Demonstration of quantum volume 64 on a superconducting quantum computing system
Petar Jurcevic, Ali Javadi-Abhari, Lev S. Bishop, Isaac Lauer, Daniela F. Bogorin, Markus Brink, Lauren Capelluto, Oktay Günlük, Toshinari Itoko, Naoki Kanazawa, Abhinav Kandala, George Keefe, Kevin Krsulich, W. Landers, Eric P. Lewandowski, Douglas McClure, Giacomo Nannicini, Adinath Narasgond, Hasan M. Nayfeh, Emily Pritchett, Mary Beth Rothwell, Srikanth J. Srinivasan, Neereja Sundaresan, Cindy Wang, Ken Xuan Wei, Christopher J. Wood, Jeng-Bang Yau, Eric Zhang, Oliver Dial, Jerry M. Chow, Jay Gambetta
Abstract We improve the quality of quantum circuits on superconducting quantum computing systems, as measured by the quantum volume (QV), with a combination of dynamical decoupling, compiler optimizations, shorter two-qubit gates, and excited state promoted readout. This result shows that the path to larger QV systems requires the simultaneous increase of coherence, control gate fidelities, measurement fidelities, and smarter software which takes into account hardware details, thereby demonstrating the need to continue to co-design the software and hardware stack for the foreseeable future.
Abstract We improve the quality of quantum circuits on superconducting quantum computing systems, as measured by the quantum volume (QV), with a combination of dynamical decoupling, compiler optimizations, shorter two-qubit gates, and excited state promoted readout. This result shows that the path to larger QV systems requires the simultaneous increase of coherence, control gate fidelities, measurement fidelities, and smarter software which takes into account hardware details, thereby demonstrating the need to continue to co-design the software and hardware stack for the foreseeable future.
Quantum volume (QV) is a widely recognized metric for assessing the practical capabilities of quantum computers, as it provides an estimate of the largest quantum circuit that can be reliably executed. However, measuring QV on a real device requires comparing experimental outcomes with ideal theoretical results—a process that rapidly becomes computationally expensive. By examining the cumulative impact of errors in two-qubit gates, we present a simple, accessible `rule of thumb’ that relates the quantum volume directly to the average error rate of native gates. Our formula shows a strong agreement with experimental data from leading quantum computing platforms, including both superconducting and trapped-ion systems. This straightforward model offers a clear, intuitive guideline for predicting quantum hardware performance, enabling more informed decisions regarding circuit design and resource allocation.
Abstract As quantum processors grow in scale and reliability, the need for efficient quantum gate decomposition of circuits to a set of specific available gates, becomes ever more critical. The decomposition of a particular algorithm into a sequence of these available gates is not unique. Thus, the fidelity of an algorithm’s implementation can be increased by choosing an optimized decomposition. This is true both for noisy intermediate-scale quantum platforms as well as for implementation of quantum error correction schemes. Here we present a compilation scheme which implements a general-circuit decomposition to a sequence of Ising-type, long-range, multi-qubit (MQ) entangling gates, that are separated by layers of single qubit rotations. We use trapped ions as an example in which MQ gates naturally arise, yet any system that has connectivity beyond nearest-neighbors may gain from our approach. We evaluate our methods using the quantum volume (QV) test over N qubits. In this context, our method replaces 3 N 2 / 2 two-qubit gates with 2 N + 1 MQ gates. Furthermore, our method minimizes the magnitude of the entanglement phases, which typically enables an improved implementation fidelity, by using weaker driving fields or faster realizations. We numerically test our compilation and show that, compared to conventional realizations with sequential two-qubit gates, our compilations improves the logarithm of QV by 20% to 25%.
As the number of qubits in quantum computing increases, the scalability of existing qubit circuit structures and control systems may become insufficient for large-scale expansion and high-fidelity control. To address this challenge, we propose a behavioral-level model of a superconducting qubit and its control electronics, followed by a co-simulation to evaluate their performance. In this paper, we present the modeling process, simulation procedure, and resulting design specifications for the qubit control system. Our co-simulation approach utilizes MATLAB and Simulink, enabling us to derive critical circuit design specifications, such as the required Digital-to-Analog Converter (DAC) resolution, which should be 8 bits or higher, to achieve high-fidelity control. By taking into account factors such as DAC sampling rates, integral and differential nonlinearities, and filter characteristics, we optimize the control system for efficient and accurate qubit manipulation. Our model and simulation approach offer a promising solution to the scalability challenges in quantum computing, providing valuable insights for the design of large-scale superconducting quantum computing systems.
Quantum volume (QV) has become the de-facto standard benchmark to quantify the capability of Noisy Intermediate-Scale Quantum (NISQ) devices. While QV values are often reported by NISQ providers for their systems, we perform our own series of QV calculations on 24 NISQ devices currently offered by IBM Q, IonQ, Rigetti, Oxford Quantum Circuits, and Quantinuum (formerly Honeywell). Our approach characterizes the performances that an advanced user of these NISQ devices can expect to achieve with a reasonable amount of optimization, but without white-box access to the device. In particular, we compile QV circuits to standard gate sets of the vendor using compiler optimization routines where available, and we perform experiments across different qubit subsets. We find that running QV tests requires very significant compilation cycles, QV values achieved in our tests typically lag behind officially reported results and also depend significantly on the classical compilation effort invested.
Current quantum computers suffer from noise that stems from interactions between the quantum system that constitutes the quantum device and its environment. These interactions can be suppressed through dynamical decoupling to reduce computational errors. However, the performance of dynamical decoupling depends on the type of system-environment interactions that are present, which often lack an accurate model in quantum devices. We show that the performance of dynamical decoupling can be improved by optimizing its rotational gates to tailor them to the quantum hardware. We find that, compared to canonical decoupling sequences, such as CPMG, XY4, and UR6, the optimized dynamical decoupling sequences yield the best performance in suppressing noise in superconducting qubits. Our work thus enhances existing error suppression methods, which helps increase the circuit depth and result quality on noisy hardware.
Superconducting quantum computing hardware relies on classical radio-frequency (RF) electronic circuits to control and measure quantum states. As the complexity of quantum computers increases, the significance of the classical circuitry component grows, necessitating the development of accurate models for its design and optimization. This study presents an analytical model that quantifies the influence of circuit nonidealities on readout fidelity in superconducting quantum computing systems. This model considers a wide range of nonidealities typically encountered in the readout chain, encompassing frequency, amplitude, phase inaccuracies, impedance mismatch, quantum noise, and amplifier noise. It further predicts the combined impact of these nonidealities on fidelity. The model’s precision and efficacy are confirmed through numerical quantum-classical co-simulation. The highest observed relative mismatch is roughly 3.5%, but most errors remain below 1%. Such a model can significantly aid in designing and optimizing practical quantum computers.