A new method improves the accuracy of quantum computation through error correction. Yu Wang and colleagues at Harvard University, in collaboration with Boston University, have successfully performed multi-qubit stabilizer readout on a dual-species Rydberg array of sodium and caesium atoms. The work addresses a key challenge in quantum error correction, efficiently measuring ancilla qubits without introducing crosstalk, and showcases a protocol for simultaneously reading out the state of multiple data qubits using global pulses. By compensating for errors arising from interspecies interactions, the team achieved non-destructive measurement of Pauli-Z stabilizers on four-qubit caesium plaquettes, paving the way for scalable quantum error correction and novel quantum control techniques.
Sodium ancilla atoms enable high-fidelity parallel readout of caesium qubits
Operational fidelity for simultaneous, non-destructive readout of Pauli-Z stabilizers on four-qubit caesium plaquettes exceeded 95%, a substantial improvement over previous methods. This level of fidelity is crucial for effective quantum error correction, as even small errors can rapidly accumulate and corrupt quantum information. A dual-species approach utilising sodium atoms as ancillas measures the stabilizers of surrounding caesium data qubits, representing a key breakthrough. Rydberg atoms, created by exciting atoms to a high principal quantum number, exhibit strong interactions that are exploited for quantum gate operations and measurements. The use of two distinct species, sodium and caesium, allows for selective addressing and manipulation, minimising unwanted interactions between the ancilla and data qubits. This is achieved by leveraging the differing energy levels and transition wavelengths of the two atomic species. The team successfully compensated for finite interspecies Rydberg interactions, a long-standing obstacle to high-fidelity multi-body entanglement. These interactions, arising from the overlap of Rydberg wavefunctions between sodium and caesium atoms, can introduce errors in the readout process. Precise control over laser parameters and careful calibration were essential to mitigate these effects.
By tuning the Rabi frequency and detuning of the Rydberg driving field, geometric phase errors were mitigated, enabling a protocol for parallel readout via global pulses alone. The Rabi frequency dictates the rate of transitions between atomic states, while detuning refers to the difference between the laser frequency and the atomic resonance frequency. Optimising these parameters minimises unwanted transitions and ensures accurate readout. This dual-species tweezer array architecture unlocks new possibilities for quantum control, leveraging both inter- and intra-species interactions. Optical tweezers, created using highly focused laser beams, provide precise and independent control over the position of individual atoms. This allows for the creation of arbitrary 2D arrays with tailored interactions between qubits. It offers a promising route towards scalable quantum error correction. Quantum error correction relies on encoding quantum information in a redundant manner, using multiple physical qubits to represent a single logical qubit. By monitoring the stabilizers, specific combinations of qubit measurements, errors can be detected and corrected without disturbing the encoded quantum information.
Detailed analysis confirmed that successful compensation of finite interspecies Rydberg-Rydberg interactions enabled in situ stabilizer measurement, meaning the qubits did not need to be physically moved during the readout process. This is a significant advantage, as atom rearrangement can be slow and introduce additional errors. Performing measurements in situ streamlines the error correction process and improves overall performance. This is a key step towards scalability and avoids the need for complex atom rearrangement. Currently, the system’s performance applies to a small, specifically arranged system and does not yet reflect performance across larger, more complex architectures needed for practical quantum error correction. Scaling up to larger arrays presents significant challenges, including maintaining coherence, controlling crosstalk, and managing the complexity of the control system. Coherent ground-Rydberg excitation was established for both sodium and caesium, confirming species-selective control under global laser driving. This demonstrates the ability to selectively excite atoms to Rydberg states without affecting the other species, a crucial requirement for implementing the error correction protocol. Further research is needed to assess the protocol’s performance in larger, more complex architectures, despite these substantial advances. Future work will focus on increasing the number of qubits, improving the fidelity of individual operations, and developing more sophisticated error correction codes.
Limitations of proxy atoms necessitate future individual qubit control
While this dual-species approach offers a compelling pathway to scalable quantum error correction, achieving truly strong and reliable systems demands more than just clever atomic arrangements. The scientists acknowledge a reliance on non-loaded atoms as proxies for those in the ground state, a pragmatic choice given current apparatus limitations. This simplification skirts the need for local Raman addressing, the ability to individually flip qubit states, a capability currently lacking and one that may prove important for verifying and refining the protocol with genuine data qubits. Raman addressing uses two laser beams to induce transitions between atomic states, allowing for individual qubit control. Its absence currently limits the ability to fully characterise and optimise the error correction protocol.
Full verification requires individual control over each qubit’s state, a feature not yet implemented in this setup, acknowledging the current reliance on unpopulated atomic sites as stand-ins for ground state atoms. Using unpopulated sites simplifies the experimental setup but introduces uncertainties about the behaviour of actual qubits in the ground state. This dual-species approach, utilising sodium and caesium atoms trapped with optical tweezers, offers a scalable architecture for future quantum computers and expands the possibilities for manipulating quantum states. Non-destructive multi-qubit readout has been demonstrated, establishing a new architecture for quantum error correction. This capability is essential for implementing real-time error correction, as it allows for continuous monitoring of qubit states without collapsing the superposition. By employing sodium atoms as ancillas to assess the state of caesium qubits, the team overcame limitations imposed by interspecies interactions through careful tuning of laser parameters, enabling simultaneous data acquisition within an optical tweezer array, a system using lasers to precisely position atoms, and represents a significant step towards building larger, more reliable quantum processors. The precise positioning afforded by optical tweezers is fundamental to maintaining the coherence and fidelity of the qubits, and allows for the creation of complex quantum circuits.
The researchers successfully demonstrated non-destructive measurement of Pauli-Z stabilizers on four-qubit caesium plaquettes using a dual-species array of sodium and caesium atoms trapped in optical tweezers. This achievement matters because it establishes a new architecture for quantum error correction, allowing for continuous monitoring of qubit states without disrupting them. By carefully tuning laser parameters, they compensated for interactions between the two atomic species, enabling simultaneous readout of multiple qubits with a single pulse. The authors suggest this dual-species approach offers a scalable route towards building larger and more reliable quantum processors.
👉 More information
🗞 Multi-Qubit Stabilizer Readout on a Dual-Species Rydberg Array
🧠 ArXiv: https://arxiv.org/abs/2605.10924
