Quantum Controller Achieves 99.9% Accuracy in Qubit Operations

Scientists at the University of Chinese Academy of Sciences in collaboration with University of Chinese Academy of Sciences, Tokyo University of Science and RIKEN Center for Quantum Computing (RQC), led by Kuang Liu, have developed a new superconducting quantum controller enabling high-fidelity qubit manipulation at millikelvin temperatures. This controller demonstrates a capability for direct chip-to-chip interconnection with qubits and operates using an all-digital methodology. Randomised benchmarking reveals an average Clifford fidelity exceeding 99.9% with minimal leakage, and an estimated gate operation energy of 0.121 fJ. The controller addresses a key bottleneck currently limiting the development of larger, more powerful superconducting quantum processors, paving the way for more complex quantum computations.

Single-flux-quantum pulses and passive bias networks enable precise qubit manipulation

Superconducting circuits, built around components called Josephson junctions, form the core of this new control system. These junctions, exhibiting quantum mechanical properties such as the Josephson effect, allow for incredibly fast, low-power operation at cryogenic temperatures, typically below 4 Kelvin. The principle relies on the tunnelling of Cooper pairs, pairs of electrons, across a thin insulating barrier within the junction, creating a non-dissipative current. The generation of single-flux-quantum (SFQ) pulses is central to the controller’s success; these tiny packets of magnetic flux represent a single quantum of magnetic flux and, crucially, a single quantum of voltage when interacting with a superconducting circuit. This quantisation is fundamental to the digital nature of the control system, allowing for precise and discrete manipulation of the qubits.

This digital approach ensures clean and reliable signals, minimising errors during qubit operations, much like sending precise instructions to a computer. Unlike analogue control signals which are susceptible to noise and distortion, digital signals offer inherent robustness. A fully passive superconducting bias network is incorporated into the design, a circuit that eliminates static electrical noise and the creation of unwanted quasiparticles. Quasiparticles are energy excitations within the superconductor that can disrupt qubit coherence by introducing energy dissipation and decoherence. The passive nature of the network means it requires no external power source, further minimising noise and improving stability. A superconducting quantum controller operating at 10 millikelvins has been developed, a temperature vital for maintaining qubit coherence as thermal fluctuations are significantly reduced at such low temperatures, extending the lifetime of quantum information.

Utilising superconducting circuits and Josephson junctions, the controller enables direct connection with qubits and all-digital manipulation, avoiding the wiring complexity of previous systems which often relied on numerous coaxial cables and attenuators, introducing signal loss and noise. Randomized benchmarking demonstrated a high average Clifford fidelity of 99.9% with minimal leakage to higher energy levels, measuring approximately 10 to the power of negative four. This indicates a very low probability of the qubit transitioning to an unintended state during a gate operation. The controller achieves this performance with an average gate operation energy of 0.121 femtojoules, addressing a key limitation in scaling superconducting quantum computers; reducing energy dissipation is crucial for building large-scale systems as heat management becomes increasingly challenging.

Demonstrated high-fidelity qubit control and reduced energy dissipation via a novel superconducting

This new superconducting quantum controller reduced error rates to 0.1%, a substantial improvement over previous limitations. This performance, demonstrated through randomised benchmarking, achieves a uniformly high average Clifford fidelity of 99.9%, surpassing the critical threshold for viable quantum error correction. Quantum error correction is essential for building fault-tolerant quantum computers, as qubits are inherently susceptible to noise and decoherence. Maintaining qubit coherence long enough for complex calculations previously proved insurmountable due to environmental noise and control inaccuracies; this controller significantly extends coherence times. The fidelity achieved is a significant step towards implementing effective error correction schemes.

The system also exhibits an estimated average gate operation energy of 0.121 fJ, representing a strong step towards scalable quantum computing, engineered with a fully passive superconducting bias network. Measurements confirmed the controller generates flat-top enveloped signals with arbitrary duration, essential for coherent qubit control at millikelvin temperatures. Precise control over pulse shape and duration is critical for accurately addressing and manipulating the qubits. Pulse sequences with lengths of 80, 160, and 300 nanoseconds, at a clock frequency of 2.4 gigahertz, demonstrated this capability. Rabi experiments, used to verify drive coherence, produced a clear chevron pattern when sweeping both the control signal frequency and pulse duration, indicating strong and precise qubit manipulation. The chevron pattern confirms that the control signal is effectively driving the qubit transitions. Furthermore, Ramsey interference experiments, employing sequences of XSQC/2 gates separated by variable intervals, successfully generated interference fringes, validating the controller’s ability to perform accurate phase control; the system achieved orthogonal rotations near the subharmonic condition of ω01/N, demonstrating precise control over the qubit’s phase evolution. These experiments confirm the controller’s ability to perform the fundamental operations required for quantum computation.

Advancing qubit control through high-fidelity superconducting digital circuitry

Scientists are steadily improving the tools needed to build practical quantum computers, but scaling up these systems presents formidable engineering challenges. These challenges include managing the increasing complexity of wiring, reducing heat dissipation, and maintaining qubit coherence. This new superconducting quantum controller, offering precise digital manipulation of qubits, tackles the vital issue of reliably directing and reading information from these delicate quantum states. The work acknowledges a broader field of control methods, specifically referencing ongoing work optimising two-qubit gates using similar single-flux-quantum techniques. Two-qubit gates are essential for creating entanglement, a key resource for quantum computation.

This development represents a major step forward, even as alternative control methods, including optimisation of two-qubit gates utilising these techniques, are actively being refined. Achieving 99.9% fidelity and exceptionally low energy consumption directly addresses a key bottleneck hindering the construction of larger, more stable quantum processors. This level of fidelity opens questions regarding the scalability of this control method to multi-qubit systems and the development of techniques for managing the increased complexity of larger quantum processors. Future research will likely focus on integrating multiple controllers to address larger qubit arrays and developing automated calibration procedures to maintain high fidelity across the entire system. The ability to reliably control and interconnect many qubits is crucial for realising the full potential of quantum computing.

The researchers demonstrated a superconducting quantum controller capable of high-fidelity, all-digital manipulation of qubits at 10 mK, achieving an average Clifford fidelity of 99.9% with low energy consumption of 0.121 fJ. This is significant because efficient in-situ control methods are essential for scaling up superconducting quantum computers and overcoming current engineering limitations. The controller enables direct chip-to-chip interconnection, addressing a key bottleneck in the field. Authors suggest future work will focus on integrating multiple controllers and automated calibration for larger qubit arrays.