Quantum Algorithm Stable with 12 Qubits

Scientists have introduced RFOX (Rotated-Field Oscillatory eXchange), a parameter-free quantum algorithm for tackling combinatorial optimisation problems, representing a step towards realising the potential of quantum computers for practical applications. Brian García Sarmina and colleagues at Huzhou University designed RFOX specifically to address the challenges inherent in combinatorial optimisation, a field encompassing problems like logistics, finance, and materials discovery. Their findings demonstrate that RFOX maintains a robust spectral gap, a crucial metric for algorithm performance, through a novel combination of quantum drivers and analytically derived corrections, a significant advantage over conventional methods which often suffer from unpredictable reductions in this gap. Simulations utilising up to 12 qubits and experiments conducted on IBM Quantum processors reveal RFOX achieves near-optimal results with substantially fewer computational steps, suggesting a scalable and tuning-free pathway for the development of future quantum optimisers.

RFOX algorithm efficiently locates ground states in disordered quantum systems

The RFOX algorithm achieves near-optimal or exact ground states with up to ten times fewer Trotter slices than conventional quantum optimisation methods. The Trotter slice count directly correlates with the computational effort required; reducing this count is therefore a critical advancement. Prior techniques have struggled to reliably find solutions for disordered systems beyond a limited number of qubits, primarily due to the exponential increase in computational demands associated with simulating complex quantum states. Disordered systems, characterised by randomness in their underlying parameters, pose a particular challenge for classical algorithms and often require extensive sampling to find optimal solutions. RFOX, a new parameter-free approach, maintains a constant runtime scaling of T ∝∆−2 min, where T represents the runtime and ∆ min denotes the minimum spectral gap. This characteristic is particularly noteworthy as it contrasts with existing algorithms which typically experience performance slowdowns as the spectral gap diminishes, indicating a more stable and predictable performance profile. The spectral gap represents the energy difference between the ground state and the first excited state; a larger gap generally implies easier optimisation.

Random-field Ising model instances containing 7, 9, and 12 qubits, tested across three distinct magnetic-field strengths, were used to verify the algorithm’s effectiveness. The random-field Ising model is a standard benchmark for evaluating optimisation algorithms, particularly those designed for disordered systems. Varying the magnetic-field strengths allowed the researchers to assess the algorithm’s robustness under different conditions. Experiments conducted on IBM Quantum processors, including the Eagle r3 and Heron r1 systems with 12, 15, and 20 qubits, mirrored these performance gains, confirming its strong performance on real quantum hardware. This is crucial, as simulations, while valuable, cannot fully capture the complexities and noise inherent in actual quantum devices. While the current results focus on relatively small qubit numbers and do not yet demonstrate scalability to the hundreds or thousands of qubits needed to tackle truly practical optimisation challenges, the observed performance is highly encouraging. A stable runtime, even with increasing problem complexity, is a key advantage, and future work will focus on extending this stability to larger systems. Validating performance on actual quantum hardware goes beyond purely theoretical simulations and offers a tangible path towards more powerful quantum problem-solvers, bridging the gap between theoretical potential and practical realisation.

Maintaining spectral gap stability is important for scalable quantum optimisation

Practical quantum optimisation is steadily progressing, promising solutions to problems currently beyond the reach of classical computers. These problems span a wide range of disciplines, including drug discovery, financial modelling, and materials science. However, the success of RFOX, and indeed any quantum optimisation algorithm, critically depends on maintaining this stable spectral gap as systems grow; tests were limited to 20 qubits, a relatively modest scale in the context of quantum computing. Scaling up to the hundreds or thousands of qubits needed for real-world applications presents a formidable challenge, potentially exposing limitations not yet apparent. Maintaining qubit coherence, minimising errors, and managing the complexity of quantum circuits all become increasingly difficult as the number of qubits increases. The algorithm’s reliance on a constant runtime scaling is particularly important in this regard, as it suggests that the computational cost will not increase exponentially with problem size.

A specific quantum driver, utilising an almost constant non-stoquastic $XX$ catalyst, uniquely combines with a weak harmonic $ZX$ counter-diabatic term, a technique used to improve the performance of quantum processes by mitigating the effects of unwanted transitions. This combination is vital for its stability. The $XX$ driver facilitates efficient exploration of the solution space, while the $ZX$ counter-diabatic term helps to suppress unwanted excitations and maintain the system in the ground state. The development of RFOX introduces a new quantum algorithm designed to solve complex combinatorial optimisation problems. Using the Floquet-Magnus expansion, the researchers derived a closed-form effective Hamiltonian, allowing for analytical understanding of the algorithm’s behaviour. The first-order term of this Hamiltonian retains the full $XX$ driver, while the leading correction consists of a single qubit $Y$ field at high drive frequency. Simulations and experiments have validated the algorithm, consistently outperforming existing methods, particularly in disordered systems where traditional approaches struggle. Further development, including exploration of different qubit connectivity topologies and error mitigation strategies, could unlock solutions to complex problems currently intractable for conventional computers, and the next decade will begin to clarify its full potential and limitations within the evolving landscape of quantum computation.

Researchers developed RFOX, a new quantum algorithm for combinatorial optimisation that demonstrated near-optimal results on random-field Ising model instances with 7, 9, and 12 qubits. The algorithm utilises a unique combination of quantum drivers to maintain a stable and essentially flat spectral gap, offering a performance advantage over conventional methods, especially with increasing disorder. This stability translates to a constant runtime scaling, meaning the computational cost does not increase exponentially with problem size. The authors suggest that future work will focus on exploring different qubit arrangements and error reduction techniques.