Reset scheme achieves over 99% fidelity with transmon qubits

Researchers at the Department of Microtechnology and Nanoscience, Chalmers University of Technology, Gothenburg, Sweden; Department of Chemistry, Princeton University, Princeton, NJ, USA; Department of Electrical and Computer Engineering, Princeton University, Princeton, NJ, USA; and VTT Technical Research Centre of Finland, FI-02044 VTT, Finland have demonstrated a protocol for fixed-frequency transmon qubits, an architecture compatible with the surface code, that simultaneously addresses both qubit reset and leakage reduction. This combined capability is desirable for successful quantum error correction.

The authors state that experiments involved a pair of qubits. This work reports a complete cycle of qubit reset, leakage reduction, and coupler reset in 83 nanoseconds, enabling fixed-frequency qubit architectures as potential building blocks for future fault-tolerant quantum computers and offering a means to reduce error correction cycle runtime.

Tunable Couplers Enable Fast Qubit Reset and Leakage Reduction

Over 99% fidelity in qubit reset and leakage reduction has been demonstrated using a novel protocol with fixed-frequency transmon qubits, a result that directly addresses a critical bottleneck in building practical quantum computers. This approach allows for the swift transfer of unwanted energy from qubits to a readout resonator, where it dissipates into the feedline, effectively resetting the qubit state. The architecture employed is specifically designed for compatibility with the surface code, a leading candidate for fault-tolerant quantum computing, pairing fixed-frequency transmon qubits with these tunable couplers.

Unlike many existing reset schemes that require additional hardware or complex control signals, this protocol operates within the constraints of current fixed-frequency qubit technology. The team’s design utilizes the tunable couplers to implement a qubit-coupler (QC) SWAP gate, initially tuning the coupler on resonance with the ancilla qubit, Q_0, while leaving the data qubit, Q_1, unaffected.

This initial step is crucial for preparing the system for subsequent energy transfer and dissipation. This speed is essential for minimizing the impact of qubit decoherence and maximizing the efficiency of quantum error correction cycles. The researchers detail that their adiabatic reset can simultaneously move both the |1⟩ and |2⟩ state populations to the |0⟩ state, achieving an error of (7.87 ± 1.94)× 10^-3 within 61 nanoseconds.

The leakage reduction unit (LRU) effectively removes the |2⟩ state population without disturbing the computational subspace of the data qubits, with an error rate of 9.50× 10^-3 in just 5 nanoseconds. The experimental work was conducted on a 25-qubit device, focusing on a two-qubit subset where a pair of qubits are involved in the experiments.

The team, from the Department of Microtechnology and Nanoscience, Chalmers University of Technology, Gothenburg, Sweden; Department of Chemistry, Princeton University, Princeton, NJ, USA; Department of Electrical and Computer Engineering, Princeton University, Princeton, NJ, USA; and VTT Technical Research Centre of Finland, FI-02044 VTT, Finland, detail that the qubit-to-coupler coupling rates, are 40 MHz for Q_0 and 60 MHz for Q_1, while the qubit-resonator coupling rates, are around 50 MHz. As described in the paper, “Fast unconditional reset and leakage reduction in fixed-frequency transmon qubits,” the team’s approach minimizes impact on other qubits on the chip and requires no additional hardware resources. This streamlined design is a significant advantage for scaling up to larger, more complex quantum processors, and offers a means to both reduce error correction cycle runtime and improve algorithmic fidelity on quantum computers.

Fixed-Frequency Transmon Qubits for Surface Code Compatibility

Fixed-frequency transmon qubits are gaining traction as a viable architecture for building scalable quantum computers, particularly those leveraging the surface code for error correction. Unlike tunable qubits requiring complex control schemes, fixed-frequency designs simplify fabrication and control electronics, easing the path toward larger processor sizes. This high accuracy is critical, as maintaining qubit coherence during error correction cycles is paramount to reliable computation. The demonstrated protocol hinges on the use of tunable couplers positioned between qubit pairs.

These couplers act as intermediaries, transferring unwanted qubit excitation to the readout resonator associated with each qubit, where the energy dissipates into the feedline. This approach avoids the need for complex, individual qubit control during reset and leakage reduction, streamlining the process and reducing potential error sources. The team’s design specifically addresses the challenges of both ancilla qubit reset and data qubit leakage reduction, two intertwined issues in surface code implementation.

99% Fidelity Achieved with Passive Reset Protocols

The team’s design addresses a critical bottleneck in quantum computation: maintaining qubit coherence long enough to perform complex calculations. The protocol simultaneously tackles two key challenges. First, it rapidly resets ancilla qubits, those used for error detection, to their ground state. Second, it minimizes leakage, the loss of quantum information from data qubits into higher energy levels, which corrupts calculations. This combined operation, completed in 83 nanoseconds, represents an improvement over previous methods requiring separate, sequential steps.

An error of (1.87 ± 1.12)× 10^-3 was achieved for the semi-adiabatic reset, and an error of 1.87 × 10^-3 for the adiabatic reset, within 9 nanoseconds. The qubits themselves are fixed-frequency transmon qubits, designed and fabricated as a flip-chip architecture, with the elements in red placed on the qubit chip and those in blue on the control chip interposer, optimizing signal routing and minimizing interference.

83ns Total Time for Reset and Leakage Reduction

The ability to swiftly and accurately reset qubits and mitigate leakage is now demonstrably faster, with a newly demonstrated protocol completing the combined operation in 83 nanoseconds. A pair of qubits are involved in the experiments. The efficiency of this protocol stems from the use of tunable couplers, which act as intermediaries for energy transfer. By carefully controlling the coupling between qubits and resonators, the researchers engineered a system where excitations can be rapidly and reliably moved away from the computational qubits.

The coupler reset itself, dissipating excitation through the readout resonator, takes only 22 nanoseconds. This is particularly important as it minimizes the need for complex control electronics or additional hardware resources, simplifying the scaling process for larger quantum processors.

Semi-Adiabatic Reset Transfers |1⟩ to |0⟩ in 9ns

The conventional expectation that qubit reset and leakage reduction demand significant overhead in quantum error correction cycles is being challenged by a new protocol achieving remarkably swift and accurate performance. This speed, coupled with high fidelity, represents a substantial refinement in the control and manipulation of quantum information. This advancement centers on a specific architecture employing fixed-frequency transmon qubits, paired and interconnected via tunable couplers.

Adiabatic Reset Simultaneously Clears |1⟩ and |2⟩ States

This simultaneous reset distinguishes the approach from methods focusing on individual state management, offering a more holistic solution for maintaining qubit coherence. This technique avoids the need for additional drive signals or complex qubit architectures, simplifying implementation and scalability. An adiabatic reset clears both |1⟩ and |2⟩ states with an error of (7.94)× 10^-3 in 61 nanoseconds, while the LRU operates with an error of 9.50× 10^-3 in just 5 nanoseconds.

This carefully engineered coupling landscape allows for efficient energy transfer and dissipation. By simultaneously resetting ancilla qubits and reducing leakage in data qubits, the protocol also promises to reduce the runtime of quantum error correction cycles and improve the overall fidelity of quantum algorithms.

The researchers emphasize that this protocol’s efficiency stems from minimal impact on neighboring qubits, simplifying the design of larger quantum processors. The system, tested on a device comprised of qubits from the Department of Microtechnology and Nanoscience, Chalmers University of Technology, Gothenburg, Sweden; Department of Chemistry, Princeton University, Princeton, NJ, USA; Department of Electrical and Computer Engineering, Princeton University, Princeton, NJ, USA; and VTT Technical Research Centre of Finland, FI-02044 VTT, Finland, demonstrates that the reset and leakage reduction can be implemented without requiring additional hardware resources.

Coupler-Resonator SWAP Implements Leakage Reduction Unit

Leveraging a coupler-resonator SWAP gate to manage qubit excitation, this new protocol’s architecture was demonstrated on a device comprised of the Department of Microtechnology and Nanoscience, Chalmers University of Technology, Gothenburg, Sweden; Department of Chemistry, Princeton University, Princeton, NJ, USA; Department of Electrical and Computer Engineering, Princeton University, Princeton, NJ, USA; and VTT Technical Research Centre of Finland, FI-02044 VTT, Finland. The experimental setup utilizes a microwave setup to monitor signal transmission through the feedline, providing the data necessary to validate the protocol’s performance.

This detailed characterization of the device, including its fabrication and performance metrics, is documented elsewhere, but forms the foundation for the observed results. The device’s design features qubits arranged in a checkerboard pattern, with neighboring qubits separated in frequency. The protocol achieves five distinct objectives, as detailed by the authors. The leakage reduction unit operates with an error of 9.50× 10^-3 over 61 nanoseconds.

A coupler reset unit dissipates excitation through the qubit readout resonator in 22 nanoseconds, and the entire sequence, simultaneous ancilla reset and data qubit leakage reduction, takes a total of 83 nanoseconds. The system’s design minimizes impact on neighboring qubits, a crucial consideration for scalability. The researchers emphasize that this integrated approach is readily implementable with existing fixed-frequency qubits and tunable couplers, requiring no additional hardware resources. The experimental demonstration was performed on a pair of qubits, allowing for precise control and measurement of the reset and leakage reduction processes.