IonQ, NVIDIA et al: Accelerating Quantum Monte Carlo Simulations with Matchgate Shadow Techniques on Ion Trap Hardware

Accurate modelling of chemical reaction rates remains a significant challenge in computational chemistry, often requiring approximations that limit predictive power. Recent advances explore the potential of quantum computing to address these limitations, particularly in simulating the electronic structure of molecules. A collaborative team, comprising researchers from IonQ Inc., AstraZeneca Gothenburg, and NVIDIA Corporation, now reports a substantial advancement in this field. Their work details an end-to-end workflow integrating quantum-classical computation, utilising the quantum-classical auxiliary field Monte Carlo (QC-AFQMC) algorithm with tomography via matchgate shadows, and executed on the IonQ Forte trapped-ion quantum computer and accelerated by NVIDIA GPUs. The research, detailed in their article, “Quantum-Classical Auxiliary Field Quantum Monte Carlo with Matchgate Shadows on Trapped Ion Quantum Computers”, demonstrates a significant speedup over existing implementations and applies the method to simulate a crucial step in the nickel-catalyzed Suzuki-Miyaura reaction, utilising 24 qubits and achieving results consistent with established high-accuracy quantum chemical methods.

Quantum simulations are accelerating chemical modelling through the combined utilisation of classical and quantum computing resources. Researchers have successfully demonstrated a complete workflow for modelling chemical reaction barriers, employing the quantum-classical auxiliary field Monte Carlo (QC-AFQMC) algorithm, enhanced with tomography via matchgate shadows. This implementation operates within a high-performance computing environment, leveraging both the IonQ Forte quantum processing unit (QPU) and NVIDIA graphics processing units (GPUs) hosted on Amazon Web Services, achieving acceleration representing several orders of magnitude improvement over existing QC-AFQMC implementations.

The study applies this accelerated algorithm to simulate the oxidative addition step within the nickel-catalyzed Suzuki-Miyaura reaction, a crucial process in organic chemistry, utilising 24 qubits on the IonQ Forte. Sixteen qubits represent the trial wave function, while a further eight ancilla qubits are allocated for error mitigation. This represents the largest QC-AFQMC calculation with matchgate shadow experiments performed to date on actual quantum hardware. Researchers report significant speedups in collecting matchgate circuit measurements and a corresponding improvement in the time required for post-processing analysis through a distributed-parallel implementation, paving the way for more complex simulations. Matchgate shadows are a technique used to characterise quantum states, providing information about the probability distribution of measurement outcomes.

Evaluation of the model reaction’s chemical reaction barriers, performed using an active-space QC-AFQMC approach, yields results consistent with high-accuracy reference calculations. When matchgates are sampled using an ideal simulator, the calculated barriers fall within the uncertainty interval of kcal/mol compared to the reference Coupled Cluster Singles Doubles (CCSD(T)) result, validating the methodology. CCSD(T) is a highly accurate, albeit computationally expensive, method used in quantum chemistry to calculate the energy of molecular systems. Measurements performed directly on the QPU demonstrate a slightly larger deviation, remaining within 10 kcal/mol of the reference value, highlighting the potential for practical chemical simulations on near-term quantum devices.

The successful integration of quantum and classical computing resources, alongside algorithmic innovations and efficient parallelisation, establishes a pathway towards practical chemical simulations on near-term quantum devices. Researchers utilise matchgate shadows to effectively mitigate errors and improve the accuracy of the calculations, demonstrating a significant advancement in quantum chemistry. This work highlights the potential of QC-AFQMC as a viable method for tackling complex chemical problems that are intractable for classical computers, opening new avenues for drug discovery and materials science.

The research provides detailed coordinate data for two nickel complexes, labelled C and B, each comprising 34 atoms, facilitating further investigation into catalytic mechanisms. These coordinates, representing the three-dimensional positions of carbon, hydrogen, oxygen, and nickel atoms, are fundamental to constructing accurate molecular models and performing computational chemistry calculations, enabling researchers to analyse the geometry and bonding characteristics of these complexes. This data provides a basis for further investigation into catalytic mechanisms, allowing scientists to understand the intricate details of chemical reactions at the molecular level.

Future research will focus on extending this methodology to even larger and more complex chemical systems, exploring new algorithms for error mitigation, and developing more efficient methods for data analysis. Scientists aim to push the boundaries of quantum chemistry, unlocking the potential of quantum computers to solve some of the most challenging problems in science and technology, ultimately leading to the design of new materials, drugs, and catalysts. The team plans to investigate the use of different quantum hardware platforms and explore the potential of hybrid quantum-classical algorithms further to enhance the accuracy and efficiency of these simulations.

The authors are Luning Zhao, Joshua J. Goings, Willie Aboumrad, Andrew Arrasmith, Lazaro Calderin, Spencer Churchill, Dor Gabay, Thea Harvey-Brown, Melanie Hiles, Magda Kaja, Matthew Keesan, Karolina Kulesz, Andrii Maksymov, Mei Maruo, Mauricio Muñoz, Bas Nijholt, Rebekah Schiller, Yvette de Sereville, Amy Smidutz, Felix Tripier, Grace Yao, Trishal Zaveri, Coleman Collins, Martin Roetteler, Evgeny Epifanovsky, Arseny Kovyrshin, Lars Tornberg, Anders Broo, Jeff R. Hammond, Zohim Chandani, Pradnya Khalate, Elica Kyoseva, Yi-Ting Chen, Eric M. Kessler, Cedric Yen-Yu Lin, and Gandhi.