Haiqu and Quanscient Advance Quantum Computational Fluid Dynamics on Amazon Braket

Introduction

Haiqu and Quanscient demonstrated the first successful execution of multiple Quantum Lattice Boltzmann Method (QLBM) time steps on commercial quantum hardware in December 2024. The collaboration, which earned finalist recognition in the 2024 Airbus-BMW Quantum Mobility Challenge, achieved a breakthrough in quantum computational fluid dynamics by executing deep quantum circuits on IonQ’s Aria 1 quantum processing unit via Amazon Braket. Haiqu, a San Francisco-based quantum software company specializing in middleware solutions for near-term quantum hardware, partnered with Quanscient, a Finnish quantum simulation company developing algorithms for multiphysics problems. The partnership was announced through AWS Quantum Technologies Blog and demonstrated quantum computing’s potential to address computational bottlenecks in aerospace and automotive engineering. The collaboration focused on solving the two-dimensional advection-diffusion equation using quantum algorithms on lattices up to 64×64 grid points with 16 qubits, representing advancement toward industrial-scale quantum computational fluid dynamics applications.

Problem Statement

Classical computational fluid dynamics simulations face fundamental limitations when modeling complex systems at high resolutions required for aerospace and automotive applications. Numerical simulations of fluid dynamics, electromagnetics, and thermomechanical problems are essential for designing components such as airplane wings and automotive aerodynamics, yet classical computing approaches encounter prohibitive computational complexity when scaling to industrial-relevant grid sizes. Current classical methods require exponential increases in computational resources as simulation resolution increases, creating bottlenecks that limit engineering optimization capabilities. The aerospace industry demands simulations with billions of grid points to accurately model airflow around aircraft components, while automotive manufacturers need high-fidelity simulations for vehicle aerodynamics and fuel efficiency optimization. Traditional approaches using finite element or finite difference methods struggle with the memory and processing requirements of such large-scale simulations. Engineering teams face trade-offs between simulation accuracy and computational feasibility, often forcing them to use lower-resolution models that may miss critical fluid dynamics phenomena. The computational complexity grows as O(N³) for three-dimensional problems with N grid points per dimension, making high-resolution simulations computationally intractable on classical systems. Industry reports indicate that CFD simulations can consume weeks of computational time on high-performance computing clusters for complex geometries, delaying product development cycles and increasing costs. The limitations become particularly acute when multiple design iterations are required, as each simulation represents a significant computational investment.

Quantum Approach

The collaboration leveraged Quanscient’s Quantum Lattice Boltzmann Method algorithms combined with Haiqu’s quantum middleware technology to address classical CFD limitations through quantum computing. QLBM offers logarithmic scaling advantages over classical methods by encoding fluid distribution functions in quantum states and implementing collision and propagation operations through quantum circuits. Quanscient developed specialized algorithms that map classical partial differential equations to quantum computations, focusing on the advection-diffusion equation as a fundamental building block for more complex fluid dynamics problems. The quantum approach encodes vorticity, stream function, and temperature variables in quantum states, with quantum circuits implementing the lattice Boltzmann collision and propagation steps. Haiqu’s middleware provided critical enabling technology through subcircuit compression, state preparation optimization, and error mitigation routines that allowed execution of deep quantum circuits on near-term hardware. The team implemented a hybrid quantum-classical workflow where quantum processors handle the core lattice Boltzmann evolution while classical systems manage data preprocessing and postprocessing. Amazon Braket provided the cloud infrastructure and access to IonQ’s trapped-ion quantum hardware, with the Braket SDK facilitating seamless integration between classical and quantum components. The quantum circuits incorporated 802 gates in depth for the 12-qubit implementation, representing unprecedented complexity for quantum CFD applications. Haiqu’s approximate subcircuit compilation techniques enabled the execution of circuits that would otherwise be impossible on current noisy intermediate-scale quantum devices. The approach utilized quantum parallelism to potentially achieve exponential speedups over classical methods for certain classes of fluid dynamics problems, particularly those involving large-scale grid simulations where quantum encoding provides memory advantages.

Results and Business Impact

The collaboration achieved the first successful execution of three complete QLBM time evolution steps on quantum hardware, specifically using 16 qubits on the IonQ Aria 1 QPU through Amazon Braket. Results demonstrated quantum advantage potential for simulations involving lattices up to 64×64 grid points, representing approximately 1000 billion effective grid points when accounting for quantum encoding efficiencies. The team executed quantum circuits with 802 gates depth while maintaining fidelity sufficient for meaningful CFD calculations, surpassing previous literature reports for quantum fluid dynamics implementations. Performance metrics showed successful state preparation and measurement across multiple time steps, with quantum results closely matching ideal state vector simulator outputs for validation. The collaboration’s work was recognized as a finalist in the prestigious Airbus-BMW Quantum Mobility Challenge 2024, demonstrating industry recognition of the approach’s commercial potential. AWS Quantum Technologies Blog featured the research, highlighting the synergy between domain-specific algorithm development, quantum middleware optimization, and cloud infrastructure. Technical achievements included successful implementation of measure-re-encode loops for time-stepping, quantum state tomography for result verification, and error mitigation strategies that maintained computational accuracy despite hardware noise. The quantum approach demonstrated logarithmic scaling of qubit requirements with lattice size, compared to linear scaling in classical approaches, indicating potential for significant advantage at larger scales. Benchmark comparisons showed quantum circuits could theoretically simulate systems with over 100 qubits enabling beyond 1000 billion grid points, though current hardware limitations restrict immediate realization of this potential. Industry stakeholders from aerospace and automotive sectors expressed interest in the technology’s potential for design optimization applications, particularly for complex geometries where classical simulation becomes computationally prohibitive. The collaboration established a framework for quantum-classical hybrid workflows that could be adapted to other multiphysics simulation challenges beyond fluid dynamics.

Future Directions

The partnership plans to extend quantum CFD capabilities to three-dimensional problems and more complex fluid dynamics phenomena including turbulence modeling and multi-phase flows. Scaling efforts will focus on implementing the approach on larger quantum processors as they become available, with particular interest in systems offering 100+ qubits for industrial-scale demonstrations. Algorithm development roadmap includes extending QLBM to handle Reynolds stress modeling, heat transfer coupling, and chemical reaction kinetics relevant to combustion applications. Integration with existing engineering workflows represents a key development priority, with plans to create interfaces between quantum CFD tools and standard computer-aided design platforms used in aerospace and automotive industries. The collaboration will pursue fault-tolerant quantum computing implementations as error correction technologies mature, potentially enabling quantum advantage for practical engineering problems within the next decade. Hardware requirements analysis suggests optimal quantum CFD performance on trapped-ion or neutral atom systems offering high-fidelity multi-qubit gates and long coherence times. Research expansion includes investigating quantum algorithms for other partial differential equations commonly encountered in engineering applications, such as structural mechanics and electromagnetic field problems. Commercial deployment strategies focus on quantum cloud services integration, allowing engineering teams to access quantum CFD capabilities without requiring on-premises quantum hardware. Partnership development with major engineering software vendors aims to incorporate quantum algorithms into existing simulation suites, providing seamless adoption pathways for industry users. Long-term vision encompasses quantum-accelerated digital twins for real-time engineering optimization, particularly valuable for applications requiring rapid design iteration such as Formula 1 aerodynamics or aircraft wing optimization.

Conclusion

The Haiqu-Quanscient collaboration represents a significant advancement in practical quantum computing applications for engineering simulation. The successful demonstration of quantum computational fluid dynamics on commercial quantum hardware marks a milestone toward quantum advantage in industrial applications. The partnership’s recognition as finalists in the Airbus-BMW Quantum Mobility Challenge validates the approach’s relevance to aerospace and automotive industry needs. Technical achievements including execution of deep quantum circuits and demonstration of quantum CFD algorithms establish a foundation for future scaling toward industrial-relevant problem sizes. The collaboration’s emphasis on hybrid quantum-classical workflows provides a practical pathway for quantum technology adoption in engineering organizations. Industry implications extend beyond fluid dynamics to broader multiphysics simulation challenges where quantum computing could provide computational advantages. The partnership demonstrates the importance of combining domain expertise in algorithm development with quantum middleware optimization and cloud infrastructure support. Competitive dynamics in quantum CFD suggest early movers will establish significant advantages in next-generation engineering simulation capabilities. The transformative potential for engineering design optimization, particularly in aerospace and automotive applications, positions quantum CFD as a key driver for quantum computing commercialization in the simulation software market.