Periodic Reporting for period 1 - Q-CIRC (Superconducting qubits with 1 second coherence time using rotation codes)
Reporting period: 2022-08-01 to 2025-01-31
In the pursuit of fault-tolerant quantum computing, the ability to store quantum information in qubits for extended durations is essential. Superconducting qubits have emerged as a leading technology due to their scalability and compatibility with existing fabrication techniques. However, improving the coherence times of these qubits remains a significant challenge, as they are typically susceptible to various sources of noise. Achieving longer coherence times is critical because it enables higher gate fidelities, which in turn reduces the hardware overhead required for fault-tolerant quantum computing.
To date, the most long-lived superconducting qubits are bosonic qubits encoded as single-photon states in three-dimensional microwave cavities. While these cavities have demonstrated impressive coherence times of up to 2 milliseconds, they face several limitations—notably, errors induced by the on-chip superconducting qubits used to control the bosonic qubit constrain their performance.
Q-CIRC aims to push the boundaries of superconducting bosonic qubit coherence by innovating across multiple fronts. First, we will design and produce advanced superconducting cavities that better protect quantum information from decoherence, leveraging novel materials, surface treatments, and configurations. Second, we will develop new methods based on active feedback control to mitigate errors in real time, enhancing the coherence of the encoded information. Third, we plan to implement quantum error detection and correction directly at the physical hardware level, thereby further reducing the impact of errors on qubit coherence.
By addressing these challenges, Q-CIRC seeks to set the stage for a new generation of high-coherence superconducting qubits with improved gate fidelities and enhanced quantum error correction performance.
To date, the most long-lived superconducting qubits are bosonic qubits encoded as single-photon states in three-dimensional microwave cavities. While these cavities have demonstrated impressive coherence times of up to 2 milliseconds, they face several limitations—notably, errors induced by the on-chip superconducting qubits used to control the bosonic qubit constrain their performance.
Q-CIRC aims to push the boundaries of superconducting bosonic qubit coherence by innovating across multiple fronts. First, we will design and produce advanced superconducting cavities that better protect quantum information from decoherence, leveraging novel materials, surface treatments, and configurations. Second, we will develop new methods based on active feedback control to mitigate errors in real time, enhancing the coherence of the encoded information. Third, we plan to implement quantum error detection and correction directly at the physical hardware level, thereby further reducing the impact of errors on qubit coherence.
By addressing these challenges, Q-CIRC seeks to set the stage for a new generation of high-coherence superconducting qubits with improved gate fidelities and enhanced quantum error correction performance.
During the first reporting period, we made significant steps toward achieving the objectives of Q-CIRC. We successfully encoded a bosonic qubit in a newly designed cavity with an intrinsic lifetime exceeding 100 milliseconds. Using an integrated on-chip superconducting ancilla, we demonstrated a single-photon cavity qubit with a relaxation time of 25.6 milliseconds and a coherence time of 34 milliseconds, making it the longest coherence times achieved for any superconductivity-based qubit to date. Beyond merely storing quantum information, this long-lived cavity qubit enabled us to encode a Schrödinger cat state with a record size of 1,024 photons.
Furthermore, we addressed the crucial issue of ancilla-induced noise propagating to the cavity qubit. By continuously monitoring the ancilla qubit and implementing active feedback control, we effectively suppressed this source of errors. This approach led to a fivefold increase in the pure dephasing time of a cavity qubit.
Finally, we explored the theory of bosonic qubits in the presence photon loss and dephasing errors. We mapped a "phase space" of bosonic error-correcting codes optimized for specific error rates. These findings provide valuable insights for tailoring error correction strategies to different types of noise.
Furthermore, we addressed the crucial issue of ancilla-induced noise propagating to the cavity qubit. By continuously monitoring the ancilla qubit and implementing active feedback control, we effectively suppressed this source of errors. This approach led to a fivefold increase in the pure dephasing time of a cavity qubit.
Finally, we explored the theory of bosonic qubits in the presence photon loss and dephasing errors. We mapped a "phase space" of bosonic error-correcting codes optimized for specific error rates. These findings provide valuable insights for tailoring error correction strategies to different types of noise.
The progress achieved in Q-CIRC thus far already surpasses the current state of the art, with qubit relaxation times and pure dephasing times improved by an order of magnitude compared to previous demonstrations in superconducting bosonic qubits.