Harvard-led Research Team Achieves Continuous Operation of a 3,000-Qubit System ​

In a significant leap forward for quantum computing, a team of researchers has successfully demonstrated the continuous operation and reloading of a large-scale 3,000-qubit system. This breakthrough addresses the critical challenge of atom loss in neutral atom quantum computers, opening the door for the development of deep-circuit, fault-tolerant quantum computation.

​

Key Takeaways

• Continuous Qubit Replenishment: The research introduces a novel architecture that can continuously reload atomic qubits at a rate of up to 30,000 per second. This is nearly two orders of magnitude faster than previous technologies.
  
  
• Large-Scale, Long-Duration Operation: The system successfully maintained an array of over 3,000 qubits for more than two hours, a task previously hindered by the limited lifetime of trapped atoms.
  
  
• Preservation of Quantum Coherence: A key achievement is the ability to introduce new qubits into the system without disturbing the delicate quantum states (coherence) of existing qubits.
  
  
• Zoned Architecture for Scalability: The system employs a zoned architecture, separating the atom reservoir, preparation zone, and storage zone to minimize interference and enable parallel operations.
  
  
• Path to Fault-Tolerant Quantum Computing: These advancements are directly applicable to building fault-tolerant quantum computers, which are necessary for solving complex problems beyond the reach of classical computers. The high qubit-reloading rate is sufficient to overcome atom loss in a quantum processor with approximately 10,000 physical qubits.
  
  
• Broad Applications: Beyond quantum computing, this technology has significant implications for advancing atomic clocks, quantum sensors, and quantum networking.

A New Architecture for Quantum Stability

Neutral atom quantum computers are a leading platform for quantum science due to their potential for scalability and strong qubit interactions. However, a major obstacle has been the inevitable loss of atoms over time due to operational errors and finite trap lifetimes. This loss necessitates pulsed operation, limiting the complexity of quantum circuits and the overall performance of these systems.

To overcome this, the research team from Harvard University, MIT, and ETH Zurich developed an innovative architecture based on two serial optical lattice conveyor belts. This system transports a continuous supply of laser-cooled Rubidium-87 atoms into a "science region" where they are loaded into optical tweezers.

The architecture is divided into three distinct zones: a reservoir, a preparation zone, and a storage zone. This separation is crucial for shielding the fragile qubits in the storage zone from the disruptive light and magnetic fields used in the atom-loading and preparation processes.

The Three-Zone System:

Why It's Important

The Dawn of Continuous Quantum Operation

The transition from pulsed to continuous operation represents a shift for neutral atom quantum technologies. For quantum computing, this enables the implementation of quantum error correction (QEC) codes, which are essential for building fault-tolerant systems. 

QEC requires repeated cycles of error detection and correction, a process that is hampered by atom loss. The high-rate qubit replenishment demonstrated in this research provides a direct solution to this problem.

The researchers estimate that with current entangling gate fidelities, a rate of 15,000 rearranged qubits per second is sufficient to sustain a quantum processor of about 10,000 physical qubits. 

With realistic improvements, this could enable the operation of several hundred logical qubits with significantly reduced failure rates, bringing practical quantum computation closer to reality.

Beyond computation, this work has potential implications for other fields. In quantum metrology, continuously operating atomic clocks can achieve higher stability and precision by reducing dead time. Similarly, quantum networking applications may benefit from a continuous stream of atomic qubits for faster generation of entanglement between remote nodes.

Figure 4. Continuous reloading while maintaining storage qubit coherence. a, Time sequence visualizing our continuous reloading protocol (see also ED Fig. 9 and Supplementary Movie). Following the initial storage array assembly, the longest-stored subarray is ejected and refilled with a preloaded set of qubits from the preparation zone every ∼ 80 ms, while storage zone shielding is applied throughout. For subfigures (c-d), storage qubits are placed in the equal superposition state and undergo an XY16-64 decoupling sequence during each reloading cycle. b, Continuously reloading storage zone qubits, we maintain a high degree of storage array polarization (red) for, in principle, unbounded duration. For reference, we provide a T1 measurement without qubit replenishment (gray). c, Similar to (b), now additionally applying an Xπ/2 − (XY16-64) − X−π/2 dynamical decoupling sequence during each subarray replenishment. We probe coherence of each subarray at various times during the replenishment cycle by reading out qubits in state |0⟩ (blue) or |1⟩ (red) as detailed in Methods. Individual subarrays (color shading) are unaffected by adjacent qubit reloading, and their dephasing is offset in time due to the cyclic subarray reloading protocol. The exponential sawtooth overlays are guides to the eye. For reference, we provide the T2 measurement of a single subarray under the same cyclic decoupling sequence without qubit replenishment (gray). d, After multiple rounds of reloading under dynamical decoupling, we apply a final DD sequence and vary the phase of the last π/2-pulse to read out in different qubit bases. Complementary to (c), the observed coherence contrast varies for each subarray (color shading) due to the time-offset in subarray replenishment. (Credit: Paper)

Summary of Results

A key innovation is the use of two angled optical lattice conveyor belts to transport atoms from a magneto-optical trap (MOT) to the science chamber. This dual-lattice scheme prevents a direct line-of-sight between the MOT and the qubit array, successfully shielding the storage qubits from scattered light from the initial atom capture process.

The system achieves an impressive qubit flux of over 30,000 qubits per second without rearrangement, and 15,000 qubits per second when sorting atoms into defect-free arrays. This is made possible by a rapid, 20-millisecond qubit preparation sequence.

The team successfully assembled and maintained a large-scale array of over 3,000 atoms for more than two hours, far exceeding the natural 60-second lifetime of the atoms in the optical tweezers. This was achieved by iteratively replacing one of six subarrays every 80 milliseconds.

Critically, the coherence of the storage qubits was preserved during the continuous reloading process. By applying a "shielding" light, the researchers were able to protect the storage qubits from scattered light during preparation zone operations, maintaining a coherence time (T2​) of 1.09 seconds, close to the reference value of 1.34 seconds.

The team also demonstrated the ability to maintain a large array of qubits in a coherent superposition state while continuously reloading the array. This showcases the system's capability to perform quantum operations in parallel with qubit replenishment, a crucial requirement for complex quantum algorithms.

Conclusion

This research marks a pivotal achievement in the field of quantum computing. By demonstrating the continuous, high-rate, and coherent operation of a large-scale qubit system, the authors have provided a viable path toward overcoming the longstanding challenge of atom loss. The novel architecture and techniques presented in this work are expected to accelerate the development of fault-tolerant quantum computers and enhance a wide range of quantum technologies.