemplary implementation described above relates to a square grid of tiles and an equal number of horizontal and vertical qubits in each tile, other implementations of the systems and methods described in this application may have at least one of: a) a non-square grid of tiles, and b) tiles with an unequal number of horizontal and vertical qubits.
[0053] Examples of systems of apparatus implementing shift registers in quantum processors are provided, for example, by U.S. Patent Application No. 15 / 726,239 (also published as U.S. Patent Application Publication No. 20180101786A1) and International PCT Application Publication No. WO 2019 / 070935A2, both entitled “QUANTUM FLUX PARAMETRON BASED STRUCTURES (EG MUXES, DEMUXES, SHIFT REGISTERS), ADDRESSING LINES AND RELATED METHODS.”
[0054] Some quantum processors address the shift register-controlled DACs using an XYZ scheme, where the X and Y signals are designated the address (ADDR) and trigger (TRIG) lines, respectively, and the Z signal is designated the power (PWR) line. The DAC stages within a partition of a quantum processor can be electrically coupled in series so that they can communicate with each other. By initiating a partition, asserting the ADDR line, and toggling the TRIG line a number of times, a corresponding number of pulses can be written to the selected DAC.
[0055] Techniques useful for operating scalable processors with relatively large numbers of logic devices (e.g., qubits) can use QFPs to implement, for example, shift registers, multiplexers, demultiplexers, and permanent magnetic memory. Such techniques can use XY or XYZ addressing schemes and can use control lines that run in a "twisted" pattern across a series of devices. In some implementations, the DAC can store energy using thin-film superconducting materials and / or a series of Josephson junctions using mechanical inductance, can use single or multiple loop designs, and can include serpentine structures. The DAC can be galvanically and / or inductively coupled to other devices.
[0056] This application describes systems and methods that include a) a shift register controlled DAC (i.e., a DAC controlled by one or more shift registers) and b) a shift register configured communicatively coupled to a qubit readout device (e.g., a qubit readout quantum flux parametron [QFP]).
[0057] The shift register can be communicatively coupled to one or more frequency and sensitivity tunable resonator (FASTR) detectors (also referred to herein as FASTRs or FASTR devices). See, e.g., JD Whittaker et al., "A FREQUENCY AND SENSITIVITY TUNABLE MICRORESONATOR ARRAY FOR HIGH-SPEED QUANTUM PROCESSOR READOUT," Journal of Applied Physics 119, 014506 (2016).
[0058] One implementation of the systems and methods described herein includes a regular grid (e.g., a rectangular grid) of shift register controlled DACs with two readout axes, one axis for reading out horizontal qubits and another axis for reading out vertical qubits.
[0059] The shift registers for routing data to and from the FASTR detector can be communicatively coupled in at least a tree-like fashion. In this application, a "tree" refers to a set of nodes connected by a set of edges such that there is a path from any node to any other node, and there are no cycles at each node. In this application, the term "at least tree-like" refers to a relaxation of the condition that there are no cycles at each node. The FASTR detector can be the root of the tree, and each shift register stage can be connected as desired. In one implementation, the system can move data from the FASTR detector to a desired shift register stage and back.
[0060] In another implementation of the systems and methods described in this application, there are three readout axes and hexagonal regions of the DAC.
[0061] In yet another implementation, the I / O area includes only a portion of a tile. In yet another implementation, the I / O area includes two or more tiles. In yet another implementation, the I / O area includes only a partial or complete row of one or more tiles. In yet another implementation, the I / O area includes only a partial or complete column of one or more tiles.
[0062] One advantage of the systems and methods described herein is the reduced programming and / or readout time of a quantum processor. In one implementation, the programming time is 2 and the readout time can be proportional to M. In one implementation, the lines to a quantum processor comprising an N by N grid of tiles include N microwave lines. In other implementations, the lines to a quantum processor comprising an N by N grid of tiles include more than N microwave lines. In yet other implementations, the lines to a quantum processor comprising an N by N grid of tiles include less than N microwave lines. The lines to the quantum processor may also include a number of shift register control lines.
[0063] In one implementation, each qubit of the quantum processor is located on one integrated circuit (also referred to in this application as a chip) and the FMR readout is located on a separate chip. An advantage of using a separate chip for the FMR readout is that it improves isolation of the qubits of the quantum processor from the microwave FMRR lines. Another advantage is that it reduces the area required for the layout of the quantum processor. Yet another advantage is that it is possible to test and pre-select the FMR readout devices in a cryogenic environment and then integrate them with the quantum processor, thereby improving the reliability of the integrated system.
[0064] In another implementation, the quantum processor may be equipped with a conventional readout system, including peripheral FMRR and / or non-destructive readout (NDRO) modules for input and output, in which both the FMR readout and the quantum processor can be tested independently before integration.
[0065] In another embodiment, the FMR readout device is integrated into the quantum processor, for example in-situ with a tiled arrangement of qubits.
[0066] In one implementation, the FMR readout devices are arranged in a rectangular grid, which is controlled by microwave lines arranged in parallel, see for example the implementation of Figure 1 and the following description.
[0067] The rectangular grid of FMR readout devices can be communicatively coupled into a unified input / output network, for example as shown in and described below with reference to FIG. 2. In this configuration, the input / output (I / O) to each internal tile can be performed independently. The control I / O to each tile and the lines for controlling this I / O can be identical. The input / output to each tile can be performed in parallel to parallelize the input and / or output operations in a tiled manner. In this context, performing two or more operations (e.g., two or more inputs / outputs) in parallel refers to performing the two or more operations simultaneously or at least partially overlapping in time. An input operation may include programming the device. An output operation may include reading out the device.
[0068] It may be desirable to calibrate one portion of a quantum processor independently of another portion. The term "calibration" in this context includes measuring characteristics of the quantum processor and / or each device within the quantum processor. It may be beneficial for calibration to be able to control the local environment (e.g., qubits) of each device in a quantum processor. Typically, the local environment refers to the device and its associated DAC, its nearest neighbor and its associated DAC, and its next nearest neighbor and its associated DAC.
[0069] One approach to calibrating is to split the calibration process into a set of local measurements that can be performed in parallel with each other. This approach may be applicable to various types of quantum computers, including but not limited to adiabatic quantum computers and gate model quantum computers.
[0070] To perform calibration in parallel, it may be advantageous to be able to program the processor and read out the qubits in a manner that does not add time proportional to the number of devices being measured (which would otherwise have little or no advantage over calibrating each local region one at a time). One approach uses a parallel programming scheme that matches the scheme of locality (e.g., where each of N×M blocks of devices shares a set of programming lines). In this case, it is possible to parallelize the programming of devices that are not local to each other. Reading can be performed by a parallel reading scheme that can read the non-local devices in parallel. Programming can be performed by a parallel programming scheme that uses DACs near the devices that are not local to each other.
[0071] Control of the local environment (e.g., qubits) of each device can be parallelized if parallel programming interfaces exist. Annealing of qubits can be parallelized if qubits share analog lines and / or if the currents on separate analog lines can be altered in parallel. Measuring the state of a qubit (also referred to herein as reading out the qubit) can be parallelized using parallel readout techniques (e.g., FMR readout techniques). The systems and methods described herein include FMR readout techniques.
[0072] The term "parallelizable" in this context refers to an operation whose cost of performing the operation does not scale linearly with the number of devices used. In some implementations, it may be advantageous for one aspect of an operation (e.g., the most time-consuming aspect) to be parallelizable even if other aspects are not.
[0073] FIG. 1 is a schematic diagram illustrating an exemplary implementation of a portion 100 of a superconducting processor.
[0074] The portion 100 comprises a series of FMR readout devices 102-1, 102-2, 102-3, and 102-4 (collectively referred to herein as FMR readout devices 102), FMR readout devices 104-1, 104-2, 104-3, and 104-4 (collectively referred to herein as FMR readout devices 104), FMR readout devices 106-1, 106-2, 106-3, and 106-4 (collectively referred to herein as FMR readout devices 106), and FMR readout devices 108-1, 108-2, 108-3, and 108-4 (collectively referred to herein as FMR readout devices 108). The FMR readout device 102 is communicatively coupled to an input line 110. The FMR readout device 104 is communicatively coupled to an input line 112. The FMR readout device 106 is communicatively coupled to an input line 114. The FMR readout device 108 is communicatively coupled to an input line 116.
[0075] In one implementation, input lines 110, 112, 114, and 116 are separate input lines. In another implementation, some or all of input lines 110, 112, 114, and 116 are communicatively coupled to one another, e.g., wired to one another.
[0076] The FMR readout device 102 is communicatively coupled to other elements of the superconducting processor (not shown in FIG. 1 ) at 118-1, 118-2, 118-3, and 118-4, respectively. The FMR readout device 104 is communicatively coupled to other elements of the superconducting processor at 120-1, 120-2, 120-3, and 120-4, respectively. The FMR readout device 106 is communicatively coupled to other elements of the superconducting processor at 122-1, 122-2, 122-3, and 122-4, respectively. The FMR readout device 108 is communicatively coupled to other elements of the superconducting processor at 124-1, 124-2, 124-3, and 124-4, respectively.
[0077] In one implementation, the FMR readout device is communicatively coupled in situ to other elements of the superconducting processor. In another implementation, the FMR readout device is communicatively coupled to other elements of the superconducting processor by superconducting vias. In this application, a via (vertical interconnect access) is an electrical connection between layers in a physical multi-layer electronic circuit (e.g., integrated circuit) that passes through the plane of one or more adjacent layers. In yet another implementation, the FMR readout device is communicatively coupled to other elements of the superconducting processor on a separate chip using bump bonds, solder bonds, or another suitable electrical communication coupling.
[0078] 2 is a schematic diagram illustrating another exemplary implementation of a portion 200 of a superconducting processor. The portion 200 of the superconducting processor includes a plurality of shift registers.
[0079] The portion 200 of the superconducting processor includes a shift register 202 communicatively coupleable to an FMRR module (e.g., FMRR module 104-1 of FIG. 1). In some implementations, the shift register 202 is a single shift register stage. In some implementations, the shift register 202 includes two or more shift register stages. The portion 200 of the superconducting processor can be communicatively coupleable to an FMRR module, for example, by multiple shift register stages.
[0080] The portion 200 of the superconducting processor includes a tee stage 204 communicatively coupling a shift register 202 to three-phase internal streets 206 and 208. In this application, a street is a data path, i.e., a path through the superconducting processor along which data may travel. Data traveling along a street may be input data and / or output data. In one implementation, the tee stage 204 is communicatively coupled to an FMRR module (e.g., FMRR module 104-1 of FIG. 1) by the shift register 202. In another implementation, the tee stage 204 is communicatively coupled to an FMRR module by the shift register 202, where the shift register 202 includes multiple shift register stages.
[0081] A street may have fewer than the three phases shown in Figure 2. A street may have more than the three phases shown in Figure 2. In some implementations, a street has four phases.
[0082] Portion 200 of superconducting processor includes qubit readout devices 210-1 and 210-2. Qubit readout devices 210-1 and 210-2 are components of a row of qubit readout devices. The qubit readout devices in a row of qubit readout devices, including qubit readout devices 210-1 and 210-2, are collectively referred to in this application as qubit readout device 210. Each qubit readout device of qubit readout device 210 may include two or more stages. For clarity of illustration, only one stage of each qubit readout device of qubit readout device 210 is shown in FIG. 2.
[0083] Superconducting processor portion 200 includes qubit readout devices 212-1 and 212-2, which are members of a column of qubit readout devices, each qubit of which is collectively referred to in this application as qubit readout device 212. Superconducting processor portion 200 includes a three-phase internal street 214 communicatively coupled to qubit readout device 212.
[0084] The portion 200 of the superconducting processor includes a DAC buffer stage 216. In some implementations, the DAC buffer stage 216 is in phase with a qubit readout stage, such as the qubit readout device 212.
[0085] Portion 200 of the superconducting processor includes shift-register-loadable DACs 218-1 and 218-2. Shift-register-loadable DACs 218-1 and 218-2 are members of a grid of shift-register-loadable DACs. The shift-register-loadable DACs in the grid of shift-register-loadable DACs are collectively referred to in this application as shift-register-loadable DACs 218. In some implementations, the DACs include QFP-DACs.
[0086] A portion 200 of a superconducting processor includes three interior streets 206, 208, and 214: a) a horizontal interior street 206 communicatively coupled to a qubit readout device 210 for vertically aligned qubits, b) a vertical interior street 214 communicatively coupled to a qubit readout device 212 for horizontally aligned qubits, and c) a vertical interior street 208 communicatively coupled to a shift-register loadable DAC buffer stage 216. In this application, the term "shift-register loadable DAC buffer stage" refers to a DAC buffer stage that can be loaded from a shift register. The shift-register loadable DAC buffer stage 216 is communicatively coupled to a shift-register loadable DAC 218.
[0087] 2, the internal streets 206 and 208 are three-phase internal streets, i.e., there are three interleaved families of stages: ABCABCABC.... In some implementations, stages in one family (e.g., A stage) are controlled by a first global annealing line, stages in another family (e.g., B stage) are controlled by a second global annealing line, and stages in yet another family (e.g., C stage) are controlled by a third global annealing line. In practice, there may be one or more global DC bias lines in addition to the global annealing lines.
[0088] The purpose of the global annealing line is to support two operations, referred to in this application as "inhibit" and "latch." The inhibit operation can set and hold the current in each shift register stage to zero (or near zero, i.e., below a defined threshold). The latch operation can anneal each shift register stage. In one implementation where the shift register stages are bistable QFPs, each QFP can have a normalized current of either +1 or -1 at the end of the latch operation, depending on the value of the local bias that each shift register stage receives. For example, to move data through the shift register, the inhibit and latch operations may be performed simultaneously.
[0089] In an exemplary situation, the system includes data in A shift register stages, with each A stage having a flux state of either +1 or -1. For example, an A stage may include a sequence of data represented as (+1, +1, -1) as follows: A(+1)-BCA(+1)-BCA(-1)-BC
[0090] If stages B and C are inhibited and their states are set to zero, the sequence becomes: A(+1)-B(0)-C(0)-A(+1)-B(0)-C(0)-A(-1)-B(0)-C(0)
[0091] If stage B is annealed while stage A is latched and stage C is inhibited, the most important signal in its environment may come from the adjacent stage A. At the end of the anneal, stage B contains the inverse content of the adjacent stage A, as follows: A(+1)-B(-1)-C(0)-A(+1)-B(-1)-C(0)-A(-1)-B(+1)-C(0)
[0092] By inhibiting stage A, holding stage B latched, and annealing stage C, we can copy the state from B to C as follows: A(0)-B(-1)-C(+1)-A(0)-B(-1)-C(+1)-A(0)-B(+1)-C(-1)
[0093] At this point in the process, stage C contains the original sequence of data from stage A, i.e. (+1, +1, -1).
[0094] As the above example situation illustrates, with an appropriate series of inhibit and / or latch operations, data can move in either direction along the shift register.
[0095] In some implementations, the interior streets 206, 208, and 214 include four or more interleaved families in each stage.
[0096] Other configurations can be used, for example some implementations include one or more "crossover" stages, which refer to places in the configuration where the phase patterns can be interrupted.
[0097] For example, the crossover stage can be a "tee." A tee can be used to redirect data around a corner, as follows: A B CPQRABC X Y Z A B C
[0098] Shift registers (A, B, C) can be operated simultaneously with shift registers (P, Q, R). Shift registers A and P can be driven by the same first signal, B and Q can be driven by the same second signal, and C and R can be driven by the same third signal, while suppressing X throughout. In this way data can be shifted from (A, B, C) on the top, through (P, Q, R), turning around at the corner, to (A, B, C) on the right side of the previous diagram.
[0099] Next, shift registers (A, B, C) can be operated simultaneously with (X, Y, Z) while suppressing P, so data can be shifted downward from (A, B, C) on the top, through (X, Y, Z) to (A, B, C) on the bottom of the previous diagram.
[0100] In each of the above cases, there are "controlled" data paths, i.e., data paths along which data is shifted, and "uncontrolled" data paths, i.e., data paths along which the contents of the shift registers are not controlled. Shift registers in uncontrolled data paths can be ignored if they are separated from the controlled paths by at least one inhibited stage (e.g., X in the first case, P in the second case).
[0101] In the exemplary implementation shown in FIG. 2, there are 12 horizontal qubit readout devices, 12 vertical qubit readout devices, and a 12×12 grid of DACs. In other implementations, the processor may have a non-square grid of DACs. In other implementations, the dimensions of the grid are not the same as the number of qubit readout devices. An exemplary implementation has 12 horizontal qubit readout devices, 12 vertical qubit readout devices, and a 28×14 grid of DACs. In other implementations, other suitable tile dimensions are used. In an exemplary implementation, the shift-register-loadable DAC has six stages.
[0102] In this application, the variable m is used to represent the maximum number of qubit readout devices, the width of the DAC grid, and the height of the DAC grid. To program a system (e.g., a system including the superconducting processor portion 100 of FIG. 1 or the superconducting processor portion 200 of FIG. 2), a state can first be loaded into the vertical street. For a three-phase street, one-third of the data lines can be transferred to the buffer stage at a time.
[0103] Each load operation can take approximately m steps, i.e., each load operation can take 0(m) steps, where m is the maximum of the total read size, the DAC grid width, and the DAC grid height. The load operation can be repeated 0(m) times to load a DAC stage, and the time it takes to program a tile is 0(m 2 Multiple tiles with the same control structure can be programmed in parallel, so the total programming time for such tiles can also be 0(m 2 ) can be used.
[0104] In one implementation, at least a subset of the tiles have the same control structure. In some implementations, the control structure can be operated in parallel. In this case, the subset of tiles has at least approximately the same programming time as a single tile. Similarly, a processor readout may take 0(m) time because the shift register length is limited in reading the processor. A single qubit readout may take 0(m) time. A single tile readout of multiple qubits may take 0(m) time because multiple qubits in a single tile can be readout in parallel. Additionally, multiple tiles readout of multiple qubits may take 0(m) time because multiple tiles can be readout in parallel.
[0105] In some implementations, the shift registers may be communicatively coupled between adjacent tiles in the processor grid to improve redundancy and, for example, reduce the risk of having an inoperative unit.
[0106] Exemplary Readout System 3 illustrates a readout system 300 for a superconducting circuit 302 according to at least one exemplary implementation. In the illustrated implementation of FIG. 3, the superconducting circuit 302 comprises one or more superconducting resonators. In one implementation, the superconducting circuit 302 comprises a superconducting quantum processor. In another implementation, the superconducting circuit 302 comprises a classical superconducting processor. In other implementations, the superconducting circuit 302 comprises a superconducting device.
[0107] The readout system 300 comprises a digital board 304 and a microwave board 306. The digital board 304 comprises a field programmable gate array (FPGA) 308, two digital-to-analog converters (DACs) 310a and 310b, and two analog-to-digital converters (ADCs) 312a and 312b. In some implementations, the digital board 304 comprises two FPGAs, one providing output to the DACs 310a and 310b, and the other providing output to the ADCs 312a and 312b. In one implementation, each of the DACs 310a and 310b may comprise a dual-channel 14-bit DAC operating at up to about 5.6 Gsps (giga samples per second). The ADCs 312a and 312b may be implemented using a multi-channel device, for example a quad-channel 10-bit ADC that can operate in dual-channel mode at up to about 2.5 Gsps.
[0108] The readout system 300 advantageously enables independent addressing of two sidebands of the Frequency Multiplexed Readout (FMR) spectrum. The complex received signal can be expressed as: x(n)=I(n)+jQ(n) where I(n) is the output of ADC 312a and Q(n) is the output of ADC 312b.
[0109] The FMR spectrum can be calculated as follows:
number
[0110] The digital board 304 further comprises two loopback lines 314a and 314b, and a sync / clock connection 316a. The loopback line 314a communicatively couples the output of the DAC 310a to the input of the ADC 312a. The loopback line 314b communicatively couples the output of the DAC 310b to the input of the ADC 312b.
[0111] The microwave board 306 (also referred to in this application as the microwave subsystem 306 ) further includes a loopback line 317 .
[0112] Loopback lines 314 a and 314 b on the digital board 304 and loopback line 317 on the microwave board 306 are optional and can be used to bypass other elements of the readout system 300 .
[0113] The readout system 300 further comprises two reconstruction filters 318a and 318b and two anti-aliasing filters 320a and 320b. The reconstruction filters 318a and 318b are low-pass analog filters that can be used to generate a band-limited analog signal from a digital input. The anti-aliasing filters 320a and 320b are low-pass analog filters that can be used to band-limit the received signal to at least approximately satisfy the sampling theorem over the band of interest.
[0114] The microwave board 306 includes a voltage controlled oscillator (VCO) / phase locked loop (PLL) 322 that provides a reference microwave signal, mixers 324 and 326, and a programmable attenuator 328. The microwave board 306 further includes amplifiers 330, 332, 334, and 336. The amplifiers 330, 332, 334, and 336 can be used to provide level control of the signal received from the superconducting circuit 302. The microwave board 306 further includes a microwave switch 338 that is controllable by a signal from the FPGA 308 on the digital board 304. In one implementation, the mixers 324 and 326 are complex mixers.
[0115] The readout system 300 further comprises an amplifier 340, attenuators 342 and 344, circulators 346 and 348, and DC blocks 350 and 352. The DC blocks 350 and 352 can be used as thermal insulation layers in the input and output lines, respectively, to and from the superconducting circuit 302.
[0116] In one implementation, amplifier 340 and attenuator 342 may operate at 4 K. Attenuator 344 may operate at 0.6 K. Circulators 346 and 348, and DC blocks 350 and 352 may operate at 8 mK.
[0117] In one exemplary implementation, using 60 resonators and a bandwidth of 2.5 GHz, a data rate of nearly 600 Mbps can be achieved with a shift register stage operation time of 25 ns.
[0118] The method of operation of the readout system 300 of FIG. 3 is described in PCT Patent Application No. WO 2016US31885 (published as International Patent Application Publication No. WO2016183213A1).
[0119] A cryogenic subsystem (not shown in FIG. 3) may be used to cool the superconducting circuit 302 to temperatures as low as a few mK (millikelvin).
[0120] Frequency multiplexed readout (FMR) technique for superconducting qubits 4 illustrates at least one exemplary implementation of a hybrid computing system 400 that includes a digital computer 402 that may incorporate FMR techniques as described above, and a quantum computer 404. The digital computer 402 is also referred to as a digital processor in this application.
[0121] Digital computer 402 includes a CPU 406, user interface elements 408, 410, 412, and 414, a disk 416, a controller 418, a bus 420, and a memory 422. Memory 422 includes a BIOS 424, an operating system 426, a server module 428, a computation module 430, a quantum processor module 432, a readout module 434, and other modules that can be used to operate hybrid computing system 400.
[0122] Although digital computer 402 will sometimes be referred to in the singular herein, this is not intended to limit its application to a single digital computer. The system and method may also be practiced in a distributed computing environment where tasks or sets of instructions are performed or executed by remote processing devices that are linked through a communications network. In a distributed computing environment, computer-readable and / or processor-readable instructions (sometimes known as program modules), application programs and / or data may be stored in local and / or remote memory storage devices (e.g., computer-readable and / or processor-readable persistent media).
[0123] Quantum computer 404 includes a quantum processor 436, a readout control system 438, a qubit control system 440, and a coupler control system 442. Quantum computer 404 may incorporate FMR technology including one or more superconducting resonators. Computing system 400 may include a readout system, such as readout system 300 of FIG.
[0124] Figure 5 is a flow diagram of an exemplary method 500 for programming a hybrid computing system, such as computing system 400 of Figure 4. Method 500 includes operations 502-516, although one of ordinary skill in the art will appreciate that certain operations may be omitted and / or additional operations may be added in alternative implementations. One of ordinary skill in the art will appreciate that the order of operations is presented for illustrative purposes only and may be altered in alternative implementations.
[0125] At 502, method 500 begins. At 504, a processor-based system, e.g., a digital processor-based system, programs a quantum processor (e.g., quantum processor 436 of FIG. 4). Programming the quantum processor may include programming multiple tiles of the quantum processor in parallel. For example, as shown in FIG. 5, programming the quantum processor may include programming a first tile at 506 and programming a second tile in parallel at 508. The duration of programming the first tile may be at least approximately the same as the duration of programming the second tile. In this context, the term "approximately" refers to the duration of programming the first tile being within 10% of the duration of programming the second tile. Programming the quantum processor may occur in preparation for computation by the quantum processor (e.g., quantum annealing) and / or during calibration of the quantum processor.
[0126] At 510, the processor-based system reads out the qubits of the quantum processor, e.g., via components of the quantum processor-based system. The readout may occur, e.g., after computation by the quantum processor (e.g., after quantum annealing) and / or during calibration of the quantum processor. Reading out the qubits of the quantum processor may include reading out multiple qubits in a single tile in parallel and / or reading out multiple qubits in multiple tiles in parallel. For example, as shown in FIG. 5, reading out the qubits may include reading out a qubit in a first tile at 512 and in parallel reading out a qubit in a second tile at 514. At 516, method 500 ends.
[0127] The duration of reading out the qubit in the first tile may be at least approximately the same as the duration of reading out the qubit in the second tile, where in this context the term "approximately" refers to the duration of reading out the qubit in the first tile being within 10% of the duration of reading out the qubit in the second tile.
[0128] The systems and methods described in this application allow for scalable integration of FMRR technology into the structure of a quantum processor. In conventional applications of FMRR technology, FASTR detectors are typically placed around the quantum processor. The time required to move data from quantum devices (e.g., qubits) in the processor to the FASTR detectors at the periphery of the quantum processor can create an inherent readout latency. This latency can be proportional to the number of shift register stages between the quantum device and the FASTR detector. The quantum device with the longest latency sets a lower bound on the overall readout time of the quantum processor. When FMRR technology is placed around the quantum processor, the dimensions (e.g., width) of the quantum processor can establish a lower bound on the readout time of the quantum processor.
[0129] The techniques described in this application involve integrating FASTR detectors with quantum devices (eg, qubits) within a quantum processor more tightly than conventional configurations.
[0130] In this application, the technique includes a construct called a tile (also referred to in this application as a unit tile). The tile can have a substantially constant latency and a substantially constant readout time. Each grid of tiles can have the same total latency and the same total readout time as an individual tile.
[0131] An advantage of this technique is that it provides a scalable readout system. Depending on the implementation, the readout system can be varied without incurring run-time costs that vary with the number of devices being read out. In one implementation, varying the readout system is limited by the number of FASTR detectors that can be attached to a single microwave line and the number of microwave lines in the readout system.
[0132] The systems and methods described herein integrate a FASTR detector with a DAC that can load states from a shift register. In conventional methods, programming a quantum processor can use a large number of analog lines to control a grid of DACs. See, for example, U.S. Patent Application No. 15 / 726,239 (also published as U.S. Patent Application Publication No. 20180101786A1). In practice, the number of available analog lines may be constrained. In one exemplary implementation, 185 analog lines may be used to program the DACs in the quantum processor. By using fewer analog lines, programming time can be increased.
[0133] Making changes to conventional technology can be very complicated (e.g., if a relatively large number of analog lines need to be added) or very time consuming (e.g., if the change increases programming time).
[0134] The systems and methods described herein can program and control large processors using a reduced number of analog lines compared to conventional input / output techniques. In one exemplary implementation, 10-30 analog lines can be used to program the DACs in the quantum processor. One implementation includes a collection of tiles, each with a respective set of shift registers that can be collectively controlled by a single set of analog lines. The programming time of a tile can be substantially fixed, and the programming time of a collection of tiles can be substantially the same as the programming time of a single tile. The scalability of a system including a quantum processor can be determined at least in part by the number of FASTR detectors that can be attached to a single microwave line and by the number of microwave lines in the system.
[0135] The various embodiments described above can be combined to realize further embodiments. 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 Sheet include, but are not limited to, the following: PCT Patent Application No. PCT / US2019 / 18792, filed February 20, 2018, entitled "SYSTEMS AND METHODS FOR COUPLING A SUPERCONDUCTING TRANSMISSION LINE TO AN ARRAY OF RESONATORS"; PCT Patent Application No. PCT / US2019 / 18792, filed May 11, 2016, entitled "FREQUENCY MULTIPLEXED RESONATOR INPUT AND / OR OUTPUT FOR A SUPERCONDUCTING TRANSMISSION LINE TO AN ARRAY OF RESONATORS"; PCT Patent Application No. PCT / US2019 / 18792, filed May 11, 2016, entitled "SYSTEMS AND METHODS FOR COUPLING ... No. PCT / US2016 / 031885 (also published as International Patent Application Publication No. WO2016183213A1), entitled “SYSTEMS AND METHODS FOR SUPERCONDUCTING FLUX QUBIT READOUT”, granted on October 7, 2014, and U.S. Patent No. 8,169,231, entitled “SYSTEMS, METHODS, AND APPARATUS FOR QUBIT STATE READOUT”, granted on May 1, 2012, both of which are incorporated herein by reference in their entirety, and which are incorporated herein by reference in their entirety.No. 15 / 726,239, filed October 5, 2017 (also published as U.S. Patent Application Publication No. US20180101786A1), entitled “METHODS AND APPARATUS FOR EFFICIENT INPUT AND OUTPUT TO QUANTUM PROCESSORS,” and U.S. Provisional Patent Application No. 62 / 851,377, filed May 22, 2019, entitled “METHODS AND APPARATUS FOR EFFICIENT INPUT AND OUTPUT TO QUANTUM PROCESSORS,” and U.S. Provisional Patent Application No. 62 / 851,377, filed June 11, 2019, ... No. 62 / 860,098, entitled "A HIGH-SPEED MICROPHASE DEVICE," all of which are incorporated herein by reference. If necessary, aspects of each embodiment can be modified to utilize the systems, circuits, and concepts of the various patents, applications, and publications to realize further embodiments.
[0136] These and other changes 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 scope of the claims to the specific embodiments disclosed in the specification and claims, but should be construed to include all possible embodiments along with the full scope of equivalents to which such claims are entitled. Thus, the claims are not limited by this disclosure.
Claims
1. 1. A quantum processor comprising a plurality of tiles, the plurality of tiles being arranged in a first grid, a first tile of the plurality of tiles comprising: A first qubit; and a shift register including at least one shift register stage communicatively coupled to a frequency multiplexed resonant (FMR) readout device; a qubit readout device; and a plurality of digital-to-analog converter (DAC) buffer stages; a plurality of shift register loadable DACs arranged in a second grid; A quantum processor comprising:
2. The quantum processor of claim 1 , wherein the first qubit is communicatively coupled to the qubit readout device.
3. 2. The quantum processor of claim 1, wherein at least one of the plurality of DAC buffer stages is communicatively coupled to at least one of the plurality of shift register loadable DACs.
4. The quantum processor of claim 1 , wherein the first qubit is a superconducting flux qubit.
5. 10. The quantum processor of claim 1, further comprising a microwave transmission line, the microwave transmission line communicatively coupled to an FMR readout device.
6. The quantum processor of claim 1 , wherein the FMR readout device comprises a superconducting resonator.
7. The quantum processor of claim 1 , wherein the first grid and the second grid are interspersed with one another on a superconducting integrated circuit.
8. 10. The quantum processor of claim 1, wherein the FMR readout device and the multiple shift register loadable DACs reside on the same superconducting integrated circuit.
9. a transmission line including at least one transmission line inductance; A superconducting resonator; a coupling capacitance communicatively coupling the superconducting resonator to the transmission line; The quantum processor of claim 1 further comprising:
10. 10. The quantum processor of claim 1, wherein the FMR readout device is communicatively coupled to at least one other element of the quantum processor by at least one of a superconducting via, a bump bond, and a solder bond.
11. a first superconducting integrated circuit including a quantum processor, the quantum processor including a plurality of superconducting flux qubits; A second superconducting integrated circuit includes an input / output system, said input / output system comprising: a shift register including at least one shift register stage communicatively coupled to a frequency multiplexed resonant readout (FMRR) module; Qubit readout device, A plurality of DAC buffer stages; and A plurality of shift register loadable digital-to-analog converters (DACs) arranged in a grid a second superconducting integrated circuit comprising: A quantum computer equipped with
12. 1. A method of programming a hybrid computing system, the hybrid computing system comprising: a quantum processor and a digital processor, the quantum processor comprising a plurality of tiles, the plurality of tiles being arranged in a first grid; a first tile of the plurality of tiles; A first qubit; and a shift register including at least one shift register stage communicatively coupled to a frequency multiplexed resonant (FMR) readout device; a qubit readout device; and a plurality of digital-to-analog converter (DAC) buffer stages; a plurality of shift register loadable DACs arranged in a second grid; Equipped with The method further comprising: programming the quantum processor with the digital processor; reading out the quantum processor by the digital processor; A method comprising:
13. The method of claim 12 , wherein the programming of the quantum processor by the digital processor comprises programming the first tile of the plurality of tiles.
14. 14. The method of claim 13, wherein said programming said first tile of said plurality of tiles comprises programming at least one of said plurality of shift register loadable DACs.
15. The method of claim 12 , wherein the programming of the quantum processor by the digital processor further comprises programming a second tile of the plurality of tiles.
16. 16. The method of claim 15, wherein the programming the first tile and the second tile of the plurality of tiles comprises programming the first tile and the second tile of the plurality of tiles in parallel.
17. 17. The method of claim 15, wherein programming the first tile of the plurality of tiles has a first duration and programming the second tile of the plurality of tiles has a second duration, the first duration being at least approximately the same as the second duration.
18. 13. The method of claim 12, wherein the reading out of the quantum processor by the digital processor comprises reading out the first qubit of the first tile of the plurality of tiles in parallel with a second qubit of the first tile of the plurality of tiles.
19. 19. The method of claim 12, wherein the reading out of the quantum processor by the digital processor comprises reading out the first qubit of the first tile of the plurality of tiles in parallel with a third qubit of a second tile of the plurality of tiles.
20. 13. The method of claim 12, wherein the reading out of the quantum processor by the digital processor includes reading out the first qubit of the first tile of the plurality of tiles and reading out a second qubit of a second tile of the plurality of tiles, and wherein a first time taken to read out the first qubit of the first tile of the plurality of tiles is at least about the same as a second time taken to read out the second qubit of the second tile of the plurality of tiles.
21. 13. The method of claim 12, further comprising calibrating at least a first portion of the quantum processor and a second portion of the quantum processor in parallel operation, the first portion of the quantum processor being non-local with respect to the second portion of the quantum processor.
Citation Information
Patent Citations
Frequency multiplexing read-out device and design method of the same
JP2018194518A
Assemblies and methods of construction for quantum computing devices
JP2018524667A
Frequency multiplexed resonator input and / or output for a superconducting device
US20180145631A1
Systems and methods for superconducting devices used in superconducting circuits and scalable computing
WO2017192733A2
