Session A42: Multi-Qubit Characterizations and Cross-talk For Superconducting Qubits

A blueprint for demonstrating quantum supremacy with superconducting qubits

Here, using nine superconducting qubits, we experimentally illustrate a blueprint for demonstrating quantum supremacy. We present data characterizing two basic ingredients required for any supremacy experiment: complexity and fidelity. First, we demonstrate that the qubits can uniformly explore the Hilbert-space, providing a direct measure of algorithm complexity. Next, we compare the measurement results with the expected behavior and show that the algorithm can be implemented with high fidelity. Experiments for probing complexity and fidelity provide a foundation for demonstrating quantum supremacy.   

Operating and Characterizing of a 72 Superconducting Qubit Processor “Bristlecone”: Part 1

Quantum computing has recently entered a new regime where a number of quantum processors with tens of qubits are being built and operated around the world. With approximately 50 qubits and high performance, it is possible to outperform classical computers in limited cases, a feat known as “Quantum Supremacy.” However, achieving supremacy requires high-fidelity operations in all aspects of the algorithm, including single-qubit gates, two-qubit gates, and readout. While these milestones have been achieved in small scale devices, challenges in calibration, precision control, and reproducibility are amplified at scale. In this talk, we will discuss the 72 qubit “Bristlecone” processor we have been developing at Google, and our work towards calibrating, operating, and characterizing this device. Part 1 of 2.   

Operating and Characterizing a 72 Superconducting Qubit Processor “Bristlecone”: Part 2

Quantum computing has recently entered a new regime where a number of quantum processors with tens of qubits are being built and operated around the world. With approximately 50 qubits and high performance, it is possible to outperform classical computers in limited cases, a feat known as “Quantum Supremacy.” However, achieving supremacy requires high-fidelity operations in all aspects of the algorithm, including single-qubit gates, two-qubit gates, and readout. While these milestones have been achieved in small scale devices, challenges in calibration, precision control, and reproducibility are amplified at scale. In this talk, we will discuss the 72 qubit “Bristlecone” processor we have been developing at Google, and our work towards calibrating, operating, and characterizing this device. Part 2 of 2.   

Characterization and mitigation of noise and crosstalk in a five-qutrit transmon processor

The multi-level structure of the transmon circuit makes it an attractive choice for implementing multi-valued quantum logic, i.e. qudit-based protocols with d > 2. To date, while transmon levels outside of the qubit computational subspace have been used for control and sensing, experimental implementation of a quantum algorithm using multi-level transmons has not been demonstrated. Specifically, we address practical issues which arise when running a multi-qutrit algorithm. These include classical microwave crosstalk between the individual transmon control lines; an always-on ZZ interaction due to fixed off-resonant coupling between neighboring qutrits; conditionality of single-qutrit pulses due to the cross-resonance effect; and challenges due to the sensitivity of higher transmon levels to charge noise. For each of these issues, we present methods of characterization and mitigation, demonstrating their efficacy on our hardware and the future design choices they inform.   

Automatic calibration of arrays of superconducting qubits and couplers

High-fidelity operation of superconducting qubits requires fine-tuning numerous experimental parameters. As superconducting qubit systems scale up to large arrays, automatic calibration becomes essential, both for bringing up systems from scratch and for maintaining calibrations. These algorithms depend strongly on the qubit architecture. For example, consider arrays of qubits connected by adjustable couplers, where the coupling can be turned off to isolate qubits and turned on to facilitate fast interactions. These couplers substantially complicate automatic calibration. At the start, the coupling may inadvertently be very strong, which can disrupt qubit bringup, and we can only learn about the coupler configuration indirectly through qubit measurements. In this presentation, we describe algorithms and experiments to automatically calibrate superconducting qubits and couplers.   

Coupler characterization of transmons for cross-resonance

We characterize two-qubit cross-resonance gates and unintended residual coupling on various coupled qubit arrangements. Direct coupling versus coupling via a quantum bus are studied on the basis of cross-resonance gate rate and fall-off of non-nearest neighbor coupling. We experimentally extract coupling rates using Hamiltonian tomography methods, and compare with microwave simulations.   

Deletrious Effects of Spectator Qubits in Multiqubit Circuits

For a number of quantum technologies, such as superconducting qubits, multiqubit algorithms are performed by concatenating a series of one- and two-qubit gates. However, in such an architecture there will be no true one- and two-qubit gates because of spectator qubits, i.e., qubits which are connected, but should be idle during the gate. For drive activated two-qubit gates, such as the cross-resonance gate, the role of these spectator qubits can be non-trivial. Here we will discuss our studies on these spectator qubits, for example, by measuring the entanglement dynamics of the spectator qubit with the active qubits. Understanding spectator dynamics will lead to improved designs of multiqubit gates for quantum computing technologies.   

Measuring and quantifying classical crosstalk in multi-qubit superconducting circuits

In the most general sense, crosstalk is used to describe the unwanted interactions on elements of a quantum processor due to targeted drives on non-local elements. For superconducting circuits, we use "classical crosstalk” specifically to describe crosstalk that is a result of the microwave environment of the chip, and is not a direct consequence of the qubit frequency arrangement or the drive Hamiltonians. Such crosstalk can be an important source of gate infidelity. Here, we shall present techniques to measure and quantify classical crosstalk, and use them to probe the microwave environment of our chips. These techniques could serve as a useful guide for the design and packaging of large multi-qubit devices.
  

Multi-qubit circuit characterization through physics-based statistical inference

We develop new physical models for realistic two-qubit gates in superconducting qubit architectures to account for specific shape of qubit control pulses and effects of noise and decoherence, including those induced by two-level defects prevalent in the fabrications of superconducting chip. We then formulate statistical inference method to efficiently estimate the physical model parameters for an ensemble of quasi-random circuits that contain non-Clifford gates. For a family of circuit ensembles, we are able to obtain analytically the circuit fidelities and their variances by averaging the log-likelihood distribution of the model parameters over Haar measure. Our inference method applies to quantum systems with arbitrary number of qubits and thus serves as valuable tools for characterizing large scale quantum circuit that will likely outperform the most powerful classical computers existing to date.   

Mode Hybridization Analysis of Bus Resonators for a Superconducting Multi-Qubit Chip

We present an effective numerical method to analyze the mode hybridization in a multi-transmon circuit QED chip. Surface code, a promising architecture for fault-tolerant quantum computing, requires qubits with connectivity to all nearest neighbors. This extensive interconnectivity together with strong coupling between qubits and resonators causes mode hybridization. A complete analysis of the chip is needed in these conditions to accurately predict the loaded frequency of the bus resonators and thereby also the two-qubit gate time. We present and experimentally verify a simulation method for analyzing the complete chip combining finite-element electromagnetic simulation with numerical circuit simulation for accurate and fast computation. This research is funded by Intel Corporation and IARPA (U.S. Army Research Office grant W911NF-16-1-0071).   

A many-body coupler for coherent 4-local interaction of superconducting flux qubits

Interactions of more than two bodies simultaneously rarely appear in nature. Even in dense systems, forces usually act pairwise. However, n-local terms with n>3 frequently arise when mapping physical or mathematical problems to an Ising spin Hamiltonian. Superconducting flux qubits have emerged as a promising, versatile platform to simulate such spin systems and find their ground state, but the implementation of n-local superconducting qubit couplers has so far proven elusive. Here we present a circuit that enables large 4-local interaction between four flux qubits without spurious 2-local terms and without relying on an effective low-energy description as in Hamiltonian gadgets. We demonstrate numerically how 4-local coupling of up to several hundred MHz arises from a coupler circuit with tailored spectral properties. These properties are engineered by combining known superconducting circuit physics with suggestions by an inverse design algorithm that we have developed previously. In addition, we show first experimental results probing a fabricated coupler prototype.   

Robust And Versatile Superconducting Coupler

Having access to a large collection of fast and high-fidelity two-qubit gates can significantly reduce the gate count of large-scale quantum computations. In this talk, we present a novel tunable superconducting dipole coupling element that enables a variety of strong two-qubit interactions by modulating external fields. The proposed device can realize both high-fidelity and fast SWAP and controlled two-qubit gates.   

Implementing optimized time-varying coupling and dissipation in small logical qubit architectures.

Tunable couplings between high coherence quantum objects and lossy resonators is a promising approach to state stabilization. Using an additional lossy resonator to provide a mechanism for the fast reset of the first resonator, a steady state residual error scaling of about 1 / T₁ can be achieved. While this technique has been demonstrated for a single qubit coupled to a single resonator, this work looks at applying the same resonator reset technique for an idealized three qubit bit flip code and the Very Small Logical Qubit (VSLQ), a promising route to passive error correction in superconducting architectures. Optimal device and signal parameters for the VSLQ are well understood for a numerically optimized, continuous coupling strength, so we explore the effects of using an optimized, time-varying coupling strength between the primary qubits and shadow resonators and report a best possible scaling logical qubit lifetime T_L that scales as T₁², where T₁ is the lifetime for a single primary qubit.