Hybrid Quantum Computing Boosts Atom Simulations

Jihyeon Park and colleagues at the University of Wisconsin, Madison, present CANOE, the Classically Assisted Non-Orthogonal Eigensolver, which distributes computational workload between quantum and classical hardware. The method uses the strengths of quantum states, supplementing them with readily available classical basis states to improve ground-state representation. Validated through simulations of a 76-qubit chromium atom system, CANOE offers a pathway toward achieving chemical accuracy with moderate computational expense and incorporates a stabilisation procedure to address potential numerical instability. These findings suggest CANOE provides a practical set of tools for using both limited quantum resources and expansive classical computation for simulations in the near term. Achieving chemical accuracy in quantum simulations via hybrid quantum-classical determinants Ground-state energy errors were reduced to chemical accuracy, defined as 1 kcal/mol, utilising the CANOE method. Crossing this threshold enables accurate simulations of complex molecular systems, such as the 76-qubit chromium atom used for validation, previously intractable due to the computational cost of classically representing quantum states. Traditional full configuration interaction (FCI) methods, while capable of achieving chemical accuracy, suffer from factorial scaling with system size, rendering them impractical for all but the smallest molecules. CANOE achieves this by combining a few quantum-generated basis states with a much larger pool of classically computed determinants, distributing the computational workload and minimising the need for extensive quantum resources. This hybrid approach represents a significant departure from purely quantum or purely classical methods, aiming to leverage the benefits of both paradigms. The quantum component generates a compact set of highly entangled states that capture essential correlation effects, while the classical component provides a flexible and efficient means of expanding the basis set to achieve the desired level of accuracy. The complex 76-qubit chromium atom system validated the method, demonstrating its ability to model strongly correlated systems. This benchmark involved simulating a molecule with 76 spin orbitals, a substantial increase in complexity over previous demonstrations. Such systems, characterised by strong electron-electron interactions, pose a significant challenge for conventional electronic structure methods. Adding each additional basis state demonstrably improves the accuracy of ground-state representability, highlighting the method’s sensitivity and potential for refinement. The rate of convergence with respect to the number of quantum basis states is a crucial metric for assessing the efficiency of the algorithm. A new histogram-based protocol efficiently estimates overlaps between quantum and classical states, reducing the computational burden compared to full state tomography, which scales exponentially with qubit number. Full state tomography, requiring a complete characterisation of the quantum state, becomes prohibitively expensive for even moderately sized systems. The histogram-based approach provides a practical alternative by focusing on the relevant degrees of freedom needed for eigenvalue computation. To address potential numerical instability caused by linear dependencies within the hybrid basis, CANOE incorporates a Schur-complement-based stabilisation procedure, ensuring stable eigenvalue computation. Linear dependencies can arise when the basis states are not sufficiently linearly independent, leading to ill-conditioned matrices and inaccurate results. The Schur complement effectively removes these dependencies, ensuring a robust and reliable solution. Current simulations assume ideal conditions and do not yet demonstrate performance with realistic quantum hardware limitations such as noise or limited coherence; future work will focus on mitigating these effects and exploring the method’s durability in noisy environments. Investigating the impact of gate errors, decoherence, and other sources of noise is essential for assessing the practical viability of CANOE on real quantum devices. Hybrid quantum-classical algorithms circumvent near-term quantum hardware limitations CANOE offers a pragmatic path toward utilising quantum computers for molecular simulations, acknowledging a fundamental tension inherent in hybrid approaches. Several competing methods, such as the variational quantum eigensolver (VQE) and quantum phase estimation (QPE), attempt to map the entire quantum problem onto the available hardware. However, these strategies demand increasingly complex quantum circuits as system size grows, potentially exceeding the capabilities of near-term devices. VQE, while relatively shallow in circuit depth, often requires many variational parameters, leading to optimisation challenges. QPE, theoretically capable of achieving exponential speedups, requires deep circuits and long coherence times, which are currently beyond the reach of most quantum computers. The limitations of near-term quantum hardware necessitate the development of algorithms that can effectively utilise limited resources and mitigate the effects of noise. Distributing computational load efficiently utilises available resources, addressing the challenge of complex quantum circuits.

The Classically Assisted Non-Orthogonal Eigensolver presents a new strategy for tackling molecular simulations by intelligently combining the strengths of both classical and quantum computing. By offloading a significant portion of the computation to classical hardware, CANOE reduces the demands on the quantum processor, allowing it to focus on generating the essential correlated states. The successful implementation of a Schur-complement-based stabilisation procedure addresses potential numerical instability inherent in hybrid calculations, paving the way for more robust simulations and enabling the exploration of larger, more complex molecular systems. This stabilisation procedure is crucial for maintaining the accuracy and reliability of the results, particularly when dealing with large and complex systems. Further research could explore the application of CANOE to other areas of quantum chemistry, such as excited-state calculations and molecular dynamics simulations. The ability to accurately and efficiently simulate molecular systems has broad implications for fields such as drug discovery, materials science, and fundamental chemical research. 👉 More information 🗞 CANOE: Classically Assisted Non-Orthogonal Eigensolver 🧠 ArXiv: https://arxiv.org/abs/2603.13188 Tags: