Quantum Calculations Become Far Simpler with New Operator Weighting Method

Jialiang Tang and colleagues at Instituto de Ciencia de Materiales de Madrid ICMM-CSIC, in a collaboration between the institute, the Joint Quantum Institute and Joint Centre for Quantum Information and Computer Science, and NIST/University of Maryland, present a new approach using a weighted nested-commutator ansatz to approximate adiabatic gauge potentials with local operators. The technique expands the potential variational space, enabling more effective optimisation and demonstrably improved performance compared to standard nested-commutator methods when preparing one-dimensional matrix product states and the ground state of a quantum Ising model. Numerical results indicate a sharp acceleration in the preparation of quantum states, successfully demonstrating the method on systems containing up to 1000 qubits and a two-dimensional lattice with 30 sites.

Scalable quantum state preparation via locally approximated counterdiabatic driving

A scalable counterdiabatic (CD) driving technique now accelerates quantum state preparation for systems containing up to 1000 qubits, a substantial increase over previous limitations. Counterdiabatic driving is a method for manipulating quantum systems that aims to suppress unwanted transitions during state evolution, effectively guiding the system towards a desired target state. However, its implementation relies on calculating and applying highly nonlocal adiabatic gauge potentials (AGP), which represent the ‘forces’ needed to counteract these unwanted transitions. These AGPs are notoriously difficult to compute and implement, particularly in large many-body systems, due to their nonlocal nature, meaning they require interactions between distant parts of the system. This new technique overcomes these computational barriers through a weighted nested-commutator (WNC) ansatz. This WNC ansatz approximates complex potentials using only locally operating components, expanding the variational space and enabling more efficient optimisation than standard nested-commutator methods.

The method’s effectiveness in preparing complex quantum states and sharply accelerating the process was demonstrated using one-dimensional matrix product states and a two-dimensional lattice with 30 sites. Matrix product states are a powerful tool for representing the quantum states of one-dimensional systems, allowing for efficient simulation of many-body interactions. Simulations extended to systems of this size, utilising a local optimisation scheme that reduces computational demands and scales with the number of qubits. This local optimisation is crucial; traditional optimisation algorithms can become intractable as system size increases, but by focusing on local adjustments, the computational burden is significantly reduced. While these results confirm substantial acceleration, current work focuses on specific model systems and does not yet demonstrate efficacy across a broader range of physically relevant Hamiltonians or durability to experimental noise. The WNC ansatz consistently outperformed the standard nested-commutator approach when preparing one-dimensional matrix product states (MPS) and the ground state of a complex quantum Ising model, offering a valuable tool for materials science and fundamental physics research. The quantum Ising model, for example, is a fundamental model in condensed matter physics used to study magnetic phenomena and phase transitions.

Variational optimisation of adiabatic gauge potentials using weighted nested commutators

The weighted nested-commutator (WNC) ansatz addresses the challenge of approximating adiabatic gauge potentials (AGP), the extra ‘forces’ applied to a quantum system to facilitate smooth evolution, using only locally operating components. The core difficulty lies in the fact that accurately calculating AGPs often requires considering correlations between distant particles, making the computation exponentially complex with system size. Unlike standard nested-commutator methods, the WNC approach assigns independent variational weights to each step in building these AGP approximations; this is akin to adding multiple, adjustable stabilisers to a bicycle to ensure a smoother ride. Increasing the number of adjustable parameters within a defined range expands the possibilities for accurately representing the AGP, enabling more effective optimisation of quantum states. The weights themselves become parameters that are optimised during the quantum state preparation process.

This expanded ‘variational space’ is vital for efficiently preparing complex quantum states, particularly in many-body systems where particle interactions are significant. Many-body systems present a significant challenge to simulation due to the exponential growth in computational complexity with the number of particles. Numerical demonstrations involved systems of up to 1000 qubits for one-dimensional matrix product states and hexagonal lattices containing up to 30 sites, allowing for accelerated quantum state preparation. The ability to accurately represent AGPs is crucial for tackling complex quantum simulations, and this method provides a promising avenue for achieving that goal. The hexagonal lattice structure was chosen to explore the method’s performance in two-dimensional systems, providing a more realistic testbed for potential applications in materials science. The use of 30 sites, while still relatively small, represents a significant step towards simulating larger and more complex systems.

Accelerating quantum state preparation using weighted nested-commutator approximations

Simulating quantum systems demands ever more computational power as scientists strive to model increasingly complex materials and phenomena. This new work represents a major step towards that goal, demonstrating a method for preparing quantum states with up to 1000 qubits, a scale previously hampered by the difficulty of accurately calculating ‘adiabatic gauge potentials’. Some experts question whether simulating systems of 1000 qubits represents a genuinely practical advance, given the limitations of current quantum hardware. While current quantum computers are still limited in the number of qubits and their coherence times, the ability to efficiently simulate larger systems on classical computers is a crucial stepping stone towards developing and validating algorithms for future quantum computers. Achieving this scale required approximating complex calculations using a weighted nested-commutator ansatz, a clever shortcut for simplifying the necessary computations. The approximation allows for a reduction in computational cost without sacrificing too much accuracy, making it possible to simulate larger systems.

Despite this approximation, the method demonstrably speeds up state preparation for these systems. Accurately preparing complex quantum states is computationally expensive, and this development offers a pathway to overcome this key limitation in quantum simulation. Scalable counterdiabatic driving was achieved for systems containing up to 1000 qubits by approximating calculations using this method, expanding the range of possible solutions considered during optimisation and leading to improved performance compared to previous approaches. The improvement in performance is attributed to the increased flexibility afforded by the weighted nested-commutator ansatz, which allows for a more accurate representation of the adiabatic gauge potentials. This, in turn, leads to a faster and more efficient preparation of the desired quantum state, opening up new possibilities for exploring complex quantum phenomena and designing novel materials.

Researchers successfully demonstrated a new method for efficiently preparing quantum states containing up to 1000 qubits. This achievement matters because accurately simulating quantum systems is computationally demanding, hindering progress in materials science and quantum algorithm development. By employing a weighted nested-commutator ansatz to approximate complex calculations, they significantly accelerated state preparation for one-dimensional systems and a two-dimensional model. Future work could focus on extending this approach to even larger systems and exploring its application to more complex physical models, potentially aiding the design of new materials with tailored properties.